protein_name
stringlengths
4
4
structure_title
stringlengths
3
322
main_text
stringlengths
0
302k
3M2H
Crystallographic and Single Crystal Spectral Analysis of the Peroxidase Ferryl Intermediate
Crystallographic and Single Crystal Spectral Analysis of the Peroxidase Ferryl Intermediate Yergalem T. Meharenna, Tzanko Doukovb, Huiying Lia, S. Michael Soltisb,*, and Thomas L. Poulosa,* aDepartments of Molecular Biology and Biochemistry, Pharmaceutical Sciences, and Chemistry, University of California, Irvine, California 92697-3900 bMacromolecular Crystallographic Group, The Stanford Synchrotron Radiation Lightsource, SLAC, Stanford University, Stanford, California 94025 Abstract The ferryl (Fe(IV)O) intermediate is important in many heme enzymes and thus the precise nature of the Fe(IV)-O bond is critical in understanding enzymatic mechanisms. The 1.40 Å crystal structure of cytochrome c peroxidase Compound I has been solved as a function of x-ray dose while monitoring the visible spectrum. The Fe-O bond increases linearly from 1.73 Å in the low x- ray dose structure to 1.90 Å in the high dose structure. The low dose structure correlates well with a Fe(IV)=O bond while we postulate that the high dose structure is the cryo-trapped Fe(III)-OH species previously thought to be Fe(IV)-OH. The ferryl, Fe(IV)O, species is a critically important intermediate in a number of metalloproteins and especially heme enzymes. The high redox potential enables Fe(IV)O to serve as a potent oxidant utilized by several heme enzymes including cytochromes P450, nitric oxide synthase (NOS), cytochrome oxidase, and peroxidases. Since the ferryl intermediate is quite stable in peroxidases, most of what we know about Fe(IV)O in heme enzymes derives from studies with peroxidases. In most heme peroxidases one H2O2 oxidizing equivalent is used to oxidize Fe(III) to Fe(IV)O and the second is used to oxidize an organic group to give Fe(IV)R.+ (1) and this activated intermediate is called Compound I. In most heme peroxidases such as horse radish peroxidase (HRP) R is the porphyrin (2) although in yeast cytochome c peroxidase (CCP) R is the active site Trp191 (3). A majority of studies find that the Fe(IV)-O bond is short, somewhat less than 1.7 Å, thus indicating a Fe(IV)=O bond as opposed to a Fe(IV)-OH bond (4). An empirical formula called Badger’s rule relates the calculated Fe-O bond with the calculated vibrational frequency (5) and the experimental frequencies and EXAFS bond distances fit very well to these plots (5) further supporting a Fe(IV)=O double bond. However, a majority of x-ray crystal structures are distinct outliers giving distances closer to 1.8-1.9 Å (4, 6) with one exception being the HRP Compound I structure (7). These differences are not trivial since the longer bond predicts that the ferryl species should be protonated to give Fe(IV)-OH, while the shorter bond gives Fe(IV)=O. The chemistry of each of these species is quite different (8) and knowing the correct structure is essential if we are to understand details of heme enzyme mechanisms. *To whom correspondences should be addressed. T.L.P.: poulos@uci.edu; phone (494) 824-7020; FAX, (949) 824-3280. SUPPORTING INFORMATION AVAILABLE Experimental details and Tables 1S and 2S . This material is available free of charge at http://pubs.acs.org. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 October 27. Published in final edited form as: Biochemistry. 2010 April 13; 49(14): 2984–2986. doi:10.1021/bi100238r. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript A serious problem encountered at high intensity synchrotron x-ray sources is rapid reduction of metal centers, particularly high potential metal centers such as Fe(IV). As a result great care must be taken to minimize reduction and the redox state should be verified during data collection (for example with UV/VIS spectroscopy). We recently found that crystals of the CCP N184R mutant diffract unusually well (9) and thus might provide an opportunity to obtain a low x-ray dose Compound I structure but at sufficiently high resolution to resolve the discrepancies between crystal structures and solution studies. Here we present single crystal spectroscopy together with a composite data collection strategy that has allowed the Fe-O bond distance to be measured as a function of x-ray dose. Fig. 1A shows the single crystal spectrum of CCP Compound I as a function of x-ray dose. Before data collection the spectrum in the 500-700 nm region is identical to the solution spectrum of Compound I. After extensive x-ray exposure (inset to Fig. 1A) the spectrum clearly is no longer that of Compound I nor is this similar to the Fe(III) high spin solution spectrum of CCP. The nature of this species will be discussed further on. Fig. 1B shows the estimated percentage of Compound I remaining in the crystal as a function of x-ray exposure as monitored by changes in the visible spectrum. Based on this plot ~90% of Compound I remains after receiving an estimated x-ray dose of 0.035 MGy (calculations were performed using RADDOSE (10)) or just ~0.1% of the theoretical radiation damage limit for protein crystals, ≈30 MGy (11). Therefore, a data collection strategy for obtaining predominantly Compound I was employed using multiple crystals, none of which received more than 0.035 MGy. With this maximum dose, we estimate that the resulting “integrated” structure is comprised of ~90% Compound I. Crystallographic data collection was carried out at 65 K on SSRL BL9-2 (~4×1011 photons/s at 13.0 KeV). Nearly 100 crystals were mounted and indexed in an automated fashion. Exposures used for indexing were attenuated by 99% and did not significantly contribute to reduction of Compound I. For each crystal, data collections were carried out in 15 separate runs. Run 1 consisted of 5° of data, representing the first 0.035 MGy of x-ray exposure. Then the same 5° of scanning angle were recollected 12 more times giving runs 2 through 13 with increased x-ray dose. In run 14 a full 120° of data were collected in order to fully reduce the crystal followed by run 15 which again repeated the same 5° representing the highest x-ray dose. The same 15-run data collection protocol was adopted for similarly sized crystals and the scanning angles were chosen to optimize the completeness of the data. Each composite data set was assembled by merging 5° of data with identical run numbers from 19 crystals. A total of 15 structures at 1.40 Å resolution were refined providing a picture of the structural changes associated with increasing x-ray dose (Table S1). In Fig. 2A we compare the structures of the low dose (set 1) and the ferric resting state 1.06 Å structure of the N184R mutant (3E2O) (9). In the ferric resting state a water molecule is positioned ≈ 2.0 Å from the heme iron while in the low dose data set the Fe-O oxygen distance is 1.73 Å. In both structures a water molecule is within H-bonding distance of the Fe-linked oxygen. In the ferric state the heme iron is displaced from the porphyrin plane by 0.18 Å toward the proximal His ligand while in Compound I the iron is displaced by 0.07 Å in the opposite direction toward the distal pocket. Thus the net movement of the iron is ≈ 0.25 Å relative to the porphyrin plane owing to the oxidation of the iron from Fe(III) to Fe(IV). Note that the water molecules in the distal pocket, including the one closest to the iron, are located in nearly the same position relative to the heme while the His-Fe bond increases from 2.07 Å to 2.12 Å upon oxidation to Fe(IV). Thus, the short Fe-O bond in the Compound I structure is due in large part to motion of the iron. As in our previous work on peroxide treated CCP (12) Arg48 in the distal pocket forms a 2.78 Å H-bond with the iron linked O atom. Meharenn et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 October 27. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript We next compare the set 1 (low dose, Fig. 2C) and set 15 (high dose, Fig. 2D) structures. At the 4.0 σ contour level the electron density between the Fe and O atoms is not continuous in set 15 and the Fe-O bond length has increased from 1.73 Å to 1.90 Å. The local water structure remains largely unchanged. The changes owing to x-ray induced reduction are highlighted by examining a Fo(low dose)-Fo(high dose) electron density difference map contoured at ±5σ (Fig. 2B). This map clearly shows that the iron is positioned quite differently in each structure and is closer toward the distal pocket in the low dose structure. In addition the His-Fe bond decreases from 2.12 Å to 2.07 Å upon photo reduction again due to motion of the iron back into the porphyrin plane. The only other notable feature in the Fo(low dose)-Fo(high dose) difference map is around the carbonyl O atom of the heme ligand, His175. This group is slightly less than 0.1 Å closer to Trp191 in the low dose structure and may reflect a local tightening of the structure around the Trp191 cation radical that provides additional electrostatic stability. The various heme parameter distances are provided in Table S2. The structures of set 1 through set 13 next were used to assess how the Fe-O bond changes as a function of x-ray dose and the results are shown in Fig. 3. The fit to a simple straight line equation is remarkably good and extrapolates to zero dose at a Fe-O bond distance of 1.72 Å. Raman data (13) coupled with Badger’s rule (4) gives a Fe-O bond of 1.68 Å. Therefore, the low dose Compound I crystal structure agrees within 0.04 Å with the Raman data and the ferryl center in CCP Compound I can best be described as Fe(IV)=O and not Fe(IV)-OH. The nature of the ferryl center after extensive x-ray exposure is intriguing: the short Fe-O bond (1.90 Å) compared to the ≈ 2.0 - 2.3 Å observed in Fe(III) high spin peroxidase structures and the total lack of similarity between the high dose spectrum (Fig. 1) and the solution spectrum of Fe(III) CCP shows that the high dose structure is not that of Fe(III) high spin CCP. The spectrum is similar to that of HRP Fe(II) in both the crystal and solution except in HRP there is no ligand coordinated to the iron (7). Since we clearly see a ligand coordinated to the iron in the high dose structure we very likely have trapped either Fe(II)- OH or Fe(III)-OH. Unfortunately we cannot compare single crystal and solution spectra since formation of Fe(III)-OH, and presumably Fe(II)-OH, requires an increase in pH and CCP is not stable above pH 8.0. Our first goal in this study was to further develop the necessary methods and protocols required to obtain x-ray structures of high potential intermediates in metalloproteins. This requires isomorphous crystals that diffract well in order to have sufficient resolution to obtain the level of accuracy required for estimating subtle bond parameter differences (7). Coupling data collection with on-line single crystal spectroscopy to monitor the redox state is also essential. Our second goal was to obtain a very low dose x-ray structure of CCP Compound I at high resolution in order to reconcile the long standing differences observed in the Fe(IV)-O bond distance between most available x-ray structures and other biophysical techniques. The low dose CCP Compound I structure agrees within 0.04 Å of previous experimental estimates indicating that the ferryl species in Compound I is Fe(IV)=O and not Fe(IV)-OH. It should be noted that from the perspective of the heme, CCP Compound I is equivalent to HRP Compound II since both contain Fe(IV) with no porphyrin radical. Thus it is likely that other crystal structures where the Fe(IV)-O bond in Compound II was estimated to be 1.8 Å (7, 14) or longer may also have a significant amount of a reduced iron species. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Meharenn et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 October 27. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Acknowledgments We thank Aina Cohen, John Kovarick and Michael Hollenbeck for their contribution to the design and implementation of the single-crystal microspectrophotometer. Portions of this research were carried out at the Stanford Synchrotron Radiation Lightsource, a national user facility operated by Stanford University on behalf of the U.S. Department of Energy, Office of Basic Energy Sciences. The SSRL Structural Molecular Biology Program is supported by the Department of Energy, Office of Biological and Environmental Research, and by the National Institutes of Health, National Center for Research Resources, Biomedical Technology Program, and the National Institute of General Medical Sciences. Work at UCI was supported by NIH grant GM42614 (TLP). References 1. Poulos TL, Kraut J. J Biol Chem. 1980; 255:8199–8205. [PubMed: 6251047] 2. Dolphin D, Forman A, Borg DC, Fajer J, Felton RH. Proc Natl Acad Sci USA. 1971; 68:614–618. [PubMed: 5276770] 3. Sivaraja M, Goodin DB, Smith M, Hoffman BM. Science. 1989; 245:738–740. [PubMed: 2549632] 4. Behan RK, Green MT. J Inorg Biochem. 2006; 100:448–459. [PubMed: 16500711] 5. Green MT. J Am Chem Soc. 2006; 128:1902–1906. [PubMed: 16464091] 6. Hersleth HP, Hsiao YW, Ryde U, Gorbitz CH, Andersson KK. Chem Biodivers. 2008; 5:2067– 2089. [PubMed: 18972498] 7. Berglund GI, Carlsson GH, Smith AT, Szoke H, Henriksen A, Hajdu J. Nature. 2002; 417:463–468. [PubMed: 12024218] 8. Green MT, Dawson JH, Gray HB. Science. 2004; 304:1653–1656. [PubMed: 15192224] 9. Meharenna YT, Oertel P, Bhaskar B, Poulos TL. Biochemistry. 2008; 47:10324–10332. [PubMed: 18771292] 10. Paithankar KS, Owen RL, Garman EF. J Synchr Radiat. 2009; 16:152–162. 11. Owen RL, Rudino-Pinera E, Garman EF. Proc Natl Acad Sci U S A. 2006; 103:4912–4917. [PubMed: 16549763] 12. Bonagura CA, Bhaskar B, Shimizu H, Li H, Sundaramoorthy M, McRee D, Goodin DB, Poulos TL. Biochemistry. 2003; 42:5600–5608. [PubMed: 12741816] 13. Reczek CM, Sitter AJ, Terner J. J Molec Struc. 1989; 214:27–41. 14. Hersleth HP, Dalhus B, H GC, A KK. J Biol Inorg Chem. 2002; 7:299–304. [PubMed: 11935353] Meharenn et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 October 27. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Single crystal spectra of CCP Compound I as a function of x-ray dose. Prior to x-ray exposure the spectrum is identical to the solution spectrum of Compound I. The estimated percentage of Compound I remaining in the crystal as a function of x-ray dose in panel B was based on the decrease in the absorbance peak at 634 nm. Meharenn et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 October 27. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. A) Superposition of the low dose structure (red) on the Fe(III) structure (cyan). Note that the iron is displaced below the plane of the heme in the Fe(III) structure and above the plane of the heme in the low dose structure; B) Fo(low dose)-Fo(high dose) electron density difference map using phases obtained from the low dose structure. The map is contoured at -5.0σ (green) and +5.0σ (blue); C and D) 2Fo-Fc electron density maps contoured at 4.0σ for the dose data set 1 (panel C) and high dose data set 15 (panel D). Oxygen and water molecules are represented by the small spheres. Meharenn et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 October 27. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Plot of the Fe-O distance as a function of x-ray dose. Each of the 13 structures was refined exactly the same way using the same starting structure and two different protocols. In the first the distances between the Fe and N atoms (4 pyrrole and 1 His closed circles) were restrained while in the second protocol no restraints were applied (open circles). At no time were restraints imposed on the Fe-O distance. The estimated error in the Fe-O bond distance is ≈0.017Å (see Supporting Information). Meharenn et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 October 27. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
3M2I
Crystallographic and Single Crystal Spectral Analysis of the Peroxidase Ferryl Intermediate
Crystallographic and Single Crystal Spectral Analysis of the Peroxidase Ferryl Intermediate Yergalem T. Meharenna, Tzanko Doukovb, Huiying Lia, S. Michael Soltisb,*, and Thomas L. Poulosa,* aDepartments of Molecular Biology and Biochemistry, Pharmaceutical Sciences, and Chemistry, University of California, Irvine, California 92697-3900 bMacromolecular Crystallographic Group, The Stanford Synchrotron Radiation Lightsource, SLAC, Stanford University, Stanford, California 94025 Abstract The ferryl (Fe(IV)O) intermediate is important in many heme enzymes and thus the precise nature of the Fe(IV)-O bond is critical in understanding enzymatic mechanisms. The 1.40 Å crystal structure of cytochrome c peroxidase Compound I has been solved as a function of x-ray dose while monitoring the visible spectrum. The Fe-O bond increases linearly from 1.73 Å in the low x- ray dose structure to 1.90 Å in the high dose structure. The low dose structure correlates well with a Fe(IV)=O bond while we postulate that the high dose structure is the cryo-trapped Fe(III)-OH species previously thought to be Fe(IV)-OH. The ferryl, Fe(IV)O, species is a critically important intermediate in a number of metalloproteins and especially heme enzymes. The high redox potential enables Fe(IV)O to serve as a potent oxidant utilized by several heme enzymes including cytochromes P450, nitric oxide synthase (NOS), cytochrome oxidase, and peroxidases. Since the ferryl intermediate is quite stable in peroxidases, most of what we know about Fe(IV)O in heme enzymes derives from studies with peroxidases. In most heme peroxidases one H2O2 oxidizing equivalent is used to oxidize Fe(III) to Fe(IV)O and the second is used to oxidize an organic group to give Fe(IV)R.+ (1) and this activated intermediate is called Compound I. In most heme peroxidases such as horse radish peroxidase (HRP) R is the porphyrin (2) although in yeast cytochome c peroxidase (CCP) R is the active site Trp191 (3). A majority of studies find that the Fe(IV)-O bond is short, somewhat less than 1.7 Å, thus indicating a Fe(IV)=O bond as opposed to a Fe(IV)-OH bond (4). An empirical formula called Badger’s rule relates the calculated Fe-O bond with the calculated vibrational frequency (5) and the experimental frequencies and EXAFS bond distances fit very well to these plots (5) further supporting a Fe(IV)=O double bond. However, a majority of x-ray crystal structures are distinct outliers giving distances closer to 1.8-1.9 Å (4, 6) with one exception being the HRP Compound I structure (7). These differences are not trivial since the longer bond predicts that the ferryl species should be protonated to give Fe(IV)-OH, while the shorter bond gives Fe(IV)=O. The chemistry of each of these species is quite different (8) and knowing the correct structure is essential if we are to understand details of heme enzyme mechanisms. *To whom correspondences should be addressed. T.L.P.: poulos@uci.edu; phone (494) 824-7020; FAX, (949) 824-3280. SUPPORTING INFORMATION AVAILABLE Experimental details and Tables 1S and 2S . This material is available free of charge at http://pubs.acs.org. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 October 27. Published in final edited form as: Biochemistry. 2010 April 13; 49(14): 2984–2986. doi:10.1021/bi100238r. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript A serious problem encountered at high intensity synchrotron x-ray sources is rapid reduction of metal centers, particularly high potential metal centers such as Fe(IV). As a result great care must be taken to minimize reduction and the redox state should be verified during data collection (for example with UV/VIS spectroscopy). We recently found that crystals of the CCP N184R mutant diffract unusually well (9) and thus might provide an opportunity to obtain a low x-ray dose Compound I structure but at sufficiently high resolution to resolve the discrepancies between crystal structures and solution studies. Here we present single crystal spectroscopy together with a composite data collection strategy that has allowed the Fe-O bond distance to be measured as a function of x-ray dose. Fig. 1A shows the single crystal spectrum of CCP Compound I as a function of x-ray dose. Before data collection the spectrum in the 500-700 nm region is identical to the solution spectrum of Compound I. After extensive x-ray exposure (inset to Fig. 1A) the spectrum clearly is no longer that of Compound I nor is this similar to the Fe(III) high spin solution spectrum of CCP. The nature of this species will be discussed further on. Fig. 1B shows the estimated percentage of Compound I remaining in the crystal as a function of x-ray exposure as monitored by changes in the visible spectrum. Based on this plot ~90% of Compound I remains after receiving an estimated x-ray dose of 0.035 MGy (calculations were performed using RADDOSE (10)) or just ~0.1% of the theoretical radiation damage limit for protein crystals, ≈30 MGy (11). Therefore, a data collection strategy for obtaining predominantly Compound I was employed using multiple crystals, none of which received more than 0.035 MGy. With this maximum dose, we estimate that the resulting “integrated” structure is comprised of ~90% Compound I. Crystallographic data collection was carried out at 65 K on SSRL BL9-2 (~4×1011 photons/s at 13.0 KeV). Nearly 100 crystals were mounted and indexed in an automated fashion. Exposures used for indexing were attenuated by 99% and did not significantly contribute to reduction of Compound I. For each crystal, data collections were carried out in 15 separate runs. Run 1 consisted of 5° of data, representing the first 0.035 MGy of x-ray exposure. Then the same 5° of scanning angle were recollected 12 more times giving runs 2 through 13 with increased x-ray dose. In run 14 a full 120° of data were collected in order to fully reduce the crystal followed by run 15 which again repeated the same 5° representing the highest x-ray dose. The same 15-run data collection protocol was adopted for similarly sized crystals and the scanning angles were chosen to optimize the completeness of the data. Each composite data set was assembled by merging 5° of data with identical run numbers from 19 crystals. A total of 15 structures at 1.40 Å resolution were refined providing a picture of the structural changes associated with increasing x-ray dose (Table S1). In Fig. 2A we compare the structures of the low dose (set 1) and the ferric resting state 1.06 Å structure of the N184R mutant (3E2O) (9). In the ferric resting state a water molecule is positioned ≈ 2.0 Å from the heme iron while in the low dose data set the Fe-O oxygen distance is 1.73 Å. In both structures a water molecule is within H-bonding distance of the Fe-linked oxygen. In the ferric state the heme iron is displaced from the porphyrin plane by 0.18 Å toward the proximal His ligand while in Compound I the iron is displaced by 0.07 Å in the opposite direction toward the distal pocket. Thus the net movement of the iron is ≈ 0.25 Å relative to the porphyrin plane owing to the oxidation of the iron from Fe(III) to Fe(IV). Note that the water molecules in the distal pocket, including the one closest to the iron, are located in nearly the same position relative to the heme while the His-Fe bond increases from 2.07 Å to 2.12 Å upon oxidation to Fe(IV). Thus, the short Fe-O bond in the Compound I structure is due in large part to motion of the iron. As in our previous work on peroxide treated CCP (12) Arg48 in the distal pocket forms a 2.78 Å H-bond with the iron linked O atom. Meharenn et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 October 27. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript We next compare the set 1 (low dose, Fig. 2C) and set 15 (high dose, Fig. 2D) structures. At the 4.0 σ contour level the electron density between the Fe and O atoms is not continuous in set 15 and the Fe-O bond length has increased from 1.73 Å to 1.90 Å. The local water structure remains largely unchanged. The changes owing to x-ray induced reduction are highlighted by examining a Fo(low dose)-Fo(high dose) electron density difference map contoured at ±5σ (Fig. 2B). This map clearly shows that the iron is positioned quite differently in each structure and is closer toward the distal pocket in the low dose structure. In addition the His-Fe bond decreases from 2.12 Å to 2.07 Å upon photo reduction again due to motion of the iron back into the porphyrin plane. The only other notable feature in the Fo(low dose)-Fo(high dose) difference map is around the carbonyl O atom of the heme ligand, His175. This group is slightly less than 0.1 Å closer to Trp191 in the low dose structure and may reflect a local tightening of the structure around the Trp191 cation radical that provides additional electrostatic stability. The various heme parameter distances are provided in Table S2. The structures of set 1 through set 13 next were used to assess how the Fe-O bond changes as a function of x-ray dose and the results are shown in Fig. 3. The fit to a simple straight line equation is remarkably good and extrapolates to zero dose at a Fe-O bond distance of 1.72 Å. Raman data (13) coupled with Badger’s rule (4) gives a Fe-O bond of 1.68 Å. Therefore, the low dose Compound I crystal structure agrees within 0.04 Å with the Raman data and the ferryl center in CCP Compound I can best be described as Fe(IV)=O and not Fe(IV)-OH. The nature of the ferryl center after extensive x-ray exposure is intriguing: the short Fe-O bond (1.90 Å) compared to the ≈ 2.0 - 2.3 Å observed in Fe(III) high spin peroxidase structures and the total lack of similarity between the high dose spectrum (Fig. 1) and the solution spectrum of Fe(III) CCP shows that the high dose structure is not that of Fe(III) high spin CCP. The spectrum is similar to that of HRP Fe(II) in both the crystal and solution except in HRP there is no ligand coordinated to the iron (7). Since we clearly see a ligand coordinated to the iron in the high dose structure we very likely have trapped either Fe(II)- OH or Fe(III)-OH. Unfortunately we cannot compare single crystal and solution spectra since formation of Fe(III)-OH, and presumably Fe(II)-OH, requires an increase in pH and CCP is not stable above pH 8.0. Our first goal in this study was to further develop the necessary methods and protocols required to obtain x-ray structures of high potential intermediates in metalloproteins. This requires isomorphous crystals that diffract well in order to have sufficient resolution to obtain the level of accuracy required for estimating subtle bond parameter differences (7). Coupling data collection with on-line single crystal spectroscopy to monitor the redox state is also essential. Our second goal was to obtain a very low dose x-ray structure of CCP Compound I at high resolution in order to reconcile the long standing differences observed in the Fe(IV)-O bond distance between most available x-ray structures and other biophysical techniques. The low dose CCP Compound I structure agrees within 0.04 Å of previous experimental estimates indicating that the ferryl species in Compound I is Fe(IV)=O and not Fe(IV)-OH. It should be noted that from the perspective of the heme, CCP Compound I is equivalent to HRP Compound II since both contain Fe(IV) with no porphyrin radical. Thus it is likely that other crystal structures where the Fe(IV)-O bond in Compound II was estimated to be 1.8 Å (7, 14) or longer may also have a significant amount of a reduced iron species. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Meharenn et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 October 27. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Acknowledgments We thank Aina Cohen, John Kovarick and Michael Hollenbeck for their contribution to the design and implementation of the single-crystal microspectrophotometer. Portions of this research were carried out at the Stanford Synchrotron Radiation Lightsource, a national user facility operated by Stanford University on behalf of the U.S. Department of Energy, Office of Basic Energy Sciences. The SSRL Structural Molecular Biology Program is supported by the Department of Energy, Office of Biological and Environmental Research, and by the National Institutes of Health, National Center for Research Resources, Biomedical Technology Program, and the National Institute of General Medical Sciences. Work at UCI was supported by NIH grant GM42614 (TLP). References 1. Poulos TL, Kraut J. J Biol Chem. 1980; 255:8199–8205. [PubMed: 6251047] 2. Dolphin D, Forman A, Borg DC, Fajer J, Felton RH. Proc Natl Acad Sci USA. 1971; 68:614–618. [PubMed: 5276770] 3. Sivaraja M, Goodin DB, Smith M, Hoffman BM. Science. 1989; 245:738–740. [PubMed: 2549632] 4. Behan RK, Green MT. J Inorg Biochem. 2006; 100:448–459. [PubMed: 16500711] 5. Green MT. J Am Chem Soc. 2006; 128:1902–1906. [PubMed: 16464091] 6. Hersleth HP, Hsiao YW, Ryde U, Gorbitz CH, Andersson KK. Chem Biodivers. 2008; 5:2067– 2089. [PubMed: 18972498] 7. Berglund GI, Carlsson GH, Smith AT, Szoke H, Henriksen A, Hajdu J. Nature. 2002; 417:463–468. [PubMed: 12024218] 8. Green MT, Dawson JH, Gray HB. Science. 2004; 304:1653–1656. [PubMed: 15192224] 9. Meharenna YT, Oertel P, Bhaskar B, Poulos TL. Biochemistry. 2008; 47:10324–10332. [PubMed: 18771292] 10. Paithankar KS, Owen RL, Garman EF. J Synchr Radiat. 2009; 16:152–162. 11. Owen RL, Rudino-Pinera E, Garman EF. Proc Natl Acad Sci U S A. 2006; 103:4912–4917. [PubMed: 16549763] 12. Bonagura CA, Bhaskar B, Shimizu H, Li H, Sundaramoorthy M, McRee D, Goodin DB, Poulos TL. Biochemistry. 2003; 42:5600–5608. [PubMed: 12741816] 13. Reczek CM, Sitter AJ, Terner J. J Molec Struc. 1989; 214:27–41. 14. Hersleth HP, Dalhus B, H GC, A KK. J Biol Inorg Chem. 2002; 7:299–304. [PubMed: 11935353] Meharenn et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 October 27. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Single crystal spectra of CCP Compound I as a function of x-ray dose. Prior to x-ray exposure the spectrum is identical to the solution spectrum of Compound I. The estimated percentage of Compound I remaining in the crystal as a function of x-ray dose in panel B was based on the decrease in the absorbance peak at 634 nm. Meharenn et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 October 27. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. A) Superposition of the low dose structure (red) on the Fe(III) structure (cyan). Note that the iron is displaced below the plane of the heme in the Fe(III) structure and above the plane of the heme in the low dose structure; B) Fo(low dose)-Fo(high dose) electron density difference map using phases obtained from the low dose structure. The map is contoured at -5.0σ (green) and +5.0σ (blue); C and D) 2Fo-Fc electron density maps contoured at 4.0σ for the dose data set 1 (panel C) and high dose data set 15 (panel D). Oxygen and water molecules are represented by the small spheres. Meharenn et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 October 27. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Plot of the Fe-O distance as a function of x-ray dose. Each of the 13 structures was refined exactly the same way using the same starting structure and two different protocols. In the first the distances between the Fe and N atoms (4 pyrrole and 1 His closed circles) were restrained while in the second protocol no restraints were applied (open circles). At no time were restraints imposed on the Fe-O distance. The estimated error in the Fe-O bond distance is ≈0.017Å (see Supporting Information). Meharenn et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 October 27. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
3M2K
Crystal Structure of fluorescein-labeled Class A -beta lactamase PenP in complex with cefotaxime
RESEARCH ARTICLE Open Access Structural studies of the mechanism for biosensing antibiotics in a fluorescein- labeled β-lactamase Wai-Ting Wong, Ho-Wah Au, Hong-Kin Yap, Yun-Chung Leung*, Kwok-Yin Wong* and Yanxiang Zhao* Abstract Background: β-lactamase conjugated with environment-sensitive fluorescein molecule to residue 166 on the Ω-loop near its catalytic site is a highly effective biosensor for β-lactam antibiotics. Yet the molecular mechanism of such fluorescence-based biosensing is not well understood. Results: Here we report the crystal structure of a Class A β-lactamase PenP from Bacillus licheniformis 749/C with fluorescein conjugated at residue 166 after E166C mutation, both in apo form (PenP-E166Cf) and in covalent complex form with cefotaxime (PenP-E166Cf-cefotaxime), to illustrate its biosensing mechanism. In the apo structure the fluorescein molecule partially occupies the antibiotic binding site and is highly dynamic. In the PenP- E166Cf-cefatoxime complex structure the binding and subsequent acylation of cefotaxime to PenP displaces fluorescein from its original location to avoid steric clash. Such displacement causes the well-folded Ω-loop to become fully flexible and the conjugated fluorescein molecule to relocate to a more solvent exposed environment, hence enhancing its fluorescence emission. Furthermore, the fully flexible Ω-loop enables the narrow-spectrum PenP enzyme to bind cefotaxime in a mode that resembles the extended-spectrum β-lactamase. Conclusions: Our structural studies indicate the biosensing mechanism of a fluorescein-labelled β-lactamase. Such findings confirm our previous proposal based on molecular modelling and provide useful information for the rational design of β-lactamase-based biosensor to detect the wide spectrum of β-lactam antibiotics. The observation of increased Ω-loop flexibility upon conjugation of fluorophore may have the potential to serve as a screening tool for novel β-lactamase inhibitors that target the Ω-loop and not the active site. Background β-Lactamase is one of the major mechanisms of antibio- tic resistance in bacteria. Enzymes of this family deacti- vate β-lactam antibiotics by hydrolyzing the conserved β-lactam moiety in the antibiotics and rendering them ineffective to bind to their target proteins, the penicillin- binding proteins (PBPs), which are essential for bacterial cell wall synthesis and survival [1,2]. Detailed mechanis- tic studies of these enzymes over the past decades have revealed a conserved mechanism of β-lactam hydrolysis that consists of two steps, the acylation step in which the β-lactam ring is “opened” and acylated to the side chain hydroxyl group of Ser70 through nucleophilic attack to form the enzyme-substrate acyl adduct ES*; followed by the deacylation step in which the ES* inter- mediate is hydrolyzed and released as E + P facilitated by Glu166 (residue numbering according to the most conserved Class A β-lactamases) [3]. The substrate profile of a β-lactamase in hydrolyzing diverse β-lactam antibiotics is strongly influenced by a structural element termed Ω-loop, a short stretch of residues on the surface of the β-lactamase structure that forms part of the outer part of the antibiotic binding site [4-9]. For narrow-spectrum β-lactamases such as the PenP used in this study and the clinically significant TEM-1 or SHV-1 enzymes, Ω-loop is tightly packed onto the enzyme active site through hydrophobic and electrostatic interactions with residues lining the * Correspondence: bctleung@inet.polyu.edu.hk; bckywong@inet.polyu.edu.hk; bcyxzhao@inet.polyu.edu.hk Department of Applied Biology and Chemical Technology, Central Laboratory of the Institute of Molecular Technology for Drug Discovery and Synthesis, The Hong Kong Polytechnic University, Hung Hom, Hong Hong, China Wong et al. BMC Structural Biology 2011, 11:15 http://www.biomedcentral.com/1472-6807/11/15 © 2011 Wong et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. catalytic site, posing as steric hindrance for binding of second- or third-generation antibiotics with bulky side chains attached onto the β-lactam nucleus. Many mutant strains of TEM- and SHV-like β-lactamases overcome this inefficiency and broaden their hydrolytic profile by acquiring mutations in the Ω-loop region to render this region more flexible to accommodate large- sized antibiotics [4-10]. Many extended-spectrum β-lactamases have significantly extended Ω-loop, result- ing in an enlarged active site that readily binds to and hydrolyzes almost all antibiotics [11-13]. Exploiting the proximity of Ω-loop to the antibiotic binding site and its structural flexibility, we have suc- cessfully converted a β-lactamase PenPC from Bacillus cereus 569/H into a biosensor for β-lactam antibiotics by mutating the catalytically critical residue Glu166 on the Ω-loop to cysteine and conjugating an environment- sensitive fluorescein molecule to its reactive side chain thiol group to form PenPC-E166Cf as reported in pre- vious studies [14-16]. Fluorescein is an environment- sensitive fluorophore with suppressed fluorescence in a hydrophobic environment but fluoresces strongly in a polar aqueous environment [17]. The mutation of Glu166 to cysteine severely reduces the efficiency of the deacylation step of β-lactamase catalysis, rendering the enzyme to stall at the acylation step and form a stable ES* acyl adduct that enhances the fluorescence emission of the conjugated fluorescein [16]. We have speculated that the fluorescein molecule is positioned near the cat- alytic site so that the binding and subsequent acylation of β-lactam antibiotics would displace it to a more polar environment, enhancing its fluorescence intensity [16]. Here we report structural studies of fluorescein- conjugated PenP β-lactamase from Bacillus licheniformis 749/C to validate our proposed biosensing mechanism. The structural findings suggest an important role of Ω- loop in the biosensing process, which will help the rational design of improved biosensors for β-lactam detection as well as for novel antibiotics discovery. Results and Discussion The biosensing profile of PenP-E166Cf The biosensing profile of fluorescein conjugated PenP (PenP-E166Cf) for detecting β-lactam antibiotics have never been reported before. In our previous study, a highly similar enzyme, PenPC from Bacillus cereus 569/ H with 58% amino acid sequence identity to PenP, was successfully engineered into a biosensor using the same design scheme (PenPC-E166Cf) [15,16]. We chose to work with PenP in this study for the advantage of its easy propensity for crystallization, which would enable structural studies to understand its biosensing mechan- ism at atomic resolution. PenPC, on the other hand, has poor thermal stability and is difficult to crystallize. Because of the high sequence similarity between these two proteins, as well as the general sequence conserva- tion among all Class A β-lactamase enzymes we expect that PenP can serve as a good model system to under- stand the biosensing mechanism of fluorescein-based biosensing. Indeed the biosensing profile of PenP-E166Cf is highly similar to that of PenPC-E166Cf. The conjugation of fluorescein to the mutated Cys166 residue through thiol linkage is highly efficient for PenP. The ESI-MS profile confirmed that over 90% of PenP was labelled by the fluorophore and converted to PenP-E166Cf, with little unlabelled PenP remaining (Figure 1a). The fluorescence scanning spectrum of PenP-E166Cf shows an increase of ~25% in emitted intensity when the antibiotic cefotaxime is present at 10 μM concentration (Figure 1b). A variety of β-lactam antibiotics, including the first-, second- and third-generation compounds with diverse chemical struc- tures in addition to the conserved β-lactam core, induce significant fluorescence enhancement in PenP-E166Cf at concentration as low as 1 μM (Figure 1c). Lastly the time-dependent spectra of PenP-E166Cf in the presence of cefotaxime at different concentrations ranging from 0.01 μM to 10 μM shows that PenP-E166Cf can detect cefotaxime at concentration as low as 0.01 μM and the fluorescence response is saturated at 1 μM (Figure 1d). The structure of PenP-E166Cf in apo form PenP-E166Cf readily crystallized in the form of clustered needles. These crystals were tinted in bright yellow col- our, indicating the presence of fluorescein (data not shown YZ). To confirm that fluorescein remaining con- jugated to the protein in the crystal form we harvested and thoroughly washed these yellow-coloured crystals and analyzed the dissolved crystals on SDS-PAGE gel under both visible and UV light. A band corresponding to PenP (~30.5 kDa) is clearly visible under both condi- tions, confirming that the crystals are indeed of PenP- E166Cf (data not shown YZ). The structure of PenP-E166Cf was solved by molecu- lar replacement using the known structure of PenP (PDB ID 4BLM) as search model. Two molecules of PenP-E166Cf are found in each asymmetric unit. Struc- ture rebuilding and refinement were done in CCP4 pro- gram [18]. The overall structure of PenP-E166Cf is largely identical to that of the wild-type unlabeled PenP. The RMSD of all 4011 protein atoms between the labeled and wild-type structures is just ~1.5 Å. For main chain atoms, the RMSD is only 0.8 Å. Key residues lin- ing the catalytic site, including Ser70 and mutated Cys166 are virtually identical between the labeled and wild-type structures (Figure 2a). In summary the conju- gation of fluorescein to PenP does not alter its overall structural folding. Wong et al. BMC Structural Biology 2011, 11:15 http://www.biomedcentral.com/1472-6807/11/15 Page 2 of 8 The fluorescein molecule was modeled onto the PenP structure after careful inspection of the fo-fc and 2fo-fc electron density map. These maps are not of high qual- ity at regions around the fluorescein conjugation site, with only pieces of discontinuous density visible at 2.0 s contour level in the fo-fc map (Figure 2a). We tried our best to fit fluorescein into these pieces of electron den- sity, particularly matching the melaimide group to a piece of electron density near the thiol side chain of Cys166, as well as matching the xanthene group at the end of the fluorescein molecule to a large piece of elec- tron density near the catalytic site (Figure 2a). This modeled structure is stable after rounds of structural refinement, showing good electron density for the Ω-loop residues and the fluorescein molecule at 1.0 s contour level in the 2fo-fc map, suggesting that our fitting is reasonable (Figure 2b). However, no electron density was visible for the benzoic group in the 30053 Da (a) (d) PenP(29608Da)+fluorescein(427Da)+water(18Da)=30053Da (c) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 505 525 545 565 585 Wavelength (nm) Relative Fluorescence Intensity E166Cf only 10-5M cefotaxime (b) 10-5 10-6 10-7 5x10-8 10-8 E166Cf only Figure 1 Biosensing of b-lactam antibiotics by fluorescein-labelled PenP. (a) De-convoluted ESI mass spectrum of PenP-E166Cf. The add-up at the bottom confirms the correct mass of the labelled protein. (b) Fluorescence scanning spectra of PenP-E166Cf in the presence of 10-5M cefotaxime in 50 mM phosphate buffer (pH 7.0). (c) Change in fluorescence emission of PenP-E166Cf after incubation with different antibiotics (cefotaxime, ceftriaxone, ceftazidime, cephaloridine, cephalothin, cefoxitin, cefuroxime, penicillin G and ampicillin) at 10-6 M for 100 s. (d) Time- dependent fluorescence spectra in the presence of different concentrations (1 × 10-8 M - 1 × 10-5 M) of cefotaxime monitored at 515 nm. Wong et al. BMC Structural Biology 2011, 11:15 http://www.biomedcentral.com/1472-6807/11/15 Page 3 of 8 mid-region of the fluorophore molecule, indicating that this region is more disordered as compared to other parts of the fluorophore molecule. In our PenP-E166Cf structure the fluorescein molecule partially occupies the outer edge of the antibiotic binding region and is in close contact with several residues at the catalytic site. The maleimide moiety near the thiol link- age site is inserted into the catalytic core, located within 2.5 Å from the side chain of Ser70 on one side and 3.5 Å away from Ω-loop on the other side. The xanthene group near the other end of the fluorescein molecule extends toward the solvent (Figure 2b), loosely packed against β-strand B3 that forms part of the extended substrate binding area involved in coordinating antibiotics as shown in the extended-spectrum class A β-lactamase, Toho-1, in complex with cefotaxime, cephalothin, and benzylpenicillin [19]. No specific interactions were observed between fluorescein and the protein. Total sol- vent accessible area is 188 Å2, 33% of the total surface area, indicating that fluorescein is partially packed against the PenP molecule and not fully solvent exposed. The fluorescein molecule is highly dynamic, as reflected by the poor electron density map as well as high average temperature factor (~72.3). In contrast, the rest of the structure shows excellent electron density and low aver- age temperature factor (~23.5) that is typical of the 2.2 Å data set. The Ω-loop, on which the fluorescein molecule is conjugated, was little affected by the dynamic fluoro- phore and adopts the same conformation as that of the unlabelled PenP (Figure 2b). The structure of PenP-E166Cf in complex with cefotaxime We chose to determine the PenP-E166Cf-cefotaxime structure, using cefotaxime as a representative of the many β-lactam antibiotics because of its positive fluores- cence response induced in PenP-E166Cf as well as its chemical structure that contains functional groups typi- cal of both second- and third-generation antibiotics. Cefotaxime was soaked into the PenP-E166Cf crystals by incubating the crystals in the reservoir solution with 0.01 M cefotaxime added for 20 minutes. The PenP- E166C structure, without the conjugated fluorescein molecule, was used as the starting model for structure determination. After initial rounds of refinement both the fo-fc and 2fo-fc electron density maps were carefully inspected for evidence of cefotaxime and fluorescein, as well as for any structural changes on PenP. The cefotaxime was clearly visible in fo-fc electron density map as covalently bonded through its carbonyl carbon atom C7 to the Og atom of Ser70, which repre- sents the acylated ES* adduct (Figure 3a). But we could not identify any electron density in either fo-fc or 2fo-fc :-loop fluorescein Ser70 Cys166 2.5 Å (b) Ser70 Cys166 :-loop (a) Figure 2 Crystal structure of PenP-E166Cf. (a) The fo-fc omit map of fluorescein-5-maleimide contoured at 2.0 s. (b) The 2fo-fc map of Phe165 to Asn170 and fluorescein-5-maleimide contoured at 1.0 s. Side chains of Phe165 to Asn170 and Ser70 are shown in cpk cylinder model. Fluorescein is shown in green cylinder model. Wong et al. BMC Structural Biology 2011, 11:15 http://www.biomedcentral.com/1472-6807/11/15 Page 4 of 8 map that would be accountable for fluorescein molecule around the location seen in PenP-E166Cf or anywhere nearby. Furthermore, the fo-fc map showed strong nega- tive signal for a large segment of Ω-loop (residues 164 to 174) and the 2fo-fc map showed no electron density for this region at all, indicating this region became highly disordered upon acylation of cefotaxime (Figure 3b). Based on these observations we did not include fluorescein molecule or the disordered region of Ω-loop in our final refined structure of PenP-E166Cf- cefotaxime. The overall structure folding of fluorescein-labeled and cefotaxime-bound PenP is nearly identical to that of the wild-type unlabeled PenP and the fluorescein-labeled PenP-E166Cf. From the calculation result by the CCP4 program, it was found that the B factor of Glu163, (c) :-loop GC1 Toho-1 PenP-E166Cf (c) :-loop GC1 Toho-1 PenP-E166Cf Ser70 cefotaxime Ω-loop Cys166 Ser70 cefotaxime fluorescein :-loop Cys166 Ser70 cefotaxime fluorescein :-loop (a) (b) (c) Figure 3 Crystal structure of PenP-E166Cf-cefotaxime. (a) The fo-fc map of cefotaxime in PenP-E166Cf-cefotaxime complex contoured at 2.0 s. The light blue dash line represents the disordered Arg164 to Pro174 due to the poor electron density. (b) Comparison of PenP-E166Cf- cefotaxime complex with apo PenP-E166Cf structure. The two structures are superimposed by main chain atoms. Key residues including Cys166, Ser70 and cefotaxime are also shown in cpk cylinder model. (c) Comparison of binding mode of cefotaxime in PenP-E166Cf with that of Toho-1 and GC-1. PenP-E166Cf, Toho-1 and GC1 are superimposed by aligning on overall main chain atoms. Cefotaxime is in cylinder model colored in cpk (PenP-E166Cf-cefotaxime), golden (Toho-1) and red (GC1) respectively. Wong et al. BMC Structural Biology 2011, 11:15 http://www.biomedcentral.com/1472-6807/11/15 Page 5 of 8 Gly175, Glu176 on Ω-loop, which are next to the disor- dered region, is significantly higher (~65 Å2) than other parts of the protein (~20 Å2). The refinement statistic for this set of crystal structure has different values from that of the apo PenP-E166Cf structure due to the cefo- taxime and the difference in Ω-loop. To investigate why the binding and acylation of cefotax- ime causes the Ω-loop and the conjugated fluorescein molecule to become highly flexible and structural disor- dered, we superposed the PenP-E166Cf structure onto the PenP-E166Cf-cefotaxime complex structure. Fluorescein is seen as occupying a site that partially overlaps with the acylated cefotaxime; particularly the benzoic group of fluorescein molecule is in direct steric clash with the 7-amino substituent of cefotaxime (Figure 3b). Thus the binding and acylation of cefotaxime to PenP would dis- place fluorescein from its original position to avoid steric clash. It is likely that the Ω-loop, in order to accommodate such displacement, loses its well-folded structure and becomes highly flexible. As a consequence the fluorescein molecule conjugated to the flexible Ω-loop becomes fully exposed to the polar aqueous environment, leading to enhanced fluorescence. Thus our structural findings con- firmed our initial proposal of a biosensing mechanism based on displacement of fluorescein [15,16]. To understand the impact of conjugated fluorescein molecule on the substrate binding kinetics of PenP we compared the PenP-E166Cf-cefotaxime structure to two other β-lactamase structures in complex with cefotaxime, including the narrow-spectrum Toho-1 and the extended-spectrum GC1 [19,20]. In Toho-1 structure the methoxyimino side chain points away from the active site and is solvent-exposed (Figure 3c). Such an orientation packs the methoxyimino side chain tightly against the thiozolyl ring, leading to a distorted configuration of the cephem nucleus that is catalytically incompetent for dea- cylation [19]. In GC1 structure the transition analog of cefotaxime binds to GC1 in a fully extended conforma- tion, with oxyimino group inserted to active site and extended away from the thiozolyl ring (Figure 3c). This conformation is regarded as catalytically competent to facilitate deacylation because the distortion on the cephem nucleus is released [20]. Importantly, the binding mode of cefotaxime in our PenP-E166Cf-cefotaxime structure closely resembles that of GC1 (Figure 3c), sug- gesting that with its Ω-loop fully flexible the naturally narrow-spectrum PenP can accommodate cefotaxime in a manner that resembles the extended-spectrum GC1. Conclusions Our structural studies indicate the molecular mechan- ism how fluorescein-labeled β-lactamase detects β-lac- tam antibiotics. The conjugated fluorescein molecule is located near the catalytic site and partially occupies the antibiotic binding region. The binding and acylation of β-lactam antibiotics such as cefotaxime would expel the fluorescein molecule from its original position and leads to increased flexibility of the Ω-loop, to which the fluor- ophore is linked. As a result, the fluorophore is relo- cated from its original position with partial solvent exposure to become fully solvent exposed, leading to enhanced fluorescence emission. These findings confirm our previous proposal based on structural modeling. Furthermore the Ω-loop demonstrates the propensity of becoming highly flexible and unstructured if its tight packing against the catalytic site is disturbed. Such increased flexibility enables PenP to bind and acylate cefotaxime, a naturally poor substrate, in a manner that resembles the extended-spectrum cefotaxime-resistant β-lactamases. This finding could be valuable in the future design of novel antibiotics that resist the binding or hydrolysis by β-lactamases. Methods Protein expression and purification Two constructs of PenP protein were used for our experiments, the maltose binding protein (MBP)-fusion construct for time-dependent fluorescence measure- ments and the His6-tagged construct for crystallization and structural studies, as well as scanning fluorescence spectra. The MBP fusion has been shown not to inter- fere with fluorescence measurements in our previous studies (data not shown). The MBP-fusion construct was cloned into pMAL-c2X vector (NEB). The His6- tagged PenP enzyme was cloned into a modified pRset- A vector (Invitrogen) with a TEV protease cleavage site upstream of the PenP gene. The E166C mutation was constructed using QuikChange Site-Directed Mutagen- esis Kits (Strategene). The MBP-fusion construct was expressed in E. coli strain BL21 (DE3) at 37°C for overnight after induction by 300 μM IPTG when A600 reached 0.5-0.7. The har- vested cells were centrifuged and lysed by sonication. The supernatant after sonication was passed through amylose affinity chromatography. The eluted fractions were pooled and buffer exchanged to 20 mM ammo- nium bicarbonate. The protein was freeze-dried for sto- rage afterwards. The His6-tagged PenP protein was expressed in E. coli strain BL21 (DE3) at 37°C for overnight after induction by 200 μM IPTG when A600 reached 0.8-1.2. The har- vested cells were centrifuged and the supernatant was passed through Nickel affinity chromatography, followed by DEAE anion exchange chromatography. The frac- tions containing the target protein were pooled and con- centrated by Amicon® Ultra-15 Centrifugal Filter Devices (Millipore NMWL = 10,000). The His6-tag was cleaved by adding the TEV protease in 1:20 molar ratio Wong et al. BMC Structural Biology 2011, 11:15 http://www.biomedcentral.com/1472-6807/11/15 Page 6 of 8 to the concentrated PenP-E166C protein (2 mg/ml). The mixture was incubated at 30°C for 6 hours and was further purified by Nickel affinity chromatography to remove uncleaved protein. Fluorescein labeling of PenP-E166C to form PenP-E166Cf A ten-fold molar excess of fluorescein, with concentra- tion of 20 mM, was dissolved in DMF (Dimethyl forma- mide) and added to the concentrated PenP-E166C protein solution drop by drop. The labelling reaction was allowed to proceed in darkness with stirring for 1 hour, and then dialysed against 50 mM potassium phos- phate buffer (pH 7.0) at 4°C for several times in order to remove excess fluorescein. The labelled PenP-E166Cf pro- tein was concentrated to less than 1 ml and further puri- fied by Superdex™75 gel filtration column (GE Healthcare). The running buffer contains 20 mM Tris- HCl, 50 mM NaCl, pH 7.5. The target fractions were pooled and concentrated by Amicon Ultra to 25 mg/ml. The labelling efficiency was confirmed by ESI-MS. Fluorescence spectra of PenP-E166Cf for antibiotic detection Fluorescence profile of PenP-E166Cf alone, as well as in presence of various β-lactams were measured using Per- kin-Elmer LS50B spectrofluorimeter. Both scanning spectra and time-dependent spectra were measured. Dif- ferent β-lactam antibiotics, including cefotaxime, cef- triaxone, ceftazidime, cephaloridine, cephalothin, cefoxitin, cefuroxime, penicillin G, and ampicillin, were incubated with PenP-E166Cf for 100 s at 1 μM to allow sufficient acylation of the antibiotic to form ES* adduct. The product after acylation was subjected to fluores- cence measurement as previously described [15]. Crystallization, structure determination and refinement Crystals of PenP-E166Cf were grown by hanging-drop vapour diffusion method after mixing 1 μl of protein and 1 μl of reservoir solution containing 25% (w/v) PEG 4000, 0.1 M Hepes pH 7.2, 0.4 M NH4Acetate and 0.2 M K2HPO4. Small crystals in the form of clustered nee- dles appeared readily. For data collection, single crystals were obtained after separating them from the clustered needles. Crystals were harvested and cryoprotected in its reservoir solution supplemented with 20% ethylene gly- col for one minute prior to flash freeze and data collec- tion on the Rigaku MicroMax™-007HF x-ray machine. For PenP-E166Cf-cefotaxime data set, crystals were soaked in its growth solution added with 0.01 M of cefotaxime for 15 minutes and then mounted to the x-ray machine. Data were integrated and scaled by Crys- talClear™1.3.5 SP2 (Rigaku Inc.). The crystals belong to the monoclinic group P21 with cell parameter: a = 43.43 Å, b = 92.3 Å, c = 66.43 Å and β = 104°. The PenP-E166Cf crystals diffracted to 2.15 Å resolution, while the PenP-E166Cf-cefotaxime crystal diffracted to 2.8 Å. Both structures were determined by molecular replacement using PenP structure as the search model (PDB ID 4BLM) [21]. The program COOT was used for inspection of electron density maps and model building [22]. There are two molecules per asymmetric unit. The fluorescein and cefotaxime mole- cules were built by PRODRG [23] and appended to the PenP structure for refinement. Structure determination and refinement of PenP-E166Cf and PenP-E166Cf- cefotaxime were done using the CCP4 program suite [18]. A summary of the crystallographic data and refine- ment statistics are given in Table 1. The coordinates and structure factors from this study have been Table 1 X-ray data-collection and structure refinement statistics. E166Cf E166Cf+cefotaxime PDB code 3M2J 3M2K Data collection Space group P21 P21 Unit cell parameters (Å) a 43.3 43.5 b 92.3 91.4 c 66.3 66.1 b 104.82 104.52 Resolution range (Å) 52-2.15 (2.24-2.15) 45-2.80 (2.95-2.80) No. of total reflections 79750 40611 No. of unique reflections 29537 12412 I/s 7.1 (2.7) 6.3 (2.4) Completeness (%) 97.0 (99.5) 99.8 (99.9) Rmerge (%) 9.7 (27.1) 11.8 (32.0) Structure refinement Resolution (Å) 50.0-2.20 45.0-2.80 Rcryst/Rfree (%) 20.0/23.2 21.2/27.7 r.m.s.d. bonds (Å)/angles (°) 0.018/1.784 0.010/1.672 No. of reflections Working set 24217 11749 Test set 1291 647 No. of atoms Protein atoms 4011 3706 Water molecules 254 29 Average B-factor (Å2) Main chain 24.7 16.96 Ligand molecules 48.4 42.46 Water 32.7 10.7 Wong et al. BMC Structural Biology 2011, 11:15 http://www.biomedcentral.com/1472-6807/11/15 Page 7 of 8 deposited into Protein Data Bank (PDB) under accession codes 3M2J (PenP-E166Cf apo structure) and 3M2K (PenP-E166Cf-cefotaxime). Acknowledgements This work was supported by the Research Grants Council (PolyU 5463/05 M, PolyU 5017/06P, PolyU 5641/08 M, and PolyU 5639/09M), the Area of Excellence Fund of the University Grants Committee (AoE/P-10/01) and the Research Committee of the Hong Kong Polytechnic University. We thank Shanghai Synchrotron Radiation Facility (SSRF) for access to beam time. Mr. C.H. Cheng is acknowledged for technical assistance with in-house x-ray crystallography facility. Authors’ contributions WTW performed experiments, analyzed data and drafted manuscript. HWA and HKY assisted in experiments. YXZ, KYW and YCL designed project, analyzed data and drafted manuscript. All authors read and approved the final manuscript. Received: 21 September 2010 Accepted: 28 March 2011 Published: 28 March 2011 References 1. Fisher JF, Mobashery S: Three decades of the class A beta-lactamase acyl- enzyme. Curr Protein Pept Sci 2009, 10(5):401-407. 2. Neu HC: The crisis in antibiotic resistance. Science 1992, 257(5073):1064-1073. 3. Strynadka NC, Adachi H, Jensen SE, Johns K, Sielecki A, Betzel C, Sutoh K, James MN: Molecular structure of the acyl-enzyme intermediate in beta- lactam hydrolysis at 1.7 A resolution. Nature 1992, 359(6397):700-705. 4. Palzkill T, Botstein D: Identification of amino acid substitutions that alter the substrate specificity of TEM-1 beta-lactamase. J Bacteriol 1992, 174(16):5237-5243. 5. Palzkill T, Le QQ, Venkatachalam KV, LaRocco M, Ocera H: Evolution of antibiotic resistance: several different amino acid substitutions in an active site loop alter the substrate profile of beta-lactamase. Mol Microbiol 1994, 12(2):217-229. 6. Petrosino J, Cantu C, Palzkill T: beta-Lactamases: protein evolution in real time. Trends Microbiol 1998, 6(8):323-327. 7. Petrosino JF, Palzkill T: Systematic mutagenesis of the active site omega loop of TEM-1 beta-lactamase. J Bacteriol 1996, 178(7):1821-1828. 8. Trehan I, Beadle BM, Shoichet BK: Inhibition of AmpC beta-lactamase through a destabilizing interaction in the active site. Biochemistry 2001, 40(27):7992-7999. 9. Banerjee S, Pieper U, Kapadia G, Pannell LK, Herzberg O: Role of the omega-loop in the activity, substrate specificity, and structure of class A beta-lactamase. Biochemistry 1998, 37(10):3286-3296. 10. Knox JR: Extended-spectrum and inhibitor-resistant TEM-type beta- lactamases: mutations, specificity, and three-dimensional structure. Antimicrob Agents Chemother 1995, 39(12):2593-2601. 11. Powers RA, Caselli E, Focia PJ, Prati F, Shoichet BK: Structures of ceftazidime and its transition-state analogue in complex with AmpC beta-lactamase: implications for resistance mutations and inhibitor design. Biochemistry 2001, 40(31):9207-9214. 12. Nukaga M, Kumar S, Nukaga K, Pratt RF, Knox JR: Hydrolysis of third- generation cephalosporins by class C beta-lactamases. Structures of a transition state analog of cefotoxamine in wild-type and extended spectrum enzymes. J Biol Chem 2004, 279(10):9344-9352. 13. Crichlow GV, Kuzin AP, Nukaga M, Mayama K, Sawai T, Knox JR: Structure of the extended-spectrum class C beta-lactamase of Enterobacter cloacae GC1, a natural mutant with a tandem tripeptide insertion. Biochemistry 1999, 38(32):10256-10261. 14. Chan PH, Chan KC, Liu HB, Chung WH, Leung YC, Wong KY: Fluorescein- labeled beta-lactamase mutant for high-throughput screening of bacterial beta-lactamases against beta-lactam antibiotics. Anal Chem 2005, 77(16):5268-5276. 15. Chan PH, Liu HB, Chen YW, Chan KC, Tsang CW, Leung YC, Wong KY: Rational design of a novel fluorescent biosensor for beta-lactam antibiotics from a class A beta-lactamase. J Am Chem Soc 2004, 126(13):4074-4075. 16. Chan PH, So PK, Ma DL, Zhao Y, Lai TS, Chung WH, Chan KC, Yiu KF, Chan HW, Siu FM, et al: Fluorophore-labeled beta-lactamase as a biosensor for beta-lactam antibiotics: a study of the biosensing process. J Am Chem Soc 2008, 130(20):6351-6361. 17. Klonis N, Clayton AH, Voss EW Jr, Sawyer WH: Spectral properties of fluorescein in solvent-water mixtures: applications as a probe of hydrogen bonding environments in biological systems. Photochem Photobiol 1998, 67(5):500-510. 18. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 1994, 50(Pt 5):760-763. 19. Shimamura T, Ibuka A, Fushinobu S, Wakagi T, Ishiguro M, Ishii Y, Matsuzawa H: Acyl-intermediate structures of the extended-spectrum class A beta-lactamase, Toho-1, in complex with cefotaxime, cephalothin, and benzylpenicillin. J Biol Chem 2002, 277(48):46601-46608. 20. Crichlow GV, Nukaga M, Doppalapudi VR, Buynak JD, Knox JR: Inhibition of class C beta-lactamases: structure of a reaction intermediate with a cephem sulfone. Biochemistry 2001, 40(21):6233-6239. 21. Knox JR, Moews PC: Beta-lactamase of Bacillus licheniformis 749/C. Refinement at 2 A resolution and analysis of hydration. J Mol Biol 1991, 220(2):435-455. 22. Emsley P, Cowtan K: Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 2004, 60(Pt 12 Pt 1):2126-2132. 23. Schuttelkopf AW, van Aalten DM: PRODRG: a tool for high-throughput crystallography of protein-ligand complexes. Acta Crystallogr D Biol Crystallogr 2004, 60(Pt 8):1355-1363. doi:10.1186/1472-6807-11-15 Cite this article as: Wong et al.: Structural studies of the mechanism for biosensing antibiotics in a fluorescein-labeled β-lactamase. BMC Structural Biology 2011 11:15. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color figure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Wong et al. BMC Structural Biology 2011, 11:15 http://www.biomedcentral.com/1472-6807/11/15 Page 8 of 8
3M2L
Crystal structure of the M113F mutant of alpha-hemolysin
Molecular bases of cyclodextrin adapter interactions with engineered protein nanopores Arijit Banerjeea, Ellina Mikhailovaa, Stephen Cheleya, Li-Qun Gub,2, Michelle Montoyac,3, Yasuo Nagaokaa,4, Eric Gouauxd, and Hagan Bayleya,1 aDepartment of Chemistry, University of Oxford, Oxford, OX1 3TA, United Kingdom; bDepartment of Medical Biochemistry & Genetics, Texas A&M University System Health Science Center, College Station, TX 77843-1114; cDepartment of Biochemistry and Molecular Biophysics and Howard Hughes Medical Institute, Columbia University, New York, NY 10032; and dVollum Institute and Howard Hughes Medical Institute, Oregon Health and Science University, Portland, OR 97239 Edited by Gregory A. Petsko, Brandeis University, Waltham, MA, and approved March 3, 2010 (received for review December 15, 2009) Engineered protein pores have several potential applications in biotechnology: as sensor elements in stochastic detection and ultrarapid DNA sequencing, as nanoreactors to observe single- molecule chemistry, and in the construction of nano- and micro- devices. One important class of pores contains molecular adapters, which provide internal binding sites for small molecules. Mutants of the α-hemolysin (αHL) pore that bind the adapter β-cyclodextrin (βCD) ∼104 times more tightly than the wild type have been ob- tained. We now use single-channel electrical recording, protein en- gineering including unnatural amino acid mutagenesis, and high- resolution x-ray crystallography to provide definitive structural in- formation on these engineered protein nanopores in unparalleled detail. alpha-hemolysin ∣single molecule ∣stochastic sensing ∣structure ∣ unnatural amino acid M any research groups have used protein engineering to obtain enzymes and antibodies with new properties suited for specific tasks (1–6). Fewer groups have taken on the difficult problem of engineering membrane proteins (7). We have engi- neered the α-hemolysin protein pore, mindful of several potential applications in biotechnology, including its ability to act as a de- tector in stochastic sensing (8) and ultrarapid DNA sequencing (9), to serve as a nanoreactor for the observation of single- molecule chemistry (10) and to act as a component for the con- struction of nano- and microdevices (11). An important breakthrough in this area, which enabled the sto- chastic sensing of organic molecules including the detection of DNA bases in the form of nucleoside monophosphates (12, 13), was the discovery of internal molecular adapters, a form of non- covalent protein modification (14). Most useful have been cyclo- dextrin (CD) adapters, which have until now been used in the absence of detailed structural information about how they work. The present paper is a definitive investigation, which provides such information through the application of a wide variety of technical approaches: single-channel electrical recording, protein engineering including unnatural amino acid mutagenesis, and x-ray crystallography. The studies employing mutagenesis show that the striking interactions seen in the crystal structures also occur in individual pores in lipid bilayers. We reveal that the tight-binding αHL mutants (15) M113N7 and M113F7 bind βCD in different orientations within the hep- tameric pore. In the case of M113N7, the top (primary hydroxyls) of the CD ring faces the trans entrance of the pore. In the case of M113F7, the bottom (secondary hydroxyls) of the CD ring faces the trans entrance, while the top of the ring is bonded to the pore through remarkable CH-π interactions. Another tight-binding mutant, M113V7, can bind the CD in both orientations. These results illustrate the exquisite level of engineering that can be achieved with protein nanopores, which is, for example, far be- yond what is possible with solid-state pores. The work also pro- vides information valuable for the design of new binding sites within the lumen of the αHL pore or within other β-barrel pro- teins. Our results will be of interest to others exploring the inter- actions of CDs with the αHL pore (16, 17), including groups involved in computational studies (18, 19). In addition CDs bind to a variety of other pores, including porins (20, 21) and connex- ins (22), and are being tested in vivo as blockers of the anthrax protective antigen pore (23, 24). The CD adapter concept has also been incorporated into other formats, e.g., with glass nano- pores (25), and artificial pores based on CDs have been made by several groups (26–28). Our work is pertinent to these studies. Results Kinetics and Thermodynamics of the Interactions of βCD with αHL Pores Containing Met, Phe and Asn at Position 113. We showed earlier that position 113 in the αHL pore (Fig. 1A) is critical for the bind- ing of βCD (14). Subsequently, residue 113, which is Met in the WT protein, was changed to each of the remaining 19 naturally occurring amino acids by site-directed mutagenesis (15). We found that 11 of these mutants, expressed as homoheptamers, bound βCD with a similar affinity and with similar kinetics to the WT homoheptamer. Two mutants (P, W) bound βCD about 10 times more strongly than the WT homoheptamer, while six of them (V, H, Y, D, N, F) bound with high affinity, i.e., with a Kd value 103 to 104 times lower than the WT. Remarkably, the side chains of the latter six amino acids bear little resemblance to one another, and this issue is addressed in the present paper. We first examined the two amino acids with the most disparate side chains (Fand N) by making heteromeric pores containing WT (Met-113), M113F, and M113N subunits. Three series of heteroheptamers were produced: WT7−nM113Nn, WT7−nM113Fn, and M113F7−nM113Nn. The heteroheptamers were separated by SDS-polyacrylamide gel electrophoresis aided by an oligoaspartate (D8) tail on the first of the two types of sub- unit (Fig. 1B) (29). All 21 combinations of WT, M113F, and M113N subunits formed αHL pores that interacted with βCD as shown by single-channel current recordings, which revealed the extent of block by βCD (Fig. S1), the association and dissociation Author contributions: A.B., S.C., E.G., and H.B. designed research; A.B., E.M., S.C., L.-Q.G., M.M., and Y.N. performed research; A.B., E.G., and H.B. analyzed data; and A.B. and H.B. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. 1To whom correspondence should be addressed. E-mail: hagan.bayley@chem.ox.ac.uk. 2Present address: Department of Biological Engineering and Dalton Cardiovascular Research Center, University of Missouri, Columbia, MO 65211. 3Present address: Nature Structural & Molecular Biology, 75 Varick Street, 9th Floor, New York NY 10013-1917. 4Present address: Department of Biotechnology, Faculty of Engineering, Kansai University, 3-3-35 Yamate-cho, Suita, Osaka 564-8680, Japan. This article contains supporting information online at www.pnas.org/cgi/content/full/ 0914229107/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.0914229107 PNAS ∣May 4, 2010 ∣vol. 107 ∣no. 18 ∣8165–8170 BIOCHEMISTRY rate constants for βCD (kon and koff), and (from the latter) the equilibrium dissociation constant for βCD (Kd ¼ koff∕kon) (15). The kon values for βCD for the 21 combinations of subunits were all similar at ∼5 × 105 M−1 s−1 (Fig. 1C, Upper). By contrast, the koff values differed widely, ranging from ∼5 × 10−2 s−1 to ∼103 s−1. For WT7−nM113Nn and WT7−nM113Fn, the koff values decreased as M113N or M113F subunits were added. In the case of M113N, there was a steep drop in the value of koff after the fifth subunit had been incorporated. In the case of M113F, the decrease in the value of koff occurred less precipitously as the M113F subunits were added (Fig. 1C, Lower). Intriguingly, with M113F7−nM113Nn, koff first increased as M113N subunits were added to M113F7 until n ¼ 4 (M113F3M113N4) and then de- creased for larger values of n (Fig. 1C, Lower). We recognize that there is more than one permutation of heteromers containing two to five mutant subunits (Fig. 1B), but we have ignored this fact here because no significant differences in the properties of indi- vidual heteromers were observed. For example, 42 recordings were made of WT5M113N2, which has three permutations. Because, kon showed little variation with subunit composition, the variation in Kd was similar to the variation in koff (Fig. 1C). While these studies were in progress, the crystal structures of βCD complexed to M113N7 (Fig. 2B) and M113F7 (Fig. 2C) were solved (Table S1) (30). High-resolution structures could be obtained because the CD and the αHL pore have the same C7 symmetry. In the case of M113N7, βCD is bound with the second- ary hydroxyl face “upward” (Fig. 2B). In each glucose unit of the βCD, the 2-hydroxyl is hydrogen bonded to the side-chain amide of an Asn-113 (the residue introduced by mutagenesis) and the 3-hydroxyl is hydrogen bonded to the ϵ-amino group of Lys-147. In the case of M113F7, two βCDs are bound to the αHL pore (Fig. 2C). It is the top βCD in the structure that concerns us, be- cause it is in contact with the Phe-113 residues introduced by mu- tagenesis. It is immediately apparent that the top βCD in M113F7 is in the opposite orientation to the βCD in M113N7 with each 6-hydroxyl group in a CH-π bonding interaction (31–35) with a Phe-113 side chain. The opposite orientations of the βCDs in M113N7 and M113F7 immediately explain why heteromers formed from similar numbers of M113N and M113F subunits (e.g., M113N4M113F3) bind βCD weakly (see also Discussion). Unnatural Amino Acid Mutagenesis. To further explore the range of noncovalent interactions that are available when βCD binds to the αHL pore, five unnatural amino acids (Fig. 3A and Fig. S2) were incorporated at position 113, by using the in vitro nonsense codon suppression method (36). In particular, we had noted that M113V7 containing the β-branched Val binds βCD tightly (15), and therefore we compared cyclopropylglycine (Cpg) and cyclo- propylalanine (Cpa). We also further examined the means by which M113F7 binds βCD tightly, by comparing the properties of 4-fluorophenylalanine (f1F), pentafluorophenylalanine (f5F), and cyclohexylalanine (Cha) at position 113. The five homomeric pores all produced single-channel cur- rents with unitary conductance values in the range expected for properly assembled heptamers (Fig. S3). All five bound βCD (Fig. 3B, Level 2), either tightly (f1F, Cpg) or weakly (f5F, Cha, Cpa) as described in detail below. During the long βCD binding events, additional current spikes were seen (Fig. 3B). Similar Fig. 1. Binding of βCD by heteromeric αHL pores formed by WT, M113F and M113N subunits. (A) Crystal structure of WT-αHL (61) showing residue 113 (Met, yellow). Left panel, side view and right panel, top view. (B) Separation of 35S-labeled αHL heteroheptamers by SDS-polyacrylamide electrophoresis. The separation of the M113F7-nM113Nn heteromers is shown as detected by autoradiography of a dried gel. The M113F subunits carried a D8 tail. Lane 1, M113N7; lane 2, M113F7−nM113Nn (the heteromers formed from several preparations made with differing ratios of M113F and M113N subunits were mixed to give roughly equal amounts of each subunit combination); lane 3, M113F7. A diagram of the eight different combinations of subunits and their permuta- tions is shown to the right of the autoradiogram. The various permutations are not separated by electrophoresis. (C) Kinetics of the interaction of βCD with single heteromeric αHL pores as determined by bilayer recording. Values of kon were calculated by using kon ¼ 1∕ðτon½βCDÞ, where τon is the mean interevent interval. Values of koff were determined by using koff ¼ 1∕τoff, where τoff is the mean dwell time of βCD in the pore. Values of Kd were calculated by using Kd ¼ koff∕kon. Each point represents the mean  s:d: for three or more determinations. Where they cannot be seen, the s.d. values lie within the symbol. Black squares, WT7−nM113Nn; gray squares, M113F7−nM113Nn; empty squares, M113F7−nWTn. 8166 ∣ www.pnas.org/cgi/doi/10.1073/pnas.0914229107 Banerjee et al. events had been observed previously with certain Met-113 repla- cement mutants and may represent movement of the βCD at its binding site (e.g., rotation about axes perpendicular to the C7 axis) (15). The additional current spikes were more prevalent for M113V7 and M113Cpg7, which may take part in more con- formationally labile interactions with βCD, compared with say M113F7 (Fig. S4). Interactions of βCD with Homoheptamers Bearing Aromatic Residues at Position 113. To further understand the nature of the binding of βCD to aromatic side chains, we examined the kinetics of βCD binding to the homoheptamers containing f1F or f5F at position 113, M113f1F7 and M113f5F7 (Fig. 3C). For both mutants, the value of kon was very similar to that of WT7, but the values of koff and therefore Kd for M113f1F7 differed dramatically from WT7 and were close to the values for the tight-binding mutant M113F7 (Table S2A). By contrast, koff and Kd for M113f5F7 were similar to the values for WT7 (Table S2A). To determine whether M113f1F7 binds βCD in the same orien- tation as M113F7 (Fig. 2C), we made heteromers of the M113f1F subunit with M113N or M113F and examined M113F4M113f1F3 and M113N4M113f1F3. M113F4M113f1F3 binds βCD as strongly as either M113F7 or M113f1F7, but M113N4M113f1F3 binds βCD weakly with a similar affinity to WT7 (Fig. 3D and Table S3). Therefore, it is reasonable to infer that M113F7 and M113f1F7 bind βCD in the same orientation with the 6- hydroxyl groups of the CD in proximity to the aromatic rings on the protein. Cyclohexylalanine (Cha) was used to replace the aromatic side chains with a roughly isosteric hydrophobic group. Again the va- lue of kon for βCD was little changed, but koff for M113Cha7 had an intermediate value of 42  6 s−1. Therefore, M113Cha7 binds βCD more weakly than M113F7 but distinctly more strongly than the WT7 pore (Table S2A and Fig. 3C). Interactions of βCD with Homoheptamers Bearing Hydrophobic Resi- dues at Position 113. M113V7 binds βCD very strongly, and there- fore we compared αHL pores with Cpg or Cpa at position 113. Cpg is roughly isosteric with Val, and like Val has a β-branched side chain. Gratifyingly, M113Cpg7 has a kon value similar to the other αHL pores, and koff and Kd values close to those of M113V7 (Table S2B and Fig. 3E). Cyclopropylalanine (Cpa), with an additional methylene group compared to Cpg, is roughly isosteric with Leu, a weak binder, and M113Cpa7 also binds βCD weakly with kon, koff and Kd values similar to those of WT7 (Table S2B and Fig. 3E). M113I7 and M113T7, which are β-branched, are also weak binders, but Ile and Thr are less closely related to Val than Cpg. To determine whether M113V7 binds βCD in the same orien- tation as M113F7 or M113N7 (Fig. 2), we made heteromers of M113V and the M113N or M113F subunits. M113V3M113F4, M113V4M113F3, M113V3M113N4, and M113V4M113N3 were examined in detail. All four heteroheptamers bound βCD more weakly than M113V7, M113F7 or M113N7 (Fig. 3F and Table S4), suggesting that Val at position 113 interacts with βCD strongly but in a different manner to either Phe or Asn. Each heteromer exhibited a range of Kd values, perhaps reflecting the various pos- sible permutations of the two different subunits around the cen- tral axis of the heptamer, although this heterogeneity was not seen for heteromers made from WT, M113F and M113N (Fig. 1). Discussion Soon after we discovered that βCD binds to the WT-αHL pore for around a millisecond, we found a mutant pore, M113N7, that re- leases βCD ∼104 times more slowly (14). This prompted us to examine all 19 mutants in which residue 113 is replaced by a nat- ural amino acid, with the surprising result that a collection of ami- no acids with structurally unrelated side chains (V, H, Y, D, N, F) are tight binders (15). Here, we have examined the nature of the binding interactions more closely by single-channel electrical re- cording, protein engineering including unnatural amino acid mu- tagenesis, and high-resolution x-ray crystallography, and we provide the first definitive structural information on an engi- neered protein nanopore. We find that βCD can bind tightly to the αHL pore in three different ways depending on the residue at 113, as exemplified by Asn, Phe, and Val. Because Asn and Phe have quite different side chains, we first compared the ability of M113N and M113F subunits to take part in binding the CD. The examination of het- eromeric proteins containing WT (Met-113), M113N and M113F subunits showed that the replacement of WT subunits in WT7 with M113N or M113F subunits led to increased affinity for βCD. The more M113N or M113F subunits that were added, the tighter binding became. By contrast, when subunits in M113N7 were replaced with M113F subunits, binding became weaker, reaching a minimum at three to four M113F subunits, and then increasing in strength with five M113F subunits or more (Fig. 1C). Parallel structural studies (30) revealed the basis of the “oppos- ing” effects of the M113N and M113F subunits. βCD binds to M113N7 in the opposite orientation to that in which it binds to M113F7. In M113N7, the secondary hydroxyls in the βCD ring are hydrogen bonded to Lys-147 and Asn-113 (Fig. 2A). By con- trast, βCD interacts with M113F7 through its primary hydroxyl face (Fig. 2B). It seemed likely that M113V7, bound βCD in yet another way, and this was examined by forming heteromers between M113V and M113N or M113F. The presence of three or four subunits of either M113N or M113F greatly decreases the affinity of the pore for βCD (Fig. 3F), with an average koff of 7.3 × 101 s−1, indicating that a third binding mode is indeed operating Fig. 2. X-ray structures of M113N and M113F homoheptamers with βCD bound. (A) Side view of heptameric αHL. βCD binds in the blue highlighted region. (B) βCD bound to M113N7 (dotted lines indicate hy- drogen bonding). The side chains of Lys-147 are in pale brown and the side chains of Asn- 113 in yellow. (C) βCD bound to M113F7 (dotted lines indicate CH-π bonding). The side chains of Phe-113 are in yellow. The sec- ond βCD in the M113F7 · ðβCDÞ2 structure is hydrogen bonded to the top βCD in a head- to-head arrangement and has no apparent interactions with the protein. For both (B) and (C), four β strands were omitted from the barrel to give a better view. Banerjee et al. PNAS ∣ May 4, 2010 ∣ vol. 107 ∣ no. 18 ∣ 8167 BIOCHEMISTRY (Table S4). In summary, the three groups of tight-binding mutants comprise αHL pores incorporating, at position 113: (i) the hydro- gen-bonding amino acids N, D (the latter would have to be largely in the protonated form), and possibly H; (ii) the aromatics F, Y, f1F, and possibly H, and more weakly W; (iii) the β-branched ami- no acids V, Cpg. There may be yet other means by which CDs can bind to the αHL pore. For example, we earlier found that hepta- 6-sulfato-βCD can bind tightly to αHL pores containing the N139Q mutation (37). Presumably, this CD is bound at a site low- er down in the β barrel in a fashion that includes hydrogen bond- ing to the Gln at position 139. While the various mutants exhibited widely different koff values, the value of kon was almost invariant and averaged ∼2.3 × 105 M−1 s−1 (Table S2) (15). Ap- parently, transport to the binding site is rate limiting, through a route unaffected by mutagenesis. koff increased precipitously with the addition of WTsubunits to M113N7 (Fig. 1C). Crystal structures of M113N7 show that resi- dues 111, 113, and 147 are reorganized by compari- son with WT7 and then undergo a more limited rearrangement when βCD binds (Fig. S5). For example, the side chain of Lys-147 shifts position to form a bifurcated hydrogen bond with a 3-hydroxyl group of βCD and the side chain carbonyl of an Asn- 113 (Fig. S6). Therefore, the side chains of residues 111, 113, and 147 might be in a variety of conformations in WT7−nM113Nn het- eromers and offer less well preorganized binding sites for βCD than they do in M113N7. Further, the intramolecular hydrogen bonds of the secondary hydroxyls in βCD (38) must be disrupted upon binding as both hydroxyls on each glucose ring form hydro- gen bonds to the mutant subunits (Fig. 2B). Because the hydrogen bonds that are broken in βCD are arranged in a circle, the break- age of bonds involving a single glucose (three bonds in all) will be relatively more disruptive than those involving adjoining glucose residues or the entire circle. The overall binding cooperativity in M113N7 could be attributed to enthalpic cooperativity outweigh- ing entropic penalties to binding (39). Positive cooperativity has been observed previously in fairly rigid model systems (40). By contrast with M113N7, there is little movement of side chains in ðM113FÞ7 by comparison with WT7 and little move- ment, including Phe-113, upon binding βCD (Fig. S7A). Further, the crystal structure of M113F7 · βCD suggests that each Phe re- sidue interacts independently with the βCD through what appear to be CH-π interactions (Fig. S7B). These interactions are ex- pected to be weak and not strongly directional and hence offer less enthalpic cooperativity, as supported by the B-factors (crys- tallographic temperatures factors) at the primary βCD binding site, which are between ∼40 and 50. Positive cooperativity is ob- served, but it is less pronounced than in the case of M113N7 (Table S5). In the case of M113N7, the B-factors of the residues that bind βCD are in the 20s implying that the βCD is more rigidly held than it is in M113F7. The binding of sugars to aromatic residues in proteins can in- clude CH-π bonding (41) or OH-π bonding or a finely balanced complement of both (42, 43). However, we have dismissed the possibility of an OH-π interaction between Phe-113 and the 6-hydroxyl groups of βCD as the distance between the center of the phenyl rings to the nearest hydroxyl oxygen is higher (5.2  0.65 Å, n ¼ 7) than that expected for a favorable OH-π interaction (33). While we propose that βCD binds to Phe-113 through a C-6 CH-π interaction (Fig. S7B), the distances between the center of the Phe-113 ring and the nearest C-6 of βCD ob- served in the M113F7 · βCD structure (4.66  0.24 Å, n ¼ 7) are in the upper range of the expected distance for a strong inter- action, which is ∼4.5 Å (33). The angle between the normal to the aromatic rings and the line connecting the C-6 atoms to the aro- matic midpoint is 8.0  5.6°, which is well within the expected range (44). The measurements with M113f5F7 argue against a hydrophobic interaction between Phe residues at position 113 and the βCD ring. In f5F, the hydrophobicity of the phenyl ring is significantly increased (45) yet M113f5F7 binds βCD weakly, like WT7 (Fig. 3C and Table S2A). By contrast with F, f1F, Y and N, homomeric αHL pores with f5F and W at position 113 bound βCD relatively weakly (Fig. 3C and Table S2A). In the case of f5F, the powerful electron with- drawing action of the five fluorine atoms leaves a highly increased positive charge at the center of the ring (46, 47), mitigating against a hydrogen-bonding interaction. The electron-rich Trp Fig. 3. Properties of pores containing natural and unnatural amino acid sub- stitutions at position 113. The data were recorded at þ40 mV in 1.0 M NaCl, 10 mM sodium phosphate, pH 7.5. (A) Unnatural amino acids used in this study: 4-fluorophenylalanine, f1F; pentafluorophenylalanine, f5F; cyclohex- ylalanine, Cha; cyclopropylglycine, Cpg; cyclopropylalanine, Cpa. (B) Repre- sentative current traces from single homoheptameric αHL pores, containing unnatural amino acids at position 113, in the presence of βCD. βCD (40 μM final) was added to the trans chamber. Level 1, open pore current; level 2, pore occupied by βCD. The broken line indicates zero current. (C) In- teraction of βCD with homomeric αHL pores containing aromatic amino acids at position 113. Kd values for the interaction between βCD and the αHL pore were calculated by using Kd ¼ koff∕kon. Each column represents the mean  s:d: for 10 or more determinations: dark gray, natural amino acids; light gray, unnatural amino acids. Data adapted from Gu and colleagues (15) are marked (*). (D) Representative current traces from single-channel recordings of βCD binding to M113F4M113f1F3 and M113N4M113f1F3. βCD (40 μM final) was added to the trans chamber. The broken line indicates zero current. (E) Interaction of βCD with homomeric αHL pores containing hydrophobic amino acids at position 113. Kd values for the interaction between βCD and the αHL pore were calculated by using Kd ¼ koff∕kon. Each column represents the mean  s:d: for ten or more determinations: dark gray, natural amino acids; light gray, unnatural amino acids. Data adapted from Gu and colleagues (15) are marked (*). (F) koff values for βCD from heteroheptamers formed with M113F and M113V subunits and with M113N and M113V subunits. βCD (40 μM final) was added to the trans chamber. The kon values for βCD for all these mutants are similar, at ∼3 × 105 M−1 s−1. Empty square: average koff values for the mutant (bar is s:d). Filled square: M113V3M113F4; filled circle: M113V4M113F3; filled upright triangle: M113V3M113N4; filled in- verted triangle: M113V4M113N3. 8168 ∣ www.pnas.org/cgi/doi/10.1073/pnas.0914229107 Banerjee et al. ring (44, 46, 47) should favor hydrogen bonding, but here we can- not make a direct comparison with the crystal structure of M113F7 as the indole ring is far larger than benzene. It is possible that it cannot become oriented in the same manner and that it is misaligned for hydrogen bonding. Our experiments suggest that M113V7 and M113Cpg7 bind βCD in a third way. In heteromers with M113V, both M113F and M113N reduce the affinity of the pore for βCD suggesting that neither the CH-π interaction with Phe-113 nor the hydrogen- bonding interactions with Asn-113 and Lys-147 are compatible with binding to Val. Close interactions of Val with glucose rings have been noted previously (48). Therefore, we propose that the Val side-chain interacts with the side of the glucose ring. This in- teraction might occur in one or both orientations of the CD ring (Fig. 4). Conclusion We provide structural information on engineered protein nano- pores and describe three distinct ways in which βCD can bind within the lumen of mutant αHL pores in atomic detail. Our re- sults will be useful in several areas of basic science and biotech- nology. By using host molecules lodged within the αHL pore, host-guest interactions can be investigated in fine detail at the single-molecule level (17, 49). The present work will now permit us to examine binding events at a known face of a CD. The work also provides information for designing new binding sites within the lumen of the αHL pore (37) or within other β barrel proteins (21, 50) and for using molecular design to devise ways in which to covalently attach CDs within pores (13, 51). These areas impact practical applications of nanopore technology including stochas- tic sensing (8), single-molecule DNA sequencing (9, 12, 13, 52), the use of nanoreactors for the observation of single-molecule chemistry (10), and the construction of nano- and microdevices (11, 53), as well as the design of CDs as therapeutic agents (23, 24). Methods Full details of the experimental procedures can be found in SI Appendix. Materials L-Amino acids were obtained as follows: 4-fluorophenylalanine (f1F) (Fluka); pentafluorophenylalanine (f5F) (PepTech Corp.); cyclopropylglycine (Cpg) (Ty- ger); cyclopropylalanine (Cpa) (Tyger). 4-N-benzoyl-5′-O-(4,4′-dimethoxytri- tyl)-2′-deoxycytidine-3′-[(2-cyanoethyl)-(N,N-diisopropyl)]-phosphoramidite and bis(2-cyanoethyl)-N,N-diisopropylphosphoramidite for the synthesis of pdCpA were purchased from Glen Research and Toronto Research Chemicals, respectively. Preparation of NVOC-Protected Aminoacyl-pdCpA. NVOC-protected aminoacyl-pdCpAs were prepared as reported previously by reacting the dinucleotide pdCpA with N-protected, carboxylic acid-activated, amino acids (54–56). Preparation of NVOC-Protected Aminoacyl-tRNA. NVOC-protected aminoacyl- pdCpAs were ligated enzymatically with a truncated tRNA, prepared by using methods described elsewhere (57, 58). Genetic Constructs and Mutagenesis. All new αHL constructs were verified by DNA sequencing. Details of each construct can be found in SI Appendix. Synthesis, Assembly, and Purification of Mutant αHL pores. αHL monomers (WT and mutants) were prepared in vitro by coupled transcription and translation (IVTT) and assembled into homoheptamers on rabbit red blood cell membranes followed by purification by SDS–PAGE as described earlier (59). Heteroheptamers were prepared by mixing the two required DNAs (one encoding an αHL with a D8 tail) before IVTT and then oligomerizing the mixed translation products on rabbit red blood cell membranes. Pores with the desired combinations of subunits were purified by SDS–PAGE (59). Synthesis, Assembly, and Purification of αHL Mutants Containing Unnatural Ami- no Acids. αHL polypeptides containing unnatural amino acids were synthe- sized by IVTT in the presence of rabbit red blood cell membranes. The plasmid with a stop codon (TAG) at position 113 was used. Deprotected ami- noacyl-tRNAs (SI Appendix) were added to the IVTT mixtures. For heterohep- tamers with subunits containing unnatural amino acids in combination with M113N or M113F, monomers were first made, which were then coassembled on rabbit red blood cell membranes and subsequently purified by SDS–PAGE. Single-Channel Current Recordings in Planar Lipid Bilayers. (15, 60) Recordings were made with 1.0 M NaCl, 10 mM sodium phosphate, pH 7.5, in both cham- bers, at an applied potential of þ40 mV. Data were recorded at 22  2°C. The bilayer was formed from 1,2-diphytanoyl-sn-glycero-phosphocholine (Avanti Polar Lipids). Proteins were added to the cis chamber, and βCD to the trans chamber. Single-channel currents were recorded with an Axopatch 200B patch-clamp amplifier (Axon Instruments) and filtered at 2 kHz with a built-in 4-pole low-pass Bessel Filter. The data were acquired at a sampling rate of 10 kHz. For mutants that bind βCD strongly, the data were acquired for at least 30 min and for weak-binding mutants for at least 10 min. Kinetic Data Analysis. Current amplitude and dwell-time histograms were made by using ClampFit 9.0. The mean dwell times, τoff, were determined by fitting the dwell-time histograms to single exponentials. Values of kon and koff were obtained by using the mean dwell times and mean interevent intervals, as described previously (15, 60). This analysis assumes a binary in- teraction, which was supported in all cases examined by the finding of only one major blockade level and a single exponential distribution of dwell times (τoff). Fig. 4. Molecular model showing the three classes of interaction between the αHL pore and βCD identified in this work. The model identifies the region of βCD responsible for each interaction (H atoms interacting with Phe-113 or Asn-113 and Lys-147: gray). The first class of interaction is with aromatic residues and involves the seven -CH2OH groups of the βCD. The second class is typified by the interactions with Asn at position 113, which involve hydro- gen-bonds to the secondary 2-hydroxyls of the βCD. Structural studies show that this interaction is supported by hydrogen bonding between Lys-147 and the secondary 3-hydroxyls of the βCD. Structural studies and experiments with heteromers suggest that the βCD in M113F7 is in the opposite orienta- tion to the βCD in M113N7, in support of the model shown here. As the inter- action with Val is hydrophobic, it is not directional and βCD may not bind at the same position inside the β barrel as it does in M113F7 or M113N7. Banerjee et al. PNAS ∣ May 4, 2010 ∣ vol. 107 ∣ no. 18 ∣ 8169 BIOCHEMISTRY Protein Crystallography. Details can be found in SI Appendix. Protein Data Bank: The coordinates and structure factors of the described structures have been deposited with accession codes 3M2L ðM113F7Þ, 3M3R ðM113F7 · βCDÞ, 3M4D ðM113N7Þ, 3M4E ðM113N7 · βCDÞ. ACKNOWLEDGMENTS. We thank Dennis Dougherty for the plasmid pTHG73. This work was funded by a Royal Society Wolfson Research Merit Award (to H.B.), the Medical Research Council (H.B.), the National Institutes of Health (H.B.), and the Howard Hughes Medical Institute (E.G.). 1. Lu Y, Yeung N, Sieracki N, Marshall NM (2009) Design of functional metalloproteins. Nature 460:855–862. 2. Gebauer M, Skerra A (2009) Engineered protein scaffolds as next-generation antibody therapeutics. Curr Opin Chem Biol 13:245–255. 3. Gronwall C, Stahl S (2009) Engineered affinity proteins—Generation and applications. J Biotechnol 140:254–269. 4. Arnold U (2009) Incorporation of non-natural modules into proteins: Structural features beyond the genetic code. Biotechnol Lett 31:1129–1139. 5. Tracewell CA, Arnold FH (2009) Directed enzyme evolution: Climbing fitness peaks one amino acid at a time. Curr Opin Chem Biol 13:3–9. 6. Fruk L, Kuo CH, Torres E, Niemeyer CM (2009) Apoenzyme reconstitution as a chemical tool for structural enzymology and biotechnology. Angew Chem Int Ed Engl 48:1550–1574. 7. Bayley H, Jayasinghe L (2004) Functional engineered channels and pores. Mol Membr Biol 21:209–220. 8. Bayley H, Cremer PS (2001) Stochastic sensors inspired by biology. Nature 413:226–230. 9. Branton D, et al. (2008) The potential and challenges of nanopore sequencing. Nature Biotechnol 26:1146–1153. 10. Bayley H, Luchian T, Shin S-H, Steffensen MB (2008) Single Molecules and Nanotech- nology, eds R Rigler and H Vogel (Springer, Heidelberg), pp 251–277. 11. Maglia G, et al. (2009) Droplet networks with incorporated protein diodes show collective properties. Nat Nanotechnol 4:437–440. 12. Astier Y, Braha O, Bayley H (2006) Toward single molecule DNA sequencing: Direct identification of ribonucleoside and deoxyribonucleoside 5'-monophosphates by using an engineered protein nanopore equipped with a molecular adapter. J Am Chem Soc 128:1705–1710. 13. Clarke J, et al. (2009) Continuous base identification for single-molecule nanopore DNA sequencing. Nature Nanotechnol 4:265–270. 14. Gu L-Q, Braha O, Conlan S, Cheley S, Bayley H (1999) Stochastic sensing of organic analytes by a pore-forming protein containing a molecular adapter. Nature 398:686–690. 15. Gu L-Q, Cheley S, Bayley H (2001) Prolonged residence time of a noncovalent molecular adapter, β-cyclodextrin, within the lumen of mutant α-hemolysin pores. J Gen Physiol 118:481–494. 16. Ervin EN, Kawano R, White RJ, White HS (2008) Simultaneous alternating and direct current readout of protein ion channel blocking events using glass nanopore membranes. Anal Chem 80:2069–2076. 17. Gurnev PA, Harries D, Parsegian VA, Bezrukov SM (2009) The dynamic side of the Hof- meister effect: A single-molecule nanopore study of specific complex formation. ChemPhysChem 10:1445–1449. 18. Mamonova T, Kurnikova M (2006) Structure and energetics of channel-forming protein-polysaccharide complexes inferred via computational statistical thermody- namics. J Phys Chem B 110:25091–25100. 19. Egwolf B, Luo Y, Walters DE, Roux B (2010) Ion selectivity of alpha-hemolysin with beta-cyclodextrin adapter. II. Multi-ion effects studied with grand canonical monte carlo/brownian dynamics simulations. J Phys Chem B 114:2901–2909. 20. Orlik F, et al. (2003) CymA of Klebsiella oxytoca outer membrane: Binding of cyclodex- trins and study of the current noise of the open membrane. Biophys J 85:876–885. 21. Chen M, Khalid S, Sansom MSP, Bayley H (2008) OmpG: Engineering a quiet pore for biosensing. Proc Natl Acad Sci USA 105:6272–6277. 22. Locke D, Koreen IV, Liu JY, Harris AL (2004) Reversible pore block of connexin channels by cyclodextrins. J Biol Chem 279:22883–22892. 23. Karginov VA, Nestorovich EM, Moayeri M, Leppla SH, Bezrukov SM (2005) Blocking anthrax lethal toxin at the protective antigen channel by using structure-inspired drug design. Proc Natl Acad Sci USA 102:15075–15080. 24. Moayeri M, Robinson TM, Leppla SH, Karginov VA (2008) In vivo efficacy of beta- cyclodextrin derivatives against anthrax lethal toxin. Antimicrob Agents Chemother 52:2239–2241. 25. Gao C, Ding S, Tan Q, Gu LQ (2009) Method of creating a nanopore-terminated probe for single-molecule enantiomer discrimination. Anal Chem 81:80–86. 26. Pregel MJ, Jullien L, Lehn J-M (1992) Towards artificial ion channels: Transport of alkali metal ions across liposomal membranes by “bouquet” molecules. Angew Chem Int Edit 31:1637–1639. 27. Bacri L, Benkhaled A, Guegan P, Auvray L (2005) Ionic channel behavior of modified cyclodextrins inserted in lipid membranes. Langmuir 21:5842–5846. 28. Jog PV, Gin MS (2008) A light-gated synthetic ion channel. Org Lett 10:3693–3696. 29. Howorka S, Cheley S, Bayley H (2001) Sequence-specific detection of individual DNA strands using engineered nanopores. Nat Biotechnol 19:636–639. 30. Montoya M (2004) Insights into Membrane Association and Bioengineering of a Pore- Forming Toxin: Structural Studies of Staphylococcal α-Hemolysin (Columbia University, New York). 31. Steiner T (2002) The hydrogen bond in the solid state. Angew Chem Int Ed 41:49–76. 32. Steiner T (2002) Hydrogen bonds from water molecules to aromatic acceptors in very high-resolution protein crystal structures. Biophys Chem 95:195–201. 33. Steiner T, Koellner G (2001) Hydrogen bonds with pi-acceptors in proteins: Frequencies and role in stabilizing local 3D structures. J Mol Biol 305:535–557. 34. Brandl M, Weiss MS, Jabs A, Sühnel J, Hilgenfeld R (2001) CH...π-interactions in proteins. J Mol Biol 307:357–377. 35. Weiss MS, Brandl M, Suhnel J, Pal D, Hilgenfeld R (2001) More hydrogen bonds for the (structural) biologist. Trends Biochem Sci 26:521–523. 36. Wang L, Xie J, Schultz PG (2006) Expanding the genetic code. Annu Rev Biophys Biomol Struct 35:225–249. 37. Gu L-Q, Cheley S, Bayley H (2001) Capture of a single molecule in a nanocavity. Science 291:636–640. 38. Saenger W, et al. (1998) Structures of the common cyclodextrins and their larger analogues—Beyond the doughnut. Chem Rev 98:1787–1802. 39. Hunter CA, Tomas S (2003) Cooperativity, partially bound states, and enthalpy-entropy compensation. Chem Biol 10:1023–1032. 40. Bisson AP, Hunter CA (1996) Cooperativity in the assembly of zipper complexes. Chem Commun 1723–1724. 41. del Carmen Fernandez-Alonso M, Canada FJ, Jimenez-Barbero J, Cuevas G (2005) Molecular recognition of saccharides by proteins. Insights on the origin of the carbohydrate-aromatic interactions. J Am Chem Soc 127:7379–7386. 42. Jimenez-Barbero J, Asensio JL, Canada FJ, Poveda A (1999) Free and protein-bound carbohydrate structures. Curr Opin Struct Biol 9:549–555. 43. Stanca-Kaposta EC, et al. (2007) Carbohydrate molecular recognition: A spectroscopic investigation of carbohydrate-aromatic interactions. Phys Chem Chem Phys 9:4444–4451. 44. Brandl M, Weiss MS, Jabs A, Suhnel J, Hilgenfeld R (2001) CH...π-interactions in proteins. J Mol Biol 307:357–377. 45. Woll MG, Hadley EB, Mecozzi S, Gellman SH (2006) Stabilizing and destabilizing effects of phenylalanine –>F5-phenylalanine mutations on the folding of a small protein. J Am Chem Soc 128:15932–15933. 46. Mecozzi S, West AP, Jr, Dougherty DA (1996) Cation-pi interactions in aromatics of biological and medicinal interest: Electrostatic potential surfaces as a useful qualitative guide. Proc Natl Acad Sci USA 93:10566–10571. 47. Dougherty DA (2008) Cys-loop neuroreceptors: Structure to the rescue?. Chem Rev 108:1642–1653. 48. Hondoh H, et al. (2008) Substrate recognition mechanism of alpha-1,6-glucosidic linkage hydrolyzing enzyme, dextran glucosidase from Streptococcus mutans. J Mol Biol 378:913–922. 49. Kang XF, Cheley S, Guan X, Bayley H (2006) Stochastic detection of enantiomers. J Am Chem Soc 128:10684–10685. 50. Chen M, Li QH, Bayley H (2008) Orientation of the monomeric porin OmpG in planar lipid bilayers. ChemBioChem 9:3029–3036. 51. Wu H-C, Astier Y, Maglia G, Mikhailova E, Bayley H (2007) Protein nanopores with covalently attached molecular adapters. J Am Chem Soc 129:16142–16148. 52. Bayley H (2006) Sequencing single molecules of DNA. Curr Opin Chem Biol 10:628–637. 53. Astier Y, Bayley H, Howorka S (2005) Protein components for nanodevices. Curr Opin Chem Biol 9:576–584. 54. Robertson SA, Noren CJ, Anthony-Cahill SJ, Griffith MC, Schultz PG (1989) The use of 5'-phospho-2 deoxyribocytidylylriboadenosine as a facile route to chemical aminoacy- lation of tRNA. Nucleic Acids Res 17:9649–9660. 55. Ellman J, Mendel D, Anthony-Cahill S, Noren CJ, Schultz PG (1991) Biosynthetic method for introducing unnatural amino acids site-specifically into proteins. Method Enzymol 202:301–336. 56. Kearney PC, et al. (1996) Dose-response relations for unnatural amino acids at the agonist binding site of the nicotinic acetylcholine receptor: Tests with novel side chains and with several agonists. Mol Pharmacol 50:1401–1412. 57. England TE, Bruce AG, Uhlenbeck OC (1980) Specific labeling of 3′ termini of RNA with T4 RNA ligase. Method Enzymol 65:65–74. 58. Nowak MW, et al. (1998) In vivo incorporation of unnatural amino acids into ion channels in xenopus oocyte expression system. Method Enzymol 293:504–529. 59. Cheley S, Braha O, Lu X, Conlan S, Bayley H (1999) A functional protein pore with a “retro” transmembrane domain. Protein Sci 8:1257–1267. 60. Gu L-Q, et al. (2000) Reversal of charge selectivity in transmembrane protein pores by using non-covalent molecular adapters. Proc Natl Acad Sci USA 97:3959–3964. 61. Song L, et al. (1996) Structure of staphylococcal α-hemolysin, a heptameric transmem- brane pore. Science 274:1859–1865. 8170 ∣ www.pnas.org/cgi/doi/10.1073/pnas.0914229107 Banerjee et al.
3M2R
Structural Insight into Methyl-Coenzyme M Reductase Chemistry using Coenzyme B Analogues
Structural Insight into Methyl-Coenzyme M Reductase Chemistry using Coenzyme B Analogues,†,‡ Peder E. Cedervall§,#, Mishtu Dey||,#,⊥, Arwen R. Pearson§,⊥, Stephen W. Ragsdale||, and Carrie M. Wilmot*,||,§ § Department of Biochemistry, Molecular Biology and Biophysics, University of Minnesota, Minneapolis, Minnesota 55455 || Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109 Abstract Methyl-coenzyme M reductase (MCR) catalyzes the final and rate-limiting step in methane biogenesis; the reduction of methyl-coenzyme M (methyl-SCoM) by coenzyme B (CoBSH) to methane and a heterodisulfide (CoBS-SCoM). Crystallographic studies show that the active site is deeply buried within the enzyme, and contains a highly reduced nickel-tetrapyrrole, coenzyme F430. Methyl-SCoM must enter the active site prior to CoBSH, as species derived from analogues of methyl-SCoM are always observed bound to the F430 nickel in the deepest part of the 30 Å long substrate channel that leads from the protein surface to the active site. The seven-carbon mercaptoalkanoyl chain of CoBSH binds within a 16 Å predominantly hydrophobic part of the channel close to F430, with the CoBSH thiolate lying closest to the nickel at a distance of 8.8 Å. It has previously been suggested that binding of CoBSH initiates catalysis by inducing a conformational change that moves methyl-SCoM closer to the nickel promoting cleavage of the C- S bond of methyl-SCoM. In order to better understand the structural role of CoBSH early in the MCR mechanism, we have determined crystal structures of MCR in complex with four different CoBSH analogues; pentanoyl-, hexanoyl-, octanoyl- and nonanoyl- derivatives of CoBSH (CoB5SH, CoB6SH, CoB8SH and CoB9SH respectively). The data presented here reveal that the shorter CoB5SH mercaptoalkanoyl chain overlays with that of CoBSH, but terminates two units short of the CoBSH thiolate position. In contrast, the mercaptoalkanoyl chain of CoB6SH adopts a different conformation, such that its thiolate is coincident with the position of the CoBSH thiolate. †This work was supported by a Department of Energy grant DE-FG02-08ER15931 to S.W.R and supplement to C.M.W. and a Minnesota Partnership for Biotechnology and Medical Genomics grant SPAP-05-0013-P-FY06. ‡Co-ordinates and structure factors have been deposited in the Protein Data Bank as entries 3m1v (MCRHSCoM); 3m2r (MCRCoB5SH); 3m2u (MCRCoB6SH); 3m2v (MCRCoB8SH); 3m30 (MCRCoB9SH). *Address correspondence to: Carrie M. Wilmot, Tel: (612) 624-2406, wilmo004@umn.edu. ⊥Current addresses: Mishtu Dey, Department of Chemistry, Massachusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, Massachusetts 02139; Arwen Pearson, Astbury Centre for Structural Molecular Biology, Institute for Molecular and Cellular Biology, Astbury Building. Leeds, LS2 9JT, U.K. #These authors contributed equally to this work. Supporting Informational material is available free of charge via the Internet at http://pubs.acs.org and contains the following: MATERIAL AND METHODS for single crystal UV-visible microspectrophotometry, X-ray photoreduction experiment and X-ray crystallography of the MCR-heterodisulfide product complex (MCRCoBSH + methyl-SCoM); RESULTS AND DISCUSSION for redox changes and MCRCoBSH + methyl-SCoM crystal structure; Table S1, X-ray Data Collection, Processing and Refinement Statistics for MCRCoBSH + methyl-SCoM; Figure S1, use of Fo-Fc electron density in modelling MCRCoB8SH; Figure S2, illustration of electron density quality of MCRCoB5SH; Figure S3, solution and single crystal UV-visible spectra; Figure S4, modeling of CoB6SH and CoBSH into the electron density of MCRCoB6SH; Figure S5, partially occupied HSCoM in MCRCoB6SH; Figure S6, the active site and substrate channel of MCRCoBSH + methyl-SCoM; Figure S7, alternative conformation of Valα482 in MCRHSCoM; Figure S8, the two conformations of the gly-rich loop in MCRHSCoM; Figure S9, propagation of conformational changes in MCRHSCoM; Figure S10, the two conformations of CoB9SH in MCRCoB9SH; Figure S11, EPR spectra of MCRred1 sample; Scheme S1, scheme of the characterized forms of MCR. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 September 7. Published in final edited form as: Biochemistry. 2010 September 7; 49(35): 7683–7693. doi:10.1021/bi100458d. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript This is consistent with the observation that CoB6SH is a slow substrate. A labile water in the substrate channel was found to be a sensitive indicator for the presence of CoBSH and HSCoM. The longer CoB8SH and CoB9SH analogues can be accommodated in the active site through exclusion of this water. These analogues react with Ni(III)-methyl; a proposed MCR catalytic intermediate of methanogenesis. The CoB8SH thiolate is 2.6 Å closer to the nickel than that of CoBSH, but the additional carbon of CoB9SH only decreases the nickel thiolate distance a further 0.3 Å. Although the analogues did not induce any structural changes in the substrate channel, the thiolates appeared to preferentially bind at two distinct positions in the channel; one being the previously observed CoBSH thiolate position, and the other being at a hydrophobic annulus of residues that lines the channel proximal to the nickel. INTRODUCTION Methanogenic archaea are organisms that under strict anaerobic conditions derive energy by reducing compounds such as carbon dioxide, methylamine, acetate and methanol, to methane (1, 2). The global production of methane by these organisms is estimated at one billion tons annually. Microbially produced methane is not only a potential source of renewable energy but also a potent greenhouse gas, and as such study of this process has environmental ramifications. In these microorganisms, methyl-coenzyme M reductase (MCR)1 is the enzyme that catalyzes the final step in methanogenesis, in which the substrates methyl-coenzyme M (methyl-SCoM, 2-(methylthio)ethanesulfonate) and coenzyme B (CoBSH, N-7-mercaptoheptanoylthreonine phosphate) are converted to methane and a heterodisulfide (CoBS-SCoM) (Scheme 1) (3). MCR is a 300 kDa protein with six subunits arranged in a α2β2γ2 oligomer (4). The known crystal structures show that MCR has two active sites approximately 50 Å apart that are deeply buried within the enzyme (5). The active site pocket is comprised of residues from subunits α, α′, β and γ, with a 30 Å long substrate channel leading to the enzyme surface (Figure 1). At the heart of the active site pocket is coenzyme F430, which is a highly reduced nickel-containing tetrapyrrole (6–8). Currently sixteen distinct enzymatic and complexed states of MCR have been spectroscopically characterized (Supporting Information, Scheme S1) (9–31). In the resting active state of the enzyme, denoted MCRred1, the redox-active nickel of F430 is present in the Ni(I) state (9, 16, 32). MCR is extremely oxygen sensitive and upon oxygen exposure the enzyme enters an inactive Ni(II) state, denoted MCRred1-silent (6). In this state it cannot be converted back to the active Ni(I) form by any known reducing agent making this a challenging system to study. Additional complications involve the tight association of coenzymes to purified MCR that are not easily displaced as demonstrated by X-ray crystallographic and kinetic studies (5, 33–35). Despite the fact that MCR has been studied for decades, no true catalytic intermediate has been observed, and the actual mechanism remains elusive. Currently three general mechanistic schemes for the enzymatic reaction have been proposed, each of which posit different chemistry to initiate catalysis. Mechanism I involves Ni(I) acting as a nucleophile in an SN2-type reaction that generates Ni(III)-methyl as an intermediate (Scheme 2A) (35– 38). Mechanism II starts with methyl-SCoM undergoing homolytic cleavage at the Ni(I) to generate a methyl radical and a Ni(II)-SCoM species (Scheme 2B) (39–41). A more recently proposed mechanism III suggests protonation of coenzyme F430 promotes reductive cleavage of the methyl-SCoM thioether bond (42). 1Abbreviations: MCR, Methyl-coenzyme M reductase; methyl-SCoM, methyl-coenzyme M; CoBSH, coenzyme B; HSCoM, coenzyme M; CoBS-SCoM, heterodisulfide of coenzyme B and coenzyme M; APS, Advanced Photon Source; ASU, asymmetric unit; BPS, bromopropanesulfonate. Cedervall et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Due to the stringent requirement to exclude O2, the available MCR crystal structures are all in the inactive Ni(II) state. All but one contain CoBSH and HSCoM (demethylated methyl- SCoM, an inhibitor and substrate analogue) in the active site (PDB codes 1hbn, 1hbo, 1hbu, 1e6y, 1e6v) (5, 33, 34). Another crystal structure has bound heterodisulfide product, CoBS- SCoM (MCRsilent, PDB code 1hbm, Scheme 1 and Supporting Information, Scheme S1) (5, 33). All these structures reveal that both substrates access the active site through the same channel (Figure 1). The binding site of HSCoM (and presumably methyl-SCoM) is more deeply buried within the enzyme, and so it must enter prior to CoBSH for productive chemistry to occur. As binding of CoBSH in the absence of co-substrate would be inhibitory, it was suggested that a conformational change upon methyl-SCoM binding might lower the Kd for CoBSH, and thus promote an ordered mechanism. Furthermore it has been suggested that CoBSH binding induces a conformational change that brings the methyl- SCoM substrate into closer proximity to the nickel, and this promotes C-S bond cleavage. To investigate the proposed structural role of CoBSH in initiating catalysis, we have solved the X-ray crystal structures of MCR in complex with four different CoBSH analogues. CoBSH has a heptanoyl moiety linked to the thiol group, and the analogues are pentanoyl-, hexanoyl-, octanoyl- or nonanoyl-containing derivatives of CoBSH (CoB5SH, CoB6SH, CoB8SH and CoB9SH respectively; Figure 2) (3, 35, 43–47). In addition, we present a structure in which the substrate channel predominantly lacks either CoBSH or heterodisulfide product. MATERIALS AND METHODS Materials The organism Methanothemobacter marburgensis (catalog OCM82) was obtained from the Oregon Collection of Methanogens (Portland, OR). All buffers and media reagents were obtained from Sigma-Aldrich (St. Louis, MO). The gases N2 (99.98%), H2/CO2 (80%/20%), and ultra high purity H2 (99.999%) were obtained from Cryogenic Gases (Grand Rapids, MI). Ti(III) citrate solutions were prepared from a stock solution of 246 mM Ti(III) citrate, which was synthesized by adding sodium citrate to Ti(III) trichloride (30 wt % solution in 2 N hydrochloric acid) (Acros Organics, Morris Plains, NJ) under anaerobic conditions and adjusting pH to 7.0 with sodium bicarbonate (48). The concentration of Ti(III) citrate was determined by titrating against a solution of methyl viologen. Synthesis of methyl-SCoM, CoB5SH, CoB6SH, CoBSH, CoB8SH, and CoB9SH Methyl-SCoM was prepared from HSCoM and methyl iodide (49). The homodisulfides, CoB5S-SCoB5, CoB6S-SCoB6, CoBS-SCoB, CoB8S-SCoB8, CoB9S-SCoB9 were prepared as described from 5-bromovaleric acid, 6-bromohexanoic acid (Sigma-Aldrich, St. Louis, MO), 7-bromoheptanoic acid (Karl Industries, Aurora, OH), 8-bromooctanoic acid, and 9- bromononanoic acid (Matrix Scientific, Columbia, SC), respectively (43, 46). The free thiol forms of CoB5SH, CoB6SH, CoBSH, CoB8SH, and CoB9SH were generated by the reduction of the homodisulfides as previously described (45). The purity of the CoBSH analogues was determined by 1H NMR spectroscopy. All compounds synthesized were stored in a Vacuum Atmospheres chamber maintained at an oxygen level below 1 ppm, as monitored continually with an oxygen analyzer (model 317, Teledyne Analytical Instruments, City of Industry, CA) until use. M. marburgensis Growth and MCRred1 Purification Buffer preparations and all manipulations were performed under strict anaerobic conditions in a Vacuum Atmospheres chamber maintained at an oxygen level below 1 ppm, as monitored continually with an oxygen analyzer (model 317, Teledyne Analytical Instruments, City of Industry, CA). MCRred1 was isolated from M. marburgensis cultured on Cedervall et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript H2/CO2 (80%/20%) at 65 °C in a 14-L fermentor (New Brunswick Scientific Co., Inc. New Brunswick, NJ). Culture media were prepared as previously described (20, 24). MCRred1 was generated in vivo and purified as described previously (20). The purification procedure routinely generates 65–75% MCRred1 as determined by UV-visible and EPR spectroscopy. Spectroscopy of MCR UV-visible spectra of the Ni(I)-containing MCRred1 were recorded in the anaerobic chamber using a spectrophotometer (Model USB4000-UV-VIS, Ocean Optics, Dunedin, FL). EPR spectra were recorded on a Bruker EMX spectrometer (Bruker Biospin Corp., Billerica, MA), equipped with an Oxford ITC4 temperature controller, a Hewlett-Packard Model 5340 automatic frequency counter and Bruker gaussmeter. The EPR spectroscopic parameters included: temperature, 70 K; microwave power, 10 mW; microwave frequency, 9.43 GHz; receiver gain, 2 × 104; modulation amplitude, 10.0 G; modulation frequency, 100 kHz. Double integrations of the EPR spectra were performed and referenced to a 1 mM copper perchlorate standard. The NMR data were acquired at 298 K on a Bruker Avance DRX 500 MHz instrument equipped with a TXI cryoprobe. Preparation of MCRred1 for Crystallization All crystallization experiments were performed in the anaerobic chamber in which MCR was purified unless otherwise noted. MCRred1 was prepared in 50 mM Tris, pH 7.6 and excess HSCoM was removed by buffer exchange using an Amicon Ultra centrifuge filter with a 50 kDa cut-off membrane (Millipore). Typically, 2–3 ml of MCRred1 was exchanged with 10–15 ml of 50 mM Tris, pH 7.6. The enzyme was concentrated to 500–600 μl, and this process was repeated three times. The fraction of MCRred1 in the purified MCR sample was calculated from the UV-visible spectrum using extinction coefficients of 27.0 mM−1cm−1 at 385 nm for Ni(I)-MCRred1, and 9.15 mM−1cm−1 at 420 nm for Ni(II)- MCRred1-silent (20). The amount of MCRred1 in samples used for crystallization was determined to be ~80% and the concentration of total enzyme used was in the range of about 120–150 μM (~32–40 mg/ml). All crystallization experiments were performed anaerobically by incubating 2.0 μl of enzyme solution in 50 mM Tris, pH 7.5 and 2.0 μl of reservoir solution (100 mM Hepes-Na, pH 7.3/7.5/8.0; 150 mM magnesium acetate (Mg(CH3COO)2), and 20/22% (w/v) PEG 400) in a sitting drop over 1 ml reservoir solution at 9 ºC. Triangular and rectangular prismatic crystals with a bright yellowish-green color confirmed the presence of nickel coenzyme F430. The crystals grew to a size of approximately 100–200 μm in 4–5 days. CoBSH-depleted crystals were obtained by incubating 2 μl of a reaction mixture containing 139 μM MCRred1 and 13 mM HSCoM with 2 μl of reservoir solution (100 mM Hepes-Na pH 7.5, 150 mM magnesium acetate (Mg(CH3COO)2), 22% PEG 400). Crystals of MCR complexed with the CoBSH analogues were grown by co-crystallization. The CoB5SH co-crystals were obtained by incubating 2μl enzyme solution containing 124 μM MCRred1, 10 mM methyl-SCoM and 1 mM CoB5SH with 2μl of reservoir solution (100 mM Hepes-Na, pH 7.5, 150 mM (Mg(CH3COO)2), 22 % PEG 400). The crystals with bound CoB6SH and CoB9SH were obtained by co-crystallization of 1 mM of analogue with 142 μM MCRred1 and equilibrated with 2μl of reservoir solution (100 mM Hepes-Na, pH 7.5, 150 mM Mg(CH3COO)2, 20 % PEG 400 for CoB6SH and 100 mM Hepes-Na, pH 7.3, 150 mM Mg(CH3COO)2), 22 % PEG 400 for CoB9SH). Crystals were cryoprotected in reservoir solution containing 25 % (v/v) PEG 400 by soaking for 2–5 minutes before cryocooling in liquid nitrogen in the anaerobic chamber. Crystals of CoB8SH bound to MCR were obtained by incubating 2 μl of a mixture of 119 μM MCRred1 and 1 mM CoB8SH with 2 μl of reservoir solution (100 mM Hepes-Na, pH 7.3, 150 mM Mg(CH3COO)2), 20 % PEG 400). Before cryoprotection, the crystals were soaked for 5–10 minutes in a 100 mM solution of methyl iodide2. The methyl iodide solution used for soaking was prepared by adding a concentrated stock of methanolic solution of methyl iodide to the reservoir Cedervall et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript solution. Soaked crystals were quickly cryoprotected as described above and cryocooled in the anaerobic chamber. X-ray Diffraction Data Collection, Processing and Refinement X-ray diffraction data were collected at 100 K on a ADSC Quantum-315 detector at the APS Beamline 14-BM-C (BioCARS). The wavelength of X-rays was 0.979 Å. Data were processed using HKL2000 (50). As in the previous X-ray crystallographic studies, the crystals belong to the monoclinic space group P21 (a = 82 Å, b = 118 Å, c = 122 Å, β = 92°), with one MCR molecule (two active sites) per asymmetric unit (5, 33). For refinement, REFMAC in the Collaborative Computational Project Number 4 (CCP4) program suite was used (51). A random sample of 5 % of the data across all resolution shells was chosen to check refinement progress through calculation of an Rfree. The same reflections were used to calculate Rfree for all structures, thus preventing bias due to high structural identity. The remaining reflections were used in refinement (Rwork). Model building was done using the Crystallographic Object-Oriented Toolkit (COOT) (52). The diffraction data and their models are designated as MCRCoBXSH, where X is the number of carbons in the alkanoyl portion of the analogue. Library files in CCP4 for F430 and CoBSH were incorrect, and these were modified in Monomer Library Sketcher in the CCP4 program suite by comparison with schematic drawings from Grabarse et al. (33). Co-ordinate and CCP4 library files for the different CoBSH analogues were created in Monomer Library Sketcher. The general model building and refinement strategy for all structures was as follows. It was clear from the electron density in the substrate channel and at the active site that mixtures of species were present in all datasets. These could be visualized with Fo-Fc and Fo-Fo difference electron density maps (Supporting Information, Figure S1). The known positions of CoBSH and HSCoM from the published Ni(II)-MCR crystal structures (PDB codes 1hbn, 1hbo, 1hbu (33)) were used as guides to determine which species could be present in each dataset, and these were then simultaneously modeled into the electron density. By alteration of their relative occupancies (in 10% increments) followed by refinement, the ratio of occupancy between different species was determined using the assumption that the average B-factors for all molecular species bound should be similar to that of F430 and adjacent well-ordered protein atoms within the active site and substrate channel. The combinations of modeled ligands were constantly reassessed throughout refinement based on the remaining difference electron density. This included test refinements of different ligand combinations during the latter stages, thus using the optimized phases to check whether a different combination of ligands could also explain the electron density. Sensible chemical structures and interactions, along with keeping the combined occupancies of sterically mutually exclusive species ≤ 100%, were maintained throughout refinement. The model was finally accepted when the difference electron density map was minimal and the B-factors for the models converged. In practice the first structure refined was that of MCRCoB5SH. Initial phases were generated by difference Fourier using a previously determined crystal structure (PDB code 1mro (5)) but with all non-bonded molecules, including water, removed from the model except F430. Initial rigid body refinement followed by restrained refinement of MCRCoB5SH reduced the Rwork to 26.5%. After model building and subsequent rounds of restrained refinement the Rwork was 14.3% (Rfree 16.6%). Of the five structures only the CoB5SH analogue is completely coincident with CoBSH, and so particular care had to be used in teasing apart the ratios of the two species in modeling the MCRCoB5SH electron density. This was done by 2Methyl iodide was added with the intention of creating a Ni(III)-methyl species, which CoB8SH stabilizes. This was not achieved, but the diffraction quality of this crystal was significantly better than crystals co-crystallized with CoB8SH alone, and so has been included in this study. Cedervall et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript initially refining 100% CoBSH in the substrate channel. Positive Fo-Fc difference density located at the carbon where the shorter CoB5SH thiol might be expected to be indicated the presence of a more electron-rich species than carbon, which is consistent with the presence of the CoB5SH sulfur. The refinement converged at a model containing a 50:50 mix of CoBSH and the analogue. However, positive Fo-Fc difference density was still present at the position of the CoBSH thiol, therefore a water molecule was added to the CoB5SH model at 50% occupancy and upon refinement this accounted for the electron density. An illustration of the electron density quality from this structure is shown in Supporting Information, Figure S2. An HSCoM-, CoBSH- and CoBSH analogue-free version of the refined MCRCoB5SH structure was used as the starting model to generate initial phases for the four other structures. After the initial round of restrained refinement the Rwork for these structures were reduced to 14.5–15.6 %. RESULTS AND DISCUSSION Crystal Structures of MCR Five crystal structures were determined, four of which are in complex with CoBSH analogues differing in the number of carbons atoms in the alkanoyl portion of the molecule. CoBSH is an N-7-mercaptoheptanoyl-containing molecule, whereas the four CoBSH analogues contain N-5-mercaptopentanoyl-, N-6-mercaptohexanoyl-, N-8-mercaptooctanoyl- or N-9-mercaptononanoyl-moieties (Figure 2). The corresponding crystal structures are designated as MCRCoBXSH, where X is the number of carbons in the alkanoyl portion of the analogue. The other crystal structure is of MCRred1c-silent (MCR in the Ni(II) state in complex with HSCoM, designated here as MCRHSCoM) that is CoBSH-depleted. The datasets have resolutions in the range from 1.35 – 1.8 Å. Although the crystallizations were set up with the MCR solution predominantly in the Ni(I)-MCRred1 state, by the time X-ray diffraction data were collected they had been oxidized to the Ni(II)-MCRred1-silent state (Supporting Information). Following data collection there was no evidence for photoreduction of the Ni(II) back to Ni(I) in any of the crystals, as assessed by single crystal UV-visible microspectrophotometry (Supporting Information and Figure S3). Attempts to photoreduce the crystals using different wavelengths and temperatures were unsuccessful (Supporting Information). Overall, the resulting structures are very similar to each other and to the previously published structures of MCR, with differences mainly localized to the active site and substrate channel. The two active sites in the ASU were refined independently. Unless otherwise stated there was no difference between them. All five datasets contain a mixture of species bound to the enzyme. There is always a background of CoBSH and HSCoM, which co-purify with MCR and cannot be fully removed by extensive buffer exchange or by the addition of a CoBSH analogue. HSCoM is added during purification of MCR, as it stabilizes the resting active Ni(I) state (unpublished data), and this leads to HSCoM occupancies between 50–100% amongst the structures (Table 1). In contrast CoBSH, which is not added during purification, has occupancies ranging from 30–50%. As these confounding species have all been described at high occupancy in other crystallographic studies, the structural data of interest could be isolated (5, 33). In each case, the additional electron density could be explained by inclusion of the appropriate CoBXSH model used in that experiment at 50% or higher occupancy. The resulting models, along with 2Fo-Fc electron density, are shown in Figure 3. The Rwork for the final structures range from 13.0 to 15.0 % (Rfree 15.5 to 19.5 %). The X-ray data collection, processing, refinement and model building statistics are given in Table 1. Cedervall et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Analogues shorter than CoBSH; CoB5SH and CoB6SH CoB5SH is two methylene groups shorter than CoBSH, the MCR substrate. The MCRCoB5SH structure is to 1.35 Å resolution. As expected the pentanoyl chain follows the path of the CoBSH heptanoyl carbons down the substrate channel, and thus its thiol is positioned in the same place as the second carbon preceding the CoBSH thiol (Figure 3A and 4). There are no published MCR kinetic studies using CoB5SH, but as it binds in the substrate channel, it is likely to be an inhibitor. CoB6SH is one methylene shorter than CoBSH, and is a slow substrate of MCR. In this case the 1.4 Å resolution electron density of MCRCoB6SH indicates that the analogue unexpectedly binds in the substrate channel such that its thiol is virtually in the same position as the thiol of CoBSH (Figure 3B and 4). The hexanoyl chain is oriented so that it takes a shorter route down the substrate channel between carbons 2 and 5 (with the carbonyl carbon labeled as carbon 1) than CoBSH (Figure 4, Supporting Information, Figure S4). This short-cut is not seen in any of the other CoBXSH complex crystal structures, but presumably arises because this CoB6SH binding conformer is energetically more favorable, although it is not clear from the structure why this might be the case. CoB6SH binds very tightly to MCR, with an apparent Ki value of 0.1 μM (3). Water structure in the absence of HSCoM The electron density for the MCRCoB6SH crystal structure only supported the modeling of 50 % bound HSCoM. In the fraction of MCR molecules where HSCoM is absent, the HSCoM binding site is occupied by a network of four water molecules (Supporting Information, Figure S5). Two waters are positioned close to the absent sulfonate oxygen positions of HSCoM. Based on the presence of positive difference electron density, a third water was modeled ligated to the Ni, and refined to a distance of ~2 Å (2.0 Å and 2.1 Å in the two active sites of the ASU) with no distance restraint imposed between the Ni and water. This water is in a similar position as the Ni coordinating sulfonate oxygen of the heterodisulfide product in MCRsilent (Supporting Information, Figure S6 and PDB codes 3m32, 1hbm) (5, 33). The fourth water was in the vicinity of the expected position of a bridging water (W1) seen in other structures (Figure 1, 3A and 3C). Water structure in the absence of CoBSH The 1.45 Å resolution electron density obtained for MCRHSCoM indicates that the substrate channel contains only 30 % CoBSH. Nine ordered waters (W1–W9), along with an acetate ion from the crystallization solution occupy the channel, with the acetate positioned where the phosphothreonine linkage of CoBSH would be (Figure 3C). Presumably 1–2 further waters would replace the acetate under physiological conditions. Other than W3 and W7, the waters form hydrogen bonds with protein (Figure 5). One water (W2) occupies the same site as the CoBSH thiol. Presumably due to the loss of favorable interactions that exist when CoBSH is present, the hydrophobic side-chain of Valα482 adopts a second conformation modeled at 60 % occupancy (Supporting Information, Figure S7). Position of the “bridging” water, W1 The equivalent of W1 has previously been observed in MCRred1-silent and MCRox1-silent crystal structures where, in the presence of CoBSH and HSCoM, it is sited equidistant (3.2 Å) between the two coenzyme thiols (PDB codes: 1hbn, 1hbo, 1hbu) and thus been termed the “bridging water” (Figure 1) (5, 33). However, in the MCRHSCoM structure, due to the presence of the more polarized W2 water, W1 is displaced away from HSCoM to maximize the hydrogen bond interaction with W2 (2.9 Å to W2; 3.5 Å to HSCoM thiol, Figure 5). In the MCRCoB5SH structure that also contained W2, the electron density indicated that this Cedervall et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript repositioning of W1 towards W2 also occurred. In contrast, the MCRCoB6SH structure contained 100% thiol at the CoBSH position, but a partial occupancy of HSCoM (50%). In this case the electron density for W1 indicated it had moved towards the nickel to form an optimal hydrogen bond with a Ni-ligating water that was only present in the absence of HSCoM (3.7 Å to CoB(6)SH thiol; 3.0 Å to Ni-ligating water, Supporting Information, Figure S5). In all structures reported here, W1 (if present) appears to be a sensitive indicator of the relative electronegativity of the Ni-ligated atom to that occupying the position of the CoBSH thiol, and was a useful check in the crystallographic modeling and refinement process. Flexibility in the substrate channel: Alternative protein conformers The binding site of HSCoM (and presumably methyl-SCoM) is more deeply buried within the enzyme, and so it must enter prior to CoBSH for productive chemistry to occur. As binding of CoBSH in the absence of co-substrate would be inhibitory, it was suggested that a conformational change upon methyl-SCoM binding might lower the Kd for CoBSH, and thus promote an ordered mechanism. Compared to the 1hbn MCRox1-silent and 1hbu MCRred1-silent crystal structures, which both have full occupancy HSCoM, the lower occupancy of HSCoM in the 1hbo MCRred1-silent structure was associated with significantly greater flexibility within the channel, and the ability to model a second conformation of a Gly-rich amino acid stretch that formed part of the CoBSH channel. This suggested that methyl-SCoM binding might cause the channel to become more ordered, increasing the affinity of MCR for CoBSH by conformational restriction rather than a switch mechanism where the structure reorganizes from one well-defined conformer to another (33). In the MCRHSCoM data containing 30 % CoBSH and 100% HSCoM, the Fo-Fc difference electron density map at one of the two independent active sites in the ASU contained positive peaks that suggested the presence of an alternate conformation also involving this part of the polypeptide (Supporting Information, Figure S8). Using this as a guide, a similar second conformation involving seven contiguous amino acid residues of the same Gly-rich amino acid stretch (β366–372) could be modeled and refined at 20 % occupancy leaving no residual difference density. Parts of the α′ subunit (α′111–129 and α′237–242) that are in close proximity to this stretch of amino acids also exhibit second conformations, with the main-chain carbonyl of α′243 in van der Waals contact with the B ring of F430 tetrapyrrole (Supporting Information, Figure S9). Modeling these at 20% occupancy accounted for the weak positive Fo-Fc difference electron density peaks observed in these areas. The evidence of alternate conformers in these areas lends support to the proposal that increased flexibility in the substrate channel propagates through the protein (33). The MCRCoB6SH crystal structure contains 50 % CoB6SH, 50% CoBSH and 50% HSCoM. In this case there is no evidence of an alternate loop conformation in either active site of the ASU. However, as CoBSH and CoB6SH combined are at 100% occupancy, it is not surprising their favorable interactions with the substrate channel would reduce conformational disorder, despite the partial occupancy of HSCoM. Analogues longer than CoBSH; CoB8SH and CoB9SH Both analogues could be accommodated in the MCR substrate channel (Figures 3D and 3E). The electron density supported final models containing 50% CoB8SH for MCRCoB8SH (1.8 Å resolution) and 60 % CoB9SH for MCRCoB9SH (1.45 Å resolution). The phosphate head- groups are in identical positions to those of CoBSH, CoB5SH, CoB6SH (Figure 4) (5, 33). Both analogues follow the crystallographically observed chain path of bound CoBSH, with the extra atoms displacing the W1 water and placing the thiols closer to the nickel (Figure 6). CoB9SH does have a second conformer that deviates from the CoBSH path, but the thiol position for this conformer and the CoBSH-tracking conformer are identical (Figure 3E and Cedervall et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supporting Information, Figure S10). Interestingly, the thiol of CoB8SH is not coincident with the CoB9SH carbon that precedes the CoB9SH thiol (Figure 6). CoB8SH is an MCR inhibitor with an apparent Ki of 15 μM (3). CoB9SH has never been tested for inhibition of MCR-catalyzed methane formation, but it is reasonable to assume that it would be an inhibitor. CoBXSH thiol-to-nickel spatial relationship The position of CoBSH in previous crystal structures poses a conundrum (5, 33). In all the proposed catalytic mechanisms, CoBSH must interact with species generated at the nickel. Perplexingly, the sulfur of the CoBSH substrate is 8.8 Å from the Ni(II) in the MCRox1-silent and both MCRred1-silent crystal structures, and 6.4 Å from the thiol of the substrate analogue HSCoM (Figure 1). Modeling studies demonstrated that the addition of a methyl group to HSCoM did not bridge this gap (35, 45, 53). Therefore, a conformational change has been postulated that would enable CoBSH to penetrate deeper into the substrate channel, and thus approach closer to any nickelbound species. The heterodisulfide product in the MCRsilent crystal structure has the CoBSH portion in virtually the same place as in MCRox1-silent, giving no clue to possible structural changes that might occur to facilitate CoBSH reacting with nickel-associated intermediates (5, 33). Trigonometry suggests that if the alkanoyl chain of CoBSH or its analogue is in an extended conformation, each additional unit in the chain would lead to the thiol moving ~1.2 Å towards the Ni. Until this study there have been no crystal structures of CoBSH analogues in complex with MCR, so mechanistic studies using different chain length analogues of CoBSH assumed that shorter analogues would trace the observed path of CoBSH, and longer analogues would penetrate about ~1.2 Å deeper per additional chain unit into the channel. In the case of the shortest analogue CoB5SH, it does indeed follow the path of CoBSH, with the thiol of CoB5SH being 2.8 Å away from the thiol position of CoBSH. However, due to the conformation CoBSH adopts when bound in the substrate channel, the difference in the S-Ni distance is small; the CoB5SH thiol being only 0.5 Å farther from the Ni than CoBSH (8.8 Å for CoBSH vs. 9.3 Å for CoB5SH) (Table 2). This is due to the alkanoyl chain of CoBSH not being in an extended conformation from carbons 4 to 6 (carbon 1 is the carbonyl carbon). CoB6SH on the other hand adopts a conformation that places its thiol in virtually the same position as the thiol of CoBSH (Figure 4 and Table 2). This is consistent with CoB6SH being a substrate. However, the kcat is 1000-fold lower than for CoBSH (3, 35) although its Km value (180 μM, Dey & Ragsdale, in preparation) is similar to that of CoBSH (Km = 75 μM (3)). The reason for this may be that the shorter alkanoyl chain may not enable the analogue thiol to approach the nickel close enough for efficient catalysis, and thus explain why CoB6SH is such a poor substrate. In the case of the longer CoBXSH analogues, the sulfur of CoB8SH is 2.6 Å closer to the Ni ion of F430 than that of CoBSH, and 2.5 Å closer to the thiol of HSCoM (Figure 6 and Table 2). The CoB9SH molecule follows the path of CoBSH, and reaches only a little further into the substrate channel than CoB8SH, with the CoB9SH thiol positioned 2.9 Å closer to the Ni than the thiol of CoBSH (Figure 6 and Table 2). This is only 0.3 Å closer than the distance observed for the CoB8SH thiol, even though they are non-coincident. The distance to the thiol of HSCoM is 2.6 Å closer than that of the substrate, CoBSH; only 0.1 Å closer than the CoB8SH thiol. The two analogue thiols are above an annular hydrophobic aromatic environment created by Pheα330, Tyrα333, Pheα443, Pheβ361 and Tyrβ367 that lies between them and F430 (Figure 6). As a result, penetrating further into the channel may be energetically unfavorable, consistent with the small difference in relative distances between the CoB8SH/CoB9SH thiols and the HSCoM thiol/F430 nickel. The annulus is proposed to be catalytically important in positioning methyl-SCoM and stabilizing the methane product, Cedervall et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript and the tyrosines have been proposed to be proton donors associated with mechanism II (Scheme 2B) (5, 33). Thus, there appear to be three preferential distances for thiols (including that of HSCoM) within the MCR substrate channel; HSCoM at 2.4 Å, CoB6SH/CoBSH at 8.7 – 8.8 Å and CoB8SH/CoB9SH at 5.9 – 6.2 Å from the nickel of F430 (Table 2). Recent ENDOR, high-field continuous and pulse EPR work has identified changes in nickel co-ordination when CoBSH is added to MCRred1c (active Ni(I)-MCRred1 + HSCoM) (14, 15, 18, 31). This generates up to 50% MCRred2 which is comprised of two distinct nickel co- ordination geometries; an axial MCRred2a formally assigned as a Ni(III)-hydride, and a rhombic MCRred2r in which the thiol of HSCoM is a Ni(I) ligand (Supporting Information, Scheme S1). Formation of MCRred2 could also be induced by addition of CoBSH substrate analogues CoBS-CH3 and CoBS-CF3, which have a chain length one unit longer than substrate CoBSH (18, 53). The CoBS-CF3 enabled 19F-ENDOR studies to be performed, and demonstrated that following CoBS-CF3 addition the remaining MCRred1 species had Ni(I)-19F distances of 6.2 – 7.7 Å. This distance range agreed with a CoBS-CF3 model created using the CoBSH position observed in the MCRox1-silent crystal structure (53). However, in the MCRred2 species the Ni(I)-19F distances had shortened indicating a movement of ~ 2 Å towards the nickel (MCRred2a Ni(I)-19 F, 4.0 – 5.5 Å; MCRred2r Ni(I)-19F, 4.5 – 5.7 Å). In the case of the CoB8SH analogue (the closest equivalent to CoBS- CF3 used in this study) the thiol to Ni(II) distance lies between the distance ranges observed in the CoBS-CF3 studies, and so the fluorine(s) of the CoBS-CF3 in MCRred2 might penetrate a little further into the hydrophobic annulus in the MCRred2 species. As the alkanoyl chain of CoBSH is not fully extended it could easily undergo a similar conformation change to that observed in the MCRred2 state. Reaction of MCR Ni(III)-alkyl species’ with CoB8SH and CoB9SH The two longer CoBXSH analogues have been shown to undergo alkylation when reacted with MCRPS, a [Ni(III)-alkyl ↔ Ni(II)-alkylsulfonate radical] formed from reaction of Ni(I)-MCRred1 with bromopropanesulfonate (BPS) (Supporting Information, Scheme S1) (20, 23, 30, 45, 54). BPS is a substrate of MCRred1 in a reaction that involves a rapid CoBSH-independent nucleophilic attack by Ni(I) on BPS to displace bromide and generate MCRPS at a rate ~60-fold faster than generation of methane from CoBSH and methyl- HSCoM (20, 45). Certain thiols can eliminate the propylsulfonate to yield a thioether product and regenerate MCRred1, although at a rate 1000-fold slower than methane formation (45). Both CoB8SH and CoB9SH can react with MCRPS to regenerate MCRred1, but CoBSH cannot. The overall second-order rate constant for the reactivation of MCR by CoB8SH is 160 M−1s−1, whereas for CoB9SH the reaction is more sluggish (12 M−1s−1). CoB9SH might be expected to be closer to the proximal Ni ligand. It was therefore proposed that this caused steric interference and explained why CoB9SH was a poorer reactivator of MCR than CoB8SH. Our study has shown that the thiols of these two analogues are placed such that they are approximately the same distance (~3.7 Å) from the thiol of HSCoM ligated to the Ni atom (Table 2). The Ni(II)-HSCoM bond is 2.4 Å, whereas an Ni(III)-alkyl bond is expected to be ~2 Å (24, 33), thus indicating that a conformational change is required to effect the nucleophilic attack of the CoB8SH and CoB9SH thiols on an alkyl- bound species. It would thus appear that a conformational change, such as observed in MCRred2, is required for this chemistry also (53). A Ni(III)-alkyl species is akin to the first intermediate in mechanism I of MCR-catalyzed methane formation, Ni(III)-methyl (MCRMe, Supporting Information, Scheme S1, Scheme 2A) (11, 27). MCRMe has been shown to be capable of generating MCRred1 and methyl- SCoM upon addition of HSCoM (which is the reverse of mechanism I, step 1, Scheme 2A); Cedervall et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript similar chemistry to the observed formation of a thioether product from the Ni(III)-alkyl. Further addition of CoBSH following HSCoM treatment of MCRMe led to methane and heterodisulfide formation, the natural products of methanogenesis. Although this lends credence to mechanism I, it should be noted that like MCRPS, MCRMe in these experiments was generated artificially. However, the MCRCoB9SH crystal structure demonstrates that the two additional methylene units in the alkanoyl chain c.f. CoBSH, do not necessarily translate into direct interaction of the thiol with the nickel proximal ligand. However, this could represent the favorable position for a CoBSH thiol interacting with the methyl group of methyl-SCoM. Just as the alkanoyl chain of CoB6SH has a more extended conformation than CoBSH in the substrate channel, CoBSH could also adopt a more extended conformation so that its thiol was in a similar position as the thiol of CoB8SH, priming it for reaction with a nickel bound species. If a significant conformational change is required early in MCR-catalyzed chemistry, which would be a requirement of mechanism I, catalysis may well involve a rearrangement of the aromatic amino acid annulus due to the presence of the methyl of methyl-SCoM, and this might enable deeper penetration of CoBSH into MCR (Figure 6). All the crystal structures in this study, and those solved previously, are of the inactive Ni(II)-MCR, which disfavors close approach to the nickel in the absence of Ni(I)-bound methyl-SCoM, even in the case of CoB9SH. Conclusion The goal of this study was to induce structural changes within the substrate channel and active site of MCR using analogues of coenzyme CoBSH. It was hoped that this would shed light on the nature of conformational changes that have been proposed to occur in MCR catalysis. We have shown that that the CoBXSH analogues do not lead to any significant conformational changes within the context of inactive Ni(II)-MCR. Therefore, it may be that methyl-SCoM is the key coenzyme, in combination with a nickel oxidation state of 1 (and 3), that triggers a conformational change bringing the thiol of CoBSH closer to the nickel. Thus, the crystal structure of the Ni(I)-methyl-SCoM/MCR complex may be required to structurally define the conformational changes required for MCR-mediated chemistry. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments X-ray data were collected at beam-line 14-BM-C and photoreduction studies at 14-ID-B, BioCARS at the Advanced Photon Source (APS), Argonne National Laboratory, Argonne, IL. We thank Vukica Srajer and Yu- Sheng Chen for valuable assistance during data collection. Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Basic Energy Sciences, Office of Science under Contract No. DE- AC02-06CH11357. Use of the BioCARS Sector 14 was supported by the National Institutes of Health, National Center for Research Resources, under grant number RR007707. Computer resources were provided by the Basic Sciences Computing Laboratory of the University of Minnesota Supercomputing Institute, and we thank Can Ergenekan for his support. We also thank Ed Hoeffner at the Kahlert Structural Biology Laboratory (KSBL) at the University of Minnesota. Use of the KSBL was supported by a Minnesota Partnership for Biotechnology and a Medical Genomics Grant SPAP-05-0013-P-FY06. References 1. Thauer RK. Biochemistry of methanogenesis: a tribute to Marjory Stephenson. Microbiology. 1998; 144:2377–2406. [PubMed: 9782487] 2. Thauer RK, Shima S. Methane as fuel for anaerobic microorganisms. Ann N Y Acad Sci. 2008; 1125:158–170. [PubMed: 18096853] Cedervall et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 3. Ellermann J, Hedderich R, Bocher R, Thauer RK. The final step in methane formation. Investigations with highly purified methyl-CoM reductase (component C) from Methanobacterium thermoautotrophicum (strain Marburg). Eur J Biochem. 1988; 172:669–677. [PubMed: 3350018] 4. Ellefson WL, Wolfe RS. Component C of the methylreductase system of Methanobacterium. J Biol Chem. 1981; 256:4259–4262. [PubMed: 6783657] 5. Ermler U, Grabarse W, Shima S, Goubeaud M, Thauer RK. Crystal structure of methyl-coenzyme M reductase: the key enzyme of biological methane formation. Science. 1997; 278:1457–1462. [PubMed: 9367957] 6. Diekert G, Gilles HH, Jaenchen R, Thauer RK. Incorporation of 8 succinate per mol nickel into factors F430 by Methanobacterium thermoautotrophicum. Arch Microbiol. 1980; 128:256–262. [PubMed: 7212929] 7. Diekert G, Jaenchen R, Thauer RK. Biosynthetic evidence for a nickel tetrapyrrole structure of factor F430 from Methanobacterium thermoautotrophicum. FEBS Letters. 1980; 119:118–120. [PubMed: 7428919] 8. Whitman WB, Wolfe RS. Presence of nickel in Factor F430 from Methanobacterium bryantii. Biochem Biophys Res Comm. 1980; 92:1196–1201. [PubMed: 7370029] 9. Albracht SPJ, Ankel-Fuchs D, Böcher R, Ellermann J, Moll J, van der Zwann JW, Thauer RK. Five new EPR signals assigned to nickel in methyl-coenzyme M reductase from Methanobacterium thermoautotrophicum, strain Marburg. Biochim Biophys Acta. 1988; 955:86–102. 10. Dey M, Kunz RC, Lyons DM, Ragsdale SW. Characterization of alkyl-nickel adducts generated by reaction of methyl-coenzyme m reductase with brominated acids. Biochemistry. 2007; 46:11969– 11978. [PubMed: 17902704] 11. Dey M, Telser J, Kunz RC, Lees NS, Ragsdale SW, Hoffman BM. Biochemical and spectroscopic studies of the electronic structure and reactivity of a methyl-Ni species formed on methyl- coenzyme M reductase. J Am Chem Soc. 2007; 129:11030–11032. [PubMed: 17711283] 12. Duin EC, Cosper NJ, Mahlert F, Thauer RK, Scott RA. Coordination and geometry of the nickel atom in active methyl-coenzyme M reductase from Methanothermobacter marburgensis as detected by X-ray absorption spectroscopy. J Biol Inorg Chem. 2003; 8:141–148. [PubMed: 12459909] 13. Duin EC, Signor L, Piskorski R, Mahlert F, Clay MD, Goenrich M, Thauer RK, Jaun B, Johnson MK. Spectroscopic investigation of the nickel-containing porphinoid cofactor F(430). Comparison of the free cofactor in the (+)1, (+)2 and (+)3 oxidation states with the cofactor bound to methyl- coenzyme M reductase in the silent, red and ox forms. J Biol Inorg Chem. 2004; 9:563–576. [PubMed: 15160314] 14. Finazzo C, Harmer J, Bauer C, Jaun B, Duin EC, Mahlert F, Goenrich M, Thauer RK, Van Doorslaer S, Schweiger A. Coenzyme B induced coordination of coenzyme M via its thiol group to Ni(I) of F430 in active methyl-coenzyme M reductase. J Am Chem Soc. 2003; 125:4988–4989. [PubMed: 12708843] 15. Finazzo C, Harmer J, Jaun B, Duin EC, Mahlert F, Thauer RK, Van Doorslaer S, Schweiger A. Characterization of the MCRred2 form of methyl-coenzyme M reductase: a pulse EPR and ENDOR study. J Biol Inorg Chem. 2003; 8:586–593. [PubMed: 12624730] 16. Goubeaud M, Schreiner G, Thauer RK. Purified methyl-coenzyme-M reductase is activated when the enzyme-bound coenzyme F430 is reduced to the nickel(I) oxidation state by titanium(III) citrate. Eur J Biochem. 1997; 243:110–114. [PubMed: 9030728] 17. Harmer J, Finazzo C, Piskorski R, Bauer C, Jaun B, Duin EC, Goenrich M, Thauer RK, Van Doorslaer S, Schweiger A. Spin density and coenzyme M coordination geometry of the ox1 form of methyl-coenzyme M reductase: a pulse EPR study. J Am Chem Soc. 2005; 127:17744–17755. [PubMed: 16351103] 18. Harmer J, Finazzo C, Piskorski R, Ebner S, Duin EC, Goenrich M, Thauer RK, Reiher M, Schweiger A, Hinderberger D, Jaun B. A nickel hydride complex in the active site of methyl- coenzyme m reductase: implications for the catalytic cycle. J Am Chem Soc. 2008; 130:10907– 10920. [PubMed: 18652465] Cedervall et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 19. Hinderberger D, Ebner S, Mayr S, Jaun B, Reiher M, Goenrich M, Thauer RK, Harmer J. Coordination and binding geometry of methyl-coenzyme M in the red1m state of methyl- coenzyme M reductase. J Biol Inorg Chem. 2008; 13:1275–1289. [PubMed: 18712421] 20. Kunz RC, Horng YC, Ragsdale SW. Spectroscopic and kinetic studies of the reaction of bromopropanesulfonate with methyl-coenzyme M reductase. J Biol Chem. 2006; 281:34663– 34676. [PubMed: 16966321] 21. Mahlert F, Bauer C, Jaun B, Thauer RK, Duin EC. The nickel enzyme methyl-coenzyme M reductase from methanogenic archaea: In vitro induction of the nickel-based MCR-ox EPR signals from MCR-red2. J Biol Inorg Chem. 2002; 7:500–513. [PubMed: 11941508] 22. Mahlert F, Grabarse W, Kahnt J, Thauer RK, Duin EC. The nickel enzyme methyl-coenzyme M reductase from methanogenic archaea: in vitro interconversions among the EPR detectable MCR- red1 and MCR-red2 states. J Biol Inorg Chem. 2002; 7:101–112. [PubMed: 11862546] 23. Rospert S, Voges M, Berkessel A, Albracht SP, Thauer RK. Substrate-analogue-induced changes in the nickel-EPR spectrum of active methyl-coenzyme-M reductase from Methanobacterium thermoautotrophicum. Eur J Biochem. 1992; 210:101–107. [PubMed: 1332856] 24. Sarangi R, Dey M, Ragsdale SW. Geometric and electronic structures of the Ni(I) and methyl- Ni(III) intermediates of methyl-coenzyme M reductase. Biochemistry. 2009; 48:3146–3156. [PubMed: 19243132] 25. Tang Q, Carrington PE, Horng YC, Maroney MJ, Ragsdale SW, Bocian DF. X-ray absorption and resonance Raman studies of methyl-coenzyme M reductase indicating that ligand exchange and macrocycle reduction accompany reductive activation. J Am Chem Soc. 2002; 124:13242–13256. [PubMed: 12405853] 26. Telser J, Davydov R, Horng YC, Ragsdale SW, Hoffman BM. Cryoreduction of methyl-coenzyme M reductase: EPR characterization of forms, MCR(ox1) and MCR (red1). J Am Chem Soc. 2001; 123:5853–5860. [PubMed: 11414817] 27. Yang N, Reiher M, Wang M, Harmer J, Duin EC. Formation of a nickel-methyl species in methyl- coenzyme M reductase, an enzyme catalyzing methane formation. J Am Chem Soc. 2007; 129:11028–11029. [PubMed: 17711279] 28. Albracht SPJ, Ankelfuchs D, Vanderzwaan JW, Fontijn RD, Thauer RK. A New Electron- Paramagnetic-Res Signal of Nickel in Methanobacterium-Thermoautotrophicum. Biochim Biophys Acta. 1986; 870:50–57. 29. Telser J, Horng YC, Becker DF, Hoffman BM, Ragsdale SW. On the assignment of nickel oxidation states of the Ox1, Ox2 forms of methyl-coenzyme M reductase. J Am Chem Soc. 2000; 122:182–183. 30. Hinderberger D, Piskorski RR, Goenrich M, Thauer RK, Schweiger A, Harmer J, Jaun B. A nickel- alkyl bond in an inactivated state of the enzyme catalyzing methane formation. Angewandte Chemie-International Ed. 2006; 45:3602–3607. 31. Kern DI, Goenrich M, Jaun B, Thauer RK, Harmer J, Hinderberger D. Two sub-states of the red2 state of methyl-coenzyme M reductase revealed by high-field EPR spectroscopy. J Biol Inorg Chem. 2007; 12:1097–1105. [PubMed: 17690920] 32. Becker DF, Ragsdale SW. Activation of methyl-SCoM reductase to high specific activity after treatment of whole cells with sodium sulfide. Biochemistry. 1998; 37:2639–2647. [PubMed: 9485414] 33. Grabarse W, Mahlert F, Duin EC, Goubeaud M, Shima S, Thauer RK, Lamzin V, Ermler U. On the mechanism of biological methane formation: structural evidence for conformational changes in methyl-coenzyme M reductase upon substrate binding. J Mol Biol. 2001; 309:315–330. [PubMed: 11491299] 34. Grabarse W, Mahlert F, Shima S, Thauer RK, Ermler U. Comparison of three methyl-coenzyme M reductases from phylogenetically distant organisms: unusual amino acid modification, conservation and adaptation. J Mol Biol. 2000; 303:329–344. [PubMed: 11023796] 35. Horng YC, Becker DF, Ragsdale SW. Mechanistic studies of methane biogenesis by methyl- coenzyme M reductase: evidence that coenzyme B participates in cleaving the C-S bond of methyl-coenzyme M. Biochemistry. 2001; 40:12875–12885. [PubMed: 11669624] Cedervall et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 36. Berkessel A. Methyl-Coenzyme-M Reductase - Model Studies on Pentadentate Nickel-Complexes and a Hypothetical Mechanism. Bioorg Chem. 1991; 19:101–115. 37. Jaun B. Coenzyme-F430 from Methanogenic Bacteria - Oxidation of F430 Pentamethyl Ester to the Ni(Iii) Form. Helvetica Chimica Acta. 1990; 73:2209–2217. 38. Signor L, Knuppe C, Hug R, Schweizer B, Pfaltz A, Jaun B. Methane formation by reaction of a methyl thioether with a photo-excited nickel thiolate - A process mimicking methanogenesis in archaea. Chemistry-a European Journal. 2000; 6:3508–3516. 39. Chen SL, Pelmenschikov V, Blomberg MR, Siegbahn PE. Is there a Ni-methyl intermediate in the mechanism of methyl-coenzyme M reductase? J Am Chem Soc. 2009; 131:9912–9913. [PubMed: 19569621] 40. Pelmenschikov V, Blomberg MRA, Siegbahn PEM, Crabtree RH. A mechanism from quantum chemical studies for methane formation in methanogenesis. J Am Chem Soc. 2002; 124:4039– 4049. [PubMed: 11942842] 41. Pelmenschikov V, Siegbahn PE. Catalysis by methyl-coenzyme M reductase: a theoretical study for heterodisulfide product formation. J Biol Inorg Chem. 2003; 8:653–662. [PubMed: 12728361] 42. Duin EC, McKee ML. A new mechanism for methane production from methyl-coenzyme M reductase as derived from density functional calculations. J Phys Chem. 2008; B 112:2466–2482. 43. Bobik TA, Wolfe RS. Physiological importance of the heterodisulfide of coenzyme M and 7- mercaptoheptanoylthreonine phosphate in the reduction of carbon dioxide to methane in Methanobacterium. Proc Natl Acad Sci U S A. 1988; 85:60–63. [PubMed: 3124103] 44. Goenrich M, Duin EC, Mahlert F, Thauer RK. Temperature dependence of methyl-coenzyme M reductase activity and of the formation of the methyl-coenzyme M reductase red2 state induced by coenzyme B. J Biol Inorg Chem. 2005; 10:333–342. [PubMed: 15846525] 45. Kunz RC, Dey M, Ragsdale SW. Characterization of the Thioether Product Formed from the Thiolytic Cleavage of the Alkyl-Nickel Bond in Methyl-Coenzyme M Reductase. Biochemistry. 2008; 47:2661–2667. [PubMed: 18220418] 46. Noll KM, Donnelly MI, Wolfe RS. Synthesis of 7-mercaptoheptanoylthreonine phosphate and its activity in the methylcoenzyme M methylreductase system. J Biol Chem. 1987; 262:513–515. [PubMed: 3100513] 47. Olson KD, McMahon CW, Wolfe RS. Photoactivation of the 2-(methylthio)ethanesulfonic acid reductase from Methanobacterium. Proc Natl Acad Sci U S A. 1991; 88:4099–4103. [PubMed: 1903534] 48. Zehnder AJ, Wuhrmann K. Titanium (III) citrate as a nontoxic oxidation-reduction buffering system for the culture of obligate anaerobes. Science. 1976; 194:1165–1166. [PubMed: 793008] 49. Gunsalus RP, Romesser JA, Wolfe RS. Preparation of coenzyme M analogues and their activity in the methyl coenzyme M reductase system of Methanobacterium thermoautotrophicum. Biochemistry. 1978; 17:2374–2377. [PubMed: 98178] 50. Otwinowski Z, Minor W. Processing of X-ray Diffraction Data Collected in Oscillation Mode. Methods in Enzymology: Macromolecular Crystallography, part A. 1997; 276:307–326. 51. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr. 1994; 50:760–763. [PubMed: 15299374] 52. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] 53. Ebner S, Jaun B, Goenrich M, Thauer RK, Harmer J. Binding of coenzyme B induces a major conformational change in the active site of methyl-coenzyme M reductase. J Am Chem Soc. 2010; 132:567–575. [PubMed: 20014831] 54. Goenrich M, Mahlert F, Duin EC, Bauer C, Jaun B, Thauer RK. Probing the reactivity of Ni in the active site of methyl-coenzyme M reductase with substrate analogues. J Biol Inorg Chem. 2004; 9:691–705. [PubMed: 15365904] Cedervall et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. The active site and substrate channel in the MCRox1-silent crystal structure (PDB code 1hbn) (9). Coenzyme F430, CoBSH and HSCoM are drawn as stick colored by atom (carbon: dark grey). The nickel is displayed as a green sphere, water as a red sphere. Interactions are drawn as dashed lines, and the corresponding distance is indicted in Angstroms (Å). The path of the substrate channel was defined in the absence of F430, CoBSH, HSCoM and water, with the surface closest to the viewer cut away. The figure was generated using PyMOL (http://www.pymol.org). Cedervall et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Drawing of CoBSH analogues; (A) N-5-mercaptopentanoylthreonine phosphate (CoB5SH); (B) N-6-mercaptohexanoylthreonine phosphate (CoB6SH); (C) N-8- mercaptooctanoylthreonine phosphate (CoB8SH); (D) N-9-mercaptononanoylthreonine phosphate (CoB9SH). Cedervall et al. Page 16 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. The active sites and substrate channels of the MCR crystal structures.; (A) MCRCoB5SH; (B) MCRCoB6SH; (C) MCRHSCoM; (D) MCRCoB8SH; (E) MCRCoB9SH. 2Fo-Fc electron density map around the CoBXSH analogues, waters in the CoBSH binding part of the channel and the acetate ion (contoured at 1σ) is shown as a blue mesh. The protein is drawn as cartoon. CoBXSH and acetate are drawn as stick and colored by atom (carbon: CoB5SH orange; CoB6SH pale yellow; CoB8SH light blue; CoB9SH magenta; acetate white). Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere, and waters as red spheres. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 17 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. Overlay of CoBSH (from PDB code 1hbn) and the different CoBSH analogues. CoBXSH are drawn as stick with the thiol represented by a sphere, and colored CoB5SH orange; CoB6SH pale yellow; CoBSH light green; CoB8SH light blue; CoB9SH magenta. The protein is drawn as cartoon. Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 18 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Hydrogen bonding diagram for the water structure modeled in MCRHSCoM. The water molecules are named as in Figure 3C (W1–W9); WA, WB and WC are water molecules that are present in all structures (i.e. in concert with the substrate CoBSH and the CoBSH analogues). Interactions between surrounding residues and the water molecules are drawn as dashed lines, and the corresponding distance is indicated in Angstroms (Å). Cedervall et al. Page 19 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. Stereo image of the annulus of aromatic amino acids distal of coenzyme F430. The protein is drawn as cartoon with the side-chains of the aromatic residues drawn as white stick. CoBSH, (from PDB code 1hbn (9)), CoB8SH and CoB9SH are drawn as stick with the thiols represented by spheres, and colored CoBSH light green; CoB8SH light blue; CoB9SH magenta. Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 20 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. Reaction catalyzed by methyl-coenzyme M reductase Cedervall et al. Page 21 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 2. Two of the proposed catalytic mechanisms for methyl-coenzyme M reductase; (A) mechanism I; (B) mechanism II. Cedervall et al. Page 22 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 23 Table 1 X-ray Data Collection, Processing and Refinement Statistics Data collection and processing statistics Name of data set MCRCoB5SH MCRCoB6SH MCRHSCoM MCRCoB8SH MCRCoB9SH Measured reflections 1969388 2427498 1440665 1160543 1425506 Unique reflections 553755 446253 405349 211803 401701 Resolution (Å) a 50.0–1.30 (1.35–1.30) 50.0–1.40 (1.45–1.40) 50.0–1.45 (1.50–1.45) 50.0–1.80 (1.86–1.80) 50.0–1.45 (1.50–1.45) Completeness (%) a 97.1 (78.1) 99.9 (100.0) 99.5 (99.7) 99.8 (100.0) 98.1 (95.4) R-sym (%) a,b 5.5 (32.9) 7.3 (44.7) 6.2 (44.0) 8.4 (47.7) 5.6 (42.5) I/σI a 22.3 (3.6) 20.4 (4.0) 20.2 (3.2) 21.8 (3.9) 24.3 (3.2) Space group P21 P21 P21 P21 P21 Refinement and model building statistics Resolution (Å) a 20.49–1.30 (1.33–1.30) 19.89–1.40 (1.44–1.40) 20.15–1.45 (1.49–1.45) 19.93–1.80 (1.84–1.80) 20.07–1.45 (1.48–1.45) No. of reflection in working set a 525817 (30239) 423854 (25833) 384868 (25791) 201128 (11193) 381474 (23611) No. of reflection in test set a 27777 (1576) 22348 (1331) 20362 (1319) 10625 (557) 20163 (1210) R-work (%) c 14.32 13.04 13.47 14.95 13.58 R-free (%) d 16.56 15.53 16.22 19.54 16.44 ESU (Å) R-work/R-free 0.044/0.046 0.049/0.051 0.056/0.059 0.121/0.119 0.057/0.060 No. protein atoms 20087 19960 20265 19750 20036 No. coenzyme atoms 218 220 180 224 272 No. ligand atoms 37 62 52 26 49 No. water molecules 2443 2352 2516 1893 2432 RMS bond lengths (Å) 0.033 0.033 0.032 0.028 0.032 bond angles (deg.) 2.693 2.625 2.468 2.059 2.549 Ramachandran plot (%) favored 97.8 97.5 97.6 97.2 97.7 allowed 2.1 2.4 2.3 2.7 2.1 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 24 disallowed 0.1 0.1 0.1 0.1 0.1 Average B-factor (Å2) protein 12.42 13.35 12.12 17.22 12.73 coenzymes 8.20 9.24 7.25 11.24 8.27 ligands 31.95 35.48 28.29 33.76 32.92 waters 22.95 24.89 23.85 26.79 24.09 over all 13.54 14.57 13.40 18.02 13.93 Occupancy of HSCoM per active site (%)e 90/90 50/50 100/100 90/90 90/85 Occupancy of CoBSH per active site (%) e 50/50 50/50 30/30 50/50 40/40 CoBSH analogue, occupancy per active site (%) e CoB5SH, 50/50 CoB6SH, 50/50 CoB8SH, 50/50 CoB9SH, 60/60 Other molecule, occupancy per active site (%) e Acetate, 70/70 aValues in brackets correspond to the highest resolution shell. bR-sym = ΣhklΣNj=1|Ihkl−Ihkl(j)|/ΣhklN*Ihkl, sum over all reflections and all observations N, with Ihkl(j) intensity of the jth observation of reflection hkl and Ihkl mean intensity of the reflection hkl. cR-work = Σ||Fo| − |Fc||/Σ|Fo|, where |Fo| = observed structure factor amplitude and |Fc| = calculated structure factor amplitude. dR-free, R-factor based on 5% of the data excluded from refinement. eOccupancy of model in each of the two crystallographically independent active sites in the ASU Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 25 Table 2 Distances from analogue thiols. CoBXS - SCoM distance (Å) CoBXS - Ni distance (Å) CoB5SH 7.11/7.11a 9.30/9.30 CoB6SH 6.26/6.26 8.70/8.70 CoB7SH (substrate) b 6.37/6.39 8.73/8.77 CoB8SH 3.75/3.78 6.16/6.17 CoB9SH 3.71/3.68 5.96/5.91 aDistances in the two crystallographically independent active sites in the ASU bDistances in the 1.16 Å resolution MCRox1-silent structure (PDB code 1hbn)(33) Biochemistry. Author manuscript; available in PMC 2011 September 7.
3M2U
Structural Insight into Methyl-Coenzyme M Reductase Chemistry using Coenzyme B Analogues
Structural Insight into Methyl-Coenzyme M Reductase Chemistry using Coenzyme B Analogues,†,‡ Peder E. Cedervall§,#, Mishtu Dey||,#,⊥, Arwen R. Pearson§,⊥, Stephen W. Ragsdale||, and Carrie M. Wilmot*,||,§ § Department of Biochemistry, Molecular Biology and Biophysics, University of Minnesota, Minneapolis, Minnesota 55455 || Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109 Abstract Methyl-coenzyme M reductase (MCR) catalyzes the final and rate-limiting step in methane biogenesis; the reduction of methyl-coenzyme M (methyl-SCoM) by coenzyme B (CoBSH) to methane and a heterodisulfide (CoBS-SCoM). Crystallographic studies show that the active site is deeply buried within the enzyme, and contains a highly reduced nickel-tetrapyrrole, coenzyme F430. Methyl-SCoM must enter the active site prior to CoBSH, as species derived from analogues of methyl-SCoM are always observed bound to the F430 nickel in the deepest part of the 30 Å long substrate channel that leads from the protein surface to the active site. The seven-carbon mercaptoalkanoyl chain of CoBSH binds within a 16 Å predominantly hydrophobic part of the channel close to F430, with the CoBSH thiolate lying closest to the nickel at a distance of 8.8 Å. It has previously been suggested that binding of CoBSH initiates catalysis by inducing a conformational change that moves methyl-SCoM closer to the nickel promoting cleavage of the C- S bond of methyl-SCoM. In order to better understand the structural role of CoBSH early in the MCR mechanism, we have determined crystal structures of MCR in complex with four different CoBSH analogues; pentanoyl-, hexanoyl-, octanoyl- and nonanoyl- derivatives of CoBSH (CoB5SH, CoB6SH, CoB8SH and CoB9SH respectively). The data presented here reveal that the shorter CoB5SH mercaptoalkanoyl chain overlays with that of CoBSH, but terminates two units short of the CoBSH thiolate position. In contrast, the mercaptoalkanoyl chain of CoB6SH adopts a different conformation, such that its thiolate is coincident with the position of the CoBSH thiolate. †This work was supported by a Department of Energy grant DE-FG02-08ER15931 to S.W.R and supplement to C.M.W. and a Minnesota Partnership for Biotechnology and Medical Genomics grant SPAP-05-0013-P-FY06. ‡Co-ordinates and structure factors have been deposited in the Protein Data Bank as entries 3m1v (MCRHSCoM); 3m2r (MCRCoB5SH); 3m2u (MCRCoB6SH); 3m2v (MCRCoB8SH); 3m30 (MCRCoB9SH). *Address correspondence to: Carrie M. Wilmot, Tel: (612) 624-2406, wilmo004@umn.edu. ⊥Current addresses: Mishtu Dey, Department of Chemistry, Massachusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, Massachusetts 02139; Arwen Pearson, Astbury Centre for Structural Molecular Biology, Institute for Molecular and Cellular Biology, Astbury Building. Leeds, LS2 9JT, U.K. #These authors contributed equally to this work. Supporting Informational material is available free of charge via the Internet at http://pubs.acs.org and contains the following: MATERIAL AND METHODS for single crystal UV-visible microspectrophotometry, X-ray photoreduction experiment and X-ray crystallography of the MCR-heterodisulfide product complex (MCRCoBSH + methyl-SCoM); RESULTS AND DISCUSSION for redox changes and MCRCoBSH + methyl-SCoM crystal structure; Table S1, X-ray Data Collection, Processing and Refinement Statistics for MCRCoBSH + methyl-SCoM; Figure S1, use of Fo-Fc electron density in modelling MCRCoB8SH; Figure S2, illustration of electron density quality of MCRCoB5SH; Figure S3, solution and single crystal UV-visible spectra; Figure S4, modeling of CoB6SH and CoBSH into the electron density of MCRCoB6SH; Figure S5, partially occupied HSCoM in MCRCoB6SH; Figure S6, the active site and substrate channel of MCRCoBSH + methyl-SCoM; Figure S7, alternative conformation of Valα482 in MCRHSCoM; Figure S8, the two conformations of the gly-rich loop in MCRHSCoM; Figure S9, propagation of conformational changes in MCRHSCoM; Figure S10, the two conformations of CoB9SH in MCRCoB9SH; Figure S11, EPR spectra of MCRred1 sample; Scheme S1, scheme of the characterized forms of MCR. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 September 7. Published in final edited form as: Biochemistry. 2010 September 7; 49(35): 7683–7693. doi:10.1021/bi100458d. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript This is consistent with the observation that CoB6SH is a slow substrate. A labile water in the substrate channel was found to be a sensitive indicator for the presence of CoBSH and HSCoM. The longer CoB8SH and CoB9SH analogues can be accommodated in the active site through exclusion of this water. These analogues react with Ni(III)-methyl; a proposed MCR catalytic intermediate of methanogenesis. The CoB8SH thiolate is 2.6 Å closer to the nickel than that of CoBSH, but the additional carbon of CoB9SH only decreases the nickel thiolate distance a further 0.3 Å. Although the analogues did not induce any structural changes in the substrate channel, the thiolates appeared to preferentially bind at two distinct positions in the channel; one being the previously observed CoBSH thiolate position, and the other being at a hydrophobic annulus of residues that lines the channel proximal to the nickel. INTRODUCTION Methanogenic archaea are organisms that under strict anaerobic conditions derive energy by reducing compounds such as carbon dioxide, methylamine, acetate and methanol, to methane (1, 2). The global production of methane by these organisms is estimated at one billion tons annually. Microbially produced methane is not only a potential source of renewable energy but also a potent greenhouse gas, and as such study of this process has environmental ramifications. In these microorganisms, methyl-coenzyme M reductase (MCR)1 is the enzyme that catalyzes the final step in methanogenesis, in which the substrates methyl-coenzyme M (methyl-SCoM, 2-(methylthio)ethanesulfonate) and coenzyme B (CoBSH, N-7-mercaptoheptanoylthreonine phosphate) are converted to methane and a heterodisulfide (CoBS-SCoM) (Scheme 1) (3). MCR is a 300 kDa protein with six subunits arranged in a α2β2γ2 oligomer (4). The known crystal structures show that MCR has two active sites approximately 50 Å apart that are deeply buried within the enzyme (5). The active site pocket is comprised of residues from subunits α, α′, β and γ, with a 30 Å long substrate channel leading to the enzyme surface (Figure 1). At the heart of the active site pocket is coenzyme F430, which is a highly reduced nickel-containing tetrapyrrole (6–8). Currently sixteen distinct enzymatic and complexed states of MCR have been spectroscopically characterized (Supporting Information, Scheme S1) (9–31). In the resting active state of the enzyme, denoted MCRred1, the redox-active nickel of F430 is present in the Ni(I) state (9, 16, 32). MCR is extremely oxygen sensitive and upon oxygen exposure the enzyme enters an inactive Ni(II) state, denoted MCRred1-silent (6). In this state it cannot be converted back to the active Ni(I) form by any known reducing agent making this a challenging system to study. Additional complications involve the tight association of coenzymes to purified MCR that are not easily displaced as demonstrated by X-ray crystallographic and kinetic studies (5, 33–35). Despite the fact that MCR has been studied for decades, no true catalytic intermediate has been observed, and the actual mechanism remains elusive. Currently three general mechanistic schemes for the enzymatic reaction have been proposed, each of which posit different chemistry to initiate catalysis. Mechanism I involves Ni(I) acting as a nucleophile in an SN2-type reaction that generates Ni(III)-methyl as an intermediate (Scheme 2A) (35– 38). Mechanism II starts with methyl-SCoM undergoing homolytic cleavage at the Ni(I) to generate a methyl radical and a Ni(II)-SCoM species (Scheme 2B) (39–41). A more recently proposed mechanism III suggests protonation of coenzyme F430 promotes reductive cleavage of the methyl-SCoM thioether bond (42). 1Abbreviations: MCR, Methyl-coenzyme M reductase; methyl-SCoM, methyl-coenzyme M; CoBSH, coenzyme B; HSCoM, coenzyme M; CoBS-SCoM, heterodisulfide of coenzyme B and coenzyme M; APS, Advanced Photon Source; ASU, asymmetric unit; BPS, bromopropanesulfonate. Cedervall et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Due to the stringent requirement to exclude O2, the available MCR crystal structures are all in the inactive Ni(II) state. All but one contain CoBSH and HSCoM (demethylated methyl- SCoM, an inhibitor and substrate analogue) in the active site (PDB codes 1hbn, 1hbo, 1hbu, 1e6y, 1e6v) (5, 33, 34). Another crystal structure has bound heterodisulfide product, CoBS- SCoM (MCRsilent, PDB code 1hbm, Scheme 1 and Supporting Information, Scheme S1) (5, 33). All these structures reveal that both substrates access the active site through the same channel (Figure 1). The binding site of HSCoM (and presumably methyl-SCoM) is more deeply buried within the enzyme, and so it must enter prior to CoBSH for productive chemistry to occur. As binding of CoBSH in the absence of co-substrate would be inhibitory, it was suggested that a conformational change upon methyl-SCoM binding might lower the Kd for CoBSH, and thus promote an ordered mechanism. Furthermore it has been suggested that CoBSH binding induces a conformational change that brings the methyl- SCoM substrate into closer proximity to the nickel, and this promotes C-S bond cleavage. To investigate the proposed structural role of CoBSH in initiating catalysis, we have solved the X-ray crystal structures of MCR in complex with four different CoBSH analogues. CoBSH has a heptanoyl moiety linked to the thiol group, and the analogues are pentanoyl-, hexanoyl-, octanoyl- or nonanoyl-containing derivatives of CoBSH (CoB5SH, CoB6SH, CoB8SH and CoB9SH respectively; Figure 2) (3, 35, 43–47). In addition, we present a structure in which the substrate channel predominantly lacks either CoBSH or heterodisulfide product. MATERIALS AND METHODS Materials The organism Methanothemobacter marburgensis (catalog OCM82) was obtained from the Oregon Collection of Methanogens (Portland, OR). All buffers and media reagents were obtained from Sigma-Aldrich (St. Louis, MO). The gases N2 (99.98%), H2/CO2 (80%/20%), and ultra high purity H2 (99.999%) were obtained from Cryogenic Gases (Grand Rapids, MI). Ti(III) citrate solutions were prepared from a stock solution of 246 mM Ti(III) citrate, which was synthesized by adding sodium citrate to Ti(III) trichloride (30 wt % solution in 2 N hydrochloric acid) (Acros Organics, Morris Plains, NJ) under anaerobic conditions and adjusting pH to 7.0 with sodium bicarbonate (48). The concentration of Ti(III) citrate was determined by titrating against a solution of methyl viologen. Synthesis of methyl-SCoM, CoB5SH, CoB6SH, CoBSH, CoB8SH, and CoB9SH Methyl-SCoM was prepared from HSCoM and methyl iodide (49). The homodisulfides, CoB5S-SCoB5, CoB6S-SCoB6, CoBS-SCoB, CoB8S-SCoB8, CoB9S-SCoB9 were prepared as described from 5-bromovaleric acid, 6-bromohexanoic acid (Sigma-Aldrich, St. Louis, MO), 7-bromoheptanoic acid (Karl Industries, Aurora, OH), 8-bromooctanoic acid, and 9- bromononanoic acid (Matrix Scientific, Columbia, SC), respectively (43, 46). The free thiol forms of CoB5SH, CoB6SH, CoBSH, CoB8SH, and CoB9SH were generated by the reduction of the homodisulfides as previously described (45). The purity of the CoBSH analogues was determined by 1H NMR spectroscopy. All compounds synthesized were stored in a Vacuum Atmospheres chamber maintained at an oxygen level below 1 ppm, as monitored continually with an oxygen analyzer (model 317, Teledyne Analytical Instruments, City of Industry, CA) until use. M. marburgensis Growth and MCRred1 Purification Buffer preparations and all manipulations were performed under strict anaerobic conditions in a Vacuum Atmospheres chamber maintained at an oxygen level below 1 ppm, as monitored continually with an oxygen analyzer (model 317, Teledyne Analytical Instruments, City of Industry, CA). MCRred1 was isolated from M. marburgensis cultured on Cedervall et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript H2/CO2 (80%/20%) at 65 °C in a 14-L fermentor (New Brunswick Scientific Co., Inc. New Brunswick, NJ). Culture media were prepared as previously described (20, 24). MCRred1 was generated in vivo and purified as described previously (20). The purification procedure routinely generates 65–75% MCRred1 as determined by UV-visible and EPR spectroscopy. Spectroscopy of MCR UV-visible spectra of the Ni(I)-containing MCRred1 were recorded in the anaerobic chamber using a spectrophotometer (Model USB4000-UV-VIS, Ocean Optics, Dunedin, FL). EPR spectra were recorded on a Bruker EMX spectrometer (Bruker Biospin Corp., Billerica, MA), equipped with an Oxford ITC4 temperature controller, a Hewlett-Packard Model 5340 automatic frequency counter and Bruker gaussmeter. The EPR spectroscopic parameters included: temperature, 70 K; microwave power, 10 mW; microwave frequency, 9.43 GHz; receiver gain, 2 × 104; modulation amplitude, 10.0 G; modulation frequency, 100 kHz. Double integrations of the EPR spectra were performed and referenced to a 1 mM copper perchlorate standard. The NMR data were acquired at 298 K on a Bruker Avance DRX 500 MHz instrument equipped with a TXI cryoprobe. Preparation of MCRred1 for Crystallization All crystallization experiments were performed in the anaerobic chamber in which MCR was purified unless otherwise noted. MCRred1 was prepared in 50 mM Tris, pH 7.6 and excess HSCoM was removed by buffer exchange using an Amicon Ultra centrifuge filter with a 50 kDa cut-off membrane (Millipore). Typically, 2–3 ml of MCRred1 was exchanged with 10–15 ml of 50 mM Tris, pH 7.6. The enzyme was concentrated to 500–600 μl, and this process was repeated three times. The fraction of MCRred1 in the purified MCR sample was calculated from the UV-visible spectrum using extinction coefficients of 27.0 mM−1cm−1 at 385 nm for Ni(I)-MCRred1, and 9.15 mM−1cm−1 at 420 nm for Ni(II)- MCRred1-silent (20). The amount of MCRred1 in samples used for crystallization was determined to be ~80% and the concentration of total enzyme used was in the range of about 120–150 μM (~32–40 mg/ml). All crystallization experiments were performed anaerobically by incubating 2.0 μl of enzyme solution in 50 mM Tris, pH 7.5 and 2.0 μl of reservoir solution (100 mM Hepes-Na, pH 7.3/7.5/8.0; 150 mM magnesium acetate (Mg(CH3COO)2), and 20/22% (w/v) PEG 400) in a sitting drop over 1 ml reservoir solution at 9 ºC. Triangular and rectangular prismatic crystals with a bright yellowish-green color confirmed the presence of nickel coenzyme F430. The crystals grew to a size of approximately 100–200 μm in 4–5 days. CoBSH-depleted crystals were obtained by incubating 2 μl of a reaction mixture containing 139 μM MCRred1 and 13 mM HSCoM with 2 μl of reservoir solution (100 mM Hepes-Na pH 7.5, 150 mM magnesium acetate (Mg(CH3COO)2), 22% PEG 400). Crystals of MCR complexed with the CoBSH analogues were grown by co-crystallization. The CoB5SH co-crystals were obtained by incubating 2μl enzyme solution containing 124 μM MCRred1, 10 mM methyl-SCoM and 1 mM CoB5SH with 2μl of reservoir solution (100 mM Hepes-Na, pH 7.5, 150 mM (Mg(CH3COO)2), 22 % PEG 400). The crystals with bound CoB6SH and CoB9SH were obtained by co-crystallization of 1 mM of analogue with 142 μM MCRred1 and equilibrated with 2μl of reservoir solution (100 mM Hepes-Na, pH 7.5, 150 mM Mg(CH3COO)2, 20 % PEG 400 for CoB6SH and 100 mM Hepes-Na, pH 7.3, 150 mM Mg(CH3COO)2), 22 % PEG 400 for CoB9SH). Crystals were cryoprotected in reservoir solution containing 25 % (v/v) PEG 400 by soaking for 2–5 minutes before cryocooling in liquid nitrogen in the anaerobic chamber. Crystals of CoB8SH bound to MCR were obtained by incubating 2 μl of a mixture of 119 μM MCRred1 and 1 mM CoB8SH with 2 μl of reservoir solution (100 mM Hepes-Na, pH 7.3, 150 mM Mg(CH3COO)2), 20 % PEG 400). Before cryoprotection, the crystals were soaked for 5–10 minutes in a 100 mM solution of methyl iodide2. The methyl iodide solution used for soaking was prepared by adding a concentrated stock of methanolic solution of methyl iodide to the reservoir Cedervall et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript solution. Soaked crystals were quickly cryoprotected as described above and cryocooled in the anaerobic chamber. X-ray Diffraction Data Collection, Processing and Refinement X-ray diffraction data were collected at 100 K on a ADSC Quantum-315 detector at the APS Beamline 14-BM-C (BioCARS). The wavelength of X-rays was 0.979 Å. Data were processed using HKL2000 (50). As in the previous X-ray crystallographic studies, the crystals belong to the monoclinic space group P21 (a = 82 Å, b = 118 Å, c = 122 Å, β = 92°), with one MCR molecule (two active sites) per asymmetric unit (5, 33). For refinement, REFMAC in the Collaborative Computational Project Number 4 (CCP4) program suite was used (51). A random sample of 5 % of the data across all resolution shells was chosen to check refinement progress through calculation of an Rfree. The same reflections were used to calculate Rfree for all structures, thus preventing bias due to high structural identity. The remaining reflections were used in refinement (Rwork). Model building was done using the Crystallographic Object-Oriented Toolkit (COOT) (52). The diffraction data and their models are designated as MCRCoBXSH, where X is the number of carbons in the alkanoyl portion of the analogue. Library files in CCP4 for F430 and CoBSH were incorrect, and these were modified in Monomer Library Sketcher in the CCP4 program suite by comparison with schematic drawings from Grabarse et al. (33). Co-ordinate and CCP4 library files for the different CoBSH analogues were created in Monomer Library Sketcher. The general model building and refinement strategy for all structures was as follows. It was clear from the electron density in the substrate channel and at the active site that mixtures of species were present in all datasets. These could be visualized with Fo-Fc and Fo-Fo difference electron density maps (Supporting Information, Figure S1). The known positions of CoBSH and HSCoM from the published Ni(II)-MCR crystal structures (PDB codes 1hbn, 1hbo, 1hbu (33)) were used as guides to determine which species could be present in each dataset, and these were then simultaneously modeled into the electron density. By alteration of their relative occupancies (in 10% increments) followed by refinement, the ratio of occupancy between different species was determined using the assumption that the average B-factors for all molecular species bound should be similar to that of F430 and adjacent well-ordered protein atoms within the active site and substrate channel. The combinations of modeled ligands were constantly reassessed throughout refinement based on the remaining difference electron density. This included test refinements of different ligand combinations during the latter stages, thus using the optimized phases to check whether a different combination of ligands could also explain the electron density. Sensible chemical structures and interactions, along with keeping the combined occupancies of sterically mutually exclusive species ≤ 100%, were maintained throughout refinement. The model was finally accepted when the difference electron density map was minimal and the B-factors for the models converged. In practice the first structure refined was that of MCRCoB5SH. Initial phases were generated by difference Fourier using a previously determined crystal structure (PDB code 1mro (5)) but with all non-bonded molecules, including water, removed from the model except F430. Initial rigid body refinement followed by restrained refinement of MCRCoB5SH reduced the Rwork to 26.5%. After model building and subsequent rounds of restrained refinement the Rwork was 14.3% (Rfree 16.6%). Of the five structures only the CoB5SH analogue is completely coincident with CoBSH, and so particular care had to be used in teasing apart the ratios of the two species in modeling the MCRCoB5SH electron density. This was done by 2Methyl iodide was added with the intention of creating a Ni(III)-methyl species, which CoB8SH stabilizes. This was not achieved, but the diffraction quality of this crystal was significantly better than crystals co-crystallized with CoB8SH alone, and so has been included in this study. Cedervall et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript initially refining 100% CoBSH in the substrate channel. Positive Fo-Fc difference density located at the carbon where the shorter CoB5SH thiol might be expected to be indicated the presence of a more electron-rich species than carbon, which is consistent with the presence of the CoB5SH sulfur. The refinement converged at a model containing a 50:50 mix of CoBSH and the analogue. However, positive Fo-Fc difference density was still present at the position of the CoBSH thiol, therefore a water molecule was added to the CoB5SH model at 50% occupancy and upon refinement this accounted for the electron density. An illustration of the electron density quality from this structure is shown in Supporting Information, Figure S2. An HSCoM-, CoBSH- and CoBSH analogue-free version of the refined MCRCoB5SH structure was used as the starting model to generate initial phases for the four other structures. After the initial round of restrained refinement the Rwork for these structures were reduced to 14.5–15.6 %. RESULTS AND DISCUSSION Crystal Structures of MCR Five crystal structures were determined, four of which are in complex with CoBSH analogues differing in the number of carbons atoms in the alkanoyl portion of the molecule. CoBSH is an N-7-mercaptoheptanoyl-containing molecule, whereas the four CoBSH analogues contain N-5-mercaptopentanoyl-, N-6-mercaptohexanoyl-, N-8-mercaptooctanoyl- or N-9-mercaptononanoyl-moieties (Figure 2). The corresponding crystal structures are designated as MCRCoBXSH, where X is the number of carbons in the alkanoyl portion of the analogue. The other crystal structure is of MCRred1c-silent (MCR in the Ni(II) state in complex with HSCoM, designated here as MCRHSCoM) that is CoBSH-depleted. The datasets have resolutions in the range from 1.35 – 1.8 Å. Although the crystallizations were set up with the MCR solution predominantly in the Ni(I)-MCRred1 state, by the time X-ray diffraction data were collected they had been oxidized to the Ni(II)-MCRred1-silent state (Supporting Information). Following data collection there was no evidence for photoreduction of the Ni(II) back to Ni(I) in any of the crystals, as assessed by single crystal UV-visible microspectrophotometry (Supporting Information and Figure S3). Attempts to photoreduce the crystals using different wavelengths and temperatures were unsuccessful (Supporting Information). Overall, the resulting structures are very similar to each other and to the previously published structures of MCR, with differences mainly localized to the active site and substrate channel. The two active sites in the ASU were refined independently. Unless otherwise stated there was no difference between them. All five datasets contain a mixture of species bound to the enzyme. There is always a background of CoBSH and HSCoM, which co-purify with MCR and cannot be fully removed by extensive buffer exchange or by the addition of a CoBSH analogue. HSCoM is added during purification of MCR, as it stabilizes the resting active Ni(I) state (unpublished data), and this leads to HSCoM occupancies between 50–100% amongst the structures (Table 1). In contrast CoBSH, which is not added during purification, has occupancies ranging from 30–50%. As these confounding species have all been described at high occupancy in other crystallographic studies, the structural data of interest could be isolated (5, 33). In each case, the additional electron density could be explained by inclusion of the appropriate CoBXSH model used in that experiment at 50% or higher occupancy. The resulting models, along with 2Fo-Fc electron density, are shown in Figure 3. The Rwork for the final structures range from 13.0 to 15.0 % (Rfree 15.5 to 19.5 %). The X-ray data collection, processing, refinement and model building statistics are given in Table 1. Cedervall et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Analogues shorter than CoBSH; CoB5SH and CoB6SH CoB5SH is two methylene groups shorter than CoBSH, the MCR substrate. The MCRCoB5SH structure is to 1.35 Å resolution. As expected the pentanoyl chain follows the path of the CoBSH heptanoyl carbons down the substrate channel, and thus its thiol is positioned in the same place as the second carbon preceding the CoBSH thiol (Figure 3A and 4). There are no published MCR kinetic studies using CoB5SH, but as it binds in the substrate channel, it is likely to be an inhibitor. CoB6SH is one methylene shorter than CoBSH, and is a slow substrate of MCR. In this case the 1.4 Å resolution electron density of MCRCoB6SH indicates that the analogue unexpectedly binds in the substrate channel such that its thiol is virtually in the same position as the thiol of CoBSH (Figure 3B and 4). The hexanoyl chain is oriented so that it takes a shorter route down the substrate channel between carbons 2 and 5 (with the carbonyl carbon labeled as carbon 1) than CoBSH (Figure 4, Supporting Information, Figure S4). This short-cut is not seen in any of the other CoBXSH complex crystal structures, but presumably arises because this CoB6SH binding conformer is energetically more favorable, although it is not clear from the structure why this might be the case. CoB6SH binds very tightly to MCR, with an apparent Ki value of 0.1 μM (3). Water structure in the absence of HSCoM The electron density for the MCRCoB6SH crystal structure only supported the modeling of 50 % bound HSCoM. In the fraction of MCR molecules where HSCoM is absent, the HSCoM binding site is occupied by a network of four water molecules (Supporting Information, Figure S5). Two waters are positioned close to the absent sulfonate oxygen positions of HSCoM. Based on the presence of positive difference electron density, a third water was modeled ligated to the Ni, and refined to a distance of ~2 Å (2.0 Å and 2.1 Å in the two active sites of the ASU) with no distance restraint imposed between the Ni and water. This water is in a similar position as the Ni coordinating sulfonate oxygen of the heterodisulfide product in MCRsilent (Supporting Information, Figure S6 and PDB codes 3m32, 1hbm) (5, 33). The fourth water was in the vicinity of the expected position of a bridging water (W1) seen in other structures (Figure 1, 3A and 3C). Water structure in the absence of CoBSH The 1.45 Å resolution electron density obtained for MCRHSCoM indicates that the substrate channel contains only 30 % CoBSH. Nine ordered waters (W1–W9), along with an acetate ion from the crystallization solution occupy the channel, with the acetate positioned where the phosphothreonine linkage of CoBSH would be (Figure 3C). Presumably 1–2 further waters would replace the acetate under physiological conditions. Other than W3 and W7, the waters form hydrogen bonds with protein (Figure 5). One water (W2) occupies the same site as the CoBSH thiol. Presumably due to the loss of favorable interactions that exist when CoBSH is present, the hydrophobic side-chain of Valα482 adopts a second conformation modeled at 60 % occupancy (Supporting Information, Figure S7). Position of the “bridging” water, W1 The equivalent of W1 has previously been observed in MCRred1-silent and MCRox1-silent crystal structures where, in the presence of CoBSH and HSCoM, it is sited equidistant (3.2 Å) between the two coenzyme thiols (PDB codes: 1hbn, 1hbo, 1hbu) and thus been termed the “bridging water” (Figure 1) (5, 33). However, in the MCRHSCoM structure, due to the presence of the more polarized W2 water, W1 is displaced away from HSCoM to maximize the hydrogen bond interaction with W2 (2.9 Å to W2; 3.5 Å to HSCoM thiol, Figure 5). In the MCRCoB5SH structure that also contained W2, the electron density indicated that this Cedervall et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript repositioning of W1 towards W2 also occurred. In contrast, the MCRCoB6SH structure contained 100% thiol at the CoBSH position, but a partial occupancy of HSCoM (50%). In this case the electron density for W1 indicated it had moved towards the nickel to form an optimal hydrogen bond with a Ni-ligating water that was only present in the absence of HSCoM (3.7 Å to CoB(6)SH thiol; 3.0 Å to Ni-ligating water, Supporting Information, Figure S5). In all structures reported here, W1 (if present) appears to be a sensitive indicator of the relative electronegativity of the Ni-ligated atom to that occupying the position of the CoBSH thiol, and was a useful check in the crystallographic modeling and refinement process. Flexibility in the substrate channel: Alternative protein conformers The binding site of HSCoM (and presumably methyl-SCoM) is more deeply buried within the enzyme, and so it must enter prior to CoBSH for productive chemistry to occur. As binding of CoBSH in the absence of co-substrate would be inhibitory, it was suggested that a conformational change upon methyl-SCoM binding might lower the Kd for CoBSH, and thus promote an ordered mechanism. Compared to the 1hbn MCRox1-silent and 1hbu MCRred1-silent crystal structures, which both have full occupancy HSCoM, the lower occupancy of HSCoM in the 1hbo MCRred1-silent structure was associated with significantly greater flexibility within the channel, and the ability to model a second conformation of a Gly-rich amino acid stretch that formed part of the CoBSH channel. This suggested that methyl-SCoM binding might cause the channel to become more ordered, increasing the affinity of MCR for CoBSH by conformational restriction rather than a switch mechanism where the structure reorganizes from one well-defined conformer to another (33). In the MCRHSCoM data containing 30 % CoBSH and 100% HSCoM, the Fo-Fc difference electron density map at one of the two independent active sites in the ASU contained positive peaks that suggested the presence of an alternate conformation also involving this part of the polypeptide (Supporting Information, Figure S8). Using this as a guide, a similar second conformation involving seven contiguous amino acid residues of the same Gly-rich amino acid stretch (β366–372) could be modeled and refined at 20 % occupancy leaving no residual difference density. Parts of the α′ subunit (α′111–129 and α′237–242) that are in close proximity to this stretch of amino acids also exhibit second conformations, with the main-chain carbonyl of α′243 in van der Waals contact with the B ring of F430 tetrapyrrole (Supporting Information, Figure S9). Modeling these at 20% occupancy accounted for the weak positive Fo-Fc difference electron density peaks observed in these areas. The evidence of alternate conformers in these areas lends support to the proposal that increased flexibility in the substrate channel propagates through the protein (33). The MCRCoB6SH crystal structure contains 50 % CoB6SH, 50% CoBSH and 50% HSCoM. In this case there is no evidence of an alternate loop conformation in either active site of the ASU. However, as CoBSH and CoB6SH combined are at 100% occupancy, it is not surprising their favorable interactions with the substrate channel would reduce conformational disorder, despite the partial occupancy of HSCoM. Analogues longer than CoBSH; CoB8SH and CoB9SH Both analogues could be accommodated in the MCR substrate channel (Figures 3D and 3E). The electron density supported final models containing 50% CoB8SH for MCRCoB8SH (1.8 Å resolution) and 60 % CoB9SH for MCRCoB9SH (1.45 Å resolution). The phosphate head- groups are in identical positions to those of CoBSH, CoB5SH, CoB6SH (Figure 4) (5, 33). Both analogues follow the crystallographically observed chain path of bound CoBSH, with the extra atoms displacing the W1 water and placing the thiols closer to the nickel (Figure 6). CoB9SH does have a second conformer that deviates from the CoBSH path, but the thiol position for this conformer and the CoBSH-tracking conformer are identical (Figure 3E and Cedervall et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supporting Information, Figure S10). Interestingly, the thiol of CoB8SH is not coincident with the CoB9SH carbon that precedes the CoB9SH thiol (Figure 6). CoB8SH is an MCR inhibitor with an apparent Ki of 15 μM (3). CoB9SH has never been tested for inhibition of MCR-catalyzed methane formation, but it is reasonable to assume that it would be an inhibitor. CoBXSH thiol-to-nickel spatial relationship The position of CoBSH in previous crystal structures poses a conundrum (5, 33). In all the proposed catalytic mechanisms, CoBSH must interact with species generated at the nickel. Perplexingly, the sulfur of the CoBSH substrate is 8.8 Å from the Ni(II) in the MCRox1-silent and both MCRred1-silent crystal structures, and 6.4 Å from the thiol of the substrate analogue HSCoM (Figure 1). Modeling studies demonstrated that the addition of a methyl group to HSCoM did not bridge this gap (35, 45, 53). Therefore, a conformational change has been postulated that would enable CoBSH to penetrate deeper into the substrate channel, and thus approach closer to any nickelbound species. The heterodisulfide product in the MCRsilent crystal structure has the CoBSH portion in virtually the same place as in MCRox1-silent, giving no clue to possible structural changes that might occur to facilitate CoBSH reacting with nickel-associated intermediates (5, 33). Trigonometry suggests that if the alkanoyl chain of CoBSH or its analogue is in an extended conformation, each additional unit in the chain would lead to the thiol moving ~1.2 Å towards the Ni. Until this study there have been no crystal structures of CoBSH analogues in complex with MCR, so mechanistic studies using different chain length analogues of CoBSH assumed that shorter analogues would trace the observed path of CoBSH, and longer analogues would penetrate about ~1.2 Å deeper per additional chain unit into the channel. In the case of the shortest analogue CoB5SH, it does indeed follow the path of CoBSH, with the thiol of CoB5SH being 2.8 Å away from the thiol position of CoBSH. However, due to the conformation CoBSH adopts when bound in the substrate channel, the difference in the S-Ni distance is small; the CoB5SH thiol being only 0.5 Å farther from the Ni than CoBSH (8.8 Å for CoBSH vs. 9.3 Å for CoB5SH) (Table 2). This is due to the alkanoyl chain of CoBSH not being in an extended conformation from carbons 4 to 6 (carbon 1 is the carbonyl carbon). CoB6SH on the other hand adopts a conformation that places its thiol in virtually the same position as the thiol of CoBSH (Figure 4 and Table 2). This is consistent with CoB6SH being a substrate. However, the kcat is 1000-fold lower than for CoBSH (3, 35) although its Km value (180 μM, Dey & Ragsdale, in preparation) is similar to that of CoBSH (Km = 75 μM (3)). The reason for this may be that the shorter alkanoyl chain may not enable the analogue thiol to approach the nickel close enough for efficient catalysis, and thus explain why CoB6SH is such a poor substrate. In the case of the longer CoBXSH analogues, the sulfur of CoB8SH is 2.6 Å closer to the Ni ion of F430 than that of CoBSH, and 2.5 Å closer to the thiol of HSCoM (Figure 6 and Table 2). The CoB9SH molecule follows the path of CoBSH, and reaches only a little further into the substrate channel than CoB8SH, with the CoB9SH thiol positioned 2.9 Å closer to the Ni than the thiol of CoBSH (Figure 6 and Table 2). This is only 0.3 Å closer than the distance observed for the CoB8SH thiol, even though they are non-coincident. The distance to the thiol of HSCoM is 2.6 Å closer than that of the substrate, CoBSH; only 0.1 Å closer than the CoB8SH thiol. The two analogue thiols are above an annular hydrophobic aromatic environment created by Pheα330, Tyrα333, Pheα443, Pheβ361 and Tyrβ367 that lies between them and F430 (Figure 6). As a result, penetrating further into the channel may be energetically unfavorable, consistent with the small difference in relative distances between the CoB8SH/CoB9SH thiols and the HSCoM thiol/F430 nickel. The annulus is proposed to be catalytically important in positioning methyl-SCoM and stabilizing the methane product, Cedervall et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript and the tyrosines have been proposed to be proton donors associated with mechanism II (Scheme 2B) (5, 33). Thus, there appear to be three preferential distances for thiols (including that of HSCoM) within the MCR substrate channel; HSCoM at 2.4 Å, CoB6SH/CoBSH at 8.7 – 8.8 Å and CoB8SH/CoB9SH at 5.9 – 6.2 Å from the nickel of F430 (Table 2). Recent ENDOR, high-field continuous and pulse EPR work has identified changes in nickel co-ordination when CoBSH is added to MCRred1c (active Ni(I)-MCRred1 + HSCoM) (14, 15, 18, 31). This generates up to 50% MCRred2 which is comprised of two distinct nickel co- ordination geometries; an axial MCRred2a formally assigned as a Ni(III)-hydride, and a rhombic MCRred2r in which the thiol of HSCoM is a Ni(I) ligand (Supporting Information, Scheme S1). Formation of MCRred2 could also be induced by addition of CoBSH substrate analogues CoBS-CH3 and CoBS-CF3, which have a chain length one unit longer than substrate CoBSH (18, 53). The CoBS-CF3 enabled 19F-ENDOR studies to be performed, and demonstrated that following CoBS-CF3 addition the remaining MCRred1 species had Ni(I)-19F distances of 6.2 – 7.7 Å. This distance range agreed with a CoBS-CF3 model created using the CoBSH position observed in the MCRox1-silent crystal structure (53). However, in the MCRred2 species the Ni(I)-19F distances had shortened indicating a movement of ~ 2 Å towards the nickel (MCRred2a Ni(I)-19 F, 4.0 – 5.5 Å; MCRred2r Ni(I)-19F, 4.5 – 5.7 Å). In the case of the CoB8SH analogue (the closest equivalent to CoBS- CF3 used in this study) the thiol to Ni(II) distance lies between the distance ranges observed in the CoBS-CF3 studies, and so the fluorine(s) of the CoBS-CF3 in MCRred2 might penetrate a little further into the hydrophobic annulus in the MCRred2 species. As the alkanoyl chain of CoBSH is not fully extended it could easily undergo a similar conformation change to that observed in the MCRred2 state. Reaction of MCR Ni(III)-alkyl species’ with CoB8SH and CoB9SH The two longer CoBXSH analogues have been shown to undergo alkylation when reacted with MCRPS, a [Ni(III)-alkyl ↔ Ni(II)-alkylsulfonate radical] formed from reaction of Ni(I)-MCRred1 with bromopropanesulfonate (BPS) (Supporting Information, Scheme S1) (20, 23, 30, 45, 54). BPS is a substrate of MCRred1 in a reaction that involves a rapid CoBSH-independent nucleophilic attack by Ni(I) on BPS to displace bromide and generate MCRPS at a rate ~60-fold faster than generation of methane from CoBSH and methyl- HSCoM (20, 45). Certain thiols can eliminate the propylsulfonate to yield a thioether product and regenerate MCRred1, although at a rate 1000-fold slower than methane formation (45). Both CoB8SH and CoB9SH can react with MCRPS to regenerate MCRred1, but CoBSH cannot. The overall second-order rate constant for the reactivation of MCR by CoB8SH is 160 M−1s−1, whereas for CoB9SH the reaction is more sluggish (12 M−1s−1). CoB9SH might be expected to be closer to the proximal Ni ligand. It was therefore proposed that this caused steric interference and explained why CoB9SH was a poorer reactivator of MCR than CoB8SH. Our study has shown that the thiols of these two analogues are placed such that they are approximately the same distance (~3.7 Å) from the thiol of HSCoM ligated to the Ni atom (Table 2). The Ni(II)-HSCoM bond is 2.4 Å, whereas an Ni(III)-alkyl bond is expected to be ~2 Å (24, 33), thus indicating that a conformational change is required to effect the nucleophilic attack of the CoB8SH and CoB9SH thiols on an alkyl- bound species. It would thus appear that a conformational change, such as observed in MCRred2, is required for this chemistry also (53). A Ni(III)-alkyl species is akin to the first intermediate in mechanism I of MCR-catalyzed methane formation, Ni(III)-methyl (MCRMe, Supporting Information, Scheme S1, Scheme 2A) (11, 27). MCRMe has been shown to be capable of generating MCRred1 and methyl- SCoM upon addition of HSCoM (which is the reverse of mechanism I, step 1, Scheme 2A); Cedervall et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript similar chemistry to the observed formation of a thioether product from the Ni(III)-alkyl. Further addition of CoBSH following HSCoM treatment of MCRMe led to methane and heterodisulfide formation, the natural products of methanogenesis. Although this lends credence to mechanism I, it should be noted that like MCRPS, MCRMe in these experiments was generated artificially. However, the MCRCoB9SH crystal structure demonstrates that the two additional methylene units in the alkanoyl chain c.f. CoBSH, do not necessarily translate into direct interaction of the thiol with the nickel proximal ligand. However, this could represent the favorable position for a CoBSH thiol interacting with the methyl group of methyl-SCoM. Just as the alkanoyl chain of CoB6SH has a more extended conformation than CoBSH in the substrate channel, CoBSH could also adopt a more extended conformation so that its thiol was in a similar position as the thiol of CoB8SH, priming it for reaction with a nickel bound species. If a significant conformational change is required early in MCR-catalyzed chemistry, which would be a requirement of mechanism I, catalysis may well involve a rearrangement of the aromatic amino acid annulus due to the presence of the methyl of methyl-SCoM, and this might enable deeper penetration of CoBSH into MCR (Figure 6). All the crystal structures in this study, and those solved previously, are of the inactive Ni(II)-MCR, which disfavors close approach to the nickel in the absence of Ni(I)-bound methyl-SCoM, even in the case of CoB9SH. Conclusion The goal of this study was to induce structural changes within the substrate channel and active site of MCR using analogues of coenzyme CoBSH. It was hoped that this would shed light on the nature of conformational changes that have been proposed to occur in MCR catalysis. We have shown that that the CoBXSH analogues do not lead to any significant conformational changes within the context of inactive Ni(II)-MCR. Therefore, it may be that methyl-SCoM is the key coenzyme, in combination with a nickel oxidation state of 1 (and 3), that triggers a conformational change bringing the thiol of CoBSH closer to the nickel. Thus, the crystal structure of the Ni(I)-methyl-SCoM/MCR complex may be required to structurally define the conformational changes required for MCR-mediated chemistry. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments X-ray data were collected at beam-line 14-BM-C and photoreduction studies at 14-ID-B, BioCARS at the Advanced Photon Source (APS), Argonne National Laboratory, Argonne, IL. We thank Vukica Srajer and Yu- Sheng Chen for valuable assistance during data collection. Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Basic Energy Sciences, Office of Science under Contract No. DE- AC02-06CH11357. Use of the BioCARS Sector 14 was supported by the National Institutes of Health, National Center for Research Resources, under grant number RR007707. Computer resources were provided by the Basic Sciences Computing Laboratory of the University of Minnesota Supercomputing Institute, and we thank Can Ergenekan for his support. We also thank Ed Hoeffner at the Kahlert Structural Biology Laboratory (KSBL) at the University of Minnesota. Use of the KSBL was supported by a Minnesota Partnership for Biotechnology and a Medical Genomics Grant SPAP-05-0013-P-FY06. References 1. Thauer RK. Biochemistry of methanogenesis: a tribute to Marjory Stephenson. Microbiology. 1998; 144:2377–2406. [PubMed: 9782487] 2. Thauer RK, Shima S. Methane as fuel for anaerobic microorganisms. Ann N Y Acad Sci. 2008; 1125:158–170. [PubMed: 18096853] Cedervall et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 3. Ellermann J, Hedderich R, Bocher R, Thauer RK. The final step in methane formation. Investigations with highly purified methyl-CoM reductase (component C) from Methanobacterium thermoautotrophicum (strain Marburg). Eur J Biochem. 1988; 172:669–677. [PubMed: 3350018] 4. Ellefson WL, Wolfe RS. Component C of the methylreductase system of Methanobacterium. J Biol Chem. 1981; 256:4259–4262. [PubMed: 6783657] 5. Ermler U, Grabarse W, Shima S, Goubeaud M, Thauer RK. Crystal structure of methyl-coenzyme M reductase: the key enzyme of biological methane formation. Science. 1997; 278:1457–1462. [PubMed: 9367957] 6. Diekert G, Gilles HH, Jaenchen R, Thauer RK. Incorporation of 8 succinate per mol nickel into factors F430 by Methanobacterium thermoautotrophicum. Arch Microbiol. 1980; 128:256–262. [PubMed: 7212929] 7. Diekert G, Jaenchen R, Thauer RK. Biosynthetic evidence for a nickel tetrapyrrole structure of factor F430 from Methanobacterium thermoautotrophicum. FEBS Letters. 1980; 119:118–120. [PubMed: 7428919] 8. Whitman WB, Wolfe RS. Presence of nickel in Factor F430 from Methanobacterium bryantii. Biochem Biophys Res Comm. 1980; 92:1196–1201. [PubMed: 7370029] 9. Albracht SPJ, Ankel-Fuchs D, Böcher R, Ellermann J, Moll J, van der Zwann JW, Thauer RK. Five new EPR signals assigned to nickel in methyl-coenzyme M reductase from Methanobacterium thermoautotrophicum, strain Marburg. Biochim Biophys Acta. 1988; 955:86–102. 10. Dey M, Kunz RC, Lyons DM, Ragsdale SW. Characterization of alkyl-nickel adducts generated by reaction of methyl-coenzyme m reductase with brominated acids. Biochemistry. 2007; 46:11969– 11978. [PubMed: 17902704] 11. Dey M, Telser J, Kunz RC, Lees NS, Ragsdale SW, Hoffman BM. Biochemical and spectroscopic studies of the electronic structure and reactivity of a methyl-Ni species formed on methyl- coenzyme M reductase. J Am Chem Soc. 2007; 129:11030–11032. [PubMed: 17711283] 12. Duin EC, Cosper NJ, Mahlert F, Thauer RK, Scott RA. Coordination and geometry of the nickel atom in active methyl-coenzyme M reductase from Methanothermobacter marburgensis as detected by X-ray absorption spectroscopy. J Biol Inorg Chem. 2003; 8:141–148. [PubMed: 12459909] 13. Duin EC, Signor L, Piskorski R, Mahlert F, Clay MD, Goenrich M, Thauer RK, Jaun B, Johnson MK. Spectroscopic investigation of the nickel-containing porphinoid cofactor F(430). Comparison of the free cofactor in the (+)1, (+)2 and (+)3 oxidation states with the cofactor bound to methyl- coenzyme M reductase in the silent, red and ox forms. J Biol Inorg Chem. 2004; 9:563–576. [PubMed: 15160314] 14. Finazzo C, Harmer J, Bauer C, Jaun B, Duin EC, Mahlert F, Goenrich M, Thauer RK, Van Doorslaer S, Schweiger A. Coenzyme B induced coordination of coenzyme M via its thiol group to Ni(I) of F430 in active methyl-coenzyme M reductase. J Am Chem Soc. 2003; 125:4988–4989. [PubMed: 12708843] 15. Finazzo C, Harmer J, Jaun B, Duin EC, Mahlert F, Thauer RK, Van Doorslaer S, Schweiger A. Characterization of the MCRred2 form of methyl-coenzyme M reductase: a pulse EPR and ENDOR study. J Biol Inorg Chem. 2003; 8:586–593. [PubMed: 12624730] 16. Goubeaud M, Schreiner G, Thauer RK. Purified methyl-coenzyme-M reductase is activated when the enzyme-bound coenzyme F430 is reduced to the nickel(I) oxidation state by titanium(III) citrate. Eur J Biochem. 1997; 243:110–114. [PubMed: 9030728] 17. Harmer J, Finazzo C, Piskorski R, Bauer C, Jaun B, Duin EC, Goenrich M, Thauer RK, Van Doorslaer S, Schweiger A. Spin density and coenzyme M coordination geometry of the ox1 form of methyl-coenzyme M reductase: a pulse EPR study. J Am Chem Soc. 2005; 127:17744–17755. [PubMed: 16351103] 18. Harmer J, Finazzo C, Piskorski R, Ebner S, Duin EC, Goenrich M, Thauer RK, Reiher M, Schweiger A, Hinderberger D, Jaun B. A nickel hydride complex in the active site of methyl- coenzyme m reductase: implications for the catalytic cycle. J Am Chem Soc. 2008; 130:10907– 10920. [PubMed: 18652465] Cedervall et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 19. Hinderberger D, Ebner S, Mayr S, Jaun B, Reiher M, Goenrich M, Thauer RK, Harmer J. Coordination and binding geometry of methyl-coenzyme M in the red1m state of methyl- coenzyme M reductase. J Biol Inorg Chem. 2008; 13:1275–1289. [PubMed: 18712421] 20. Kunz RC, Horng YC, Ragsdale SW. Spectroscopic and kinetic studies of the reaction of bromopropanesulfonate with methyl-coenzyme M reductase. J Biol Chem. 2006; 281:34663– 34676. [PubMed: 16966321] 21. Mahlert F, Bauer C, Jaun B, Thauer RK, Duin EC. The nickel enzyme methyl-coenzyme M reductase from methanogenic archaea: In vitro induction of the nickel-based MCR-ox EPR signals from MCR-red2. J Biol Inorg Chem. 2002; 7:500–513. [PubMed: 11941508] 22. Mahlert F, Grabarse W, Kahnt J, Thauer RK, Duin EC. The nickel enzyme methyl-coenzyme M reductase from methanogenic archaea: in vitro interconversions among the EPR detectable MCR- red1 and MCR-red2 states. J Biol Inorg Chem. 2002; 7:101–112. [PubMed: 11862546] 23. Rospert S, Voges M, Berkessel A, Albracht SP, Thauer RK. Substrate-analogue-induced changes in the nickel-EPR spectrum of active methyl-coenzyme-M reductase from Methanobacterium thermoautotrophicum. Eur J Biochem. 1992; 210:101–107. [PubMed: 1332856] 24. Sarangi R, Dey M, Ragsdale SW. Geometric and electronic structures of the Ni(I) and methyl- Ni(III) intermediates of methyl-coenzyme M reductase. Biochemistry. 2009; 48:3146–3156. [PubMed: 19243132] 25. Tang Q, Carrington PE, Horng YC, Maroney MJ, Ragsdale SW, Bocian DF. X-ray absorption and resonance Raman studies of methyl-coenzyme M reductase indicating that ligand exchange and macrocycle reduction accompany reductive activation. J Am Chem Soc. 2002; 124:13242–13256. [PubMed: 12405853] 26. Telser J, Davydov R, Horng YC, Ragsdale SW, Hoffman BM. Cryoreduction of methyl-coenzyme M reductase: EPR characterization of forms, MCR(ox1) and MCR (red1). J Am Chem Soc. 2001; 123:5853–5860. [PubMed: 11414817] 27. Yang N, Reiher M, Wang M, Harmer J, Duin EC. Formation of a nickel-methyl species in methyl- coenzyme M reductase, an enzyme catalyzing methane formation. J Am Chem Soc. 2007; 129:11028–11029. [PubMed: 17711279] 28. Albracht SPJ, Ankelfuchs D, Vanderzwaan JW, Fontijn RD, Thauer RK. A New Electron- Paramagnetic-Res Signal of Nickel in Methanobacterium-Thermoautotrophicum. Biochim Biophys Acta. 1986; 870:50–57. 29. Telser J, Horng YC, Becker DF, Hoffman BM, Ragsdale SW. On the assignment of nickel oxidation states of the Ox1, Ox2 forms of methyl-coenzyme M reductase. J Am Chem Soc. 2000; 122:182–183. 30. Hinderberger D, Piskorski RR, Goenrich M, Thauer RK, Schweiger A, Harmer J, Jaun B. A nickel- alkyl bond in an inactivated state of the enzyme catalyzing methane formation. Angewandte Chemie-International Ed. 2006; 45:3602–3607. 31. Kern DI, Goenrich M, Jaun B, Thauer RK, Harmer J, Hinderberger D. Two sub-states of the red2 state of methyl-coenzyme M reductase revealed by high-field EPR spectroscopy. J Biol Inorg Chem. 2007; 12:1097–1105. [PubMed: 17690920] 32. Becker DF, Ragsdale SW. Activation of methyl-SCoM reductase to high specific activity after treatment of whole cells with sodium sulfide. Biochemistry. 1998; 37:2639–2647. [PubMed: 9485414] 33. Grabarse W, Mahlert F, Duin EC, Goubeaud M, Shima S, Thauer RK, Lamzin V, Ermler U. On the mechanism of biological methane formation: structural evidence for conformational changes in methyl-coenzyme M reductase upon substrate binding. J Mol Biol. 2001; 309:315–330. [PubMed: 11491299] 34. Grabarse W, Mahlert F, Shima S, Thauer RK, Ermler U. Comparison of three methyl-coenzyme M reductases from phylogenetically distant organisms: unusual amino acid modification, conservation and adaptation. J Mol Biol. 2000; 303:329–344. [PubMed: 11023796] 35. Horng YC, Becker DF, Ragsdale SW. Mechanistic studies of methane biogenesis by methyl- coenzyme M reductase: evidence that coenzyme B participates in cleaving the C-S bond of methyl-coenzyme M. Biochemistry. 2001; 40:12875–12885. [PubMed: 11669624] Cedervall et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 36. Berkessel A. Methyl-Coenzyme-M Reductase - Model Studies on Pentadentate Nickel-Complexes and a Hypothetical Mechanism. Bioorg Chem. 1991; 19:101–115. 37. Jaun B. Coenzyme-F430 from Methanogenic Bacteria - Oxidation of F430 Pentamethyl Ester to the Ni(Iii) Form. Helvetica Chimica Acta. 1990; 73:2209–2217. 38. Signor L, Knuppe C, Hug R, Schweizer B, Pfaltz A, Jaun B. Methane formation by reaction of a methyl thioether with a photo-excited nickel thiolate - A process mimicking methanogenesis in archaea. Chemistry-a European Journal. 2000; 6:3508–3516. 39. Chen SL, Pelmenschikov V, Blomberg MR, Siegbahn PE. Is there a Ni-methyl intermediate in the mechanism of methyl-coenzyme M reductase? J Am Chem Soc. 2009; 131:9912–9913. [PubMed: 19569621] 40. Pelmenschikov V, Blomberg MRA, Siegbahn PEM, Crabtree RH. A mechanism from quantum chemical studies for methane formation in methanogenesis. J Am Chem Soc. 2002; 124:4039– 4049. [PubMed: 11942842] 41. Pelmenschikov V, Siegbahn PE. Catalysis by methyl-coenzyme M reductase: a theoretical study for heterodisulfide product formation. J Biol Inorg Chem. 2003; 8:653–662. [PubMed: 12728361] 42. Duin EC, McKee ML. A new mechanism for methane production from methyl-coenzyme M reductase as derived from density functional calculations. J Phys Chem. 2008; B 112:2466–2482. 43. Bobik TA, Wolfe RS. Physiological importance of the heterodisulfide of coenzyme M and 7- mercaptoheptanoylthreonine phosphate in the reduction of carbon dioxide to methane in Methanobacterium. Proc Natl Acad Sci U S A. 1988; 85:60–63. [PubMed: 3124103] 44. Goenrich M, Duin EC, Mahlert F, Thauer RK. Temperature dependence of methyl-coenzyme M reductase activity and of the formation of the methyl-coenzyme M reductase red2 state induced by coenzyme B. J Biol Inorg Chem. 2005; 10:333–342. [PubMed: 15846525] 45. Kunz RC, Dey M, Ragsdale SW. Characterization of the Thioether Product Formed from the Thiolytic Cleavage of the Alkyl-Nickel Bond in Methyl-Coenzyme M Reductase. Biochemistry. 2008; 47:2661–2667. [PubMed: 18220418] 46. Noll KM, Donnelly MI, Wolfe RS. Synthesis of 7-mercaptoheptanoylthreonine phosphate and its activity in the methylcoenzyme M methylreductase system. J Biol Chem. 1987; 262:513–515. [PubMed: 3100513] 47. Olson KD, McMahon CW, Wolfe RS. Photoactivation of the 2-(methylthio)ethanesulfonic acid reductase from Methanobacterium. Proc Natl Acad Sci U S A. 1991; 88:4099–4103. [PubMed: 1903534] 48. Zehnder AJ, Wuhrmann K. Titanium (III) citrate as a nontoxic oxidation-reduction buffering system for the culture of obligate anaerobes. Science. 1976; 194:1165–1166. [PubMed: 793008] 49. Gunsalus RP, Romesser JA, Wolfe RS. Preparation of coenzyme M analogues and their activity in the methyl coenzyme M reductase system of Methanobacterium thermoautotrophicum. Biochemistry. 1978; 17:2374–2377. [PubMed: 98178] 50. Otwinowski Z, Minor W. Processing of X-ray Diffraction Data Collected in Oscillation Mode. Methods in Enzymology: Macromolecular Crystallography, part A. 1997; 276:307–326. 51. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr. 1994; 50:760–763. [PubMed: 15299374] 52. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] 53. Ebner S, Jaun B, Goenrich M, Thauer RK, Harmer J. Binding of coenzyme B induces a major conformational change in the active site of methyl-coenzyme M reductase. J Am Chem Soc. 2010; 132:567–575. [PubMed: 20014831] 54. Goenrich M, Mahlert F, Duin EC, Bauer C, Jaun B, Thauer RK. Probing the reactivity of Ni in the active site of methyl-coenzyme M reductase with substrate analogues. J Biol Inorg Chem. 2004; 9:691–705. [PubMed: 15365904] Cedervall et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. The active site and substrate channel in the MCRox1-silent crystal structure (PDB code 1hbn) (9). Coenzyme F430, CoBSH and HSCoM are drawn as stick colored by atom (carbon: dark grey). The nickel is displayed as a green sphere, water as a red sphere. Interactions are drawn as dashed lines, and the corresponding distance is indicted in Angstroms (Å). The path of the substrate channel was defined in the absence of F430, CoBSH, HSCoM and water, with the surface closest to the viewer cut away. The figure was generated using PyMOL (http://www.pymol.org). Cedervall et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Drawing of CoBSH analogues; (A) N-5-mercaptopentanoylthreonine phosphate (CoB5SH); (B) N-6-mercaptohexanoylthreonine phosphate (CoB6SH); (C) N-8- mercaptooctanoylthreonine phosphate (CoB8SH); (D) N-9-mercaptononanoylthreonine phosphate (CoB9SH). Cedervall et al. Page 16 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. The active sites and substrate channels of the MCR crystal structures.; (A) MCRCoB5SH; (B) MCRCoB6SH; (C) MCRHSCoM; (D) MCRCoB8SH; (E) MCRCoB9SH. 2Fo-Fc electron density map around the CoBXSH analogues, waters in the CoBSH binding part of the channel and the acetate ion (contoured at 1σ) is shown as a blue mesh. The protein is drawn as cartoon. CoBXSH and acetate are drawn as stick and colored by atom (carbon: CoB5SH orange; CoB6SH pale yellow; CoB8SH light blue; CoB9SH magenta; acetate white). Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere, and waters as red spheres. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 17 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. Overlay of CoBSH (from PDB code 1hbn) and the different CoBSH analogues. CoBXSH are drawn as stick with the thiol represented by a sphere, and colored CoB5SH orange; CoB6SH pale yellow; CoBSH light green; CoB8SH light blue; CoB9SH magenta. The protein is drawn as cartoon. Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 18 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Hydrogen bonding diagram for the water structure modeled in MCRHSCoM. The water molecules are named as in Figure 3C (W1–W9); WA, WB and WC are water molecules that are present in all structures (i.e. in concert with the substrate CoBSH and the CoBSH analogues). Interactions between surrounding residues and the water molecules are drawn as dashed lines, and the corresponding distance is indicated in Angstroms (Å). Cedervall et al. Page 19 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. Stereo image of the annulus of aromatic amino acids distal of coenzyme F430. The protein is drawn as cartoon with the side-chains of the aromatic residues drawn as white stick. CoBSH, (from PDB code 1hbn (9)), CoB8SH and CoB9SH are drawn as stick with the thiols represented by spheres, and colored CoBSH light green; CoB8SH light blue; CoB9SH magenta. Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 20 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. Reaction catalyzed by methyl-coenzyme M reductase Cedervall et al. Page 21 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 2. Two of the proposed catalytic mechanisms for methyl-coenzyme M reductase; (A) mechanism I; (B) mechanism II. Cedervall et al. Page 22 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 23 Table 1 X-ray Data Collection, Processing and Refinement Statistics Data collection and processing statistics Name of data set MCRCoB5SH MCRCoB6SH MCRHSCoM MCRCoB8SH MCRCoB9SH Measured reflections 1969388 2427498 1440665 1160543 1425506 Unique reflections 553755 446253 405349 211803 401701 Resolution (Å) a 50.0–1.30 (1.35–1.30) 50.0–1.40 (1.45–1.40) 50.0–1.45 (1.50–1.45) 50.0–1.80 (1.86–1.80) 50.0–1.45 (1.50–1.45) Completeness (%) a 97.1 (78.1) 99.9 (100.0) 99.5 (99.7) 99.8 (100.0) 98.1 (95.4) R-sym (%) a,b 5.5 (32.9) 7.3 (44.7) 6.2 (44.0) 8.4 (47.7) 5.6 (42.5) I/σI a 22.3 (3.6) 20.4 (4.0) 20.2 (3.2) 21.8 (3.9) 24.3 (3.2) Space group P21 P21 P21 P21 P21 Refinement and model building statistics Resolution (Å) a 20.49–1.30 (1.33–1.30) 19.89–1.40 (1.44–1.40) 20.15–1.45 (1.49–1.45) 19.93–1.80 (1.84–1.80) 20.07–1.45 (1.48–1.45) No. of reflection in working set a 525817 (30239) 423854 (25833) 384868 (25791) 201128 (11193) 381474 (23611) No. of reflection in test set a 27777 (1576) 22348 (1331) 20362 (1319) 10625 (557) 20163 (1210) R-work (%) c 14.32 13.04 13.47 14.95 13.58 R-free (%) d 16.56 15.53 16.22 19.54 16.44 ESU (Å) R-work/R-free 0.044/0.046 0.049/0.051 0.056/0.059 0.121/0.119 0.057/0.060 No. protein atoms 20087 19960 20265 19750 20036 No. coenzyme atoms 218 220 180 224 272 No. ligand atoms 37 62 52 26 49 No. water molecules 2443 2352 2516 1893 2432 RMS bond lengths (Å) 0.033 0.033 0.032 0.028 0.032 bond angles (deg.) 2.693 2.625 2.468 2.059 2.549 Ramachandran plot (%) favored 97.8 97.5 97.6 97.2 97.7 allowed 2.1 2.4 2.3 2.7 2.1 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 24 disallowed 0.1 0.1 0.1 0.1 0.1 Average B-factor (Å2) protein 12.42 13.35 12.12 17.22 12.73 coenzymes 8.20 9.24 7.25 11.24 8.27 ligands 31.95 35.48 28.29 33.76 32.92 waters 22.95 24.89 23.85 26.79 24.09 over all 13.54 14.57 13.40 18.02 13.93 Occupancy of HSCoM per active site (%)e 90/90 50/50 100/100 90/90 90/85 Occupancy of CoBSH per active site (%) e 50/50 50/50 30/30 50/50 40/40 CoBSH analogue, occupancy per active site (%) e CoB5SH, 50/50 CoB6SH, 50/50 CoB8SH, 50/50 CoB9SH, 60/60 Other molecule, occupancy per active site (%) e Acetate, 70/70 aValues in brackets correspond to the highest resolution shell. bR-sym = ΣhklΣNj=1|Ihkl−Ihkl(j)|/ΣhklN*Ihkl, sum over all reflections and all observations N, with Ihkl(j) intensity of the jth observation of reflection hkl and Ihkl mean intensity of the reflection hkl. cR-work = Σ||Fo| − |Fc||/Σ|Fo|, where |Fo| = observed structure factor amplitude and |Fc| = calculated structure factor amplitude. dR-free, R-factor based on 5% of the data excluded from refinement. eOccupancy of model in each of the two crystallographically independent active sites in the ASU Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 25 Table 2 Distances from analogue thiols. CoBXS - SCoM distance (Å) CoBXS - Ni distance (Å) CoB5SH 7.11/7.11a 9.30/9.30 CoB6SH 6.26/6.26 8.70/8.70 CoB7SH (substrate) b 6.37/6.39 8.73/8.77 CoB8SH 3.75/3.78 6.16/6.17 CoB9SH 3.71/3.68 5.96/5.91 aDistances in the two crystallographically independent active sites in the ASU bDistances in the 1.16 Å resolution MCRox1-silent structure (PDB code 1hbn)(33) Biochemistry. Author manuscript; available in PMC 2011 September 7.
3M2V
Structural Insight into Methyl-Coenzyme M Reductase Chemistry using Coenzyme B Analogues
Structural Insight into Methyl-Coenzyme M Reductase Chemistry using Coenzyme B Analogues,†,‡ Peder E. Cedervall§,#, Mishtu Dey||,#,⊥, Arwen R. Pearson§,⊥, Stephen W. Ragsdale||, and Carrie M. Wilmot*,||,§ § Department of Biochemistry, Molecular Biology and Biophysics, University of Minnesota, Minneapolis, Minnesota 55455 || Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109 Abstract Methyl-coenzyme M reductase (MCR) catalyzes the final and rate-limiting step in methane biogenesis; the reduction of methyl-coenzyme M (methyl-SCoM) by coenzyme B (CoBSH) to methane and a heterodisulfide (CoBS-SCoM). Crystallographic studies show that the active site is deeply buried within the enzyme, and contains a highly reduced nickel-tetrapyrrole, coenzyme F430. Methyl-SCoM must enter the active site prior to CoBSH, as species derived from analogues of methyl-SCoM are always observed bound to the F430 nickel in the deepest part of the 30 Å long substrate channel that leads from the protein surface to the active site. The seven-carbon mercaptoalkanoyl chain of CoBSH binds within a 16 Å predominantly hydrophobic part of the channel close to F430, with the CoBSH thiolate lying closest to the nickel at a distance of 8.8 Å. It has previously been suggested that binding of CoBSH initiates catalysis by inducing a conformational change that moves methyl-SCoM closer to the nickel promoting cleavage of the C- S bond of methyl-SCoM. In order to better understand the structural role of CoBSH early in the MCR mechanism, we have determined crystal structures of MCR in complex with four different CoBSH analogues; pentanoyl-, hexanoyl-, octanoyl- and nonanoyl- derivatives of CoBSH (CoB5SH, CoB6SH, CoB8SH and CoB9SH respectively). The data presented here reveal that the shorter CoB5SH mercaptoalkanoyl chain overlays with that of CoBSH, but terminates two units short of the CoBSH thiolate position. In contrast, the mercaptoalkanoyl chain of CoB6SH adopts a different conformation, such that its thiolate is coincident with the position of the CoBSH thiolate. †This work was supported by a Department of Energy grant DE-FG02-08ER15931 to S.W.R and supplement to C.M.W. and a Minnesota Partnership for Biotechnology and Medical Genomics grant SPAP-05-0013-P-FY06. ‡Co-ordinates and structure factors have been deposited in the Protein Data Bank as entries 3m1v (MCRHSCoM); 3m2r (MCRCoB5SH); 3m2u (MCRCoB6SH); 3m2v (MCRCoB8SH); 3m30 (MCRCoB9SH). *Address correspondence to: Carrie M. Wilmot, Tel: (612) 624-2406, wilmo004@umn.edu. ⊥Current addresses: Mishtu Dey, Department of Chemistry, Massachusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, Massachusetts 02139; Arwen Pearson, Astbury Centre for Structural Molecular Biology, Institute for Molecular and Cellular Biology, Astbury Building. Leeds, LS2 9JT, U.K. #These authors contributed equally to this work. Supporting Informational material is available free of charge via the Internet at http://pubs.acs.org and contains the following: MATERIAL AND METHODS for single crystal UV-visible microspectrophotometry, X-ray photoreduction experiment and X-ray crystallography of the MCR-heterodisulfide product complex (MCRCoBSH + methyl-SCoM); RESULTS AND DISCUSSION for redox changes and MCRCoBSH + methyl-SCoM crystal structure; Table S1, X-ray Data Collection, Processing and Refinement Statistics for MCRCoBSH + methyl-SCoM; Figure S1, use of Fo-Fc electron density in modelling MCRCoB8SH; Figure S2, illustration of electron density quality of MCRCoB5SH; Figure S3, solution and single crystal UV-visible spectra; Figure S4, modeling of CoB6SH and CoBSH into the electron density of MCRCoB6SH; Figure S5, partially occupied HSCoM in MCRCoB6SH; Figure S6, the active site and substrate channel of MCRCoBSH + methyl-SCoM; Figure S7, alternative conformation of Valα482 in MCRHSCoM; Figure S8, the two conformations of the gly-rich loop in MCRHSCoM; Figure S9, propagation of conformational changes in MCRHSCoM; Figure S10, the two conformations of CoB9SH in MCRCoB9SH; Figure S11, EPR spectra of MCRred1 sample; Scheme S1, scheme of the characterized forms of MCR. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 September 7. Published in final edited form as: Biochemistry. 2010 September 7; 49(35): 7683–7693. doi:10.1021/bi100458d. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript This is consistent with the observation that CoB6SH is a slow substrate. A labile water in the substrate channel was found to be a sensitive indicator for the presence of CoBSH and HSCoM. The longer CoB8SH and CoB9SH analogues can be accommodated in the active site through exclusion of this water. These analogues react with Ni(III)-methyl; a proposed MCR catalytic intermediate of methanogenesis. The CoB8SH thiolate is 2.6 Å closer to the nickel than that of CoBSH, but the additional carbon of CoB9SH only decreases the nickel thiolate distance a further 0.3 Å. Although the analogues did not induce any structural changes in the substrate channel, the thiolates appeared to preferentially bind at two distinct positions in the channel; one being the previously observed CoBSH thiolate position, and the other being at a hydrophobic annulus of residues that lines the channel proximal to the nickel. INTRODUCTION Methanogenic archaea are organisms that under strict anaerobic conditions derive energy by reducing compounds such as carbon dioxide, methylamine, acetate and methanol, to methane (1, 2). The global production of methane by these organisms is estimated at one billion tons annually. Microbially produced methane is not only a potential source of renewable energy but also a potent greenhouse gas, and as such study of this process has environmental ramifications. In these microorganisms, methyl-coenzyme M reductase (MCR)1 is the enzyme that catalyzes the final step in methanogenesis, in which the substrates methyl-coenzyme M (methyl-SCoM, 2-(methylthio)ethanesulfonate) and coenzyme B (CoBSH, N-7-mercaptoheptanoylthreonine phosphate) are converted to methane and a heterodisulfide (CoBS-SCoM) (Scheme 1) (3). MCR is a 300 kDa protein with six subunits arranged in a α2β2γ2 oligomer (4). The known crystal structures show that MCR has two active sites approximately 50 Å apart that are deeply buried within the enzyme (5). The active site pocket is comprised of residues from subunits α, α′, β and γ, with a 30 Å long substrate channel leading to the enzyme surface (Figure 1). At the heart of the active site pocket is coenzyme F430, which is a highly reduced nickel-containing tetrapyrrole (6–8). Currently sixteen distinct enzymatic and complexed states of MCR have been spectroscopically characterized (Supporting Information, Scheme S1) (9–31). In the resting active state of the enzyme, denoted MCRred1, the redox-active nickel of F430 is present in the Ni(I) state (9, 16, 32). MCR is extremely oxygen sensitive and upon oxygen exposure the enzyme enters an inactive Ni(II) state, denoted MCRred1-silent (6). In this state it cannot be converted back to the active Ni(I) form by any known reducing agent making this a challenging system to study. Additional complications involve the tight association of coenzymes to purified MCR that are not easily displaced as demonstrated by X-ray crystallographic and kinetic studies (5, 33–35). Despite the fact that MCR has been studied for decades, no true catalytic intermediate has been observed, and the actual mechanism remains elusive. Currently three general mechanistic schemes for the enzymatic reaction have been proposed, each of which posit different chemistry to initiate catalysis. Mechanism I involves Ni(I) acting as a nucleophile in an SN2-type reaction that generates Ni(III)-methyl as an intermediate (Scheme 2A) (35– 38). Mechanism II starts with methyl-SCoM undergoing homolytic cleavage at the Ni(I) to generate a methyl radical and a Ni(II)-SCoM species (Scheme 2B) (39–41). A more recently proposed mechanism III suggests protonation of coenzyme F430 promotes reductive cleavage of the methyl-SCoM thioether bond (42). 1Abbreviations: MCR, Methyl-coenzyme M reductase; methyl-SCoM, methyl-coenzyme M; CoBSH, coenzyme B; HSCoM, coenzyme M; CoBS-SCoM, heterodisulfide of coenzyme B and coenzyme M; APS, Advanced Photon Source; ASU, asymmetric unit; BPS, bromopropanesulfonate. Cedervall et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Due to the stringent requirement to exclude O2, the available MCR crystal structures are all in the inactive Ni(II) state. All but one contain CoBSH and HSCoM (demethylated methyl- SCoM, an inhibitor and substrate analogue) in the active site (PDB codes 1hbn, 1hbo, 1hbu, 1e6y, 1e6v) (5, 33, 34). Another crystal structure has bound heterodisulfide product, CoBS- SCoM (MCRsilent, PDB code 1hbm, Scheme 1 and Supporting Information, Scheme S1) (5, 33). All these structures reveal that both substrates access the active site through the same channel (Figure 1). The binding site of HSCoM (and presumably methyl-SCoM) is more deeply buried within the enzyme, and so it must enter prior to CoBSH for productive chemistry to occur. As binding of CoBSH in the absence of co-substrate would be inhibitory, it was suggested that a conformational change upon methyl-SCoM binding might lower the Kd for CoBSH, and thus promote an ordered mechanism. Furthermore it has been suggested that CoBSH binding induces a conformational change that brings the methyl- SCoM substrate into closer proximity to the nickel, and this promotes C-S bond cleavage. To investigate the proposed structural role of CoBSH in initiating catalysis, we have solved the X-ray crystal structures of MCR in complex with four different CoBSH analogues. CoBSH has a heptanoyl moiety linked to the thiol group, and the analogues are pentanoyl-, hexanoyl-, octanoyl- or nonanoyl-containing derivatives of CoBSH (CoB5SH, CoB6SH, CoB8SH and CoB9SH respectively; Figure 2) (3, 35, 43–47). In addition, we present a structure in which the substrate channel predominantly lacks either CoBSH or heterodisulfide product. MATERIALS AND METHODS Materials The organism Methanothemobacter marburgensis (catalog OCM82) was obtained from the Oregon Collection of Methanogens (Portland, OR). All buffers and media reagents were obtained from Sigma-Aldrich (St. Louis, MO). The gases N2 (99.98%), H2/CO2 (80%/20%), and ultra high purity H2 (99.999%) were obtained from Cryogenic Gases (Grand Rapids, MI). Ti(III) citrate solutions were prepared from a stock solution of 246 mM Ti(III) citrate, which was synthesized by adding sodium citrate to Ti(III) trichloride (30 wt % solution in 2 N hydrochloric acid) (Acros Organics, Morris Plains, NJ) under anaerobic conditions and adjusting pH to 7.0 with sodium bicarbonate (48). The concentration of Ti(III) citrate was determined by titrating against a solution of methyl viologen. Synthesis of methyl-SCoM, CoB5SH, CoB6SH, CoBSH, CoB8SH, and CoB9SH Methyl-SCoM was prepared from HSCoM and methyl iodide (49). The homodisulfides, CoB5S-SCoB5, CoB6S-SCoB6, CoBS-SCoB, CoB8S-SCoB8, CoB9S-SCoB9 were prepared as described from 5-bromovaleric acid, 6-bromohexanoic acid (Sigma-Aldrich, St. Louis, MO), 7-bromoheptanoic acid (Karl Industries, Aurora, OH), 8-bromooctanoic acid, and 9- bromononanoic acid (Matrix Scientific, Columbia, SC), respectively (43, 46). The free thiol forms of CoB5SH, CoB6SH, CoBSH, CoB8SH, and CoB9SH were generated by the reduction of the homodisulfides as previously described (45). The purity of the CoBSH analogues was determined by 1H NMR spectroscopy. All compounds synthesized were stored in a Vacuum Atmospheres chamber maintained at an oxygen level below 1 ppm, as monitored continually with an oxygen analyzer (model 317, Teledyne Analytical Instruments, City of Industry, CA) until use. M. marburgensis Growth and MCRred1 Purification Buffer preparations and all manipulations were performed under strict anaerobic conditions in a Vacuum Atmospheres chamber maintained at an oxygen level below 1 ppm, as monitored continually with an oxygen analyzer (model 317, Teledyne Analytical Instruments, City of Industry, CA). MCRred1 was isolated from M. marburgensis cultured on Cedervall et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript H2/CO2 (80%/20%) at 65 °C in a 14-L fermentor (New Brunswick Scientific Co., Inc. New Brunswick, NJ). Culture media were prepared as previously described (20, 24). MCRred1 was generated in vivo and purified as described previously (20). The purification procedure routinely generates 65–75% MCRred1 as determined by UV-visible and EPR spectroscopy. Spectroscopy of MCR UV-visible spectra of the Ni(I)-containing MCRred1 were recorded in the anaerobic chamber using a spectrophotometer (Model USB4000-UV-VIS, Ocean Optics, Dunedin, FL). EPR spectra were recorded on a Bruker EMX spectrometer (Bruker Biospin Corp., Billerica, MA), equipped with an Oxford ITC4 temperature controller, a Hewlett-Packard Model 5340 automatic frequency counter and Bruker gaussmeter. The EPR spectroscopic parameters included: temperature, 70 K; microwave power, 10 mW; microwave frequency, 9.43 GHz; receiver gain, 2 × 104; modulation amplitude, 10.0 G; modulation frequency, 100 kHz. Double integrations of the EPR spectra were performed and referenced to a 1 mM copper perchlorate standard. The NMR data were acquired at 298 K on a Bruker Avance DRX 500 MHz instrument equipped with a TXI cryoprobe. Preparation of MCRred1 for Crystallization All crystallization experiments were performed in the anaerobic chamber in which MCR was purified unless otherwise noted. MCRred1 was prepared in 50 mM Tris, pH 7.6 and excess HSCoM was removed by buffer exchange using an Amicon Ultra centrifuge filter with a 50 kDa cut-off membrane (Millipore). Typically, 2–3 ml of MCRred1 was exchanged with 10–15 ml of 50 mM Tris, pH 7.6. The enzyme was concentrated to 500–600 μl, and this process was repeated three times. The fraction of MCRred1 in the purified MCR sample was calculated from the UV-visible spectrum using extinction coefficients of 27.0 mM−1cm−1 at 385 nm for Ni(I)-MCRred1, and 9.15 mM−1cm−1 at 420 nm for Ni(II)- MCRred1-silent (20). The amount of MCRred1 in samples used for crystallization was determined to be ~80% and the concentration of total enzyme used was in the range of about 120–150 μM (~32–40 mg/ml). All crystallization experiments were performed anaerobically by incubating 2.0 μl of enzyme solution in 50 mM Tris, pH 7.5 and 2.0 μl of reservoir solution (100 mM Hepes-Na, pH 7.3/7.5/8.0; 150 mM magnesium acetate (Mg(CH3COO)2), and 20/22% (w/v) PEG 400) in a sitting drop over 1 ml reservoir solution at 9 ºC. Triangular and rectangular prismatic crystals with a bright yellowish-green color confirmed the presence of nickel coenzyme F430. The crystals grew to a size of approximately 100–200 μm in 4–5 days. CoBSH-depleted crystals were obtained by incubating 2 μl of a reaction mixture containing 139 μM MCRred1 and 13 mM HSCoM with 2 μl of reservoir solution (100 mM Hepes-Na pH 7.5, 150 mM magnesium acetate (Mg(CH3COO)2), 22% PEG 400). Crystals of MCR complexed with the CoBSH analogues were grown by co-crystallization. The CoB5SH co-crystals were obtained by incubating 2μl enzyme solution containing 124 μM MCRred1, 10 mM methyl-SCoM and 1 mM CoB5SH with 2μl of reservoir solution (100 mM Hepes-Na, pH 7.5, 150 mM (Mg(CH3COO)2), 22 % PEG 400). The crystals with bound CoB6SH and CoB9SH were obtained by co-crystallization of 1 mM of analogue with 142 μM MCRred1 and equilibrated with 2μl of reservoir solution (100 mM Hepes-Na, pH 7.5, 150 mM Mg(CH3COO)2, 20 % PEG 400 for CoB6SH and 100 mM Hepes-Na, pH 7.3, 150 mM Mg(CH3COO)2), 22 % PEG 400 for CoB9SH). Crystals were cryoprotected in reservoir solution containing 25 % (v/v) PEG 400 by soaking for 2–5 minutes before cryocooling in liquid nitrogen in the anaerobic chamber. Crystals of CoB8SH bound to MCR were obtained by incubating 2 μl of a mixture of 119 μM MCRred1 and 1 mM CoB8SH with 2 μl of reservoir solution (100 mM Hepes-Na, pH 7.3, 150 mM Mg(CH3COO)2), 20 % PEG 400). Before cryoprotection, the crystals were soaked for 5–10 minutes in a 100 mM solution of methyl iodide2. The methyl iodide solution used for soaking was prepared by adding a concentrated stock of methanolic solution of methyl iodide to the reservoir Cedervall et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript solution. Soaked crystals were quickly cryoprotected as described above and cryocooled in the anaerobic chamber. X-ray Diffraction Data Collection, Processing and Refinement X-ray diffraction data were collected at 100 K on a ADSC Quantum-315 detector at the APS Beamline 14-BM-C (BioCARS). The wavelength of X-rays was 0.979 Å. Data were processed using HKL2000 (50). As in the previous X-ray crystallographic studies, the crystals belong to the monoclinic space group P21 (a = 82 Å, b = 118 Å, c = 122 Å, β = 92°), with one MCR molecule (two active sites) per asymmetric unit (5, 33). For refinement, REFMAC in the Collaborative Computational Project Number 4 (CCP4) program suite was used (51). A random sample of 5 % of the data across all resolution shells was chosen to check refinement progress through calculation of an Rfree. The same reflections were used to calculate Rfree for all structures, thus preventing bias due to high structural identity. The remaining reflections were used in refinement (Rwork). Model building was done using the Crystallographic Object-Oriented Toolkit (COOT) (52). The diffraction data and their models are designated as MCRCoBXSH, where X is the number of carbons in the alkanoyl portion of the analogue. Library files in CCP4 for F430 and CoBSH were incorrect, and these were modified in Monomer Library Sketcher in the CCP4 program suite by comparison with schematic drawings from Grabarse et al. (33). Co-ordinate and CCP4 library files for the different CoBSH analogues were created in Monomer Library Sketcher. The general model building and refinement strategy for all structures was as follows. It was clear from the electron density in the substrate channel and at the active site that mixtures of species were present in all datasets. These could be visualized with Fo-Fc and Fo-Fo difference electron density maps (Supporting Information, Figure S1). The known positions of CoBSH and HSCoM from the published Ni(II)-MCR crystal structures (PDB codes 1hbn, 1hbo, 1hbu (33)) were used as guides to determine which species could be present in each dataset, and these were then simultaneously modeled into the electron density. By alteration of their relative occupancies (in 10% increments) followed by refinement, the ratio of occupancy between different species was determined using the assumption that the average B-factors for all molecular species bound should be similar to that of F430 and adjacent well-ordered protein atoms within the active site and substrate channel. The combinations of modeled ligands were constantly reassessed throughout refinement based on the remaining difference electron density. This included test refinements of different ligand combinations during the latter stages, thus using the optimized phases to check whether a different combination of ligands could also explain the electron density. Sensible chemical structures and interactions, along with keeping the combined occupancies of sterically mutually exclusive species ≤ 100%, were maintained throughout refinement. The model was finally accepted when the difference electron density map was minimal and the B-factors for the models converged. In practice the first structure refined was that of MCRCoB5SH. Initial phases were generated by difference Fourier using a previously determined crystal structure (PDB code 1mro (5)) but with all non-bonded molecules, including water, removed from the model except F430. Initial rigid body refinement followed by restrained refinement of MCRCoB5SH reduced the Rwork to 26.5%. After model building and subsequent rounds of restrained refinement the Rwork was 14.3% (Rfree 16.6%). Of the five structures only the CoB5SH analogue is completely coincident with CoBSH, and so particular care had to be used in teasing apart the ratios of the two species in modeling the MCRCoB5SH electron density. This was done by 2Methyl iodide was added with the intention of creating a Ni(III)-methyl species, which CoB8SH stabilizes. This was not achieved, but the diffraction quality of this crystal was significantly better than crystals co-crystallized with CoB8SH alone, and so has been included in this study. Cedervall et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript initially refining 100% CoBSH in the substrate channel. Positive Fo-Fc difference density located at the carbon where the shorter CoB5SH thiol might be expected to be indicated the presence of a more electron-rich species than carbon, which is consistent with the presence of the CoB5SH sulfur. The refinement converged at a model containing a 50:50 mix of CoBSH and the analogue. However, positive Fo-Fc difference density was still present at the position of the CoBSH thiol, therefore a water molecule was added to the CoB5SH model at 50% occupancy and upon refinement this accounted for the electron density. An illustration of the electron density quality from this structure is shown in Supporting Information, Figure S2. An HSCoM-, CoBSH- and CoBSH analogue-free version of the refined MCRCoB5SH structure was used as the starting model to generate initial phases for the four other structures. After the initial round of restrained refinement the Rwork for these structures were reduced to 14.5–15.6 %. RESULTS AND DISCUSSION Crystal Structures of MCR Five crystal structures were determined, four of which are in complex with CoBSH analogues differing in the number of carbons atoms in the alkanoyl portion of the molecule. CoBSH is an N-7-mercaptoheptanoyl-containing molecule, whereas the four CoBSH analogues contain N-5-mercaptopentanoyl-, N-6-mercaptohexanoyl-, N-8-mercaptooctanoyl- or N-9-mercaptononanoyl-moieties (Figure 2). The corresponding crystal structures are designated as MCRCoBXSH, where X is the number of carbons in the alkanoyl portion of the analogue. The other crystal structure is of MCRred1c-silent (MCR in the Ni(II) state in complex with HSCoM, designated here as MCRHSCoM) that is CoBSH-depleted. The datasets have resolutions in the range from 1.35 – 1.8 Å. Although the crystallizations were set up with the MCR solution predominantly in the Ni(I)-MCRred1 state, by the time X-ray diffraction data were collected they had been oxidized to the Ni(II)-MCRred1-silent state (Supporting Information). Following data collection there was no evidence for photoreduction of the Ni(II) back to Ni(I) in any of the crystals, as assessed by single crystal UV-visible microspectrophotometry (Supporting Information and Figure S3). Attempts to photoreduce the crystals using different wavelengths and temperatures were unsuccessful (Supporting Information). Overall, the resulting structures are very similar to each other and to the previously published structures of MCR, with differences mainly localized to the active site and substrate channel. The two active sites in the ASU were refined independently. Unless otherwise stated there was no difference between them. All five datasets contain a mixture of species bound to the enzyme. There is always a background of CoBSH and HSCoM, which co-purify with MCR and cannot be fully removed by extensive buffer exchange or by the addition of a CoBSH analogue. HSCoM is added during purification of MCR, as it stabilizes the resting active Ni(I) state (unpublished data), and this leads to HSCoM occupancies between 50–100% amongst the structures (Table 1). In contrast CoBSH, which is not added during purification, has occupancies ranging from 30–50%. As these confounding species have all been described at high occupancy in other crystallographic studies, the structural data of interest could be isolated (5, 33). In each case, the additional electron density could be explained by inclusion of the appropriate CoBXSH model used in that experiment at 50% or higher occupancy. The resulting models, along with 2Fo-Fc electron density, are shown in Figure 3. The Rwork for the final structures range from 13.0 to 15.0 % (Rfree 15.5 to 19.5 %). The X-ray data collection, processing, refinement and model building statistics are given in Table 1. Cedervall et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Analogues shorter than CoBSH; CoB5SH and CoB6SH CoB5SH is two methylene groups shorter than CoBSH, the MCR substrate. The MCRCoB5SH structure is to 1.35 Å resolution. As expected the pentanoyl chain follows the path of the CoBSH heptanoyl carbons down the substrate channel, and thus its thiol is positioned in the same place as the second carbon preceding the CoBSH thiol (Figure 3A and 4). There are no published MCR kinetic studies using CoB5SH, but as it binds in the substrate channel, it is likely to be an inhibitor. CoB6SH is one methylene shorter than CoBSH, and is a slow substrate of MCR. In this case the 1.4 Å resolution electron density of MCRCoB6SH indicates that the analogue unexpectedly binds in the substrate channel such that its thiol is virtually in the same position as the thiol of CoBSH (Figure 3B and 4). The hexanoyl chain is oriented so that it takes a shorter route down the substrate channel between carbons 2 and 5 (with the carbonyl carbon labeled as carbon 1) than CoBSH (Figure 4, Supporting Information, Figure S4). This short-cut is not seen in any of the other CoBXSH complex crystal structures, but presumably arises because this CoB6SH binding conformer is energetically more favorable, although it is not clear from the structure why this might be the case. CoB6SH binds very tightly to MCR, with an apparent Ki value of 0.1 μM (3). Water structure in the absence of HSCoM The electron density for the MCRCoB6SH crystal structure only supported the modeling of 50 % bound HSCoM. In the fraction of MCR molecules where HSCoM is absent, the HSCoM binding site is occupied by a network of four water molecules (Supporting Information, Figure S5). Two waters are positioned close to the absent sulfonate oxygen positions of HSCoM. Based on the presence of positive difference electron density, a third water was modeled ligated to the Ni, and refined to a distance of ~2 Å (2.0 Å and 2.1 Å in the two active sites of the ASU) with no distance restraint imposed between the Ni and water. This water is in a similar position as the Ni coordinating sulfonate oxygen of the heterodisulfide product in MCRsilent (Supporting Information, Figure S6 and PDB codes 3m32, 1hbm) (5, 33). The fourth water was in the vicinity of the expected position of a bridging water (W1) seen in other structures (Figure 1, 3A and 3C). Water structure in the absence of CoBSH The 1.45 Å resolution electron density obtained for MCRHSCoM indicates that the substrate channel contains only 30 % CoBSH. Nine ordered waters (W1–W9), along with an acetate ion from the crystallization solution occupy the channel, with the acetate positioned where the phosphothreonine linkage of CoBSH would be (Figure 3C). Presumably 1–2 further waters would replace the acetate under physiological conditions. Other than W3 and W7, the waters form hydrogen bonds with protein (Figure 5). One water (W2) occupies the same site as the CoBSH thiol. Presumably due to the loss of favorable interactions that exist when CoBSH is present, the hydrophobic side-chain of Valα482 adopts a second conformation modeled at 60 % occupancy (Supporting Information, Figure S7). Position of the “bridging” water, W1 The equivalent of W1 has previously been observed in MCRred1-silent and MCRox1-silent crystal structures where, in the presence of CoBSH and HSCoM, it is sited equidistant (3.2 Å) between the two coenzyme thiols (PDB codes: 1hbn, 1hbo, 1hbu) and thus been termed the “bridging water” (Figure 1) (5, 33). However, in the MCRHSCoM structure, due to the presence of the more polarized W2 water, W1 is displaced away from HSCoM to maximize the hydrogen bond interaction with W2 (2.9 Å to W2; 3.5 Å to HSCoM thiol, Figure 5). In the MCRCoB5SH structure that also contained W2, the electron density indicated that this Cedervall et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript repositioning of W1 towards W2 also occurred. In contrast, the MCRCoB6SH structure contained 100% thiol at the CoBSH position, but a partial occupancy of HSCoM (50%). In this case the electron density for W1 indicated it had moved towards the nickel to form an optimal hydrogen bond with a Ni-ligating water that was only present in the absence of HSCoM (3.7 Å to CoB(6)SH thiol; 3.0 Å to Ni-ligating water, Supporting Information, Figure S5). In all structures reported here, W1 (if present) appears to be a sensitive indicator of the relative electronegativity of the Ni-ligated atom to that occupying the position of the CoBSH thiol, and was a useful check in the crystallographic modeling and refinement process. Flexibility in the substrate channel: Alternative protein conformers The binding site of HSCoM (and presumably methyl-SCoM) is more deeply buried within the enzyme, and so it must enter prior to CoBSH for productive chemistry to occur. As binding of CoBSH in the absence of co-substrate would be inhibitory, it was suggested that a conformational change upon methyl-SCoM binding might lower the Kd for CoBSH, and thus promote an ordered mechanism. Compared to the 1hbn MCRox1-silent and 1hbu MCRred1-silent crystal structures, which both have full occupancy HSCoM, the lower occupancy of HSCoM in the 1hbo MCRred1-silent structure was associated with significantly greater flexibility within the channel, and the ability to model a second conformation of a Gly-rich amino acid stretch that formed part of the CoBSH channel. This suggested that methyl-SCoM binding might cause the channel to become more ordered, increasing the affinity of MCR for CoBSH by conformational restriction rather than a switch mechanism where the structure reorganizes from one well-defined conformer to another (33). In the MCRHSCoM data containing 30 % CoBSH and 100% HSCoM, the Fo-Fc difference electron density map at one of the two independent active sites in the ASU contained positive peaks that suggested the presence of an alternate conformation also involving this part of the polypeptide (Supporting Information, Figure S8). Using this as a guide, a similar second conformation involving seven contiguous amino acid residues of the same Gly-rich amino acid stretch (β366–372) could be modeled and refined at 20 % occupancy leaving no residual difference density. Parts of the α′ subunit (α′111–129 and α′237–242) that are in close proximity to this stretch of amino acids also exhibit second conformations, with the main-chain carbonyl of α′243 in van der Waals contact with the B ring of F430 tetrapyrrole (Supporting Information, Figure S9). Modeling these at 20% occupancy accounted for the weak positive Fo-Fc difference electron density peaks observed in these areas. The evidence of alternate conformers in these areas lends support to the proposal that increased flexibility in the substrate channel propagates through the protein (33). The MCRCoB6SH crystal structure contains 50 % CoB6SH, 50% CoBSH and 50% HSCoM. In this case there is no evidence of an alternate loop conformation in either active site of the ASU. However, as CoBSH and CoB6SH combined are at 100% occupancy, it is not surprising their favorable interactions with the substrate channel would reduce conformational disorder, despite the partial occupancy of HSCoM. Analogues longer than CoBSH; CoB8SH and CoB9SH Both analogues could be accommodated in the MCR substrate channel (Figures 3D and 3E). The electron density supported final models containing 50% CoB8SH for MCRCoB8SH (1.8 Å resolution) and 60 % CoB9SH for MCRCoB9SH (1.45 Å resolution). The phosphate head- groups are in identical positions to those of CoBSH, CoB5SH, CoB6SH (Figure 4) (5, 33). Both analogues follow the crystallographically observed chain path of bound CoBSH, with the extra atoms displacing the W1 water and placing the thiols closer to the nickel (Figure 6). CoB9SH does have a second conformer that deviates from the CoBSH path, but the thiol position for this conformer and the CoBSH-tracking conformer are identical (Figure 3E and Cedervall et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supporting Information, Figure S10). Interestingly, the thiol of CoB8SH is not coincident with the CoB9SH carbon that precedes the CoB9SH thiol (Figure 6). CoB8SH is an MCR inhibitor with an apparent Ki of 15 μM (3). CoB9SH has never been tested for inhibition of MCR-catalyzed methane formation, but it is reasonable to assume that it would be an inhibitor. CoBXSH thiol-to-nickel spatial relationship The position of CoBSH in previous crystal structures poses a conundrum (5, 33). In all the proposed catalytic mechanisms, CoBSH must interact with species generated at the nickel. Perplexingly, the sulfur of the CoBSH substrate is 8.8 Å from the Ni(II) in the MCRox1-silent and both MCRred1-silent crystal structures, and 6.4 Å from the thiol of the substrate analogue HSCoM (Figure 1). Modeling studies demonstrated that the addition of a methyl group to HSCoM did not bridge this gap (35, 45, 53). Therefore, a conformational change has been postulated that would enable CoBSH to penetrate deeper into the substrate channel, and thus approach closer to any nickelbound species. The heterodisulfide product in the MCRsilent crystal structure has the CoBSH portion in virtually the same place as in MCRox1-silent, giving no clue to possible structural changes that might occur to facilitate CoBSH reacting with nickel-associated intermediates (5, 33). Trigonometry suggests that if the alkanoyl chain of CoBSH or its analogue is in an extended conformation, each additional unit in the chain would lead to the thiol moving ~1.2 Å towards the Ni. Until this study there have been no crystal structures of CoBSH analogues in complex with MCR, so mechanistic studies using different chain length analogues of CoBSH assumed that shorter analogues would trace the observed path of CoBSH, and longer analogues would penetrate about ~1.2 Å deeper per additional chain unit into the channel. In the case of the shortest analogue CoB5SH, it does indeed follow the path of CoBSH, with the thiol of CoB5SH being 2.8 Å away from the thiol position of CoBSH. However, due to the conformation CoBSH adopts when bound in the substrate channel, the difference in the S-Ni distance is small; the CoB5SH thiol being only 0.5 Å farther from the Ni than CoBSH (8.8 Å for CoBSH vs. 9.3 Å for CoB5SH) (Table 2). This is due to the alkanoyl chain of CoBSH not being in an extended conformation from carbons 4 to 6 (carbon 1 is the carbonyl carbon). CoB6SH on the other hand adopts a conformation that places its thiol in virtually the same position as the thiol of CoBSH (Figure 4 and Table 2). This is consistent with CoB6SH being a substrate. However, the kcat is 1000-fold lower than for CoBSH (3, 35) although its Km value (180 μM, Dey & Ragsdale, in preparation) is similar to that of CoBSH (Km = 75 μM (3)). The reason for this may be that the shorter alkanoyl chain may not enable the analogue thiol to approach the nickel close enough for efficient catalysis, and thus explain why CoB6SH is such a poor substrate. In the case of the longer CoBXSH analogues, the sulfur of CoB8SH is 2.6 Å closer to the Ni ion of F430 than that of CoBSH, and 2.5 Å closer to the thiol of HSCoM (Figure 6 and Table 2). The CoB9SH molecule follows the path of CoBSH, and reaches only a little further into the substrate channel than CoB8SH, with the CoB9SH thiol positioned 2.9 Å closer to the Ni than the thiol of CoBSH (Figure 6 and Table 2). This is only 0.3 Å closer than the distance observed for the CoB8SH thiol, even though they are non-coincident. The distance to the thiol of HSCoM is 2.6 Å closer than that of the substrate, CoBSH; only 0.1 Å closer than the CoB8SH thiol. The two analogue thiols are above an annular hydrophobic aromatic environment created by Pheα330, Tyrα333, Pheα443, Pheβ361 and Tyrβ367 that lies between them and F430 (Figure 6). As a result, penetrating further into the channel may be energetically unfavorable, consistent with the small difference in relative distances between the CoB8SH/CoB9SH thiols and the HSCoM thiol/F430 nickel. The annulus is proposed to be catalytically important in positioning methyl-SCoM and stabilizing the methane product, Cedervall et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript and the tyrosines have been proposed to be proton donors associated with mechanism II (Scheme 2B) (5, 33). Thus, there appear to be three preferential distances for thiols (including that of HSCoM) within the MCR substrate channel; HSCoM at 2.4 Å, CoB6SH/CoBSH at 8.7 – 8.8 Å and CoB8SH/CoB9SH at 5.9 – 6.2 Å from the nickel of F430 (Table 2). Recent ENDOR, high-field continuous and pulse EPR work has identified changes in nickel co-ordination when CoBSH is added to MCRred1c (active Ni(I)-MCRred1 + HSCoM) (14, 15, 18, 31). This generates up to 50% MCRred2 which is comprised of two distinct nickel co- ordination geometries; an axial MCRred2a formally assigned as a Ni(III)-hydride, and a rhombic MCRred2r in which the thiol of HSCoM is a Ni(I) ligand (Supporting Information, Scheme S1). Formation of MCRred2 could also be induced by addition of CoBSH substrate analogues CoBS-CH3 and CoBS-CF3, which have a chain length one unit longer than substrate CoBSH (18, 53). The CoBS-CF3 enabled 19F-ENDOR studies to be performed, and demonstrated that following CoBS-CF3 addition the remaining MCRred1 species had Ni(I)-19F distances of 6.2 – 7.7 Å. This distance range agreed with a CoBS-CF3 model created using the CoBSH position observed in the MCRox1-silent crystal structure (53). However, in the MCRred2 species the Ni(I)-19F distances had shortened indicating a movement of ~ 2 Å towards the nickel (MCRred2a Ni(I)-19 F, 4.0 – 5.5 Å; MCRred2r Ni(I)-19F, 4.5 – 5.7 Å). In the case of the CoB8SH analogue (the closest equivalent to CoBS- CF3 used in this study) the thiol to Ni(II) distance lies between the distance ranges observed in the CoBS-CF3 studies, and so the fluorine(s) of the CoBS-CF3 in MCRred2 might penetrate a little further into the hydrophobic annulus in the MCRred2 species. As the alkanoyl chain of CoBSH is not fully extended it could easily undergo a similar conformation change to that observed in the MCRred2 state. Reaction of MCR Ni(III)-alkyl species’ with CoB8SH and CoB9SH The two longer CoBXSH analogues have been shown to undergo alkylation when reacted with MCRPS, a [Ni(III)-alkyl ↔ Ni(II)-alkylsulfonate radical] formed from reaction of Ni(I)-MCRred1 with bromopropanesulfonate (BPS) (Supporting Information, Scheme S1) (20, 23, 30, 45, 54). BPS is a substrate of MCRred1 in a reaction that involves a rapid CoBSH-independent nucleophilic attack by Ni(I) on BPS to displace bromide and generate MCRPS at a rate ~60-fold faster than generation of methane from CoBSH and methyl- HSCoM (20, 45). Certain thiols can eliminate the propylsulfonate to yield a thioether product and regenerate MCRred1, although at a rate 1000-fold slower than methane formation (45). Both CoB8SH and CoB9SH can react with MCRPS to regenerate MCRred1, but CoBSH cannot. The overall second-order rate constant for the reactivation of MCR by CoB8SH is 160 M−1s−1, whereas for CoB9SH the reaction is more sluggish (12 M−1s−1). CoB9SH might be expected to be closer to the proximal Ni ligand. It was therefore proposed that this caused steric interference and explained why CoB9SH was a poorer reactivator of MCR than CoB8SH. Our study has shown that the thiols of these two analogues are placed such that they are approximately the same distance (~3.7 Å) from the thiol of HSCoM ligated to the Ni atom (Table 2). The Ni(II)-HSCoM bond is 2.4 Å, whereas an Ni(III)-alkyl bond is expected to be ~2 Å (24, 33), thus indicating that a conformational change is required to effect the nucleophilic attack of the CoB8SH and CoB9SH thiols on an alkyl- bound species. It would thus appear that a conformational change, such as observed in MCRred2, is required for this chemistry also (53). A Ni(III)-alkyl species is akin to the first intermediate in mechanism I of MCR-catalyzed methane formation, Ni(III)-methyl (MCRMe, Supporting Information, Scheme S1, Scheme 2A) (11, 27). MCRMe has been shown to be capable of generating MCRred1 and methyl- SCoM upon addition of HSCoM (which is the reverse of mechanism I, step 1, Scheme 2A); Cedervall et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript similar chemistry to the observed formation of a thioether product from the Ni(III)-alkyl. Further addition of CoBSH following HSCoM treatment of MCRMe led to methane and heterodisulfide formation, the natural products of methanogenesis. Although this lends credence to mechanism I, it should be noted that like MCRPS, MCRMe in these experiments was generated artificially. However, the MCRCoB9SH crystal structure demonstrates that the two additional methylene units in the alkanoyl chain c.f. CoBSH, do not necessarily translate into direct interaction of the thiol with the nickel proximal ligand. However, this could represent the favorable position for a CoBSH thiol interacting with the methyl group of methyl-SCoM. Just as the alkanoyl chain of CoB6SH has a more extended conformation than CoBSH in the substrate channel, CoBSH could also adopt a more extended conformation so that its thiol was in a similar position as the thiol of CoB8SH, priming it for reaction with a nickel bound species. If a significant conformational change is required early in MCR-catalyzed chemistry, which would be a requirement of mechanism I, catalysis may well involve a rearrangement of the aromatic amino acid annulus due to the presence of the methyl of methyl-SCoM, and this might enable deeper penetration of CoBSH into MCR (Figure 6). All the crystal structures in this study, and those solved previously, are of the inactive Ni(II)-MCR, which disfavors close approach to the nickel in the absence of Ni(I)-bound methyl-SCoM, even in the case of CoB9SH. Conclusion The goal of this study was to induce structural changes within the substrate channel and active site of MCR using analogues of coenzyme CoBSH. It was hoped that this would shed light on the nature of conformational changes that have been proposed to occur in MCR catalysis. We have shown that that the CoBXSH analogues do not lead to any significant conformational changes within the context of inactive Ni(II)-MCR. Therefore, it may be that methyl-SCoM is the key coenzyme, in combination with a nickel oxidation state of 1 (and 3), that triggers a conformational change bringing the thiol of CoBSH closer to the nickel. Thus, the crystal structure of the Ni(I)-methyl-SCoM/MCR complex may be required to structurally define the conformational changes required for MCR-mediated chemistry. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments X-ray data were collected at beam-line 14-BM-C and photoreduction studies at 14-ID-B, BioCARS at the Advanced Photon Source (APS), Argonne National Laboratory, Argonne, IL. We thank Vukica Srajer and Yu- Sheng Chen for valuable assistance during data collection. Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Basic Energy Sciences, Office of Science under Contract No. DE- AC02-06CH11357. Use of the BioCARS Sector 14 was supported by the National Institutes of Health, National Center for Research Resources, under grant number RR007707. Computer resources were provided by the Basic Sciences Computing Laboratory of the University of Minnesota Supercomputing Institute, and we thank Can Ergenekan for his support. We also thank Ed Hoeffner at the Kahlert Structural Biology Laboratory (KSBL) at the University of Minnesota. Use of the KSBL was supported by a Minnesota Partnership for Biotechnology and a Medical Genomics Grant SPAP-05-0013-P-FY06. References 1. Thauer RK. Biochemistry of methanogenesis: a tribute to Marjory Stephenson. Microbiology. 1998; 144:2377–2406. [PubMed: 9782487] 2. Thauer RK, Shima S. Methane as fuel for anaerobic microorganisms. Ann N Y Acad Sci. 2008; 1125:158–170. [PubMed: 18096853] Cedervall et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 3. Ellermann J, Hedderich R, Bocher R, Thauer RK. The final step in methane formation. Investigations with highly purified methyl-CoM reductase (component C) from Methanobacterium thermoautotrophicum (strain Marburg). Eur J Biochem. 1988; 172:669–677. [PubMed: 3350018] 4. Ellefson WL, Wolfe RS. Component C of the methylreductase system of Methanobacterium. J Biol Chem. 1981; 256:4259–4262. [PubMed: 6783657] 5. Ermler U, Grabarse W, Shima S, Goubeaud M, Thauer RK. Crystal structure of methyl-coenzyme M reductase: the key enzyme of biological methane formation. Science. 1997; 278:1457–1462. [PubMed: 9367957] 6. Diekert G, Gilles HH, Jaenchen R, Thauer RK. Incorporation of 8 succinate per mol nickel into factors F430 by Methanobacterium thermoautotrophicum. Arch Microbiol. 1980; 128:256–262. [PubMed: 7212929] 7. Diekert G, Jaenchen R, Thauer RK. Biosynthetic evidence for a nickel tetrapyrrole structure of factor F430 from Methanobacterium thermoautotrophicum. FEBS Letters. 1980; 119:118–120. [PubMed: 7428919] 8. Whitman WB, Wolfe RS. Presence of nickel in Factor F430 from Methanobacterium bryantii. Biochem Biophys Res Comm. 1980; 92:1196–1201. [PubMed: 7370029] 9. Albracht SPJ, Ankel-Fuchs D, Böcher R, Ellermann J, Moll J, van der Zwann JW, Thauer RK. Five new EPR signals assigned to nickel in methyl-coenzyme M reductase from Methanobacterium thermoautotrophicum, strain Marburg. Biochim Biophys Acta. 1988; 955:86–102. 10. Dey M, Kunz RC, Lyons DM, Ragsdale SW. Characterization of alkyl-nickel adducts generated by reaction of methyl-coenzyme m reductase with brominated acids. Biochemistry. 2007; 46:11969– 11978. [PubMed: 17902704] 11. Dey M, Telser J, Kunz RC, Lees NS, Ragsdale SW, Hoffman BM. Biochemical and spectroscopic studies of the electronic structure and reactivity of a methyl-Ni species formed on methyl- coenzyme M reductase. J Am Chem Soc. 2007; 129:11030–11032. [PubMed: 17711283] 12. Duin EC, Cosper NJ, Mahlert F, Thauer RK, Scott RA. Coordination and geometry of the nickel atom in active methyl-coenzyme M reductase from Methanothermobacter marburgensis as detected by X-ray absorption spectroscopy. J Biol Inorg Chem. 2003; 8:141–148. [PubMed: 12459909] 13. Duin EC, Signor L, Piskorski R, Mahlert F, Clay MD, Goenrich M, Thauer RK, Jaun B, Johnson MK. Spectroscopic investigation of the nickel-containing porphinoid cofactor F(430). Comparison of the free cofactor in the (+)1, (+)2 and (+)3 oxidation states with the cofactor bound to methyl- coenzyme M reductase in the silent, red and ox forms. J Biol Inorg Chem. 2004; 9:563–576. [PubMed: 15160314] 14. Finazzo C, Harmer J, Bauer C, Jaun B, Duin EC, Mahlert F, Goenrich M, Thauer RK, Van Doorslaer S, Schweiger A. Coenzyme B induced coordination of coenzyme M via its thiol group to Ni(I) of F430 in active methyl-coenzyme M reductase. J Am Chem Soc. 2003; 125:4988–4989. [PubMed: 12708843] 15. Finazzo C, Harmer J, Jaun B, Duin EC, Mahlert F, Thauer RK, Van Doorslaer S, Schweiger A. Characterization of the MCRred2 form of methyl-coenzyme M reductase: a pulse EPR and ENDOR study. J Biol Inorg Chem. 2003; 8:586–593. [PubMed: 12624730] 16. Goubeaud M, Schreiner G, Thauer RK. Purified methyl-coenzyme-M reductase is activated when the enzyme-bound coenzyme F430 is reduced to the nickel(I) oxidation state by titanium(III) citrate. Eur J Biochem. 1997; 243:110–114. [PubMed: 9030728] 17. Harmer J, Finazzo C, Piskorski R, Bauer C, Jaun B, Duin EC, Goenrich M, Thauer RK, Van Doorslaer S, Schweiger A. Spin density and coenzyme M coordination geometry of the ox1 form of methyl-coenzyme M reductase: a pulse EPR study. J Am Chem Soc. 2005; 127:17744–17755. [PubMed: 16351103] 18. Harmer J, Finazzo C, Piskorski R, Ebner S, Duin EC, Goenrich M, Thauer RK, Reiher M, Schweiger A, Hinderberger D, Jaun B. A nickel hydride complex in the active site of methyl- coenzyme m reductase: implications for the catalytic cycle. J Am Chem Soc. 2008; 130:10907– 10920. [PubMed: 18652465] Cedervall et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 19. Hinderberger D, Ebner S, Mayr S, Jaun B, Reiher M, Goenrich M, Thauer RK, Harmer J. Coordination and binding geometry of methyl-coenzyme M in the red1m state of methyl- coenzyme M reductase. J Biol Inorg Chem. 2008; 13:1275–1289. [PubMed: 18712421] 20. Kunz RC, Horng YC, Ragsdale SW. Spectroscopic and kinetic studies of the reaction of bromopropanesulfonate with methyl-coenzyme M reductase. J Biol Chem. 2006; 281:34663– 34676. [PubMed: 16966321] 21. Mahlert F, Bauer C, Jaun B, Thauer RK, Duin EC. The nickel enzyme methyl-coenzyme M reductase from methanogenic archaea: In vitro induction of the nickel-based MCR-ox EPR signals from MCR-red2. J Biol Inorg Chem. 2002; 7:500–513. [PubMed: 11941508] 22. Mahlert F, Grabarse W, Kahnt J, Thauer RK, Duin EC. The nickel enzyme methyl-coenzyme M reductase from methanogenic archaea: in vitro interconversions among the EPR detectable MCR- red1 and MCR-red2 states. J Biol Inorg Chem. 2002; 7:101–112. [PubMed: 11862546] 23. Rospert S, Voges M, Berkessel A, Albracht SP, Thauer RK. Substrate-analogue-induced changes in the nickel-EPR spectrum of active methyl-coenzyme-M reductase from Methanobacterium thermoautotrophicum. Eur J Biochem. 1992; 210:101–107. [PubMed: 1332856] 24. Sarangi R, Dey M, Ragsdale SW. Geometric and electronic structures of the Ni(I) and methyl- Ni(III) intermediates of methyl-coenzyme M reductase. Biochemistry. 2009; 48:3146–3156. [PubMed: 19243132] 25. Tang Q, Carrington PE, Horng YC, Maroney MJ, Ragsdale SW, Bocian DF. X-ray absorption and resonance Raman studies of methyl-coenzyme M reductase indicating that ligand exchange and macrocycle reduction accompany reductive activation. J Am Chem Soc. 2002; 124:13242–13256. [PubMed: 12405853] 26. Telser J, Davydov R, Horng YC, Ragsdale SW, Hoffman BM. Cryoreduction of methyl-coenzyme M reductase: EPR characterization of forms, MCR(ox1) and MCR (red1). J Am Chem Soc. 2001; 123:5853–5860. [PubMed: 11414817] 27. Yang N, Reiher M, Wang M, Harmer J, Duin EC. Formation of a nickel-methyl species in methyl- coenzyme M reductase, an enzyme catalyzing methane formation. J Am Chem Soc. 2007; 129:11028–11029. [PubMed: 17711279] 28. Albracht SPJ, Ankelfuchs D, Vanderzwaan JW, Fontijn RD, Thauer RK. A New Electron- Paramagnetic-Res Signal of Nickel in Methanobacterium-Thermoautotrophicum. Biochim Biophys Acta. 1986; 870:50–57. 29. Telser J, Horng YC, Becker DF, Hoffman BM, Ragsdale SW. On the assignment of nickel oxidation states of the Ox1, Ox2 forms of methyl-coenzyme M reductase. J Am Chem Soc. 2000; 122:182–183. 30. Hinderberger D, Piskorski RR, Goenrich M, Thauer RK, Schweiger A, Harmer J, Jaun B. A nickel- alkyl bond in an inactivated state of the enzyme catalyzing methane formation. Angewandte Chemie-International Ed. 2006; 45:3602–3607. 31. Kern DI, Goenrich M, Jaun B, Thauer RK, Harmer J, Hinderberger D. Two sub-states of the red2 state of methyl-coenzyme M reductase revealed by high-field EPR spectroscopy. J Biol Inorg Chem. 2007; 12:1097–1105. [PubMed: 17690920] 32. Becker DF, Ragsdale SW. Activation of methyl-SCoM reductase to high specific activity after treatment of whole cells with sodium sulfide. Biochemistry. 1998; 37:2639–2647. [PubMed: 9485414] 33. Grabarse W, Mahlert F, Duin EC, Goubeaud M, Shima S, Thauer RK, Lamzin V, Ermler U. On the mechanism of biological methane formation: structural evidence for conformational changes in methyl-coenzyme M reductase upon substrate binding. J Mol Biol. 2001; 309:315–330. [PubMed: 11491299] 34. Grabarse W, Mahlert F, Shima S, Thauer RK, Ermler U. Comparison of three methyl-coenzyme M reductases from phylogenetically distant organisms: unusual amino acid modification, conservation and adaptation. J Mol Biol. 2000; 303:329–344. [PubMed: 11023796] 35. Horng YC, Becker DF, Ragsdale SW. Mechanistic studies of methane biogenesis by methyl- coenzyme M reductase: evidence that coenzyme B participates in cleaving the C-S bond of methyl-coenzyme M. Biochemistry. 2001; 40:12875–12885. [PubMed: 11669624] Cedervall et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 36. Berkessel A. Methyl-Coenzyme-M Reductase - Model Studies on Pentadentate Nickel-Complexes and a Hypothetical Mechanism. Bioorg Chem. 1991; 19:101–115. 37. Jaun B. Coenzyme-F430 from Methanogenic Bacteria - Oxidation of F430 Pentamethyl Ester to the Ni(Iii) Form. Helvetica Chimica Acta. 1990; 73:2209–2217. 38. Signor L, Knuppe C, Hug R, Schweizer B, Pfaltz A, Jaun B. Methane formation by reaction of a methyl thioether with a photo-excited nickel thiolate - A process mimicking methanogenesis in archaea. Chemistry-a European Journal. 2000; 6:3508–3516. 39. Chen SL, Pelmenschikov V, Blomberg MR, Siegbahn PE. Is there a Ni-methyl intermediate in the mechanism of methyl-coenzyme M reductase? J Am Chem Soc. 2009; 131:9912–9913. [PubMed: 19569621] 40. Pelmenschikov V, Blomberg MRA, Siegbahn PEM, Crabtree RH. A mechanism from quantum chemical studies for methane formation in methanogenesis. J Am Chem Soc. 2002; 124:4039– 4049. [PubMed: 11942842] 41. Pelmenschikov V, Siegbahn PE. Catalysis by methyl-coenzyme M reductase: a theoretical study for heterodisulfide product formation. J Biol Inorg Chem. 2003; 8:653–662. [PubMed: 12728361] 42. Duin EC, McKee ML. A new mechanism for methane production from methyl-coenzyme M reductase as derived from density functional calculations. J Phys Chem. 2008; B 112:2466–2482. 43. Bobik TA, Wolfe RS. Physiological importance of the heterodisulfide of coenzyme M and 7- mercaptoheptanoylthreonine phosphate in the reduction of carbon dioxide to methane in Methanobacterium. Proc Natl Acad Sci U S A. 1988; 85:60–63. [PubMed: 3124103] 44. Goenrich M, Duin EC, Mahlert F, Thauer RK. Temperature dependence of methyl-coenzyme M reductase activity and of the formation of the methyl-coenzyme M reductase red2 state induced by coenzyme B. J Biol Inorg Chem. 2005; 10:333–342. [PubMed: 15846525] 45. Kunz RC, Dey M, Ragsdale SW. Characterization of the Thioether Product Formed from the Thiolytic Cleavage of the Alkyl-Nickel Bond in Methyl-Coenzyme M Reductase. Biochemistry. 2008; 47:2661–2667. [PubMed: 18220418] 46. Noll KM, Donnelly MI, Wolfe RS. Synthesis of 7-mercaptoheptanoylthreonine phosphate and its activity in the methylcoenzyme M methylreductase system. J Biol Chem. 1987; 262:513–515. [PubMed: 3100513] 47. Olson KD, McMahon CW, Wolfe RS. Photoactivation of the 2-(methylthio)ethanesulfonic acid reductase from Methanobacterium. Proc Natl Acad Sci U S A. 1991; 88:4099–4103. [PubMed: 1903534] 48. Zehnder AJ, Wuhrmann K. Titanium (III) citrate as a nontoxic oxidation-reduction buffering system for the culture of obligate anaerobes. Science. 1976; 194:1165–1166. [PubMed: 793008] 49. Gunsalus RP, Romesser JA, Wolfe RS. Preparation of coenzyme M analogues and their activity in the methyl coenzyme M reductase system of Methanobacterium thermoautotrophicum. Biochemistry. 1978; 17:2374–2377. [PubMed: 98178] 50. Otwinowski Z, Minor W. Processing of X-ray Diffraction Data Collected in Oscillation Mode. Methods in Enzymology: Macromolecular Crystallography, part A. 1997; 276:307–326. 51. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr. 1994; 50:760–763. [PubMed: 15299374] 52. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] 53. Ebner S, Jaun B, Goenrich M, Thauer RK, Harmer J. Binding of coenzyme B induces a major conformational change in the active site of methyl-coenzyme M reductase. J Am Chem Soc. 2010; 132:567–575. [PubMed: 20014831] 54. Goenrich M, Mahlert F, Duin EC, Bauer C, Jaun B, Thauer RK. Probing the reactivity of Ni in the active site of methyl-coenzyme M reductase with substrate analogues. J Biol Inorg Chem. 2004; 9:691–705. [PubMed: 15365904] Cedervall et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. The active site and substrate channel in the MCRox1-silent crystal structure (PDB code 1hbn) (9). Coenzyme F430, CoBSH and HSCoM are drawn as stick colored by atom (carbon: dark grey). The nickel is displayed as a green sphere, water as a red sphere. Interactions are drawn as dashed lines, and the corresponding distance is indicted in Angstroms (Å). The path of the substrate channel was defined in the absence of F430, CoBSH, HSCoM and water, with the surface closest to the viewer cut away. The figure was generated using PyMOL (http://www.pymol.org). Cedervall et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Drawing of CoBSH analogues; (A) N-5-mercaptopentanoylthreonine phosphate (CoB5SH); (B) N-6-mercaptohexanoylthreonine phosphate (CoB6SH); (C) N-8- mercaptooctanoylthreonine phosphate (CoB8SH); (D) N-9-mercaptononanoylthreonine phosphate (CoB9SH). Cedervall et al. Page 16 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. The active sites and substrate channels of the MCR crystal structures.; (A) MCRCoB5SH; (B) MCRCoB6SH; (C) MCRHSCoM; (D) MCRCoB8SH; (E) MCRCoB9SH. 2Fo-Fc electron density map around the CoBXSH analogues, waters in the CoBSH binding part of the channel and the acetate ion (contoured at 1σ) is shown as a blue mesh. The protein is drawn as cartoon. CoBXSH and acetate are drawn as stick and colored by atom (carbon: CoB5SH orange; CoB6SH pale yellow; CoB8SH light blue; CoB9SH magenta; acetate white). Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere, and waters as red spheres. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 17 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. Overlay of CoBSH (from PDB code 1hbn) and the different CoBSH analogues. CoBXSH are drawn as stick with the thiol represented by a sphere, and colored CoB5SH orange; CoB6SH pale yellow; CoBSH light green; CoB8SH light blue; CoB9SH magenta. The protein is drawn as cartoon. Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 18 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Hydrogen bonding diagram for the water structure modeled in MCRHSCoM. The water molecules are named as in Figure 3C (W1–W9); WA, WB and WC are water molecules that are present in all structures (i.e. in concert with the substrate CoBSH and the CoBSH analogues). Interactions between surrounding residues and the water molecules are drawn as dashed lines, and the corresponding distance is indicated in Angstroms (Å). Cedervall et al. Page 19 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. Stereo image of the annulus of aromatic amino acids distal of coenzyme F430. The protein is drawn as cartoon with the side-chains of the aromatic residues drawn as white stick. CoBSH, (from PDB code 1hbn (9)), CoB8SH and CoB9SH are drawn as stick with the thiols represented by spheres, and colored CoBSH light green; CoB8SH light blue; CoB9SH magenta. Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 20 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. Reaction catalyzed by methyl-coenzyme M reductase Cedervall et al. Page 21 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 2. Two of the proposed catalytic mechanisms for methyl-coenzyme M reductase; (A) mechanism I; (B) mechanism II. Cedervall et al. Page 22 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 23 Table 1 X-ray Data Collection, Processing and Refinement Statistics Data collection and processing statistics Name of data set MCRCoB5SH MCRCoB6SH MCRHSCoM MCRCoB8SH MCRCoB9SH Measured reflections 1969388 2427498 1440665 1160543 1425506 Unique reflections 553755 446253 405349 211803 401701 Resolution (Å) a 50.0–1.30 (1.35–1.30) 50.0–1.40 (1.45–1.40) 50.0–1.45 (1.50–1.45) 50.0–1.80 (1.86–1.80) 50.0–1.45 (1.50–1.45) Completeness (%) a 97.1 (78.1) 99.9 (100.0) 99.5 (99.7) 99.8 (100.0) 98.1 (95.4) R-sym (%) a,b 5.5 (32.9) 7.3 (44.7) 6.2 (44.0) 8.4 (47.7) 5.6 (42.5) I/σI a 22.3 (3.6) 20.4 (4.0) 20.2 (3.2) 21.8 (3.9) 24.3 (3.2) Space group P21 P21 P21 P21 P21 Refinement and model building statistics Resolution (Å) a 20.49–1.30 (1.33–1.30) 19.89–1.40 (1.44–1.40) 20.15–1.45 (1.49–1.45) 19.93–1.80 (1.84–1.80) 20.07–1.45 (1.48–1.45) No. of reflection in working set a 525817 (30239) 423854 (25833) 384868 (25791) 201128 (11193) 381474 (23611) No. of reflection in test set a 27777 (1576) 22348 (1331) 20362 (1319) 10625 (557) 20163 (1210) R-work (%) c 14.32 13.04 13.47 14.95 13.58 R-free (%) d 16.56 15.53 16.22 19.54 16.44 ESU (Å) R-work/R-free 0.044/0.046 0.049/0.051 0.056/0.059 0.121/0.119 0.057/0.060 No. protein atoms 20087 19960 20265 19750 20036 No. coenzyme atoms 218 220 180 224 272 No. ligand atoms 37 62 52 26 49 No. water molecules 2443 2352 2516 1893 2432 RMS bond lengths (Å) 0.033 0.033 0.032 0.028 0.032 bond angles (deg.) 2.693 2.625 2.468 2.059 2.549 Ramachandran plot (%) favored 97.8 97.5 97.6 97.2 97.7 allowed 2.1 2.4 2.3 2.7 2.1 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 24 disallowed 0.1 0.1 0.1 0.1 0.1 Average B-factor (Å2) protein 12.42 13.35 12.12 17.22 12.73 coenzymes 8.20 9.24 7.25 11.24 8.27 ligands 31.95 35.48 28.29 33.76 32.92 waters 22.95 24.89 23.85 26.79 24.09 over all 13.54 14.57 13.40 18.02 13.93 Occupancy of HSCoM per active site (%)e 90/90 50/50 100/100 90/90 90/85 Occupancy of CoBSH per active site (%) e 50/50 50/50 30/30 50/50 40/40 CoBSH analogue, occupancy per active site (%) e CoB5SH, 50/50 CoB6SH, 50/50 CoB8SH, 50/50 CoB9SH, 60/60 Other molecule, occupancy per active site (%) e Acetate, 70/70 aValues in brackets correspond to the highest resolution shell. bR-sym = ΣhklΣNj=1|Ihkl−Ihkl(j)|/ΣhklN*Ihkl, sum over all reflections and all observations N, with Ihkl(j) intensity of the jth observation of reflection hkl and Ihkl mean intensity of the reflection hkl. cR-work = Σ||Fo| − |Fc||/Σ|Fo|, where |Fo| = observed structure factor amplitude and |Fc| = calculated structure factor amplitude. dR-free, R-factor based on 5% of the data excluded from refinement. eOccupancy of model in each of the two crystallographically independent active sites in the ASU Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 25 Table 2 Distances from analogue thiols. CoBXS - SCoM distance (Å) CoBXS - Ni distance (Å) CoB5SH 7.11/7.11a 9.30/9.30 CoB6SH 6.26/6.26 8.70/8.70 CoB7SH (substrate) b 6.37/6.39 8.73/8.77 CoB8SH 3.75/3.78 6.16/6.17 CoB9SH 3.71/3.68 5.96/5.91 aDistances in the two crystallographically independent active sites in the ASU bDistances in the 1.16 Å resolution MCRox1-silent structure (PDB code 1hbn)(33) Biochemistry. Author manuscript; available in PMC 2011 September 7.
3M30
Structural Insight into Methyl-Coenzyme M Reductase Chemistry using Coenzyme B Analogues
Structural Insight into Methyl-Coenzyme M Reductase Chemistry using Coenzyme B Analogues,†,‡ Peder E. Cedervall§,#, Mishtu Dey||,#,⊥, Arwen R. Pearson§,⊥, Stephen W. Ragsdale||, and Carrie M. Wilmot*,||,§ § Department of Biochemistry, Molecular Biology and Biophysics, University of Minnesota, Minneapolis, Minnesota 55455 || Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109 Abstract Methyl-coenzyme M reductase (MCR) catalyzes the final and rate-limiting step in methane biogenesis; the reduction of methyl-coenzyme M (methyl-SCoM) by coenzyme B (CoBSH) to methane and a heterodisulfide (CoBS-SCoM). Crystallographic studies show that the active site is deeply buried within the enzyme, and contains a highly reduced nickel-tetrapyrrole, coenzyme F430. Methyl-SCoM must enter the active site prior to CoBSH, as species derived from analogues of methyl-SCoM are always observed bound to the F430 nickel in the deepest part of the 30 Å long substrate channel that leads from the protein surface to the active site. The seven-carbon mercaptoalkanoyl chain of CoBSH binds within a 16 Å predominantly hydrophobic part of the channel close to F430, with the CoBSH thiolate lying closest to the nickel at a distance of 8.8 Å. It has previously been suggested that binding of CoBSH initiates catalysis by inducing a conformational change that moves methyl-SCoM closer to the nickel promoting cleavage of the C- S bond of methyl-SCoM. In order to better understand the structural role of CoBSH early in the MCR mechanism, we have determined crystal structures of MCR in complex with four different CoBSH analogues; pentanoyl-, hexanoyl-, octanoyl- and nonanoyl- derivatives of CoBSH (CoB5SH, CoB6SH, CoB8SH and CoB9SH respectively). The data presented here reveal that the shorter CoB5SH mercaptoalkanoyl chain overlays with that of CoBSH, but terminates two units short of the CoBSH thiolate position. In contrast, the mercaptoalkanoyl chain of CoB6SH adopts a different conformation, such that its thiolate is coincident with the position of the CoBSH thiolate. †This work was supported by a Department of Energy grant DE-FG02-08ER15931 to S.W.R and supplement to C.M.W. and a Minnesota Partnership for Biotechnology and Medical Genomics grant SPAP-05-0013-P-FY06. ‡Co-ordinates and structure factors have been deposited in the Protein Data Bank as entries 3m1v (MCRHSCoM); 3m2r (MCRCoB5SH); 3m2u (MCRCoB6SH); 3m2v (MCRCoB8SH); 3m30 (MCRCoB9SH). *Address correspondence to: Carrie M. Wilmot, Tel: (612) 624-2406, wilmo004@umn.edu. ⊥Current addresses: Mishtu Dey, Department of Chemistry, Massachusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, Massachusetts 02139; Arwen Pearson, Astbury Centre for Structural Molecular Biology, Institute for Molecular and Cellular Biology, Astbury Building. Leeds, LS2 9JT, U.K. #These authors contributed equally to this work. Supporting Informational material is available free of charge via the Internet at http://pubs.acs.org and contains the following: MATERIAL AND METHODS for single crystal UV-visible microspectrophotometry, X-ray photoreduction experiment and X-ray crystallography of the MCR-heterodisulfide product complex (MCRCoBSH + methyl-SCoM); RESULTS AND DISCUSSION for redox changes and MCRCoBSH + methyl-SCoM crystal structure; Table S1, X-ray Data Collection, Processing and Refinement Statistics for MCRCoBSH + methyl-SCoM; Figure S1, use of Fo-Fc electron density in modelling MCRCoB8SH; Figure S2, illustration of electron density quality of MCRCoB5SH; Figure S3, solution and single crystal UV-visible spectra; Figure S4, modeling of CoB6SH and CoBSH into the electron density of MCRCoB6SH; Figure S5, partially occupied HSCoM in MCRCoB6SH; Figure S6, the active site and substrate channel of MCRCoBSH + methyl-SCoM; Figure S7, alternative conformation of Valα482 in MCRHSCoM; Figure S8, the two conformations of the gly-rich loop in MCRHSCoM; Figure S9, propagation of conformational changes in MCRHSCoM; Figure S10, the two conformations of CoB9SH in MCRCoB9SH; Figure S11, EPR spectra of MCRred1 sample; Scheme S1, scheme of the characterized forms of MCR. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 September 7. Published in final edited form as: Biochemistry. 2010 September 7; 49(35): 7683–7693. doi:10.1021/bi100458d. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript This is consistent with the observation that CoB6SH is a slow substrate. A labile water in the substrate channel was found to be a sensitive indicator for the presence of CoBSH and HSCoM. The longer CoB8SH and CoB9SH analogues can be accommodated in the active site through exclusion of this water. These analogues react with Ni(III)-methyl; a proposed MCR catalytic intermediate of methanogenesis. The CoB8SH thiolate is 2.6 Å closer to the nickel than that of CoBSH, but the additional carbon of CoB9SH only decreases the nickel thiolate distance a further 0.3 Å. Although the analogues did not induce any structural changes in the substrate channel, the thiolates appeared to preferentially bind at two distinct positions in the channel; one being the previously observed CoBSH thiolate position, and the other being at a hydrophobic annulus of residues that lines the channel proximal to the nickel. INTRODUCTION Methanogenic archaea are organisms that under strict anaerobic conditions derive energy by reducing compounds such as carbon dioxide, methylamine, acetate and methanol, to methane (1, 2). The global production of methane by these organisms is estimated at one billion tons annually. Microbially produced methane is not only a potential source of renewable energy but also a potent greenhouse gas, and as such study of this process has environmental ramifications. In these microorganisms, methyl-coenzyme M reductase (MCR)1 is the enzyme that catalyzes the final step in methanogenesis, in which the substrates methyl-coenzyme M (methyl-SCoM, 2-(methylthio)ethanesulfonate) and coenzyme B (CoBSH, N-7-mercaptoheptanoylthreonine phosphate) are converted to methane and a heterodisulfide (CoBS-SCoM) (Scheme 1) (3). MCR is a 300 kDa protein with six subunits arranged in a α2β2γ2 oligomer (4). The known crystal structures show that MCR has two active sites approximately 50 Å apart that are deeply buried within the enzyme (5). The active site pocket is comprised of residues from subunits α, α′, β and γ, with a 30 Å long substrate channel leading to the enzyme surface (Figure 1). At the heart of the active site pocket is coenzyme F430, which is a highly reduced nickel-containing tetrapyrrole (6–8). Currently sixteen distinct enzymatic and complexed states of MCR have been spectroscopically characterized (Supporting Information, Scheme S1) (9–31). In the resting active state of the enzyme, denoted MCRred1, the redox-active nickel of F430 is present in the Ni(I) state (9, 16, 32). MCR is extremely oxygen sensitive and upon oxygen exposure the enzyme enters an inactive Ni(II) state, denoted MCRred1-silent (6). In this state it cannot be converted back to the active Ni(I) form by any known reducing agent making this a challenging system to study. Additional complications involve the tight association of coenzymes to purified MCR that are not easily displaced as demonstrated by X-ray crystallographic and kinetic studies (5, 33–35). Despite the fact that MCR has been studied for decades, no true catalytic intermediate has been observed, and the actual mechanism remains elusive. Currently three general mechanistic schemes for the enzymatic reaction have been proposed, each of which posit different chemistry to initiate catalysis. Mechanism I involves Ni(I) acting as a nucleophile in an SN2-type reaction that generates Ni(III)-methyl as an intermediate (Scheme 2A) (35– 38). Mechanism II starts with methyl-SCoM undergoing homolytic cleavage at the Ni(I) to generate a methyl radical and a Ni(II)-SCoM species (Scheme 2B) (39–41). A more recently proposed mechanism III suggests protonation of coenzyme F430 promotes reductive cleavage of the methyl-SCoM thioether bond (42). 1Abbreviations: MCR, Methyl-coenzyme M reductase; methyl-SCoM, methyl-coenzyme M; CoBSH, coenzyme B; HSCoM, coenzyme M; CoBS-SCoM, heterodisulfide of coenzyme B and coenzyme M; APS, Advanced Photon Source; ASU, asymmetric unit; BPS, bromopropanesulfonate. Cedervall et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Due to the stringent requirement to exclude O2, the available MCR crystal structures are all in the inactive Ni(II) state. All but one contain CoBSH and HSCoM (demethylated methyl- SCoM, an inhibitor and substrate analogue) in the active site (PDB codes 1hbn, 1hbo, 1hbu, 1e6y, 1e6v) (5, 33, 34). Another crystal structure has bound heterodisulfide product, CoBS- SCoM (MCRsilent, PDB code 1hbm, Scheme 1 and Supporting Information, Scheme S1) (5, 33). All these structures reveal that both substrates access the active site through the same channel (Figure 1). The binding site of HSCoM (and presumably methyl-SCoM) is more deeply buried within the enzyme, and so it must enter prior to CoBSH for productive chemistry to occur. As binding of CoBSH in the absence of co-substrate would be inhibitory, it was suggested that a conformational change upon methyl-SCoM binding might lower the Kd for CoBSH, and thus promote an ordered mechanism. Furthermore it has been suggested that CoBSH binding induces a conformational change that brings the methyl- SCoM substrate into closer proximity to the nickel, and this promotes C-S bond cleavage. To investigate the proposed structural role of CoBSH in initiating catalysis, we have solved the X-ray crystal structures of MCR in complex with four different CoBSH analogues. CoBSH has a heptanoyl moiety linked to the thiol group, and the analogues are pentanoyl-, hexanoyl-, octanoyl- or nonanoyl-containing derivatives of CoBSH (CoB5SH, CoB6SH, CoB8SH and CoB9SH respectively; Figure 2) (3, 35, 43–47). In addition, we present a structure in which the substrate channel predominantly lacks either CoBSH or heterodisulfide product. MATERIALS AND METHODS Materials The organism Methanothemobacter marburgensis (catalog OCM82) was obtained from the Oregon Collection of Methanogens (Portland, OR). All buffers and media reagents were obtained from Sigma-Aldrich (St. Louis, MO). The gases N2 (99.98%), H2/CO2 (80%/20%), and ultra high purity H2 (99.999%) were obtained from Cryogenic Gases (Grand Rapids, MI). Ti(III) citrate solutions were prepared from a stock solution of 246 mM Ti(III) citrate, which was synthesized by adding sodium citrate to Ti(III) trichloride (30 wt % solution in 2 N hydrochloric acid) (Acros Organics, Morris Plains, NJ) under anaerobic conditions and adjusting pH to 7.0 with sodium bicarbonate (48). The concentration of Ti(III) citrate was determined by titrating against a solution of methyl viologen. Synthesis of methyl-SCoM, CoB5SH, CoB6SH, CoBSH, CoB8SH, and CoB9SH Methyl-SCoM was prepared from HSCoM and methyl iodide (49). The homodisulfides, CoB5S-SCoB5, CoB6S-SCoB6, CoBS-SCoB, CoB8S-SCoB8, CoB9S-SCoB9 were prepared as described from 5-bromovaleric acid, 6-bromohexanoic acid (Sigma-Aldrich, St. Louis, MO), 7-bromoheptanoic acid (Karl Industries, Aurora, OH), 8-bromooctanoic acid, and 9- bromononanoic acid (Matrix Scientific, Columbia, SC), respectively (43, 46). The free thiol forms of CoB5SH, CoB6SH, CoBSH, CoB8SH, and CoB9SH were generated by the reduction of the homodisulfides as previously described (45). The purity of the CoBSH analogues was determined by 1H NMR spectroscopy. All compounds synthesized were stored in a Vacuum Atmospheres chamber maintained at an oxygen level below 1 ppm, as monitored continually with an oxygen analyzer (model 317, Teledyne Analytical Instruments, City of Industry, CA) until use. M. marburgensis Growth and MCRred1 Purification Buffer preparations and all manipulations were performed under strict anaerobic conditions in a Vacuum Atmospheres chamber maintained at an oxygen level below 1 ppm, as monitored continually with an oxygen analyzer (model 317, Teledyne Analytical Instruments, City of Industry, CA). MCRred1 was isolated from M. marburgensis cultured on Cedervall et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript H2/CO2 (80%/20%) at 65 °C in a 14-L fermentor (New Brunswick Scientific Co., Inc. New Brunswick, NJ). Culture media were prepared as previously described (20, 24). MCRred1 was generated in vivo and purified as described previously (20). The purification procedure routinely generates 65–75% MCRred1 as determined by UV-visible and EPR spectroscopy. Spectroscopy of MCR UV-visible spectra of the Ni(I)-containing MCRred1 were recorded in the anaerobic chamber using a spectrophotometer (Model USB4000-UV-VIS, Ocean Optics, Dunedin, FL). EPR spectra were recorded on a Bruker EMX spectrometer (Bruker Biospin Corp., Billerica, MA), equipped with an Oxford ITC4 temperature controller, a Hewlett-Packard Model 5340 automatic frequency counter and Bruker gaussmeter. The EPR spectroscopic parameters included: temperature, 70 K; microwave power, 10 mW; microwave frequency, 9.43 GHz; receiver gain, 2 × 104; modulation amplitude, 10.0 G; modulation frequency, 100 kHz. Double integrations of the EPR spectra were performed and referenced to a 1 mM copper perchlorate standard. The NMR data were acquired at 298 K on a Bruker Avance DRX 500 MHz instrument equipped with a TXI cryoprobe. Preparation of MCRred1 for Crystallization All crystallization experiments were performed in the anaerobic chamber in which MCR was purified unless otherwise noted. MCRred1 was prepared in 50 mM Tris, pH 7.6 and excess HSCoM was removed by buffer exchange using an Amicon Ultra centrifuge filter with a 50 kDa cut-off membrane (Millipore). Typically, 2–3 ml of MCRred1 was exchanged with 10–15 ml of 50 mM Tris, pH 7.6. The enzyme was concentrated to 500–600 μl, and this process was repeated three times. The fraction of MCRred1 in the purified MCR sample was calculated from the UV-visible spectrum using extinction coefficients of 27.0 mM−1cm−1 at 385 nm for Ni(I)-MCRred1, and 9.15 mM−1cm−1 at 420 nm for Ni(II)- MCRred1-silent (20). The amount of MCRred1 in samples used for crystallization was determined to be ~80% and the concentration of total enzyme used was in the range of about 120–150 μM (~32–40 mg/ml). All crystallization experiments were performed anaerobically by incubating 2.0 μl of enzyme solution in 50 mM Tris, pH 7.5 and 2.0 μl of reservoir solution (100 mM Hepes-Na, pH 7.3/7.5/8.0; 150 mM magnesium acetate (Mg(CH3COO)2), and 20/22% (w/v) PEG 400) in a sitting drop over 1 ml reservoir solution at 9 ºC. Triangular and rectangular prismatic crystals with a bright yellowish-green color confirmed the presence of nickel coenzyme F430. The crystals grew to a size of approximately 100–200 μm in 4–5 days. CoBSH-depleted crystals were obtained by incubating 2 μl of a reaction mixture containing 139 μM MCRred1 and 13 mM HSCoM with 2 μl of reservoir solution (100 mM Hepes-Na pH 7.5, 150 mM magnesium acetate (Mg(CH3COO)2), 22% PEG 400). Crystals of MCR complexed with the CoBSH analogues were grown by co-crystallization. The CoB5SH co-crystals were obtained by incubating 2μl enzyme solution containing 124 μM MCRred1, 10 mM methyl-SCoM and 1 mM CoB5SH with 2μl of reservoir solution (100 mM Hepes-Na, pH 7.5, 150 mM (Mg(CH3COO)2), 22 % PEG 400). The crystals with bound CoB6SH and CoB9SH were obtained by co-crystallization of 1 mM of analogue with 142 μM MCRred1 and equilibrated with 2μl of reservoir solution (100 mM Hepes-Na, pH 7.5, 150 mM Mg(CH3COO)2, 20 % PEG 400 for CoB6SH and 100 mM Hepes-Na, pH 7.3, 150 mM Mg(CH3COO)2), 22 % PEG 400 for CoB9SH). Crystals were cryoprotected in reservoir solution containing 25 % (v/v) PEG 400 by soaking for 2–5 minutes before cryocooling in liquid nitrogen in the anaerobic chamber. Crystals of CoB8SH bound to MCR were obtained by incubating 2 μl of a mixture of 119 μM MCRred1 and 1 mM CoB8SH with 2 μl of reservoir solution (100 mM Hepes-Na, pH 7.3, 150 mM Mg(CH3COO)2), 20 % PEG 400). Before cryoprotection, the crystals were soaked for 5–10 minutes in a 100 mM solution of methyl iodide2. The methyl iodide solution used for soaking was prepared by adding a concentrated stock of methanolic solution of methyl iodide to the reservoir Cedervall et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript solution. Soaked crystals were quickly cryoprotected as described above and cryocooled in the anaerobic chamber. X-ray Diffraction Data Collection, Processing and Refinement X-ray diffraction data were collected at 100 K on a ADSC Quantum-315 detector at the APS Beamline 14-BM-C (BioCARS). The wavelength of X-rays was 0.979 Å. Data were processed using HKL2000 (50). As in the previous X-ray crystallographic studies, the crystals belong to the monoclinic space group P21 (a = 82 Å, b = 118 Å, c = 122 Å, β = 92°), with one MCR molecule (two active sites) per asymmetric unit (5, 33). For refinement, REFMAC in the Collaborative Computational Project Number 4 (CCP4) program suite was used (51). A random sample of 5 % of the data across all resolution shells was chosen to check refinement progress through calculation of an Rfree. The same reflections were used to calculate Rfree for all structures, thus preventing bias due to high structural identity. The remaining reflections were used in refinement (Rwork). Model building was done using the Crystallographic Object-Oriented Toolkit (COOT) (52). The diffraction data and their models are designated as MCRCoBXSH, where X is the number of carbons in the alkanoyl portion of the analogue. Library files in CCP4 for F430 and CoBSH were incorrect, and these were modified in Monomer Library Sketcher in the CCP4 program suite by comparison with schematic drawings from Grabarse et al. (33). Co-ordinate and CCP4 library files for the different CoBSH analogues were created in Monomer Library Sketcher. The general model building and refinement strategy for all structures was as follows. It was clear from the electron density in the substrate channel and at the active site that mixtures of species were present in all datasets. These could be visualized with Fo-Fc and Fo-Fo difference electron density maps (Supporting Information, Figure S1). The known positions of CoBSH and HSCoM from the published Ni(II)-MCR crystal structures (PDB codes 1hbn, 1hbo, 1hbu (33)) were used as guides to determine which species could be present in each dataset, and these were then simultaneously modeled into the electron density. By alteration of their relative occupancies (in 10% increments) followed by refinement, the ratio of occupancy between different species was determined using the assumption that the average B-factors for all molecular species bound should be similar to that of F430 and adjacent well-ordered protein atoms within the active site and substrate channel. The combinations of modeled ligands were constantly reassessed throughout refinement based on the remaining difference electron density. This included test refinements of different ligand combinations during the latter stages, thus using the optimized phases to check whether a different combination of ligands could also explain the electron density. Sensible chemical structures and interactions, along with keeping the combined occupancies of sterically mutually exclusive species ≤ 100%, were maintained throughout refinement. The model was finally accepted when the difference electron density map was minimal and the B-factors for the models converged. In practice the first structure refined was that of MCRCoB5SH. Initial phases were generated by difference Fourier using a previously determined crystal structure (PDB code 1mro (5)) but with all non-bonded molecules, including water, removed from the model except F430. Initial rigid body refinement followed by restrained refinement of MCRCoB5SH reduced the Rwork to 26.5%. After model building and subsequent rounds of restrained refinement the Rwork was 14.3% (Rfree 16.6%). Of the five structures only the CoB5SH analogue is completely coincident with CoBSH, and so particular care had to be used in teasing apart the ratios of the two species in modeling the MCRCoB5SH electron density. This was done by 2Methyl iodide was added with the intention of creating a Ni(III)-methyl species, which CoB8SH stabilizes. This was not achieved, but the diffraction quality of this crystal was significantly better than crystals co-crystallized with CoB8SH alone, and so has been included in this study. Cedervall et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript initially refining 100% CoBSH in the substrate channel. Positive Fo-Fc difference density located at the carbon where the shorter CoB5SH thiol might be expected to be indicated the presence of a more electron-rich species than carbon, which is consistent with the presence of the CoB5SH sulfur. The refinement converged at a model containing a 50:50 mix of CoBSH and the analogue. However, positive Fo-Fc difference density was still present at the position of the CoBSH thiol, therefore a water molecule was added to the CoB5SH model at 50% occupancy and upon refinement this accounted for the electron density. An illustration of the electron density quality from this structure is shown in Supporting Information, Figure S2. An HSCoM-, CoBSH- and CoBSH analogue-free version of the refined MCRCoB5SH structure was used as the starting model to generate initial phases for the four other structures. After the initial round of restrained refinement the Rwork for these structures were reduced to 14.5–15.6 %. RESULTS AND DISCUSSION Crystal Structures of MCR Five crystal structures were determined, four of which are in complex with CoBSH analogues differing in the number of carbons atoms in the alkanoyl portion of the molecule. CoBSH is an N-7-mercaptoheptanoyl-containing molecule, whereas the four CoBSH analogues contain N-5-mercaptopentanoyl-, N-6-mercaptohexanoyl-, N-8-mercaptooctanoyl- or N-9-mercaptononanoyl-moieties (Figure 2). The corresponding crystal structures are designated as MCRCoBXSH, where X is the number of carbons in the alkanoyl portion of the analogue. The other crystal structure is of MCRred1c-silent (MCR in the Ni(II) state in complex with HSCoM, designated here as MCRHSCoM) that is CoBSH-depleted. The datasets have resolutions in the range from 1.35 – 1.8 Å. Although the crystallizations were set up with the MCR solution predominantly in the Ni(I)-MCRred1 state, by the time X-ray diffraction data were collected they had been oxidized to the Ni(II)-MCRred1-silent state (Supporting Information). Following data collection there was no evidence for photoreduction of the Ni(II) back to Ni(I) in any of the crystals, as assessed by single crystal UV-visible microspectrophotometry (Supporting Information and Figure S3). Attempts to photoreduce the crystals using different wavelengths and temperatures were unsuccessful (Supporting Information). Overall, the resulting structures are very similar to each other and to the previously published structures of MCR, with differences mainly localized to the active site and substrate channel. The two active sites in the ASU were refined independently. Unless otherwise stated there was no difference between them. All five datasets contain a mixture of species bound to the enzyme. There is always a background of CoBSH and HSCoM, which co-purify with MCR and cannot be fully removed by extensive buffer exchange or by the addition of a CoBSH analogue. HSCoM is added during purification of MCR, as it stabilizes the resting active Ni(I) state (unpublished data), and this leads to HSCoM occupancies between 50–100% amongst the structures (Table 1). In contrast CoBSH, which is not added during purification, has occupancies ranging from 30–50%. As these confounding species have all been described at high occupancy in other crystallographic studies, the structural data of interest could be isolated (5, 33). In each case, the additional electron density could be explained by inclusion of the appropriate CoBXSH model used in that experiment at 50% or higher occupancy. The resulting models, along with 2Fo-Fc electron density, are shown in Figure 3. The Rwork for the final structures range from 13.0 to 15.0 % (Rfree 15.5 to 19.5 %). The X-ray data collection, processing, refinement and model building statistics are given in Table 1. Cedervall et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Analogues shorter than CoBSH; CoB5SH and CoB6SH CoB5SH is two methylene groups shorter than CoBSH, the MCR substrate. The MCRCoB5SH structure is to 1.35 Å resolution. As expected the pentanoyl chain follows the path of the CoBSH heptanoyl carbons down the substrate channel, and thus its thiol is positioned in the same place as the second carbon preceding the CoBSH thiol (Figure 3A and 4). There are no published MCR kinetic studies using CoB5SH, but as it binds in the substrate channel, it is likely to be an inhibitor. CoB6SH is one methylene shorter than CoBSH, and is a slow substrate of MCR. In this case the 1.4 Å resolution electron density of MCRCoB6SH indicates that the analogue unexpectedly binds in the substrate channel such that its thiol is virtually in the same position as the thiol of CoBSH (Figure 3B and 4). The hexanoyl chain is oriented so that it takes a shorter route down the substrate channel between carbons 2 and 5 (with the carbonyl carbon labeled as carbon 1) than CoBSH (Figure 4, Supporting Information, Figure S4). This short-cut is not seen in any of the other CoBXSH complex crystal structures, but presumably arises because this CoB6SH binding conformer is energetically more favorable, although it is not clear from the structure why this might be the case. CoB6SH binds very tightly to MCR, with an apparent Ki value of 0.1 μM (3). Water structure in the absence of HSCoM The electron density for the MCRCoB6SH crystal structure only supported the modeling of 50 % bound HSCoM. In the fraction of MCR molecules where HSCoM is absent, the HSCoM binding site is occupied by a network of four water molecules (Supporting Information, Figure S5). Two waters are positioned close to the absent sulfonate oxygen positions of HSCoM. Based on the presence of positive difference electron density, a third water was modeled ligated to the Ni, and refined to a distance of ~2 Å (2.0 Å and 2.1 Å in the two active sites of the ASU) with no distance restraint imposed between the Ni and water. This water is in a similar position as the Ni coordinating sulfonate oxygen of the heterodisulfide product in MCRsilent (Supporting Information, Figure S6 and PDB codes 3m32, 1hbm) (5, 33). The fourth water was in the vicinity of the expected position of a bridging water (W1) seen in other structures (Figure 1, 3A and 3C). Water structure in the absence of CoBSH The 1.45 Å resolution electron density obtained for MCRHSCoM indicates that the substrate channel contains only 30 % CoBSH. Nine ordered waters (W1–W9), along with an acetate ion from the crystallization solution occupy the channel, with the acetate positioned where the phosphothreonine linkage of CoBSH would be (Figure 3C). Presumably 1–2 further waters would replace the acetate under physiological conditions. Other than W3 and W7, the waters form hydrogen bonds with protein (Figure 5). One water (W2) occupies the same site as the CoBSH thiol. Presumably due to the loss of favorable interactions that exist when CoBSH is present, the hydrophobic side-chain of Valα482 adopts a second conformation modeled at 60 % occupancy (Supporting Information, Figure S7). Position of the “bridging” water, W1 The equivalent of W1 has previously been observed in MCRred1-silent and MCRox1-silent crystal structures where, in the presence of CoBSH and HSCoM, it is sited equidistant (3.2 Å) between the two coenzyme thiols (PDB codes: 1hbn, 1hbo, 1hbu) and thus been termed the “bridging water” (Figure 1) (5, 33). However, in the MCRHSCoM structure, due to the presence of the more polarized W2 water, W1 is displaced away from HSCoM to maximize the hydrogen bond interaction with W2 (2.9 Å to W2; 3.5 Å to HSCoM thiol, Figure 5). In the MCRCoB5SH structure that also contained W2, the electron density indicated that this Cedervall et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript repositioning of W1 towards W2 also occurred. In contrast, the MCRCoB6SH structure contained 100% thiol at the CoBSH position, but a partial occupancy of HSCoM (50%). In this case the electron density for W1 indicated it had moved towards the nickel to form an optimal hydrogen bond with a Ni-ligating water that was only present in the absence of HSCoM (3.7 Å to CoB(6)SH thiol; 3.0 Å to Ni-ligating water, Supporting Information, Figure S5). In all structures reported here, W1 (if present) appears to be a sensitive indicator of the relative electronegativity of the Ni-ligated atom to that occupying the position of the CoBSH thiol, and was a useful check in the crystallographic modeling and refinement process. Flexibility in the substrate channel: Alternative protein conformers The binding site of HSCoM (and presumably methyl-SCoM) is more deeply buried within the enzyme, and so it must enter prior to CoBSH for productive chemistry to occur. As binding of CoBSH in the absence of co-substrate would be inhibitory, it was suggested that a conformational change upon methyl-SCoM binding might lower the Kd for CoBSH, and thus promote an ordered mechanism. Compared to the 1hbn MCRox1-silent and 1hbu MCRred1-silent crystal structures, which both have full occupancy HSCoM, the lower occupancy of HSCoM in the 1hbo MCRred1-silent structure was associated with significantly greater flexibility within the channel, and the ability to model a second conformation of a Gly-rich amino acid stretch that formed part of the CoBSH channel. This suggested that methyl-SCoM binding might cause the channel to become more ordered, increasing the affinity of MCR for CoBSH by conformational restriction rather than a switch mechanism where the structure reorganizes from one well-defined conformer to another (33). In the MCRHSCoM data containing 30 % CoBSH and 100% HSCoM, the Fo-Fc difference electron density map at one of the two independent active sites in the ASU contained positive peaks that suggested the presence of an alternate conformation also involving this part of the polypeptide (Supporting Information, Figure S8). Using this as a guide, a similar second conformation involving seven contiguous amino acid residues of the same Gly-rich amino acid stretch (β366–372) could be modeled and refined at 20 % occupancy leaving no residual difference density. Parts of the α′ subunit (α′111–129 and α′237–242) that are in close proximity to this stretch of amino acids also exhibit second conformations, with the main-chain carbonyl of α′243 in van der Waals contact with the B ring of F430 tetrapyrrole (Supporting Information, Figure S9). Modeling these at 20% occupancy accounted for the weak positive Fo-Fc difference electron density peaks observed in these areas. The evidence of alternate conformers in these areas lends support to the proposal that increased flexibility in the substrate channel propagates through the protein (33). The MCRCoB6SH crystal structure contains 50 % CoB6SH, 50% CoBSH and 50% HSCoM. In this case there is no evidence of an alternate loop conformation in either active site of the ASU. However, as CoBSH and CoB6SH combined are at 100% occupancy, it is not surprising their favorable interactions with the substrate channel would reduce conformational disorder, despite the partial occupancy of HSCoM. Analogues longer than CoBSH; CoB8SH and CoB9SH Both analogues could be accommodated in the MCR substrate channel (Figures 3D and 3E). The electron density supported final models containing 50% CoB8SH for MCRCoB8SH (1.8 Å resolution) and 60 % CoB9SH for MCRCoB9SH (1.45 Å resolution). The phosphate head- groups are in identical positions to those of CoBSH, CoB5SH, CoB6SH (Figure 4) (5, 33). Both analogues follow the crystallographically observed chain path of bound CoBSH, with the extra atoms displacing the W1 water and placing the thiols closer to the nickel (Figure 6). CoB9SH does have a second conformer that deviates from the CoBSH path, but the thiol position for this conformer and the CoBSH-tracking conformer are identical (Figure 3E and Cedervall et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supporting Information, Figure S10). Interestingly, the thiol of CoB8SH is not coincident with the CoB9SH carbon that precedes the CoB9SH thiol (Figure 6). CoB8SH is an MCR inhibitor with an apparent Ki of 15 μM (3). CoB9SH has never been tested for inhibition of MCR-catalyzed methane formation, but it is reasonable to assume that it would be an inhibitor. CoBXSH thiol-to-nickel spatial relationship The position of CoBSH in previous crystal structures poses a conundrum (5, 33). In all the proposed catalytic mechanisms, CoBSH must interact with species generated at the nickel. Perplexingly, the sulfur of the CoBSH substrate is 8.8 Å from the Ni(II) in the MCRox1-silent and both MCRred1-silent crystal structures, and 6.4 Å from the thiol of the substrate analogue HSCoM (Figure 1). Modeling studies demonstrated that the addition of a methyl group to HSCoM did not bridge this gap (35, 45, 53). Therefore, a conformational change has been postulated that would enable CoBSH to penetrate deeper into the substrate channel, and thus approach closer to any nickelbound species. The heterodisulfide product in the MCRsilent crystal structure has the CoBSH portion in virtually the same place as in MCRox1-silent, giving no clue to possible structural changes that might occur to facilitate CoBSH reacting with nickel-associated intermediates (5, 33). Trigonometry suggests that if the alkanoyl chain of CoBSH or its analogue is in an extended conformation, each additional unit in the chain would lead to the thiol moving ~1.2 Å towards the Ni. Until this study there have been no crystal structures of CoBSH analogues in complex with MCR, so mechanistic studies using different chain length analogues of CoBSH assumed that shorter analogues would trace the observed path of CoBSH, and longer analogues would penetrate about ~1.2 Å deeper per additional chain unit into the channel. In the case of the shortest analogue CoB5SH, it does indeed follow the path of CoBSH, with the thiol of CoB5SH being 2.8 Å away from the thiol position of CoBSH. However, due to the conformation CoBSH adopts when bound in the substrate channel, the difference in the S-Ni distance is small; the CoB5SH thiol being only 0.5 Å farther from the Ni than CoBSH (8.8 Å for CoBSH vs. 9.3 Å for CoB5SH) (Table 2). This is due to the alkanoyl chain of CoBSH not being in an extended conformation from carbons 4 to 6 (carbon 1 is the carbonyl carbon). CoB6SH on the other hand adopts a conformation that places its thiol in virtually the same position as the thiol of CoBSH (Figure 4 and Table 2). This is consistent with CoB6SH being a substrate. However, the kcat is 1000-fold lower than for CoBSH (3, 35) although its Km value (180 μM, Dey & Ragsdale, in preparation) is similar to that of CoBSH (Km = 75 μM (3)). The reason for this may be that the shorter alkanoyl chain may not enable the analogue thiol to approach the nickel close enough for efficient catalysis, and thus explain why CoB6SH is such a poor substrate. In the case of the longer CoBXSH analogues, the sulfur of CoB8SH is 2.6 Å closer to the Ni ion of F430 than that of CoBSH, and 2.5 Å closer to the thiol of HSCoM (Figure 6 and Table 2). The CoB9SH molecule follows the path of CoBSH, and reaches only a little further into the substrate channel than CoB8SH, with the CoB9SH thiol positioned 2.9 Å closer to the Ni than the thiol of CoBSH (Figure 6 and Table 2). This is only 0.3 Å closer than the distance observed for the CoB8SH thiol, even though they are non-coincident. The distance to the thiol of HSCoM is 2.6 Å closer than that of the substrate, CoBSH; only 0.1 Å closer than the CoB8SH thiol. The two analogue thiols are above an annular hydrophobic aromatic environment created by Pheα330, Tyrα333, Pheα443, Pheβ361 and Tyrβ367 that lies between them and F430 (Figure 6). As a result, penetrating further into the channel may be energetically unfavorable, consistent with the small difference in relative distances between the CoB8SH/CoB9SH thiols and the HSCoM thiol/F430 nickel. The annulus is proposed to be catalytically important in positioning methyl-SCoM and stabilizing the methane product, Cedervall et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript and the tyrosines have been proposed to be proton donors associated with mechanism II (Scheme 2B) (5, 33). Thus, there appear to be three preferential distances for thiols (including that of HSCoM) within the MCR substrate channel; HSCoM at 2.4 Å, CoB6SH/CoBSH at 8.7 – 8.8 Å and CoB8SH/CoB9SH at 5.9 – 6.2 Å from the nickel of F430 (Table 2). Recent ENDOR, high-field continuous and pulse EPR work has identified changes in nickel co-ordination when CoBSH is added to MCRred1c (active Ni(I)-MCRred1 + HSCoM) (14, 15, 18, 31). This generates up to 50% MCRred2 which is comprised of two distinct nickel co- ordination geometries; an axial MCRred2a formally assigned as a Ni(III)-hydride, and a rhombic MCRred2r in which the thiol of HSCoM is a Ni(I) ligand (Supporting Information, Scheme S1). Formation of MCRred2 could also be induced by addition of CoBSH substrate analogues CoBS-CH3 and CoBS-CF3, which have a chain length one unit longer than substrate CoBSH (18, 53). The CoBS-CF3 enabled 19F-ENDOR studies to be performed, and demonstrated that following CoBS-CF3 addition the remaining MCRred1 species had Ni(I)-19F distances of 6.2 – 7.7 Å. This distance range agreed with a CoBS-CF3 model created using the CoBSH position observed in the MCRox1-silent crystal structure (53). However, in the MCRred2 species the Ni(I)-19F distances had shortened indicating a movement of ~ 2 Å towards the nickel (MCRred2a Ni(I)-19 F, 4.0 – 5.5 Å; MCRred2r Ni(I)-19F, 4.5 – 5.7 Å). In the case of the CoB8SH analogue (the closest equivalent to CoBS- CF3 used in this study) the thiol to Ni(II) distance lies between the distance ranges observed in the CoBS-CF3 studies, and so the fluorine(s) of the CoBS-CF3 in MCRred2 might penetrate a little further into the hydrophobic annulus in the MCRred2 species. As the alkanoyl chain of CoBSH is not fully extended it could easily undergo a similar conformation change to that observed in the MCRred2 state. Reaction of MCR Ni(III)-alkyl species’ with CoB8SH and CoB9SH The two longer CoBXSH analogues have been shown to undergo alkylation when reacted with MCRPS, a [Ni(III)-alkyl ↔ Ni(II)-alkylsulfonate radical] formed from reaction of Ni(I)-MCRred1 with bromopropanesulfonate (BPS) (Supporting Information, Scheme S1) (20, 23, 30, 45, 54). BPS is a substrate of MCRred1 in a reaction that involves a rapid CoBSH-independent nucleophilic attack by Ni(I) on BPS to displace bromide and generate MCRPS at a rate ~60-fold faster than generation of methane from CoBSH and methyl- HSCoM (20, 45). Certain thiols can eliminate the propylsulfonate to yield a thioether product and regenerate MCRred1, although at a rate 1000-fold slower than methane formation (45). Both CoB8SH and CoB9SH can react with MCRPS to regenerate MCRred1, but CoBSH cannot. The overall second-order rate constant for the reactivation of MCR by CoB8SH is 160 M−1s−1, whereas for CoB9SH the reaction is more sluggish (12 M−1s−1). CoB9SH might be expected to be closer to the proximal Ni ligand. It was therefore proposed that this caused steric interference and explained why CoB9SH was a poorer reactivator of MCR than CoB8SH. Our study has shown that the thiols of these two analogues are placed such that they are approximately the same distance (~3.7 Å) from the thiol of HSCoM ligated to the Ni atom (Table 2). The Ni(II)-HSCoM bond is 2.4 Å, whereas an Ni(III)-alkyl bond is expected to be ~2 Å (24, 33), thus indicating that a conformational change is required to effect the nucleophilic attack of the CoB8SH and CoB9SH thiols on an alkyl- bound species. It would thus appear that a conformational change, such as observed in MCRred2, is required for this chemistry also (53). A Ni(III)-alkyl species is akin to the first intermediate in mechanism I of MCR-catalyzed methane formation, Ni(III)-methyl (MCRMe, Supporting Information, Scheme S1, Scheme 2A) (11, 27). MCRMe has been shown to be capable of generating MCRred1 and methyl- SCoM upon addition of HSCoM (which is the reverse of mechanism I, step 1, Scheme 2A); Cedervall et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript similar chemistry to the observed formation of a thioether product from the Ni(III)-alkyl. Further addition of CoBSH following HSCoM treatment of MCRMe led to methane and heterodisulfide formation, the natural products of methanogenesis. Although this lends credence to mechanism I, it should be noted that like MCRPS, MCRMe in these experiments was generated artificially. However, the MCRCoB9SH crystal structure demonstrates that the two additional methylene units in the alkanoyl chain c.f. CoBSH, do not necessarily translate into direct interaction of the thiol with the nickel proximal ligand. However, this could represent the favorable position for a CoBSH thiol interacting with the methyl group of methyl-SCoM. Just as the alkanoyl chain of CoB6SH has a more extended conformation than CoBSH in the substrate channel, CoBSH could also adopt a more extended conformation so that its thiol was in a similar position as the thiol of CoB8SH, priming it for reaction with a nickel bound species. If a significant conformational change is required early in MCR-catalyzed chemistry, which would be a requirement of mechanism I, catalysis may well involve a rearrangement of the aromatic amino acid annulus due to the presence of the methyl of methyl-SCoM, and this might enable deeper penetration of CoBSH into MCR (Figure 6). All the crystal structures in this study, and those solved previously, are of the inactive Ni(II)-MCR, which disfavors close approach to the nickel in the absence of Ni(I)-bound methyl-SCoM, even in the case of CoB9SH. Conclusion The goal of this study was to induce structural changes within the substrate channel and active site of MCR using analogues of coenzyme CoBSH. It was hoped that this would shed light on the nature of conformational changes that have been proposed to occur in MCR catalysis. We have shown that that the CoBXSH analogues do not lead to any significant conformational changes within the context of inactive Ni(II)-MCR. Therefore, it may be that methyl-SCoM is the key coenzyme, in combination with a nickel oxidation state of 1 (and 3), that triggers a conformational change bringing the thiol of CoBSH closer to the nickel. Thus, the crystal structure of the Ni(I)-methyl-SCoM/MCR complex may be required to structurally define the conformational changes required for MCR-mediated chemistry. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments X-ray data were collected at beam-line 14-BM-C and photoreduction studies at 14-ID-B, BioCARS at the Advanced Photon Source (APS), Argonne National Laboratory, Argonne, IL. We thank Vukica Srajer and Yu- Sheng Chen for valuable assistance during data collection. Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Basic Energy Sciences, Office of Science under Contract No. DE- AC02-06CH11357. Use of the BioCARS Sector 14 was supported by the National Institutes of Health, National Center for Research Resources, under grant number RR007707. Computer resources were provided by the Basic Sciences Computing Laboratory of the University of Minnesota Supercomputing Institute, and we thank Can Ergenekan for his support. We also thank Ed Hoeffner at the Kahlert Structural Biology Laboratory (KSBL) at the University of Minnesota. Use of the KSBL was supported by a Minnesota Partnership for Biotechnology and a Medical Genomics Grant SPAP-05-0013-P-FY06. References 1. Thauer RK. Biochemistry of methanogenesis: a tribute to Marjory Stephenson. Microbiology. 1998; 144:2377–2406. [PubMed: 9782487] 2. Thauer RK, Shima S. Methane as fuel for anaerobic microorganisms. Ann N Y Acad Sci. 2008; 1125:158–170. [PubMed: 18096853] Cedervall et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 3. Ellermann J, Hedderich R, Bocher R, Thauer RK. The final step in methane formation. Investigations with highly purified methyl-CoM reductase (component C) from Methanobacterium thermoautotrophicum (strain Marburg). Eur J Biochem. 1988; 172:669–677. [PubMed: 3350018] 4. Ellefson WL, Wolfe RS. Component C of the methylreductase system of Methanobacterium. J Biol Chem. 1981; 256:4259–4262. [PubMed: 6783657] 5. Ermler U, Grabarse W, Shima S, Goubeaud M, Thauer RK. Crystal structure of methyl-coenzyme M reductase: the key enzyme of biological methane formation. Science. 1997; 278:1457–1462. [PubMed: 9367957] 6. Diekert G, Gilles HH, Jaenchen R, Thauer RK. Incorporation of 8 succinate per mol nickel into factors F430 by Methanobacterium thermoautotrophicum. Arch Microbiol. 1980; 128:256–262. [PubMed: 7212929] 7. Diekert G, Jaenchen R, Thauer RK. Biosynthetic evidence for a nickel tetrapyrrole structure of factor F430 from Methanobacterium thermoautotrophicum. FEBS Letters. 1980; 119:118–120. [PubMed: 7428919] 8. Whitman WB, Wolfe RS. Presence of nickel in Factor F430 from Methanobacterium bryantii. Biochem Biophys Res Comm. 1980; 92:1196–1201. [PubMed: 7370029] 9. Albracht SPJ, Ankel-Fuchs D, Böcher R, Ellermann J, Moll J, van der Zwann JW, Thauer RK. Five new EPR signals assigned to nickel in methyl-coenzyme M reductase from Methanobacterium thermoautotrophicum, strain Marburg. Biochim Biophys Acta. 1988; 955:86–102. 10. Dey M, Kunz RC, Lyons DM, Ragsdale SW. Characterization of alkyl-nickel adducts generated by reaction of methyl-coenzyme m reductase with brominated acids. Biochemistry. 2007; 46:11969– 11978. [PubMed: 17902704] 11. Dey M, Telser J, Kunz RC, Lees NS, Ragsdale SW, Hoffman BM. Biochemical and spectroscopic studies of the electronic structure and reactivity of a methyl-Ni species formed on methyl- coenzyme M reductase. J Am Chem Soc. 2007; 129:11030–11032. [PubMed: 17711283] 12. Duin EC, Cosper NJ, Mahlert F, Thauer RK, Scott RA. Coordination and geometry of the nickel atom in active methyl-coenzyme M reductase from Methanothermobacter marburgensis as detected by X-ray absorption spectroscopy. J Biol Inorg Chem. 2003; 8:141–148. [PubMed: 12459909] 13. Duin EC, Signor L, Piskorski R, Mahlert F, Clay MD, Goenrich M, Thauer RK, Jaun B, Johnson MK. Spectroscopic investigation of the nickel-containing porphinoid cofactor F(430). Comparison of the free cofactor in the (+)1, (+)2 and (+)3 oxidation states with the cofactor bound to methyl- coenzyme M reductase in the silent, red and ox forms. J Biol Inorg Chem. 2004; 9:563–576. [PubMed: 15160314] 14. Finazzo C, Harmer J, Bauer C, Jaun B, Duin EC, Mahlert F, Goenrich M, Thauer RK, Van Doorslaer S, Schweiger A. Coenzyme B induced coordination of coenzyme M via its thiol group to Ni(I) of F430 in active methyl-coenzyme M reductase. J Am Chem Soc. 2003; 125:4988–4989. [PubMed: 12708843] 15. Finazzo C, Harmer J, Jaun B, Duin EC, Mahlert F, Thauer RK, Van Doorslaer S, Schweiger A. Characterization of the MCRred2 form of methyl-coenzyme M reductase: a pulse EPR and ENDOR study. J Biol Inorg Chem. 2003; 8:586–593. [PubMed: 12624730] 16. Goubeaud M, Schreiner G, Thauer RK. Purified methyl-coenzyme-M reductase is activated when the enzyme-bound coenzyme F430 is reduced to the nickel(I) oxidation state by titanium(III) citrate. Eur J Biochem. 1997; 243:110–114. [PubMed: 9030728] 17. Harmer J, Finazzo C, Piskorski R, Bauer C, Jaun B, Duin EC, Goenrich M, Thauer RK, Van Doorslaer S, Schweiger A. Spin density and coenzyme M coordination geometry of the ox1 form of methyl-coenzyme M reductase: a pulse EPR study. J Am Chem Soc. 2005; 127:17744–17755. [PubMed: 16351103] 18. Harmer J, Finazzo C, Piskorski R, Ebner S, Duin EC, Goenrich M, Thauer RK, Reiher M, Schweiger A, Hinderberger D, Jaun B. A nickel hydride complex in the active site of methyl- coenzyme m reductase: implications for the catalytic cycle. J Am Chem Soc. 2008; 130:10907– 10920. [PubMed: 18652465] Cedervall et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 19. Hinderberger D, Ebner S, Mayr S, Jaun B, Reiher M, Goenrich M, Thauer RK, Harmer J. Coordination and binding geometry of methyl-coenzyme M in the red1m state of methyl- coenzyme M reductase. J Biol Inorg Chem. 2008; 13:1275–1289. [PubMed: 18712421] 20. Kunz RC, Horng YC, Ragsdale SW. Spectroscopic and kinetic studies of the reaction of bromopropanesulfonate with methyl-coenzyme M reductase. J Biol Chem. 2006; 281:34663– 34676. [PubMed: 16966321] 21. Mahlert F, Bauer C, Jaun B, Thauer RK, Duin EC. The nickel enzyme methyl-coenzyme M reductase from methanogenic archaea: In vitro induction of the nickel-based MCR-ox EPR signals from MCR-red2. J Biol Inorg Chem. 2002; 7:500–513. [PubMed: 11941508] 22. Mahlert F, Grabarse W, Kahnt J, Thauer RK, Duin EC. The nickel enzyme methyl-coenzyme M reductase from methanogenic archaea: in vitro interconversions among the EPR detectable MCR- red1 and MCR-red2 states. J Biol Inorg Chem. 2002; 7:101–112. [PubMed: 11862546] 23. Rospert S, Voges M, Berkessel A, Albracht SP, Thauer RK. Substrate-analogue-induced changes in the nickel-EPR spectrum of active methyl-coenzyme-M reductase from Methanobacterium thermoautotrophicum. Eur J Biochem. 1992; 210:101–107. [PubMed: 1332856] 24. Sarangi R, Dey M, Ragsdale SW. Geometric and electronic structures of the Ni(I) and methyl- Ni(III) intermediates of methyl-coenzyme M reductase. Biochemistry. 2009; 48:3146–3156. [PubMed: 19243132] 25. Tang Q, Carrington PE, Horng YC, Maroney MJ, Ragsdale SW, Bocian DF. X-ray absorption and resonance Raman studies of methyl-coenzyme M reductase indicating that ligand exchange and macrocycle reduction accompany reductive activation. J Am Chem Soc. 2002; 124:13242–13256. [PubMed: 12405853] 26. Telser J, Davydov R, Horng YC, Ragsdale SW, Hoffman BM. Cryoreduction of methyl-coenzyme M reductase: EPR characterization of forms, MCR(ox1) and MCR (red1). J Am Chem Soc. 2001; 123:5853–5860. [PubMed: 11414817] 27. Yang N, Reiher M, Wang M, Harmer J, Duin EC. Formation of a nickel-methyl species in methyl- coenzyme M reductase, an enzyme catalyzing methane formation. J Am Chem Soc. 2007; 129:11028–11029. [PubMed: 17711279] 28. Albracht SPJ, Ankelfuchs D, Vanderzwaan JW, Fontijn RD, Thauer RK. A New Electron- Paramagnetic-Res Signal of Nickel in Methanobacterium-Thermoautotrophicum. Biochim Biophys Acta. 1986; 870:50–57. 29. Telser J, Horng YC, Becker DF, Hoffman BM, Ragsdale SW. On the assignment of nickel oxidation states of the Ox1, Ox2 forms of methyl-coenzyme M reductase. J Am Chem Soc. 2000; 122:182–183. 30. Hinderberger D, Piskorski RR, Goenrich M, Thauer RK, Schweiger A, Harmer J, Jaun B. A nickel- alkyl bond in an inactivated state of the enzyme catalyzing methane formation. Angewandte Chemie-International Ed. 2006; 45:3602–3607. 31. Kern DI, Goenrich M, Jaun B, Thauer RK, Harmer J, Hinderberger D. Two sub-states of the red2 state of methyl-coenzyme M reductase revealed by high-field EPR spectroscopy. J Biol Inorg Chem. 2007; 12:1097–1105. [PubMed: 17690920] 32. Becker DF, Ragsdale SW. Activation of methyl-SCoM reductase to high specific activity after treatment of whole cells with sodium sulfide. Biochemistry. 1998; 37:2639–2647. [PubMed: 9485414] 33. Grabarse W, Mahlert F, Duin EC, Goubeaud M, Shima S, Thauer RK, Lamzin V, Ermler U. On the mechanism of biological methane formation: structural evidence for conformational changes in methyl-coenzyme M reductase upon substrate binding. J Mol Biol. 2001; 309:315–330. [PubMed: 11491299] 34. Grabarse W, Mahlert F, Shima S, Thauer RK, Ermler U. Comparison of three methyl-coenzyme M reductases from phylogenetically distant organisms: unusual amino acid modification, conservation and adaptation. J Mol Biol. 2000; 303:329–344. [PubMed: 11023796] 35. Horng YC, Becker DF, Ragsdale SW. Mechanistic studies of methane biogenesis by methyl- coenzyme M reductase: evidence that coenzyme B participates in cleaving the C-S bond of methyl-coenzyme M. Biochemistry. 2001; 40:12875–12885. [PubMed: 11669624] Cedervall et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 36. Berkessel A. Methyl-Coenzyme-M Reductase - Model Studies on Pentadentate Nickel-Complexes and a Hypothetical Mechanism. Bioorg Chem. 1991; 19:101–115. 37. Jaun B. Coenzyme-F430 from Methanogenic Bacteria - Oxidation of F430 Pentamethyl Ester to the Ni(Iii) Form. Helvetica Chimica Acta. 1990; 73:2209–2217. 38. Signor L, Knuppe C, Hug R, Schweizer B, Pfaltz A, Jaun B. Methane formation by reaction of a methyl thioether with a photo-excited nickel thiolate - A process mimicking methanogenesis in archaea. Chemistry-a European Journal. 2000; 6:3508–3516. 39. Chen SL, Pelmenschikov V, Blomberg MR, Siegbahn PE. Is there a Ni-methyl intermediate in the mechanism of methyl-coenzyme M reductase? J Am Chem Soc. 2009; 131:9912–9913. [PubMed: 19569621] 40. Pelmenschikov V, Blomberg MRA, Siegbahn PEM, Crabtree RH. A mechanism from quantum chemical studies for methane formation in methanogenesis. J Am Chem Soc. 2002; 124:4039– 4049. [PubMed: 11942842] 41. Pelmenschikov V, Siegbahn PE. Catalysis by methyl-coenzyme M reductase: a theoretical study for heterodisulfide product formation. J Biol Inorg Chem. 2003; 8:653–662. [PubMed: 12728361] 42. Duin EC, McKee ML. A new mechanism for methane production from methyl-coenzyme M reductase as derived from density functional calculations. J Phys Chem. 2008; B 112:2466–2482. 43. Bobik TA, Wolfe RS. Physiological importance of the heterodisulfide of coenzyme M and 7- mercaptoheptanoylthreonine phosphate in the reduction of carbon dioxide to methane in Methanobacterium. Proc Natl Acad Sci U S A. 1988; 85:60–63. [PubMed: 3124103] 44. Goenrich M, Duin EC, Mahlert F, Thauer RK. Temperature dependence of methyl-coenzyme M reductase activity and of the formation of the methyl-coenzyme M reductase red2 state induced by coenzyme B. J Biol Inorg Chem. 2005; 10:333–342. [PubMed: 15846525] 45. Kunz RC, Dey M, Ragsdale SW. Characterization of the Thioether Product Formed from the Thiolytic Cleavage of the Alkyl-Nickel Bond in Methyl-Coenzyme M Reductase. Biochemistry. 2008; 47:2661–2667. [PubMed: 18220418] 46. Noll KM, Donnelly MI, Wolfe RS. Synthesis of 7-mercaptoheptanoylthreonine phosphate and its activity in the methylcoenzyme M methylreductase system. J Biol Chem. 1987; 262:513–515. [PubMed: 3100513] 47. Olson KD, McMahon CW, Wolfe RS. Photoactivation of the 2-(methylthio)ethanesulfonic acid reductase from Methanobacterium. Proc Natl Acad Sci U S A. 1991; 88:4099–4103. [PubMed: 1903534] 48. Zehnder AJ, Wuhrmann K. Titanium (III) citrate as a nontoxic oxidation-reduction buffering system for the culture of obligate anaerobes. Science. 1976; 194:1165–1166. [PubMed: 793008] 49. Gunsalus RP, Romesser JA, Wolfe RS. Preparation of coenzyme M analogues and their activity in the methyl coenzyme M reductase system of Methanobacterium thermoautotrophicum. Biochemistry. 1978; 17:2374–2377. [PubMed: 98178] 50. Otwinowski Z, Minor W. Processing of X-ray Diffraction Data Collected in Oscillation Mode. Methods in Enzymology: Macromolecular Crystallography, part A. 1997; 276:307–326. 51. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr. 1994; 50:760–763. [PubMed: 15299374] 52. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] 53. Ebner S, Jaun B, Goenrich M, Thauer RK, Harmer J. Binding of coenzyme B induces a major conformational change in the active site of methyl-coenzyme M reductase. J Am Chem Soc. 2010; 132:567–575. [PubMed: 20014831] 54. Goenrich M, Mahlert F, Duin EC, Bauer C, Jaun B, Thauer RK. Probing the reactivity of Ni in the active site of methyl-coenzyme M reductase with substrate analogues. J Biol Inorg Chem. 2004; 9:691–705. [PubMed: 15365904] Cedervall et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. The active site and substrate channel in the MCRox1-silent crystal structure (PDB code 1hbn) (9). Coenzyme F430, CoBSH and HSCoM are drawn as stick colored by atom (carbon: dark grey). The nickel is displayed as a green sphere, water as a red sphere. Interactions are drawn as dashed lines, and the corresponding distance is indicted in Angstroms (Å). The path of the substrate channel was defined in the absence of F430, CoBSH, HSCoM and water, with the surface closest to the viewer cut away. The figure was generated using PyMOL (http://www.pymol.org). Cedervall et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Drawing of CoBSH analogues; (A) N-5-mercaptopentanoylthreonine phosphate (CoB5SH); (B) N-6-mercaptohexanoylthreonine phosphate (CoB6SH); (C) N-8- mercaptooctanoylthreonine phosphate (CoB8SH); (D) N-9-mercaptononanoylthreonine phosphate (CoB9SH). Cedervall et al. Page 16 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. The active sites and substrate channels of the MCR crystal structures.; (A) MCRCoB5SH; (B) MCRCoB6SH; (C) MCRHSCoM; (D) MCRCoB8SH; (E) MCRCoB9SH. 2Fo-Fc electron density map around the CoBXSH analogues, waters in the CoBSH binding part of the channel and the acetate ion (contoured at 1σ) is shown as a blue mesh. The protein is drawn as cartoon. CoBXSH and acetate are drawn as stick and colored by atom (carbon: CoB5SH orange; CoB6SH pale yellow; CoB8SH light blue; CoB9SH magenta; acetate white). Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere, and waters as red spheres. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 17 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. Overlay of CoBSH (from PDB code 1hbn) and the different CoBSH analogues. CoBXSH are drawn as stick with the thiol represented by a sphere, and colored CoB5SH orange; CoB6SH pale yellow; CoBSH light green; CoB8SH light blue; CoB9SH magenta. The protein is drawn as cartoon. Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 18 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Hydrogen bonding diagram for the water structure modeled in MCRHSCoM. The water molecules are named as in Figure 3C (W1–W9); WA, WB and WC are water molecules that are present in all structures (i.e. in concert with the substrate CoBSH and the CoBSH analogues). Interactions between surrounding residues and the water molecules are drawn as dashed lines, and the corresponding distance is indicated in Angstroms (Å). Cedervall et al. Page 19 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. Stereo image of the annulus of aromatic amino acids distal of coenzyme F430. The protein is drawn as cartoon with the side-chains of the aromatic residues drawn as white stick. CoBSH, (from PDB code 1hbn (9)), CoB8SH and CoB9SH are drawn as stick with the thiols represented by spheres, and colored CoBSH light green; CoB8SH light blue; CoB9SH magenta. Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 20 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. Reaction catalyzed by methyl-coenzyme M reductase Cedervall et al. Page 21 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 2. Two of the proposed catalytic mechanisms for methyl-coenzyme M reductase; (A) mechanism I; (B) mechanism II. Cedervall et al. Page 22 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 23 Table 1 X-ray Data Collection, Processing and Refinement Statistics Data collection and processing statistics Name of data set MCRCoB5SH MCRCoB6SH MCRHSCoM MCRCoB8SH MCRCoB9SH Measured reflections 1969388 2427498 1440665 1160543 1425506 Unique reflections 553755 446253 405349 211803 401701 Resolution (Å) a 50.0–1.30 (1.35–1.30) 50.0–1.40 (1.45–1.40) 50.0–1.45 (1.50–1.45) 50.0–1.80 (1.86–1.80) 50.0–1.45 (1.50–1.45) Completeness (%) a 97.1 (78.1) 99.9 (100.0) 99.5 (99.7) 99.8 (100.0) 98.1 (95.4) R-sym (%) a,b 5.5 (32.9) 7.3 (44.7) 6.2 (44.0) 8.4 (47.7) 5.6 (42.5) I/σI a 22.3 (3.6) 20.4 (4.0) 20.2 (3.2) 21.8 (3.9) 24.3 (3.2) Space group P21 P21 P21 P21 P21 Refinement and model building statistics Resolution (Å) a 20.49–1.30 (1.33–1.30) 19.89–1.40 (1.44–1.40) 20.15–1.45 (1.49–1.45) 19.93–1.80 (1.84–1.80) 20.07–1.45 (1.48–1.45) No. of reflection in working set a 525817 (30239) 423854 (25833) 384868 (25791) 201128 (11193) 381474 (23611) No. of reflection in test set a 27777 (1576) 22348 (1331) 20362 (1319) 10625 (557) 20163 (1210) R-work (%) c 14.32 13.04 13.47 14.95 13.58 R-free (%) d 16.56 15.53 16.22 19.54 16.44 ESU (Å) R-work/R-free 0.044/0.046 0.049/0.051 0.056/0.059 0.121/0.119 0.057/0.060 No. protein atoms 20087 19960 20265 19750 20036 No. coenzyme atoms 218 220 180 224 272 No. ligand atoms 37 62 52 26 49 No. water molecules 2443 2352 2516 1893 2432 RMS bond lengths (Å) 0.033 0.033 0.032 0.028 0.032 bond angles (deg.) 2.693 2.625 2.468 2.059 2.549 Ramachandran plot (%) favored 97.8 97.5 97.6 97.2 97.7 allowed 2.1 2.4 2.3 2.7 2.1 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 24 disallowed 0.1 0.1 0.1 0.1 0.1 Average B-factor (Å2) protein 12.42 13.35 12.12 17.22 12.73 coenzymes 8.20 9.24 7.25 11.24 8.27 ligands 31.95 35.48 28.29 33.76 32.92 waters 22.95 24.89 23.85 26.79 24.09 over all 13.54 14.57 13.40 18.02 13.93 Occupancy of HSCoM per active site (%)e 90/90 50/50 100/100 90/90 90/85 Occupancy of CoBSH per active site (%) e 50/50 50/50 30/30 50/50 40/40 CoBSH analogue, occupancy per active site (%) e CoB5SH, 50/50 CoB6SH, 50/50 CoB8SH, 50/50 CoB9SH, 60/60 Other molecule, occupancy per active site (%) e Acetate, 70/70 aValues in brackets correspond to the highest resolution shell. bR-sym = ΣhklΣNj=1|Ihkl−Ihkl(j)|/ΣhklN*Ihkl, sum over all reflections and all observations N, with Ihkl(j) intensity of the jth observation of reflection hkl and Ihkl mean intensity of the reflection hkl. cR-work = Σ||Fo| − |Fc||/Σ|Fo|, where |Fo| = observed structure factor amplitude and |Fc| = calculated structure factor amplitude. dR-free, R-factor based on 5% of the data excluded from refinement. eOccupancy of model in each of the two crystallographically independent active sites in the ASU Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 25 Table 2 Distances from analogue thiols. CoBXS - SCoM distance (Å) CoBXS - Ni distance (Å) CoB5SH 7.11/7.11a 9.30/9.30 CoB6SH 6.26/6.26 8.70/8.70 CoB7SH (substrate) b 6.37/6.39 8.73/8.77 CoB8SH 3.75/3.78 6.16/6.17 CoB9SH 3.71/3.68 5.96/5.91 aDistances in the two crystallographically independent active sites in the ASU bDistances in the 1.16 Å resolution MCRox1-silent structure (PDB code 1hbn)(33) Biochemistry. Author manuscript; available in PMC 2011 September 7.
3M31
Structure of the C150A/C295A mutant of S. cerevisiae Ero1p
Steps in reductive activation of the disulfide-generating enzyme Ero1p Nimrod Heldman,1 Ohad Vonshak,1 Carolyn S. Sevier,2 Elvira Vitu,1 Tevie Mehlman,3 and Deborah Fass1* 1Department of Structural Biology, Weizmann Institute of Science, Rehovot 76100, Israel 2Department of Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 3Department of Biological Research Support, Weizmann Institute of Science, Rehovot 76100, Israel Received 8 March 2010; Accepted 9 July 2010 DOI: 10.1002/pro.473 Published online 28 July 2010 proteinscience.org Abstract: Ero1p is the primary catalyst of disulfide bond formation in the yeast endoplasmic reticulum (ER). Ero1p contains a pair of essential disulfide bonds that participate directly in the electron transfer pathway from substrate thiol groups to oxygen. Remarkably, elimination of certain other Ero1p disulfides by mutation enhances enzyme activity. In particular, the C150A/ C295A Ero1p mutant exhibits increased thiol oxidation in vitro and in vivo and interferes with redox homeostasis in yeast cells by hyperoxidizing the ER. Inhibitory disulfides of Ero1p are thus important for enzyme regulation. To visualize the differences between de-regulated and wild-type Ero1p, we determined the crystal structure of Ero1p C150A/C295A. The structure revealed local changes compared to the wild-type enzyme around the sites of mutation, but no conformational transitions within 25 A˚ of the active site were observed. To determine how the C150AC295 disulfide nonetheless participates in redox regulation of Ero1p, we analyzed using mass spectrometry the changes in Ero1p disulfide connectivity as a function of time after encounter with reducing substrates. We found that the C150AC295 disulfide sets a physiologically appropriate threshold for enzyme activation by guarding a key neighboring disulfide from reduction. This study illustrates the diverse and interconnected roles that disulfides can play in redox regulation of protein activity. Keywords: disulfide bond formation; enzyme activation; flavin adenine dinucleotide; lag phase Introduction A fundamental question in cell biology is how a bal- ance between thiols and disulfides is maintained in the endoplasmic reticulum (ER) to promote efficient oxidation of proteins while preventing irreversible mispairing of disulfides.1 The ER sulfhydryl oxidase Ero1 catalyzes formation of disulfide bonds2,3 and may also serve as a redox sensor, tailoring its activ- ity according to the thiol/disulfide balance or the presence of specific reduced substrates in the com- partment.4,5 Although Ero1 has certain structural and mechanistic features in common with other sulf- hydryl oxidases,6 Ero1 exhibits complex kinetics not observed in other enzyme families catalyzing the same chemical reaction. This unique behavior of Ero1, which includes a pronounced lag phase observed in assays of catalytic activity on model sub- strates performed in vitro,4,7 may be a manifestation of the regulatory feedback mechanism to prevent over-oxidation of the ER thiol pool.4 Therefore, a thorough analysis of Ero1 kinetics and their bio- chemical and structural bases is essential for Abbreviations: DTT, dithiothreitol; ER, endoplasmic reticulum; FAD, flavin adenine dinucleotide; PDI, protein disulfide isomer- ase; Pdi1pred, reduced yeast PDI; PEG, polyethylene glycol; Trx, E. coli thioredoxin I; Trxred, reduced Trx. Additional Supporting Information may be found in the online version of this article. Grant sponsors: U.S.-Israel Binational Science Foundation; Kimmelman Center for Macromolecular Assemblies. Nimrod Heldman and Ohad Vonshak contributed equally to this work. *Correspondence to: Deborah Fass, Department of Structural Biology, Weizmann Institute of Science, Rehovot 76100, Israel. E-mail: deborah.fass@weizmann.ac.il Published by Wiley-Blackwell. V C 2010 The Protein Society PROTEIN SCIENCE 2010 VOL 19:1863—1876 1863 understanding the origin of the redox balance in the ER. A previous study presented the finding that mutation of certain cysteine residues in yeast Ero1 (Ero1p) increases enzyme activity.4 We now address the mechanism by which noncatalytic disulfides tune the response of the enzyme to thiol species in the environment. Yeast Ero1p has 14 cysteine residues and a bound flavin adenine dinucleotide (FAD) cofactor [Fig. 1(A)]. Two pairs of Ero1p cysteines are on the direct electron- transfer pathway from substrate to flavin [Fig. 1(B)]. One of these pairs, a Cys-X-X-Cys motif that forms the active-site disulfide (C352AC355), abuts the isoalloxa- zine of the FAD and most likely transfers electrons directly to the cofactor.6,9 The second pair (C100A C105), the ‘‘shuttle’’ disulfide, is a Cys-X4-Cys motif on a flexible loop near the active site and appears to mediate transfer of electrons from substrate proteins to the active-site disulfide.6,9,10 A third conserved disulfide (C90AC349) is in the vicinity of the active site but is not essential for Ero1p activity in vitro or under condi- tions that have been examined in vivo.4 Its precise role and the reason for conservation are unclear. Other disulfides, more distant from the active site, are also dispensable for activity, but they may nevertheless con- tribute to regulation of the enzyme. We reported that a yeast Ero1p mutant lacking the C150AC295 disulfide, which is 35 A˚ from the FAD isoalloxazine, has a short- ened lag phase in enzyme assays on model substrates in vitro, increases the intracellular ratio of oxidized to reduced glutathione, and decreases viability of yeast.4 These observations demonstrate an important role for Ero1p redox centers off the catalytic electron transfer pathway and distant from the active site. A remaining question is how loss of the C150AC295 disulfide in Ero1p has such profound effects on enzyme activity and on redox homeostasis in cells, and whether the C150A/C295A mutant can pro- vide insight into the autoregulatory mechanism of the wild-type enzyme. The structure of Ero1p, obtained pre- viously from two crystal forms (hexagonal at 2.8 A˚ reso- lution and centered orthorhombic at 2.2 A˚ resolution),6 suggested the possibility of redox-dependent conforma- tional changes. The temperature factor distribution [Fig. 1(A)] and a comparison of the structures from the two crystal forms highlighted the flexibility of loops in what is designated the ‘‘top’’ of the Ero1p structure. A congregation of disulfides in the loop-rich region and at the interface with the 10-helix core of the enzyme, to- gether with numerous exposed hydrophobic residues and a richly featured surface containing hydrophobic pockets and grooves [Fig. 1(C)], suggests that reduction or elimination of disulfides may enable conformational rearrangement of the top portion of the enzyme, with consequent effects on enzyme activity. Conformational changes upon disulfide reduction may occur also in mammalian Ero1a, as supported by changes in intrinsic tryptophan/tyrosine fluorescence upon reduction or mutation of certain disulfides.11 To test the hypothesis that elimination of the C150AC295 disulfide in Ero1p causes conformational Figure 1. A: Ribbon diagram of wild-type Ero1p (Protein Data Bank code: 1RP4) colored according to temperature factor, with red corresponding to high temperature factor regions and blue to low. The FAD is shown in sticks. Disulfide bonds and the unpaired C208 are illustrated in ball-and-stick representation and labeled. Eleven of the 14 cysteines in full-length yeast Ero1p are present in the truncated protein that was crystallized. B: The two-electron transfer events involved in oxidation of Pdi1p by Ero1p are illustrated as paired arrows. C: The molecular surface of Ero1p is shown from two angles with cysteine sulfur atoms colored yellow and surface-exposed hydrophobic side-chains colored green. Aromatic side chains are in a darker shade. The table lists the solvent-accessible surface areas (SAS) of the side chains of the indicated cysteines as determined using the program areaimol (CCP4 package).8 1864 PROTEINSCIENCE.ORG Activation of Sulfhydryl Oxidase Ero1p changes that increase enzyme activity, we deter- mined the structure of the C150A/C295A de-regu- lated Ero1p mutant by X-ray crystallography. Inter- pretation of this structure was facilitated by enzymatic assays of wild-type and mutant Ero1p oxi- dizing either native or model substrates. Tandem mass spectrometry (LC-MS/MS) analysis of Ero1p di- sulfide connectivity during activation and turnover revealed the basis for the hyperactivity of the C150A/ C295A mutant and provided insight into the natural activation mechanism of the enzyme. Results Electron transfer kinetics of the C150A/C295A Ero1p mutant Increased turnover rates and a shortened lag phase for the C150A/C295A Ero1p mutant were previously reported based on oxygen consumption assays.4 To obtain greater precision at early time points, we per- formed similar reactions using stopped-flow fluorim- etry to monitor oxidation of reduced thioredoxin (Trxred) by Ero1p [Fig. 2(A)]. Free FAD was used as the electron acceptor for Ero1p instead of oxygen, since reduction of the flavin is detectable by a decrease in fluorescence. Two recombinant wild-type Ero1p constructs, spanning amino acid residues 10–424 or 56–424,4,6,7 were examined. These two constructs differ by approximately twofold in their activities, but their progress curves are identical in shape, as can be seen by scaling the time axes [Fig. 2(B)]. Thus, the lag phase observed during oxidation of Trxred is an inherent property of the wild-type enzyme on this substrate regardless of absolute activity. The lag phase is also independent of whether molecular oxygen or FAD is used as the electron acceptor. In contrast, the C150A/C295A mu- tant has a progress curve qualitatively different from either wild-type version of Ero1p [Fig. 2(B)]. Ero1p C150A/C295A structure The dramatic impact on enzyme kinetics of the C150A/C295A double mutation prompted us to investigate the structure of this Ero1p variant. We sought to reveal any conformational differences com- pared to wild type that may explain the differences in activity and pre-steady state kinetics. Mutant pro- tein was produced in E. coli with approximately one- third the yields obtained for wild type and was iso- lated using a comparable protocol.6 As for wild-type Ero1p, crystals of Ero1p C150A/C295A were obtained in both hexagonal and centered orthorhom- bic forms. The orthorhombic form was chosen for further study because it diffracted to higher resolu- tion. The Ero1p C150A/C295A structure was deter- mined using phases calculated from a partial model of wild-type Ero1p (see Materials and Methods sec- tion). The structure was refined using diffraction data to 1.85 A˚ resolution (Table I). Conformational differences compared to wild- type Ero1p were observed in the region of the miss- ing C150AC295 disulfide (Fig. 3, left panel). These differences occur primarily in the a-helix down- stream of C295 (designated helix a6).6 The a6 helix is five residues longer at the amino terminus in the C150A/C295A mutant, extending from L297 as opposed to I302, allowing D296 to cap the helix. In the wild-type structure, participation of C295 in the disulfide with C150 prevents D296 from capping the a6 helix, and the shorter version of the helix is instead partially capped in trans by the side chain of S293. The two Cb atoms of A150 and A295 are 9.2 A˚ apart in Ero1p C150A/C295A, in contrast to the 3.8 A˚ distance between the analogous cysteine Cb atoms of the wild-type enzyme. In summary, the region in the vicinity of residues 150 and 295 apparently Figure 2. Enzyme progress curves of various Ero1p constructs. Residue number boundaries appear in subscript in the label of each curve. A: Ero1p activity was monitored by stopped-flow fluorimetry. Ero1p, FAD, and Trxred were mixed under anaerobic conditions at final concentrations of 1, 100, and 50 lM, respectively. The exogenous FAD served as the electron acceptor for Ero1p, and fluorescence decay of FAD upon reduction reported the progress of the reaction. B: Superposition of fluorescence data by rescaling Ero110–424 and Ero156–424 C150A/C295A data along the time axis provides evidence for altered regulation of Ero156–424 C150A/C295A activity. Heldman et al. PROTEIN SCIENCE VOL 19:1863—1876 1865 relaxes in the double cysteine-to-alanine mutant to a conformation incompatible with the disulfide but containing more regular and extensive secondary structure. Despite these differences, most of the Ero1p C150A/C295A structure, including the active-site region (Fig. 3, right panel), is similar to that previ- ously observed for wild-type Ero1p. The resolution of the diffraction data was significantly better than obtained for wild type, but regions that previously gave poor electron density (i.e., the 155–165 and 108–116 loops) were also apparently flexible in the Ero1p C150A/C295A crystals. No major differences were seen in the solvent exposure of other disulfide bonds in the structure. The relatively minor differ- ences in the C150A/C295A Ero1p structure do not rule out the possibility of global changes in protein dynamics, but any putative changes in dynamics did not prohibit crystallization. The absence of a clear structural explanation for the increased activity of the C150A/C295A mutant led us to initiate a more detailed biochemical analysis of the reductive activa- tion process. Gel electrophoretic analysis of disulfide reduction during Ero1p activation Reduction of a series of disulfide bonds in Ero1p upon encounter with reducing substrate has been observed by changes in migration rate of the enzyme on denaturing gels.4 These experiments are per- formed by incubating Ero1p with substrate and blocking the reaction after various times using rapid alkylating agents to modify free thiols. For our cur- rent studies, we compared reactions blocked with the small alkylating agent N-ethyl maleimide (NEM) versus the larger maleimide derivative 4-acetamido- 40-maleimidylstilbene-2,20-disulfonic acid (AMS) and analyzed the samples on both reducing and nonre- ducing denaturing gels. These experiments extend our previous work with AMS-modified samples resolved under nonreducing conditions4 and allow for additional insight into the reductive activation process of Ero1p. The thiol trapping experiments provide informa- tion on the number of free cysteines that become reduced and the nature of any remaining disulfides in Ero1p at a given time after substrate addition. When samples are resolved under reducing condi- tions, the migration rate of Ero1p reflects the abso- lute mass of the enzyme, which depends on both the number of cysteines modified and the type of alkyl- ating agent used. AMS contributes 510 D per cys- teine modified, whereas NEM contributes 125 D. The small change in mass due to NEM alkylation of Table I. Summary of Data Collection and Refinement Statistics Ero1p mutant C150A/C295A C143A/C166A Space group C2221 P62 Unit cell parameters (A˚ ) 73.4  132.8  102.7 107.2  107.2  124.2 Resolution (A˚ ) 500.0–1.85 50.0–3.2 Completeness (%) 99.0 99.9 Redundancy 4.9 12.4 Rsym a 0.037 0.236 hI/rIi 15.4 5.7 Total reflections/test set 42,641/2864 13,131/672 Rwork/Rfree b 0.209/0.238 0.244/0.293 Rms deviations from ideality Bonds (A˚ ) 0.005 0.008 Angles () 1.314 1.847 Number of atoms Protein 2899 2842 Water 211 12 FAD 53 53 NEM 9 9 Cd2þ 1 1 a Rsym ¼ RhklRi|Ii(hkl)  hI(hkl)i|/RhklRiIi(hkl), where Ii(hkl) is the observed intensity and hI(hkl)i is the average intensity for i observations. b Rwork, Rfree ¼ R||Fobs|  |Fcalc||/R|Fobs|, where Fobs and Fcalc are the observed and calculated structure factors, respectively. A set of reflections (6.7%) were excluded from refinement and used to calculate Rfree. Figure 3. Superposition of the structures of wild-type and C150A/C295A Ero1p. In the central panel, both structures are shown in beige. In the zoom views to either side, the Ero1p C150A/C295A structure is shown in dark red. The left zoom window is rotated relative to the central view to show structural differences with greater clarity. The dotted line indicates a region of poor electron density that could not be modeled in the mutant structure and was backbone traced in the wild-type structure. The right zoom window shows the close correspondence of the wild-type and mutant enzymes in the region of the active site. 1866 PROTEINSCIENCE.ORG Activation of Sulfhydryl Oxidase Ero1p free thiols does not result in an observable shift in migration of Ero1p on a reducing, denaturing gel. In contrast, under the same gel conditions, Ero1p with AMS-modified cysteines can be distinguished from unmodified Ero1p. When samples are analyzed under nonreducing conditions, the migration rate of Ero1p is influenced not only by the change in mass from modified cysteines but also by the presence or absence of disulfide bonds between cysteines distant from one another in amino acid sequence, which affects the hydrodynamic radius of the SDS-bound, denatured protein. For AMS-modified samples, the mobility of Ero1p will be influenced both by the con- tribution of any disulfides to the hydrodynamic ra- dius and by the 510 D mass added per modified thiol. When NEM is used, the mass change due to thiol derivation is negligible, and the apparent mo- bility under nonreducing conditions will primarily reflect the increase in hydrodynamic radius upon reduction of disulfides between cysteines distant from one another in amino acid sequence. In experiments performed using Trxred as a sub- strate, Ero1p was converted stepwise to a set of more slowly migrating species (Fig. 4), as reported previously.4 Appearance of these species correlates with increased enzyme activity.4 At later times in the reaction, when most of the substrate had been oxidized, Ero1p became re-oxidized to its initial state of low activity. For the Ero1p C150A/C295A mutant, the more slowly migrating species appeared earlier in the reaction and disappeared earlier as well (Fig. 4). Furthermore, the most slowly migrating band was never the most populated species in the Ero1p C150A/C295A reaction, whereas it was the dominant species at a certain point (1 min) in the reaction with the wild-type enzyme. As explained above, the different migration rates of Ero1p species reflect different disulfide Figure 4. Disulfide reduction during catalysis by Ero1p and the de-regulated mutant C150A/C295A. Enzyme at a final concentration of 2 lM was mixed with 100 lM Trxred. Aliquots were removed from the reactions at the indicated time points, and disulfide status was preserved by blocking with maleimide reagents. The migration of Ero1p on SDS-PAGE at the various time points was examined and compared with wild-type Ero1p (WT), the indicated Ero1p mutants, and DTT-treated Ero1p (WT reduced). (A) The Trxred oxidation reaction was blocked with AMS and separated by SDS-PAGE under nonreducing conditions. The asterisk (*) indicates a species that migrates slower than Ero1p C90A/C349A when blocked with AMS, but faster than this disulfide mutant when blocked with NEM (compare with panel C). (B) The same experiment as in (A) was performed, but the samples were reduced with DTT after AMS treatment before separation by SDS-PAGE. (C) Reactions were blocked with NEM and applied to the gel under nonreducing conditions. Heldman et al. PROTEIN SCIENCE VOL 19:1863—1876 1867 connectivity. A C90A/C349A mutant, which lacks the disulfide that closes the largest loop in the protein, shows a major shift in mobility relative to wild-type Ero1p; this mutant produces the largest increase in hydrodynamic radius of any single disulfide mutant relative to wild type (Fig. 4). Notably, when the reac- tion of wild-type Ero1p with substrate was blocked with AMS, the most slowly migrating species [aster- isk in left panel of Fig. 4(A)] ran above the position of the C90A/C349A mutant. However, the same reac- tion blocked with NEM resulted only in species that migrated more quickly than the C90A/C349A mu- tant [asterisk in left panel of Fig. 4(C)]. The lack of a maximal change in the hydrodynamic radius of NEM-treated Ero1p implies that the C90AC349 di- sulfide may be intact under these reaction condi- tions. For Ero1p with an intact C90AC349 bond, the shift observed with AMS likely reflects the combined effect of the increase in mass due to thiol modifica- tion and the lesser hydrodynamic changes observed during reduction of other disulfides less distant in sequence than C90AC349. Alternatively, the slowest migrating Ero1p species could lack C90AC349, and the moderate migration shift may reflect the pres- ence of an alternate long-range disulfide formed by thiol/disulfide exchange. A catalytic intermediate with a disulfide between C105 and C352 has been previously trapped in vivo,10 and the presence of this long-range disulfide could perhaps account for faster mobility even if C90AC349 were absent. At early time points (<1 min), Ero1p blocked with NEM did not exhibit a significant shift in mo- bility on the gel [Fig. 4(C), left panel]. In contrast, Ero1p blocked with AMS did display species with lower electrophoretic mobility at these times on both nonreducing and reducing gels [Fig. 4(A,B)], sug- gesting reduction of one or more disulfides that do not greatly affect the protein hydrodynamic radius under denaturing conditions. What appeared as a single species in the NEM-blocked reaction was resolved into two distinct species by AMS treatment, which is most readily observed under nonreducing conditions but is also apparent as a closely migrat- ing pair of bands under reducing conditions. Between 0.1 and 0.5 min, which is the time-frame corresponding to enzyme activation (Fig. 2), the ratio of the upper to lower bands of the 40 kD doublet for the AMS-blocked reaction increased [Fig. 4(A)]. For early time points in reactions using the Ero1p C150A/C295A mutant, the single major band seen in the NEM-blocked reactions was similarly resolved into two species by blocking with AMS. However, in this case, the mutant Ero1p appeared maximally reduced at 0.1 min, and the ratio of the upper to lower bands of the 47 kD doublet decreased between 0.1 and 0.5 min as the Ero1p mu- tant protein was rapidly re-oxidized. Notably, signifi- cant retardation of mobility was seen for the Ero1p C150A/C295A mutant at the shortest time point. However, the shifted species migrate faster than fully reduced Ero1p or the C90A/C349A mutant and thus are not likely to reflect reduction of the C90AC349 disulfide. Mass spectrometric analysis of Ero1p activation To directly map the Ero1p disulfides that were reduced in each gel-shifted species described above, we subjected to in-gel proteolysis and tandem mass spectrometry bands from AMS-blocked reactions run under nonreducing conditions (Fig. 5). Earlier attempts at disulfide mapping of Ero1p using ma- trix-assisted laser desorption/ionization time of flight (MALDI-TOF) mass spectrometry suggested the presence of disulfides inconsistent with the X-ray crystal structure (data not shown), revealing the risk of over-interpreting data consisting of only par- ent peptide masses. LC-MS/MS data, in contrast, enable more reliable assignments of peptide identities. The first LC-MS/MS experiment [Fig. 5(B)] was performed without in-gel reduction and alkylation, such that cysteine connectivity in disulfides was pre- served. The first and most rapidly migrating par- tially reduced species observed showed C100 in AMS-modified form and the C143AC166 and C150AC295 disulfides intact. The next shifted spe- cies showed C166 alkylated and C150AC295 in di- sulfide form. The most slowly migrating species showed the active-site disulfide (C352–355) to be intact and C150, C166, and C349 to be AMS-modi- fied. Again, it should be noted that this species, when blocked with NEM rather than AMS, migrates faster than the C90A/C349A mutant [Fig. 4(C), as- terisk], suggesting that it nevertheless contains a long-range disulfide. In this set of experiments, we did not observe any non-native disulfides, that is, those not present in the Ero1p crystal structure. To complement this information, we did a sec- ond LC-MS/MS experiment, in which Ero1p species in gel slices from the same AMS-blocked experiment were reduced with dithiothreitol (DTT) and alky- lated with iodoacetamide before in-gel proteolysis. This experiment revealed which cysteines were par- ticipating in disulfides (carbamidomethylated) and which were reduced (AMS-modified) at each time point [Fig. 5(C)]. Disulfide connectivity information was lost, but peptide recovery could in principle be altered or improved by reduction of the disulfide bonded peptides. This experiment was performed both on the wild-type enzyme and on Ero1p C150A/ C295A. For both versions, the first Ero1p thiol to be detected in AMS-modified form was C143. The spe- cies showing reduced and modified C143 increased in intensity between 0.1 and 0.5 min for wild-type Ero1p but was present as the dominant band already at the 0.1 min time point of the Ero1p 1868 PROTEINSCIENCE.ORG Activation of Sulfhydryl Oxidase Ero1p C150A/C295A reaction. Analysis of the highest band in the wild-type reaction revealed populations of spe- cies with C150 and C349 reduced as well. The high- est band of the C150A/C295A mutant also showed reduction of C349. Overall, the mass spectrometry data revealed a progressive loss of disulfides in wild- type Ero1p in the order C100AC105, C143AC166, C150AC295, and perhaps C90AC349. For the C150A/C295A mutant, the most striking and consist- ent observation was the appearance of C143 as an AMS-modified thiol at early timepoints. The above mass spectrometry analyses were conducted on Ero1p after reaction with a model sub- strate. To determine how a native substrate affects noncatalytic disulfides of Ero1p, we performed com- parable experiments using yeast Protein Disulfide Isomerase (Pdi1p). Ero1p was mixed with reduced Pdi1p (Pdi1pred), and the reaction was blocked with either NEM or AMS. Because Pdi1p migrates to a similar position on electrophoretic gels as the more reduced species of Ero1p observed in the experi- ments with Trxred, Ero1p was visualized with a fluo- rescent stain specific to poly-histidine tags, and an untagged version of Pdi1p was used. Wild-type Ero1p blocked with NEM at various points in the reaction showed no species with retarded migration [Fig. 6(A), left panel], whereas blocking with AMS revealed a minor population of modified Ero1p [Fig. 6(B), left panel]. In contrast, a significant population of Ero1p C150A/C295A was shifted upon addition of Pdi1pred, after blocking with either NEM or AMS [Fig. 6(A,B), right panels]. This observation suggests that a disulfide between nonvicinal cysteines is reduced in the slowly migrating fraction. Disulfide rearrangement is also a formal possibility. In these studies, no Ero1p species similar to the most slowly migrating species seen in the reactions with Trxred were observed, though Ero1p and Ero1p C150A/ C295A did oxidize Pdi1pred under the conditions of the experiment [Fig. 6(C)]. Therefore, Ero1p is func- tional on Pdi1p without quantitative reduction of multiple regulatory disulfides. After establishing that the partially reduced species of Ero1p or Ero1p C150A/C295A obtained upon reaction with Pdi1pred do not comigrate with Pdi1p itself (55 kD), we aimed to identify the disul- fide that becomes reduced in the C150A/C295A mu- tant. We incubated Ero1p C150A/C295A for short times with Pdi1pred, blocked the reaction with either Figure 5. Mass spectrometric identification of partially reduced Ero1p species. A: The upper panels of Figure 4 (reaction time-courses of wild-type and C150A/C295A Ero1p blocked with AMS) are reproduced, with the bands that were subjected to mass spectrometry indicated by ovals. B: A summary of the mass spectrometry results from each band is indicated on a map of the Ero1p primary structure. All disulfides observed in the Ero1p crystal structures were found in the time zero band (i.e., in the absence of substrate), as indicated by the linked open circles, and no non-native disulfides were detected. The homogeneity of the time zero band, and the fact that this species was obtained from pure protein stock rather than from aliquots removed from enzymatic reactions, may explain the improved peptide coverage compared to the other bands analyzed. Reduction of C143AC166, indicated by unpaired dark circles, was seen in bands that appeared earlier than those showing reduction of C150AC295. It should also be noted that reduction of C143AC166 does not result in a mobility change in the wild-type enzyme blocked with NEM, as can be seen by comparing Figures 4(C) and 5(A). Reduction of this disulfide does, however, cause a shift when it occurs in the C150A/ C295A background. C: The indicated bands were subject to reduction and alkylation with iodoacetamide before proteolytic cleavage and mass spectrometry analysis. Open circles indicate cysteines that were detected as modified by iodoacetamide, whereas dark circles indicate cysteines that were detected as being modified by AMS, and thus had been reduced during the reaction with Trxred. Half-filled circles represent cysteines detected in both states in the same experiment. Heldman et al. PROTEIN SCIENCE VOL 19:1863—1876 1869 NEM or AMS, separated the proteins by SDS-PAGE, and stained with Coomassie (gel not shown). In the AMS-treated sample, the higher band derived from the Ero1p mutant comigrated with a Pdi1p contami- nant or degradation product and was not analyzed. In contrast, in the NEM-treated sample, the band apparently corresponding to that labeled with an as- terisk in the right panel of Figure 6(A) could be identified and excised. Tandem mass spectrometry analysis of this band showed both C143 and C166 in NEM-modified form, demonstrating that C143AC166 had been reduced by Pdi1pred. A single six-amino acid peptide containing reduced and modi- fied C90 was also observed, although the species an- alyzed migrated much more quickly in the gel than the C90A/C349A mutant and is thus likely to con- tain a long-range disulfide. Displacement of C90 by attack of C143, or more likely C166 (see below), on C349 might occur in a fraction of the Ero1p mole- cules, yielding a species with similar hydrodynamic radius to that having C143 and C166 reduced and the remaining disulfides intact. Ero1p C143A/C166A structure The identification of C143AC166 as the disulfide that is reduced much more rapidly in the hyperac- tive C150A/C295A Ero1p mutant prompted us to investigate the structural changes that occur in Ero1p upon removal of this disulfide. The Ero1p mu- tant C143A/C166A was produced in E. coli with 20% the yields obtained for wild type. Crystals were obtained in the primitive hexagonal form with unit cell dimensions similar to those of crystals obtained with wild-type Ero1p in this space group (Table I). The Ero1p C143A/C166A structure was refined using diffraction data to 3.2 A˚ resolution. Figure 6. The reaction between 2 lM Ero1p or Ero1p C150A/C295A and 75 lM Pdi1pred was blocked with (A) NEM or (B) AMS after various times and subjected to SDS-PAGE under nonreducing conditions. The band indicated by an open arrowhead in the wild-type Ero1p time course is a contaminant from the Pdi1p preparation, as it appears also in the Pdi1p control lane. A significant fraction of Ero1p C150A/C295A shows retarded mobility at early time points when blocked with either NEM or AMS (asterisks). A band corresponding to that marked with an asterisk in NEM-blocked Ero1p C150A/C295A was analyzed by LC-MS/MS, and the C143AC166 disulfide was found to be reduced. (C) Oxidation of Pdi1p by Ero1p and Ero1p C150A/C295A under similar conditions as the experiments in (A) and (B). Pdi1pred was mixed with wild-type or mutant Ero1p. Aliquots were removed at various times and reacted with PEG-maleimide of 5 kD. Mal-PEG modified Pdi1p species are indicated in the top portion of the gel. The band labeled Pdi1pox has both active sites oxidized and was thus resistant to PEGylation. 1870 PROTEINSCIENCE.ORG Activation of Sulfhydryl Oxidase Ero1p As for the C150A/C295A mutant, local rear- rangements of the polypeptide chain were seen in the C143A/C166A around the mutation site. How- ever, in contrast to the C150A/C295A mutant, in which both polypeptide segments containing the cys- teine-to-alanine mutations acquired more regular secondary structure upon mutation, the C143A/ C166A double mutant is more disordered than wild type [Fig. 7(A)]. A cluster of acidic residues (D137, D138, D140, D141, and E142) immediately upstream of C143A assumes an alternate conformation [Fig. 7(B)], without an increase in secondary structure. The region surrounding C166A loses its connection to the helical core of the enzyme, and the poly- peptide chain between residues 154 and 175 cannot be traced in electron density maps due to poor elec- tron density [Fig. 7(C)]. In the wild-type Ero1p structures and in C150A/C295A, electron density corresponding to the region around residues 155– 165 was also poor or uninterpretable. However, the disordered region is significantly extended, in the direction of the active site, in the C143A/C166A mutant. Kinetics studies of single and double mutants of the C143AC166 disulfide The structural changes seen in the Ero1p C143A/ C166A mutant are not sufficient to activate the enzyme; this mutant, when examined previously, was found not to be hyperactive but rather slightly slower than wild-type Ero1p.4 To determine whether the thiol form of either cysteine participating in this bond may have a direct role in activation, we con- structed the single cysteine-to-alanine mutants C143A and C166A. The C166A mutant behaved sim- ilarly to the C143A/C166A double mutant. However, the rate of oxidation of Trxred by the C143A mutant was found to be indistinguishable from the rate of oxidation by the de-regulated C150A/C295A mutant in a gel-based assay [Fig. 7(D)]. This observation Figure 7. C143 and C166 play different roles in the Ero1p structure. A: Stereo image of a superposition of the structures of wild-type (colored beige) and C143A/C166A Ero1p (dark red). Disulfides and the carboxy terminus are labeled. The endpoints of a large missing loop are indicated by circles and labeled with the number of the last modeled residue in each case. Mutation of C143 and C166 increases the length of the disordered loop such that an additional 10 residues cannot be modeled. To emphasize the missing loop, the view is different from in Figure 3, but it corresponds to the summary Figure 8. B: A close-up view of the region around C143 shows the local structural differences between the wild-type Ero1p structure and the C143A/C166A mutant. The Cb atom of the alanine at position 143 in the mutant is shown in blue and labeled. Side chains of three acidic residues immediately upstream of C143 are shown as blue and red sticks. To illustrate the extent of the conformational differences in this area, the side chain of D140 is labeled in both the mutant and wild-type structures. C: A 2Fo-Fc electron density map, contoured at 1r, illustrates the lack of interpretable density corresponding to the loop containing the C166A mutation. D: Activities of single mutants disrupting the C143AC166 disulfide obtained from a gel-based Trx oxidation assay. Oxidation of Trxred by the indicated Ero1p variants was blocked at various time points by addition of PEG-maleimide 5 kD. Oxidized Trx is resistant to PEG modification and thus migrates faster than modified Trx in SDS-PAGE. The band intensities of oxidized and reduced/modified Trx were quantified and plotted as fraction reduced. The C143A mutant, which retains C166, is hyperactive, but the C166A mutant, retaining C143, is not. This finding points to a role for the C166 thiol in Ero1p activation. Heldman et al. PROTEIN SCIENCE VOL 19:1863—1876 1871 suggests that liberating C166 is an important step in activating Ero1p. However, the mutagenesis experiment did not rule out the possibility that the C143A mutant is hyperactive for the trivial reason that the unpaired C166 disrupts the C150AC295 di- sulfide and produces a species that mimics C150A/ C295A. Indeed, mass spectrometry analysis of the C143A mutant showed some C295 in reduced and alkylated form, suggesting that C166 displaces it and forms a disulfide with C150 in a fraction of the molecules. It should be noted that this species was not observed in any experiment in which C143AC166 became reduced during an enzymatic reaction, suggesting that it is an artifact of lengthy exposure of C166 during enzyme preparation and purification. The activity of C143A Ero1p on Pdi1pred was greater than that of wild-type Ero1p but not as rapid as C150A/C295A (data not shown). This observation may indicate that a non-native di- sulfide bond formed when C143 is missing is reduced less effectively by Pdi1pred than by Trxred. Together, the study of single-cysteine mutants C143A and C166A supports the conclusion based on crystallographic studies that the C166 region has a large range of motion when freed from C143, and suggests a specific role for C166. Discussion The mechanism by which Ero1p activity is controlled by encounter with reducing substrates is directly related to the ability of the enzyme to maintain re- dox balance in the ER. The various Ero1p disulfides have distinct roles in catalysis and control of enzyme activity, and the experiments presented here were designed to dissect these roles. The prior observation that eliminating the C150AC295 disulfide of yeast Ero1p increases enzyme activity, coupled with the ob- servation of cascaded reduction of Ero1p disulfides in gel assays using Trxred as a substrate, suggested that reduction of C150AC295 is part of the series of events that results in activation of the enzyme.4 The reduction of C150AC295 was considered as a possible early event in the cascade, whereas the longest range disulfide (i.e., C90AC349) was surmised to open later in the reaction.4 One puzzle in this model was the implication that the C150AC295 disulfide should be reduced rapidly in the C90A/C349A Ero1p mutant to yield a fully activated enzyme. The C90A/C349A mu- tant would thus be expected to have a short lag phase and be hyperactive, but instead it has only a slightly shortened lag phase and essentially wild-type activity. The LC-MS/MS results presented here consis- tently showed reduction of the C143AC166 disulfide preceding reduction of other noncatalytic disulfides. The opening of C143AC166 first in the cascade was unexpected, considering the lack of solvent exposure of this disulfide in the Ero1p crystal structure6 and the observation that the C143A/C166A double muta- tion did not activate Ero1p.4 The disulfide mapping further indicated that reduction of C150AC295 is actually a late step in Ero1p activity assays on Trxred. This result implies that the increased activity of the C150A/C295A mutant cannot simply be ascribed to removal of an initial hurdle in the reduc- tive activation process. Whether reduction of the C150AC295 disulfide is required as the second step in the activation of Ero1p is difficult to test using Trxred as a substrate, since reduction of the C150AC295 disulfide inevitably followed reduction of C143AC166 temporally (Fig. 5). In contrast to the relative ease of eliminating a disulfide bond by mu- tagenesis, specifically stabilizing a disulfide to test the effect of a lack of its reduction presents a major experimental challenge. Remarkably, the observations presented here with the native substrate Pdi1pred suggest that reduction of multiple Ero1p disulfides may not be required before Ero1p can oxidize substrate. When Pdi1pred was mixed with wild-type Ero1p, no shifted bands corresponding to reduction of C150AC295 (or the longer range disulfide C90AC349) were observed at any point in the reaction [Fig. 6(A), left panel], although wild-type Ero1p does oxidize at least one of the domains of Pdi1pred in this time frame [Fig. 6(C), left panel]. Stoichiometric reduction of C150AC295 is thus not an obligatory step in generating active Ero1p. It is possible, however, that Pdi1pred reduces multiple disulfides in an undetectable subpopulation of Ero1p, and this fraction is then extremely active on Pdi1pred and performs all the substrate oxidation observed, while the vast majority of Ero1p molecules remain oxidized and inactive. We find this explana- tion unlikely. If Pdi1pred were inefficient at activating Ero1p but then served as an excellent substrate of the activated fraction, lag phase kinetics would be observed for oxidation of Pdi1pred. In fact, the lag phase is more pronounced during oxidation of Trxred than of Pdi1pred. ince reduction of C150AC295 does not seem to be required for Ero1p turnover, we are left with the question of why an enzyme variant lacking this disul- fide shows dramatically increased activity, on both model and native substrates.4,12 The major finding presented here is that the C150AC295 disulfide affects the reactivity of the neighboring disulfide, between C143 and C166, and that opening of the lat- ter correlates with exit from the lag phase and increased oxidase activity. The non-native substrate Trxred reduces C143AC166 in wild-type Ero1p appa- rently to completeness, whereas the native substrate Pdi1pred is inefficient at reducing this disulfide or at maintaining it reduced as Ero1p competes to re-oxi- dize it. When C150AC295 is eliminated by mutation, the C143AC166 disulfide is reduced more rapidly by Trxred and more extensively by Pdi1pred. The greater 1872 PROTEINSCIENCE.ORG Activation of Sulfhydryl Oxidase Ero1p portion of the enzyme with the C143AC166 disulfide reduced could explain the greater activity of the C150A/C295A mutant compared to wild-type Ero1p in the oxidation of Pdi1pred. Differences in substrate oxidation kinetics have also been observed for cysteine mutants in recombi- nant human Ero1a. In particular, mutation of Ero1a C131 increased slightly the rate of oxidation of Pdi1p in vitro11 and increased cellular GSSG levels.5 The explanation for increased activity of the human Ero1a C131A mutant is distinct from the mechanism outlined herein for Ero1p C150A/C295A. Ero1a C131 was proposed to form an inhibitory disulfide with Ero1a C94,11 which is one of the cysteines partici- pating in the shuttle disulfide of the mammalian enzyme. The increased activity of the Ero1a C131A mutant was proposed to reflect loss of the inhibitory C94AC131 disulfide and ability of the shuttle disul- fide between C94 and C99 to form properly. Although yeast Ero1p does not oxidize Trxred signifi- cantly at time zero of the reaction, an inhibitory di- sulfide formed between a regulatory and a shuttle cysteine cannot explain this inhibition; an intact shuttle disulfide was observed in the crystal struc- ture of yeast Ero1p and by disulfide mapping using LC-MS/MS [Fig. 5(B) and Supporting information]. If a non-native disulfide were to be present during reactions of yeast Ero1p, it would most likely be activating rather than inhibitory. However, numer- ous mass spectrometry samples of Ero1p trapped during reactions with Trxred have failed to reveal non-native disulfides, despite high peptide coverage of the protein sequence and excellent recovery of native disulfides. If such species exist, they may be transient or poorly populated. If not by engaging shuttle cysteines, then how might Ero1p regulatory disulfides restrain enzyme activity? For both C143AC166 and C150AC295, the disulfide constrains the polypeptide to conformations different from those observed when the disulfides are removed. In particular, we observed that the entire loop between residues 155 and 175 becomes disordered when C166 is not tethered to C143, as shown in the structure of the Ero1p C143A/C166A mutant. The dramatically extended range of motion of this loop may allow it to impact the active site. As the C143A/C166A double mutant is not hyperactive, however, C166 may be specifically required for these effects. What changes in the active-site region might increase the rate of catalysis? One hypothesis is that the active-site disulfide (C352–C355) of Evolp is poorly accessible for redox communication with the shuttle disulfide (C100AC105). In the structure of another yeast sulfhydryl oxidase, Erv2p,13 the shut- tle disulfide was observed within dithiol-disulfide exchange distance of the active-site disulfide. In con- trast, the active-site disulfide of Ero1p is less approachable.6 Although the shuttle disulfide is closer to the active site in one of the Ero1p crystal forms than in the other, it is still not in direct con- tact and is separated from the active site by an intervening Tyr side chain (Tyr 191). The low sol- vent accessibility of the active-site disulfide in Ero1p is not altered by in silico removal of the shuttle di- sulfide loop (data not shown), suggesting that it is buried by other parts of the structure, occluding it from solvent as well as from the shuttle disulfide. Conformational or dynamic changes, propagating from reduction of regulatory disulfides, may facili- tate redox communication between the solvent acces- sible shuttle disulfide and the buried active-site di- sulfide, forming a complete redox path from substrate to FAD. In conclusion, we determined that the C143AC166 disulfide is the first regulatory disulfide to be reduced during Ero1p activation and that the facile activation of the C150A/C295A mutant can be explained by its increased susceptibility to reduction of C143AC166 (Fig. 8). The presence or absence of the C150AC295 disulfide 35 A˚ away from the active site thus affects the fate of the key C143AC166 di- sulfide 25 A˚ from the active site, and elimination of the C143AC166 disulfide allows conformational changes that may finally propagate to the active site itself, with a consequent increase in thiol oxidase activity. Materials and Methods Enzyme and substrate preparation Ero1p, Ero1p C143A, Ero1p C166A, Ero1p C150A/ C295A, and Ero1p C143A/C166A, all spanning Figure 8. Summary of roles for regulatory disulfides distant from the Ero1p active site. Reduction of the C150AC295 disulfide is not an early step in Ero1p activation. Instead, the presence of this disulfide appears to protect the neighboring C143AC166 disulfide from reduction. Reduction of C143AC166 may be a key event in Ero1p activation, allowing the polypeptide chain in the vicinity of C166 in particular to sample new conformations (the arrow indicates schematically putative motion in the direction of the active site). Though C143AC166 is nearly 25 A˚ from the active-site (C352AC355) disulfide in the ground-state Ero1p structure, a liberated C166 would be present on a segment of polypeptide that may be loosely tethered enough to bridge this distance. Heldman et al. PROTEIN SCIENCE VOL 19:1863—1876 1873 residues amino acid 56–424 of the yeast protein, were expressed in the Origami B(DE3) plysS E. coli strain (Novagen) downstream of glutathione S-trans- ferase (GST) and an internal His6 tag using a modi- fied pGEX-4T1 plasmid (Amersham).6 Wild-type Ero1p spanning residues 10–424 was produced from a pGEX-4T1 vector without the additional His6 tag. Bacteria were grown at 37C to an optical density of 0.6 at 600 nm, at which point isopropyl-1-thio-b-D- galactopyranoside was added to a final concentration of 0.5 mM to induce protein expression. The growth temperature was shifted to 25C, and cells were har- vested 12–16 h later. After cell lysis, proteins were purified using Ni-NTA agarose (for constructs con- taining His6 tags), cleaved with thrombin to remove the GST, and re-purified over Ni-NTA. The longer Ero1p construct, lacking the His6 tag, was purified using glutathione sepharose beads, and thrombin cleavage was performed on the column to release Ero1p. The enzymes were then concentrated and run on a HiLoad 16/60 Superdex 75 prep grade size exclusion column monitored at 280 and 450 nm. The peaks corresponding to monomeric protein were col- lected. Enzymes used for crystallization were dia- lyzed against 10 mM Tris, 25 mM NaCl, pH 8 and concentrated to 400 lM (18 mg/mL). Trx was expressed, purified, and reduced as pre- viously described,6 except that Triton X-100 was not used. For experiments in which gels were stained with Invision His6-tag stain, Pdi1p was produced in BL21(DE3) plysS E. coli cells as a fusion protein with GST using the pGEX-4T1 plasmid, purified over glutathione-sepharose beads, cleaved with thrombin, and re-applied to glutathione sepharose to remove the GST. For other experiments, Pdi1p was produced with a carboxy-terminal His6-tag as described.14 For oxidation assays, Trx was reduced with 100 mM DTT for 1 h and then desalted using a PD-10 column (GE Healthcare) equilibrated in 50 mM phosphate buffer, pH 7.5, 65 mM NaCl, 0.5 mM EDTA. Pdi1p was reduced by incubating with 10 mM GSH from a stock titrated to pH 7.0. Proteins were then desalted on a PD-10 column equilibrated in 50 mM phosphate buffer, pH 7.5, 300 mM NaCl, 0.5 mM EDTA. The concentration of reduced protein thiols was determined using Ellman’s assay.15 Crystallization and structure determination Crystals of Ero1p C150A/C295A were grown by hanging drop vapor diffusion at 20C in 100 mM cac- odylic acid pH 6–6.5, 9–15 mM cadmium sulfate, 2% methanol, 2% ethanol, and 0.8–1.6M sodium acetate. These crystals were of space group C2221. Crystals of Ero1p C143A/C166A were grown by hanging drop vapor diffusion at 20C in 100 mM cacodylic acid pH 6–6.5, 12–14 mM cadmium sulfate, 2% methanol, 2% ethanol, and 1.6M sodium acetate. These crystals were of space group P62. Before flash freezing, crys- tals were soaked in mother liquor with 15% ethylene glycol for a few minutes and then transferred to a 1:1 mixture of mineral oil and paratone oil (Exxon). Diffraction data for Ero1p C150A/C295A were col- lected at the European Synchrotron Radiation Facil- ity beamline ID14-3. Data for Ero1p C143A/C166A were collected using a RU-H3R generator (Rigaku) equipped with an Raxis IVþþ image plate system and osmic mirrors. Data were processed using HKL2000.16 Phases were calculated for the C150A/ C295A mutant using the wild-type Ero1p structure after removal of residues 146–166 and 291–302, spanning the mutated cysteines. Phases were calcu- lated for the C143A/C166A mutant after removal of residues 91–121 and 136–175. Rigid body refinement was performed before generating electron density maps. The Ero1p C150A/C295A and Ero1p C143A/ C166A models was rebuilt in O17 and Coot,18 respec- tively, and refined using CNS.19 The quality of the final models were verified using MolProbity.20 Stopped-flow fluorescence Studies were conducted at 25C on an Applied Pho- tophysics stopped-flow apparatus fitted with an an- aerobic adaptor. Exogenous FAD (100 lM) rather than oxygen was used as a terminal electron acceptor, since changes in fluorescence can be moni- tored as FAD is reduced to FADH2. Excitation was at a wavelength of 450 nm, and an emission cut-off filter of 495 nm was used. Enzyme and substrate preparations were made anaerobic by N2 bubbling and pipetting in N2 atmosphere in a glove box. To insure anaerobic conditions, solutions were supple- mented with 0.1% w/v glucose and trace amounts of glucose oxidase and catalase. Samples were trans- ferred from the anaerobic chamber in gas-tight syringes (Hamilton). Equal volumes of enzyme and Trxred (prepared as described above) in 50 mM phos- phate buffer, pH 7.5, 65 mM NaCl, and 1 mM EDTA were injected into the mixing/detection chamber to achieve final concentrations of 1 lM enzyme and 50 lM substrate. Gel-based oxidation assays Oxidation reactions used to analyze Ero1p electro- phoretic mobility were performed at an enzyme (wild-type or mutant) concentration of 2 lM. Trxred was present at a final concentration of 100 lM (fully reduced protein), and Pdi1pred at a concentration of 300 lM thiols (75 lM protein). Samples from the reactions were taken at different time points and quenched by adding 1:4 v/v sample buffer (125 mM Tris, pH 6.8, 5% SDS, 50% glycerol, 0.1% bromphe- nol blue) containing 5 mM AMS (Molecular Probes), 100 mM NEM, or 1 mM PEG-maleimide of 5 kD (Nektar Therapeutics). These samples were then run 1874 PROTEINSCIENCE.ORG Activation of Sulfhydryl Oxidase Ero1p on 15% denaturing polyacrylamide gels [Figs. 4 and 5(A)], or 12% gels for the kinetics of Pdi1pred oxida- tion [Fig. 6(C)], which were then stained with Coo- massie. Gels used to visualize Ero1p or Ero1p C150A/C295A in reactions with Pdi1pred were stained with InVisionTM His-tag In-gel Stain (Invi- trogen) and scanned using a FLA-5100 fluorescent image analyzer (FUJI) with a 532 nm LPG filter. Reduced samples [Fig. 4(B)] were run with an excess of DTT. Rates of oxidation of Trxred by various mutants of Ero1p [Fig. 7(D)] were quantified using 0.5 lM enzyme and 100 lM Trxred. At each time point, 10 lL were removed and mixed with 10 lL 5 mM PEG-mal- eimide 5 kDa in 2% SDS, 50 mM Tris, pH 6.8, 0.1% bromphenol blue, 20% glycerol. Samples were applied to a 15% denaturing polyacrylamide gel, and Trxred and Trxox were visualized after separation with Coo- massie stain. Band intensities were determined using the ImageQuant 5.0 program. In-gel digestion Protein bands were excised from SDS gels that had been stained with Coomassie and destained using multiple washings with 40% methanol and 10% ace- tic acid. The protein bands were either reduced, alkylated, and in-gel digested as described,21 or else digested directly without reduction and alkylation. Digestions were performed in 50 mM ammonium bi- carbonate at 37C using various combinations of the following sequencing grade proteases (Roche Diag- nostics): bovine trypsin, bovine chymotrypsin, and Pseudomonas asp-N, all at a concentration of 12.5 ng/lL. Peptide mixtures were extracted from the gels with 80% CH3CN, 1% CF3COOH, and the or- ganic solvent was evaporated in a vacuum centri- fuge. The resulting peptide mixtures were reconsti- tuted in 80% formic acid and immediately diluted 1:10 with Milli-Q water before mass spectrometry analysis. Mass spectrometry Samples were analyzed in an LTQ Orbitrap (Thermo Fisher Scientific) operated in the positive ion mode and equipped with a nanoelectrospray ion source. Peptide mixtures were separated by online reversed- phase nanoscale capillary LC and analyzed by tan- dem mass spectrometry (LC-MS/MS). For the LC-MS/ MS, samples were injected onto a 15 cm reversed phase spraying fused-silica capillary column (inner diameter 75 lm) made in-house and packed with 3 lm ReproSil-Pur C18AQ media (Dr. Maisch GmbH, Ammerbuch-Entringen, Germany) using an UltiMate 3000 Capillary/Nano LC System (LC Packings, Dio- nex). The LC setup was connected to the Orbitrap. The flow rate through the column was 250 nL/min, and the injection volume was 5 lL. Peptides were separated with a 50 min gradient from 5 to 65% ace- tonitrile (buffer A: 5% acetonitrile, 0.1% formic acid, 0.005% TFA; buffer B: 90% acetonitrile, 0.2% formic acid, 0.005% TFA). The voltage applied to the union in order to produce an electrospray was 1.2 kV. The mass spectrometer was operated in the data-depend- ent mode. Survey MS scans were acquired in the Orbitrap with the resolution set to a value of 60,000. Up to the six most intense ions per scan were frag- mented and analyzed in the linear trap. For the anal- ysis of peptides, survey scans were recorded in the FT-mode followed by data-dependent collision-induced dissociation (CID) of the six most intense ions in the linear ion trap (LTQ). Raw spectra were processed using open-source software DTASuperCharge (http:// msquant.sourceforge.net). The data were searched with MASCOT (Matrix Science, London, UK) against a Swiss-prot database with manually incorporated Ero1p or mutant protein (Ero1p C150A/C295A). Search parameters included variable modifications of 57.02146 Da (carboxyamidomethylation) on Cys, 510.04028 Da (AMS; hydrolyzed) on Cys, 15.99491 Da (oxidation) in Met and 0.984016 Da (deamidation) on Asn and Gln. The search parameters were as fol- lows: maximum two missed cleavages, initial precur- sor ion mass tolerance 10 ppm, fragment ion mass tolerance 0.6 Da. The identity of the peptides were concluded from the detected CID products by Mascot and Sequest software and confirmed by manual inspection of the fragmentation series. The program MassMatrix was used for disulfide identification.22 Acknowledgments The authors thank members of the Weizmann Insti- tute Biological Mass Spectrometry Unit and Sarah J. Weisberg for assistance with mass spectrometry anal- ysis. Moran Bentzur and Gideon Schreiber assisted in the stopped-flow studies. Coordinates and structure factors for Ero1p C150A/C295A and Ero1p C143A/ C166A have been deposited in the Protein Data Bank with accession codes 3M31 and 3NVJ. References 1. Jessop CE, Chakravarthi S, Watkins RH, Bulleid NJ (2004) Oxidative protein folding in the mammalian endoplasmic reticulum. Biochem Soc Trans 32:655–658. 2. Frand AR, Kaiser CA (1998) The ERO1 gene of yeast is required for oxidation of protein dithiols in the endo- plasmic reticulum. Mol Cell 1:161–170. 3. Pollard MG, Travers KJ, Weissman JS (1998) Ero1p: a novel and ubiquitous protein with an essential role in oxidative protein folding in the endoplasmic reticulum. Mol Cell 1:171–182. 4. Sevier CS, Qu H, Heldman N, Gross E, Fass D, Kaiser CA (2007) Modulation of cellular disulfide bond forma- tion and the ER redox environment by feedback regula- tion of Ero1. Cell 129:333–344. 5. Appenzeller-Herzog C, Riemer J, Christensen B, Søren- sen ES, Ellgaard L (2008) A novel disulphide switch mechanism in Ero1a balances ER oxidation in human cells. EMBO J 27:2977–2987. Heldman et al. PROTEIN SCIENCE VOL 19:1863—1876 1875 6. Gross E, Kastner DB, Kaiser CA, Fass D (2004) Struc- ture of Ero1p, source of disulfide bonds for oxidative protein folding in the cell. Cell 117:601–610. 7. Gross E, Sevier CS, Heldman N, Vitu E, Bentzur M, Kaiser CA, Thorpe C, Fass D (2006) Generating disul- fides enzymatically: reaction products and electron acceptors of the endoplasmic reticulum thiol oxidase Ero1p. Proc Natl Acad Sci USA 103:299–304. 8. Lee B, Richards FM (1971) The interpretation of pro- tein structures: estimation of static accessibility. J Mol Biol 55:379–400. 9. Frand AR, Kaiser CA (2000) Two pairs of conserved cysteines are required for the oxidative activity of Ero1p in protein disulfide bond formation in the endo- plasmic reticulum. Mol Biol Cell 11:2833–2843. 10. Sevier CS, Kaiser CA (2006) Disulfide transfer between two conserved cysteine pairs imparts selectivity to pro- tein oxidation by Ero1. Mol Biol Cell 17:2256–2266. 11. Baker KM, Chakravarthi S, Langton KP, Sheppard AM, Lu H, Bulleid NJ (2008) Low reduction potential of Ero1a regulatory disulphides ensures tight control of substrate oxidation. EMBO J 27:2988–2997. 12. Vitu E, Kim S, Sevier CS, Lutzky O, Heldman N, Bent- zur M, Unger T, Yona M, Kaiser CA, Fass D (2010) Oxidative activity of yeast Ero1p on protein disulfide isomerase and related oxidoreductases of the endoplas- mic reticulum. J Biol Chem 285:18155–18165. 13. Gross E, Sevier CS, Vala A, Kaiser CA, Fass D (2002) A new FAD-binding fold and intersubunit disulfide shuttle in the thiol oxidase Erv2p. Nat Struct Biol 9: 61–67. 14. Westphal V, Darby NJ, Winther JR (1999) Functional properties of the two redox-active sites in yeast protein disulfide isomerase in vitro and in vivo. J Mol Biol 286: 1229–1239. 15. Ellman GL. (1958) A colorimetric method for determin- ing low concentrations of mercaptans. Arch Biochem Biophys 74:443–450. 16. Otwinowski Z, Minor W (1977) Processing of X-ray dif- fraction data collected in oscillation mode. Methods Enzymol 276:307–326. 17. Jones TA, Zou JY, Cowan SW, Kjeldgaard M (1991) Improved methods for building protein models in elec- tron density maps and the location of errors in these models. Acta Crystallogr A 47:110–119. 18. Emsley P, Cowtan K (2005) Coot: model-building tools for molecular graphics. Acta Crystallogr D 60: 2126–2132. 19. Brunger AT, Adams PD, Clore GM, DeLano WL, Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges M, Pannu NS, Read RJ, Rice LM, Simonson T, Warren EL (1998) Crystallography and NMR system: a new software suite for macromolecular structure determina- tion. Acta Crystallogr D 54:905–921. 20. Davis IW, Leaver-Fay A, Chen VB, Block JN, Kapral GJ, Wang X, Murray LW, Arendall WB, III, Snoeyink J, Richardson JS, Richardson DC (2007) MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nuc Acids Res 35:W375–W383. 21. Shevchenko A, Wilm M, Vorm O, Mann M (1996) Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels. Anal Chem 68:850–858. 22. Xu H, Zhang L, Freitas MA (2008) Identification and characterization of disulfide bonds in proteins and pep- tides from tandem MS data by use of the MassMatrix MS/MS search engine. J Proteome Res 7:138–144. 1876 PROTEINSCIENCE.ORG Activation of Sulfhydryl Oxidase Ero1p
3M32
Structural Insight into Methyl-Coenzyme M Reductase Chemistry using Coenzyme B Analogues
Structural Insight into Methyl-Coenzyme M Reductase Chemistry using Coenzyme B Analogues,†,‡ Peder E. Cedervall§,#, Mishtu Dey||,#,⊥, Arwen R. Pearson§,⊥, Stephen W. Ragsdale||, and Carrie M. Wilmot*,||,§ § Department of Biochemistry, Molecular Biology and Biophysics, University of Minnesota, Minneapolis, Minnesota 55455 || Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109 Abstract Methyl-coenzyme M reductase (MCR) catalyzes the final and rate-limiting step in methane biogenesis; the reduction of methyl-coenzyme M (methyl-SCoM) by coenzyme B (CoBSH) to methane and a heterodisulfide (CoBS-SCoM). Crystallographic studies show that the active site is deeply buried within the enzyme, and contains a highly reduced nickel-tetrapyrrole, coenzyme F430. Methyl-SCoM must enter the active site prior to CoBSH, as species derived from analogues of methyl-SCoM are always observed bound to the F430 nickel in the deepest part of the 30 Å long substrate channel that leads from the protein surface to the active site. The seven-carbon mercaptoalkanoyl chain of CoBSH binds within a 16 Å predominantly hydrophobic part of the channel close to F430, with the CoBSH thiolate lying closest to the nickel at a distance of 8.8 Å. It has previously been suggested that binding of CoBSH initiates catalysis by inducing a conformational change that moves methyl-SCoM closer to the nickel promoting cleavage of the C- S bond of methyl-SCoM. In order to better understand the structural role of CoBSH early in the MCR mechanism, we have determined crystal structures of MCR in complex with four different CoBSH analogues; pentanoyl-, hexanoyl-, octanoyl- and nonanoyl- derivatives of CoBSH (CoB5SH, CoB6SH, CoB8SH and CoB9SH respectively). The data presented here reveal that the shorter CoB5SH mercaptoalkanoyl chain overlays with that of CoBSH, but terminates two units short of the CoBSH thiolate position. In contrast, the mercaptoalkanoyl chain of CoB6SH adopts a different conformation, such that its thiolate is coincident with the position of the CoBSH thiolate. †This work was supported by a Department of Energy grant DE-FG02-08ER15931 to S.W.R and supplement to C.M.W. and a Minnesota Partnership for Biotechnology and Medical Genomics grant SPAP-05-0013-P-FY06. ‡Co-ordinates and structure factors have been deposited in the Protein Data Bank as entries 3m1v (MCRHSCoM); 3m2r (MCRCoB5SH); 3m2u (MCRCoB6SH); 3m2v (MCRCoB8SH); 3m30 (MCRCoB9SH). *Address correspondence to: Carrie M. Wilmot, Tel: (612) 624-2406, wilmo004@umn.edu. ⊥Current addresses: Mishtu Dey, Department of Chemistry, Massachusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, Massachusetts 02139; Arwen Pearson, Astbury Centre for Structural Molecular Biology, Institute for Molecular and Cellular Biology, Astbury Building. Leeds, LS2 9JT, U.K. #These authors contributed equally to this work. Supporting Informational material is available free of charge via the Internet at http://pubs.acs.org and contains the following: MATERIAL AND METHODS for single crystal UV-visible microspectrophotometry, X-ray photoreduction experiment and X-ray crystallography of the MCR-heterodisulfide product complex (MCRCoBSH + methyl-SCoM); RESULTS AND DISCUSSION for redox changes and MCRCoBSH + methyl-SCoM crystal structure; Table S1, X-ray Data Collection, Processing and Refinement Statistics for MCRCoBSH + methyl-SCoM; Figure S1, use of Fo-Fc electron density in modelling MCRCoB8SH; Figure S2, illustration of electron density quality of MCRCoB5SH; Figure S3, solution and single crystal UV-visible spectra; Figure S4, modeling of CoB6SH and CoBSH into the electron density of MCRCoB6SH; Figure S5, partially occupied HSCoM in MCRCoB6SH; Figure S6, the active site and substrate channel of MCRCoBSH + methyl-SCoM; Figure S7, alternative conformation of Valα482 in MCRHSCoM; Figure S8, the two conformations of the gly-rich loop in MCRHSCoM; Figure S9, propagation of conformational changes in MCRHSCoM; Figure S10, the two conformations of CoB9SH in MCRCoB9SH; Figure S11, EPR spectra of MCRred1 sample; Scheme S1, scheme of the characterized forms of MCR. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 September 7. Published in final edited form as: Biochemistry. 2010 September 7; 49(35): 7683–7693. doi:10.1021/bi100458d. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript This is consistent with the observation that CoB6SH is a slow substrate. A labile water in the substrate channel was found to be a sensitive indicator for the presence of CoBSH and HSCoM. The longer CoB8SH and CoB9SH analogues can be accommodated in the active site through exclusion of this water. These analogues react with Ni(III)-methyl; a proposed MCR catalytic intermediate of methanogenesis. The CoB8SH thiolate is 2.6 Å closer to the nickel than that of CoBSH, but the additional carbon of CoB9SH only decreases the nickel thiolate distance a further 0.3 Å. Although the analogues did not induce any structural changes in the substrate channel, the thiolates appeared to preferentially bind at two distinct positions in the channel; one being the previously observed CoBSH thiolate position, and the other being at a hydrophobic annulus of residues that lines the channel proximal to the nickel. INTRODUCTION Methanogenic archaea are organisms that under strict anaerobic conditions derive energy by reducing compounds such as carbon dioxide, methylamine, acetate and methanol, to methane (1, 2). The global production of methane by these organisms is estimated at one billion tons annually. Microbially produced methane is not only a potential source of renewable energy but also a potent greenhouse gas, and as such study of this process has environmental ramifications. In these microorganisms, methyl-coenzyme M reductase (MCR)1 is the enzyme that catalyzes the final step in methanogenesis, in which the substrates methyl-coenzyme M (methyl-SCoM, 2-(methylthio)ethanesulfonate) and coenzyme B (CoBSH, N-7-mercaptoheptanoylthreonine phosphate) are converted to methane and a heterodisulfide (CoBS-SCoM) (Scheme 1) (3). MCR is a 300 kDa protein with six subunits arranged in a α2β2γ2 oligomer (4). The known crystal structures show that MCR has two active sites approximately 50 Å apart that are deeply buried within the enzyme (5). The active site pocket is comprised of residues from subunits α, α′, β and γ, with a 30 Å long substrate channel leading to the enzyme surface (Figure 1). At the heart of the active site pocket is coenzyme F430, which is a highly reduced nickel-containing tetrapyrrole (6–8). Currently sixteen distinct enzymatic and complexed states of MCR have been spectroscopically characterized (Supporting Information, Scheme S1) (9–31). In the resting active state of the enzyme, denoted MCRred1, the redox-active nickel of F430 is present in the Ni(I) state (9, 16, 32). MCR is extremely oxygen sensitive and upon oxygen exposure the enzyme enters an inactive Ni(II) state, denoted MCRred1-silent (6). In this state it cannot be converted back to the active Ni(I) form by any known reducing agent making this a challenging system to study. Additional complications involve the tight association of coenzymes to purified MCR that are not easily displaced as demonstrated by X-ray crystallographic and kinetic studies (5, 33–35). Despite the fact that MCR has been studied for decades, no true catalytic intermediate has been observed, and the actual mechanism remains elusive. Currently three general mechanistic schemes for the enzymatic reaction have been proposed, each of which posit different chemistry to initiate catalysis. Mechanism I involves Ni(I) acting as a nucleophile in an SN2-type reaction that generates Ni(III)-methyl as an intermediate (Scheme 2A) (35– 38). Mechanism II starts with methyl-SCoM undergoing homolytic cleavage at the Ni(I) to generate a methyl radical and a Ni(II)-SCoM species (Scheme 2B) (39–41). A more recently proposed mechanism III suggests protonation of coenzyme F430 promotes reductive cleavage of the methyl-SCoM thioether bond (42). 1Abbreviations: MCR, Methyl-coenzyme M reductase; methyl-SCoM, methyl-coenzyme M; CoBSH, coenzyme B; HSCoM, coenzyme M; CoBS-SCoM, heterodisulfide of coenzyme B and coenzyme M; APS, Advanced Photon Source; ASU, asymmetric unit; BPS, bromopropanesulfonate. Cedervall et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Due to the stringent requirement to exclude O2, the available MCR crystal structures are all in the inactive Ni(II) state. All but one contain CoBSH and HSCoM (demethylated methyl- SCoM, an inhibitor and substrate analogue) in the active site (PDB codes 1hbn, 1hbo, 1hbu, 1e6y, 1e6v) (5, 33, 34). Another crystal structure has bound heterodisulfide product, CoBS- SCoM (MCRsilent, PDB code 1hbm, Scheme 1 and Supporting Information, Scheme S1) (5, 33). All these structures reveal that both substrates access the active site through the same channel (Figure 1). The binding site of HSCoM (and presumably methyl-SCoM) is more deeply buried within the enzyme, and so it must enter prior to CoBSH for productive chemistry to occur. As binding of CoBSH in the absence of co-substrate would be inhibitory, it was suggested that a conformational change upon methyl-SCoM binding might lower the Kd for CoBSH, and thus promote an ordered mechanism. Furthermore it has been suggested that CoBSH binding induces a conformational change that brings the methyl- SCoM substrate into closer proximity to the nickel, and this promotes C-S bond cleavage. To investigate the proposed structural role of CoBSH in initiating catalysis, we have solved the X-ray crystal structures of MCR in complex with four different CoBSH analogues. CoBSH has a heptanoyl moiety linked to the thiol group, and the analogues are pentanoyl-, hexanoyl-, octanoyl- or nonanoyl-containing derivatives of CoBSH (CoB5SH, CoB6SH, CoB8SH and CoB9SH respectively; Figure 2) (3, 35, 43–47). In addition, we present a structure in which the substrate channel predominantly lacks either CoBSH or heterodisulfide product. MATERIALS AND METHODS Materials The organism Methanothemobacter marburgensis (catalog OCM82) was obtained from the Oregon Collection of Methanogens (Portland, OR). All buffers and media reagents were obtained from Sigma-Aldrich (St. Louis, MO). The gases N2 (99.98%), H2/CO2 (80%/20%), and ultra high purity H2 (99.999%) were obtained from Cryogenic Gases (Grand Rapids, MI). Ti(III) citrate solutions were prepared from a stock solution of 246 mM Ti(III) citrate, which was synthesized by adding sodium citrate to Ti(III) trichloride (30 wt % solution in 2 N hydrochloric acid) (Acros Organics, Morris Plains, NJ) under anaerobic conditions and adjusting pH to 7.0 with sodium bicarbonate (48). The concentration of Ti(III) citrate was determined by titrating against a solution of methyl viologen. Synthesis of methyl-SCoM, CoB5SH, CoB6SH, CoBSH, CoB8SH, and CoB9SH Methyl-SCoM was prepared from HSCoM and methyl iodide (49). The homodisulfides, CoB5S-SCoB5, CoB6S-SCoB6, CoBS-SCoB, CoB8S-SCoB8, CoB9S-SCoB9 were prepared as described from 5-bromovaleric acid, 6-bromohexanoic acid (Sigma-Aldrich, St. Louis, MO), 7-bromoheptanoic acid (Karl Industries, Aurora, OH), 8-bromooctanoic acid, and 9- bromononanoic acid (Matrix Scientific, Columbia, SC), respectively (43, 46). The free thiol forms of CoB5SH, CoB6SH, CoBSH, CoB8SH, and CoB9SH were generated by the reduction of the homodisulfides as previously described (45). The purity of the CoBSH analogues was determined by 1H NMR spectroscopy. All compounds synthesized were stored in a Vacuum Atmospheres chamber maintained at an oxygen level below 1 ppm, as monitored continually with an oxygen analyzer (model 317, Teledyne Analytical Instruments, City of Industry, CA) until use. M. marburgensis Growth and MCRred1 Purification Buffer preparations and all manipulations were performed under strict anaerobic conditions in a Vacuum Atmospheres chamber maintained at an oxygen level below 1 ppm, as monitored continually with an oxygen analyzer (model 317, Teledyne Analytical Instruments, City of Industry, CA). MCRred1 was isolated from M. marburgensis cultured on Cedervall et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript H2/CO2 (80%/20%) at 65 °C in a 14-L fermentor (New Brunswick Scientific Co., Inc. New Brunswick, NJ). Culture media were prepared as previously described (20, 24). MCRred1 was generated in vivo and purified as described previously (20). The purification procedure routinely generates 65–75% MCRred1 as determined by UV-visible and EPR spectroscopy. Spectroscopy of MCR UV-visible spectra of the Ni(I)-containing MCRred1 were recorded in the anaerobic chamber using a spectrophotometer (Model USB4000-UV-VIS, Ocean Optics, Dunedin, FL). EPR spectra were recorded on a Bruker EMX spectrometer (Bruker Biospin Corp., Billerica, MA), equipped with an Oxford ITC4 temperature controller, a Hewlett-Packard Model 5340 automatic frequency counter and Bruker gaussmeter. The EPR spectroscopic parameters included: temperature, 70 K; microwave power, 10 mW; microwave frequency, 9.43 GHz; receiver gain, 2 × 104; modulation amplitude, 10.0 G; modulation frequency, 100 kHz. Double integrations of the EPR spectra were performed and referenced to a 1 mM copper perchlorate standard. The NMR data were acquired at 298 K on a Bruker Avance DRX 500 MHz instrument equipped with a TXI cryoprobe. Preparation of MCRred1 for Crystallization All crystallization experiments were performed in the anaerobic chamber in which MCR was purified unless otherwise noted. MCRred1 was prepared in 50 mM Tris, pH 7.6 and excess HSCoM was removed by buffer exchange using an Amicon Ultra centrifuge filter with a 50 kDa cut-off membrane (Millipore). Typically, 2–3 ml of MCRred1 was exchanged with 10–15 ml of 50 mM Tris, pH 7.6. The enzyme was concentrated to 500–600 μl, and this process was repeated three times. The fraction of MCRred1 in the purified MCR sample was calculated from the UV-visible spectrum using extinction coefficients of 27.0 mM−1cm−1 at 385 nm for Ni(I)-MCRred1, and 9.15 mM−1cm−1 at 420 nm for Ni(II)- MCRred1-silent (20). The amount of MCRred1 in samples used for crystallization was determined to be ~80% and the concentration of total enzyme used was in the range of about 120–150 μM (~32–40 mg/ml). All crystallization experiments were performed anaerobically by incubating 2.0 μl of enzyme solution in 50 mM Tris, pH 7.5 and 2.0 μl of reservoir solution (100 mM Hepes-Na, pH 7.3/7.5/8.0; 150 mM magnesium acetate (Mg(CH3COO)2), and 20/22% (w/v) PEG 400) in a sitting drop over 1 ml reservoir solution at 9 ºC. Triangular and rectangular prismatic crystals with a bright yellowish-green color confirmed the presence of nickel coenzyme F430. The crystals grew to a size of approximately 100–200 μm in 4–5 days. CoBSH-depleted crystals were obtained by incubating 2 μl of a reaction mixture containing 139 μM MCRred1 and 13 mM HSCoM with 2 μl of reservoir solution (100 mM Hepes-Na pH 7.5, 150 mM magnesium acetate (Mg(CH3COO)2), 22% PEG 400). Crystals of MCR complexed with the CoBSH analogues were grown by co-crystallization. The CoB5SH co-crystals were obtained by incubating 2μl enzyme solution containing 124 μM MCRred1, 10 mM methyl-SCoM and 1 mM CoB5SH with 2μl of reservoir solution (100 mM Hepes-Na, pH 7.5, 150 mM (Mg(CH3COO)2), 22 % PEG 400). The crystals with bound CoB6SH and CoB9SH were obtained by co-crystallization of 1 mM of analogue with 142 μM MCRred1 and equilibrated with 2μl of reservoir solution (100 mM Hepes-Na, pH 7.5, 150 mM Mg(CH3COO)2, 20 % PEG 400 for CoB6SH and 100 mM Hepes-Na, pH 7.3, 150 mM Mg(CH3COO)2), 22 % PEG 400 for CoB9SH). Crystals were cryoprotected in reservoir solution containing 25 % (v/v) PEG 400 by soaking for 2–5 minutes before cryocooling in liquid nitrogen in the anaerobic chamber. Crystals of CoB8SH bound to MCR were obtained by incubating 2 μl of a mixture of 119 μM MCRred1 and 1 mM CoB8SH with 2 μl of reservoir solution (100 mM Hepes-Na, pH 7.3, 150 mM Mg(CH3COO)2), 20 % PEG 400). Before cryoprotection, the crystals were soaked for 5–10 minutes in a 100 mM solution of methyl iodide2. The methyl iodide solution used for soaking was prepared by adding a concentrated stock of methanolic solution of methyl iodide to the reservoir Cedervall et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript solution. Soaked crystals were quickly cryoprotected as described above and cryocooled in the anaerobic chamber. X-ray Diffraction Data Collection, Processing and Refinement X-ray diffraction data were collected at 100 K on a ADSC Quantum-315 detector at the APS Beamline 14-BM-C (BioCARS). The wavelength of X-rays was 0.979 Å. Data were processed using HKL2000 (50). As in the previous X-ray crystallographic studies, the crystals belong to the monoclinic space group P21 (a = 82 Å, b = 118 Å, c = 122 Å, β = 92°), with one MCR molecule (two active sites) per asymmetric unit (5, 33). For refinement, REFMAC in the Collaborative Computational Project Number 4 (CCP4) program suite was used (51). A random sample of 5 % of the data across all resolution shells was chosen to check refinement progress through calculation of an Rfree. The same reflections were used to calculate Rfree for all structures, thus preventing bias due to high structural identity. The remaining reflections were used in refinement (Rwork). Model building was done using the Crystallographic Object-Oriented Toolkit (COOT) (52). The diffraction data and their models are designated as MCRCoBXSH, where X is the number of carbons in the alkanoyl portion of the analogue. Library files in CCP4 for F430 and CoBSH were incorrect, and these were modified in Monomer Library Sketcher in the CCP4 program suite by comparison with schematic drawings from Grabarse et al. (33). Co-ordinate and CCP4 library files for the different CoBSH analogues were created in Monomer Library Sketcher. The general model building and refinement strategy for all structures was as follows. It was clear from the electron density in the substrate channel and at the active site that mixtures of species were present in all datasets. These could be visualized with Fo-Fc and Fo-Fo difference electron density maps (Supporting Information, Figure S1). The known positions of CoBSH and HSCoM from the published Ni(II)-MCR crystal structures (PDB codes 1hbn, 1hbo, 1hbu (33)) were used as guides to determine which species could be present in each dataset, and these were then simultaneously modeled into the electron density. By alteration of their relative occupancies (in 10% increments) followed by refinement, the ratio of occupancy between different species was determined using the assumption that the average B-factors for all molecular species bound should be similar to that of F430 and adjacent well-ordered protein atoms within the active site and substrate channel. The combinations of modeled ligands were constantly reassessed throughout refinement based on the remaining difference electron density. This included test refinements of different ligand combinations during the latter stages, thus using the optimized phases to check whether a different combination of ligands could also explain the electron density. Sensible chemical structures and interactions, along with keeping the combined occupancies of sterically mutually exclusive species ≤ 100%, were maintained throughout refinement. The model was finally accepted when the difference electron density map was minimal and the B-factors for the models converged. In practice the first structure refined was that of MCRCoB5SH. Initial phases were generated by difference Fourier using a previously determined crystal structure (PDB code 1mro (5)) but with all non-bonded molecules, including water, removed from the model except F430. Initial rigid body refinement followed by restrained refinement of MCRCoB5SH reduced the Rwork to 26.5%. After model building and subsequent rounds of restrained refinement the Rwork was 14.3% (Rfree 16.6%). Of the five structures only the CoB5SH analogue is completely coincident with CoBSH, and so particular care had to be used in teasing apart the ratios of the two species in modeling the MCRCoB5SH electron density. This was done by 2Methyl iodide was added with the intention of creating a Ni(III)-methyl species, which CoB8SH stabilizes. This was not achieved, but the diffraction quality of this crystal was significantly better than crystals co-crystallized with CoB8SH alone, and so has been included in this study. Cedervall et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript initially refining 100% CoBSH in the substrate channel. Positive Fo-Fc difference density located at the carbon where the shorter CoB5SH thiol might be expected to be indicated the presence of a more electron-rich species than carbon, which is consistent with the presence of the CoB5SH sulfur. The refinement converged at a model containing a 50:50 mix of CoBSH and the analogue. However, positive Fo-Fc difference density was still present at the position of the CoBSH thiol, therefore a water molecule was added to the CoB5SH model at 50% occupancy and upon refinement this accounted for the electron density. An illustration of the electron density quality from this structure is shown in Supporting Information, Figure S2. An HSCoM-, CoBSH- and CoBSH analogue-free version of the refined MCRCoB5SH structure was used as the starting model to generate initial phases for the four other structures. After the initial round of restrained refinement the Rwork for these structures were reduced to 14.5–15.6 %. RESULTS AND DISCUSSION Crystal Structures of MCR Five crystal structures were determined, four of which are in complex with CoBSH analogues differing in the number of carbons atoms in the alkanoyl portion of the molecule. CoBSH is an N-7-mercaptoheptanoyl-containing molecule, whereas the four CoBSH analogues contain N-5-mercaptopentanoyl-, N-6-mercaptohexanoyl-, N-8-mercaptooctanoyl- or N-9-mercaptononanoyl-moieties (Figure 2). The corresponding crystal structures are designated as MCRCoBXSH, where X is the number of carbons in the alkanoyl portion of the analogue. The other crystal structure is of MCRred1c-silent (MCR in the Ni(II) state in complex with HSCoM, designated here as MCRHSCoM) that is CoBSH-depleted. The datasets have resolutions in the range from 1.35 – 1.8 Å. Although the crystallizations were set up with the MCR solution predominantly in the Ni(I)-MCRred1 state, by the time X-ray diffraction data were collected they had been oxidized to the Ni(II)-MCRred1-silent state (Supporting Information). Following data collection there was no evidence for photoreduction of the Ni(II) back to Ni(I) in any of the crystals, as assessed by single crystal UV-visible microspectrophotometry (Supporting Information and Figure S3). Attempts to photoreduce the crystals using different wavelengths and temperatures were unsuccessful (Supporting Information). Overall, the resulting structures are very similar to each other and to the previously published structures of MCR, with differences mainly localized to the active site and substrate channel. The two active sites in the ASU were refined independently. Unless otherwise stated there was no difference between them. All five datasets contain a mixture of species bound to the enzyme. There is always a background of CoBSH and HSCoM, which co-purify with MCR and cannot be fully removed by extensive buffer exchange or by the addition of a CoBSH analogue. HSCoM is added during purification of MCR, as it stabilizes the resting active Ni(I) state (unpublished data), and this leads to HSCoM occupancies between 50–100% amongst the structures (Table 1). In contrast CoBSH, which is not added during purification, has occupancies ranging from 30–50%. As these confounding species have all been described at high occupancy in other crystallographic studies, the structural data of interest could be isolated (5, 33). In each case, the additional electron density could be explained by inclusion of the appropriate CoBXSH model used in that experiment at 50% or higher occupancy. The resulting models, along with 2Fo-Fc electron density, are shown in Figure 3. The Rwork for the final structures range from 13.0 to 15.0 % (Rfree 15.5 to 19.5 %). The X-ray data collection, processing, refinement and model building statistics are given in Table 1. Cedervall et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Analogues shorter than CoBSH; CoB5SH and CoB6SH CoB5SH is two methylene groups shorter than CoBSH, the MCR substrate. The MCRCoB5SH structure is to 1.35 Å resolution. As expected the pentanoyl chain follows the path of the CoBSH heptanoyl carbons down the substrate channel, and thus its thiol is positioned in the same place as the second carbon preceding the CoBSH thiol (Figure 3A and 4). There are no published MCR kinetic studies using CoB5SH, but as it binds in the substrate channel, it is likely to be an inhibitor. CoB6SH is one methylene shorter than CoBSH, and is a slow substrate of MCR. In this case the 1.4 Å resolution electron density of MCRCoB6SH indicates that the analogue unexpectedly binds in the substrate channel such that its thiol is virtually in the same position as the thiol of CoBSH (Figure 3B and 4). The hexanoyl chain is oriented so that it takes a shorter route down the substrate channel between carbons 2 and 5 (with the carbonyl carbon labeled as carbon 1) than CoBSH (Figure 4, Supporting Information, Figure S4). This short-cut is not seen in any of the other CoBXSH complex crystal structures, but presumably arises because this CoB6SH binding conformer is energetically more favorable, although it is not clear from the structure why this might be the case. CoB6SH binds very tightly to MCR, with an apparent Ki value of 0.1 μM (3). Water structure in the absence of HSCoM The electron density for the MCRCoB6SH crystal structure only supported the modeling of 50 % bound HSCoM. In the fraction of MCR molecules where HSCoM is absent, the HSCoM binding site is occupied by a network of four water molecules (Supporting Information, Figure S5). Two waters are positioned close to the absent sulfonate oxygen positions of HSCoM. Based on the presence of positive difference electron density, a third water was modeled ligated to the Ni, and refined to a distance of ~2 Å (2.0 Å and 2.1 Å in the two active sites of the ASU) with no distance restraint imposed between the Ni and water. This water is in a similar position as the Ni coordinating sulfonate oxygen of the heterodisulfide product in MCRsilent (Supporting Information, Figure S6 and PDB codes 3m32, 1hbm) (5, 33). The fourth water was in the vicinity of the expected position of a bridging water (W1) seen in other structures (Figure 1, 3A and 3C). Water structure in the absence of CoBSH The 1.45 Å resolution electron density obtained for MCRHSCoM indicates that the substrate channel contains only 30 % CoBSH. Nine ordered waters (W1–W9), along with an acetate ion from the crystallization solution occupy the channel, with the acetate positioned where the phosphothreonine linkage of CoBSH would be (Figure 3C). Presumably 1–2 further waters would replace the acetate under physiological conditions. Other than W3 and W7, the waters form hydrogen bonds with protein (Figure 5). One water (W2) occupies the same site as the CoBSH thiol. Presumably due to the loss of favorable interactions that exist when CoBSH is present, the hydrophobic side-chain of Valα482 adopts a second conformation modeled at 60 % occupancy (Supporting Information, Figure S7). Position of the “bridging” water, W1 The equivalent of W1 has previously been observed in MCRred1-silent and MCRox1-silent crystal structures where, in the presence of CoBSH and HSCoM, it is sited equidistant (3.2 Å) between the two coenzyme thiols (PDB codes: 1hbn, 1hbo, 1hbu) and thus been termed the “bridging water” (Figure 1) (5, 33). However, in the MCRHSCoM structure, due to the presence of the more polarized W2 water, W1 is displaced away from HSCoM to maximize the hydrogen bond interaction with W2 (2.9 Å to W2; 3.5 Å to HSCoM thiol, Figure 5). In the MCRCoB5SH structure that also contained W2, the electron density indicated that this Cedervall et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript repositioning of W1 towards W2 also occurred. In contrast, the MCRCoB6SH structure contained 100% thiol at the CoBSH position, but a partial occupancy of HSCoM (50%). In this case the electron density for W1 indicated it had moved towards the nickel to form an optimal hydrogen bond with a Ni-ligating water that was only present in the absence of HSCoM (3.7 Å to CoB(6)SH thiol; 3.0 Å to Ni-ligating water, Supporting Information, Figure S5). In all structures reported here, W1 (if present) appears to be a sensitive indicator of the relative electronegativity of the Ni-ligated atom to that occupying the position of the CoBSH thiol, and was a useful check in the crystallographic modeling and refinement process. Flexibility in the substrate channel: Alternative protein conformers The binding site of HSCoM (and presumably methyl-SCoM) is more deeply buried within the enzyme, and so it must enter prior to CoBSH for productive chemistry to occur. As binding of CoBSH in the absence of co-substrate would be inhibitory, it was suggested that a conformational change upon methyl-SCoM binding might lower the Kd for CoBSH, and thus promote an ordered mechanism. Compared to the 1hbn MCRox1-silent and 1hbu MCRred1-silent crystal structures, which both have full occupancy HSCoM, the lower occupancy of HSCoM in the 1hbo MCRred1-silent structure was associated with significantly greater flexibility within the channel, and the ability to model a second conformation of a Gly-rich amino acid stretch that formed part of the CoBSH channel. This suggested that methyl-SCoM binding might cause the channel to become more ordered, increasing the affinity of MCR for CoBSH by conformational restriction rather than a switch mechanism where the structure reorganizes from one well-defined conformer to another (33). In the MCRHSCoM data containing 30 % CoBSH and 100% HSCoM, the Fo-Fc difference electron density map at one of the two independent active sites in the ASU contained positive peaks that suggested the presence of an alternate conformation also involving this part of the polypeptide (Supporting Information, Figure S8). Using this as a guide, a similar second conformation involving seven contiguous amino acid residues of the same Gly-rich amino acid stretch (β366–372) could be modeled and refined at 20 % occupancy leaving no residual difference density. Parts of the α′ subunit (α′111–129 and α′237–242) that are in close proximity to this stretch of amino acids also exhibit second conformations, with the main-chain carbonyl of α′243 in van der Waals contact with the B ring of F430 tetrapyrrole (Supporting Information, Figure S9). Modeling these at 20% occupancy accounted for the weak positive Fo-Fc difference electron density peaks observed in these areas. The evidence of alternate conformers in these areas lends support to the proposal that increased flexibility in the substrate channel propagates through the protein (33). The MCRCoB6SH crystal structure contains 50 % CoB6SH, 50% CoBSH and 50% HSCoM. In this case there is no evidence of an alternate loop conformation in either active site of the ASU. However, as CoBSH and CoB6SH combined are at 100% occupancy, it is not surprising their favorable interactions with the substrate channel would reduce conformational disorder, despite the partial occupancy of HSCoM. Analogues longer than CoBSH; CoB8SH and CoB9SH Both analogues could be accommodated in the MCR substrate channel (Figures 3D and 3E). The electron density supported final models containing 50% CoB8SH for MCRCoB8SH (1.8 Å resolution) and 60 % CoB9SH for MCRCoB9SH (1.45 Å resolution). The phosphate head- groups are in identical positions to those of CoBSH, CoB5SH, CoB6SH (Figure 4) (5, 33). Both analogues follow the crystallographically observed chain path of bound CoBSH, with the extra atoms displacing the W1 water and placing the thiols closer to the nickel (Figure 6). CoB9SH does have a second conformer that deviates from the CoBSH path, but the thiol position for this conformer and the CoBSH-tracking conformer are identical (Figure 3E and Cedervall et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supporting Information, Figure S10). Interestingly, the thiol of CoB8SH is not coincident with the CoB9SH carbon that precedes the CoB9SH thiol (Figure 6). CoB8SH is an MCR inhibitor with an apparent Ki of 15 μM (3). CoB9SH has never been tested for inhibition of MCR-catalyzed methane formation, but it is reasonable to assume that it would be an inhibitor. CoBXSH thiol-to-nickel spatial relationship The position of CoBSH in previous crystal structures poses a conundrum (5, 33). In all the proposed catalytic mechanisms, CoBSH must interact with species generated at the nickel. Perplexingly, the sulfur of the CoBSH substrate is 8.8 Å from the Ni(II) in the MCRox1-silent and both MCRred1-silent crystal structures, and 6.4 Å from the thiol of the substrate analogue HSCoM (Figure 1). Modeling studies demonstrated that the addition of a methyl group to HSCoM did not bridge this gap (35, 45, 53). Therefore, a conformational change has been postulated that would enable CoBSH to penetrate deeper into the substrate channel, and thus approach closer to any nickelbound species. The heterodisulfide product in the MCRsilent crystal structure has the CoBSH portion in virtually the same place as in MCRox1-silent, giving no clue to possible structural changes that might occur to facilitate CoBSH reacting with nickel-associated intermediates (5, 33). Trigonometry suggests that if the alkanoyl chain of CoBSH or its analogue is in an extended conformation, each additional unit in the chain would lead to the thiol moving ~1.2 Å towards the Ni. Until this study there have been no crystal structures of CoBSH analogues in complex with MCR, so mechanistic studies using different chain length analogues of CoBSH assumed that shorter analogues would trace the observed path of CoBSH, and longer analogues would penetrate about ~1.2 Å deeper per additional chain unit into the channel. In the case of the shortest analogue CoB5SH, it does indeed follow the path of CoBSH, with the thiol of CoB5SH being 2.8 Å away from the thiol position of CoBSH. However, due to the conformation CoBSH adopts when bound in the substrate channel, the difference in the S-Ni distance is small; the CoB5SH thiol being only 0.5 Å farther from the Ni than CoBSH (8.8 Å for CoBSH vs. 9.3 Å for CoB5SH) (Table 2). This is due to the alkanoyl chain of CoBSH not being in an extended conformation from carbons 4 to 6 (carbon 1 is the carbonyl carbon). CoB6SH on the other hand adopts a conformation that places its thiol in virtually the same position as the thiol of CoBSH (Figure 4 and Table 2). This is consistent with CoB6SH being a substrate. However, the kcat is 1000-fold lower than for CoBSH (3, 35) although its Km value (180 μM, Dey & Ragsdale, in preparation) is similar to that of CoBSH (Km = 75 μM (3)). The reason for this may be that the shorter alkanoyl chain may not enable the analogue thiol to approach the nickel close enough for efficient catalysis, and thus explain why CoB6SH is such a poor substrate. In the case of the longer CoBXSH analogues, the sulfur of CoB8SH is 2.6 Å closer to the Ni ion of F430 than that of CoBSH, and 2.5 Å closer to the thiol of HSCoM (Figure 6 and Table 2). The CoB9SH molecule follows the path of CoBSH, and reaches only a little further into the substrate channel than CoB8SH, with the CoB9SH thiol positioned 2.9 Å closer to the Ni than the thiol of CoBSH (Figure 6 and Table 2). This is only 0.3 Å closer than the distance observed for the CoB8SH thiol, even though they are non-coincident. The distance to the thiol of HSCoM is 2.6 Å closer than that of the substrate, CoBSH; only 0.1 Å closer than the CoB8SH thiol. The two analogue thiols are above an annular hydrophobic aromatic environment created by Pheα330, Tyrα333, Pheα443, Pheβ361 and Tyrβ367 that lies between them and F430 (Figure 6). As a result, penetrating further into the channel may be energetically unfavorable, consistent with the small difference in relative distances between the CoB8SH/CoB9SH thiols and the HSCoM thiol/F430 nickel. The annulus is proposed to be catalytically important in positioning methyl-SCoM and stabilizing the methane product, Cedervall et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript and the tyrosines have been proposed to be proton donors associated with mechanism II (Scheme 2B) (5, 33). Thus, there appear to be three preferential distances for thiols (including that of HSCoM) within the MCR substrate channel; HSCoM at 2.4 Å, CoB6SH/CoBSH at 8.7 – 8.8 Å and CoB8SH/CoB9SH at 5.9 – 6.2 Å from the nickel of F430 (Table 2). Recent ENDOR, high-field continuous and pulse EPR work has identified changes in nickel co-ordination when CoBSH is added to MCRred1c (active Ni(I)-MCRred1 + HSCoM) (14, 15, 18, 31). This generates up to 50% MCRred2 which is comprised of two distinct nickel co- ordination geometries; an axial MCRred2a formally assigned as a Ni(III)-hydride, and a rhombic MCRred2r in which the thiol of HSCoM is a Ni(I) ligand (Supporting Information, Scheme S1). Formation of MCRred2 could also be induced by addition of CoBSH substrate analogues CoBS-CH3 and CoBS-CF3, which have a chain length one unit longer than substrate CoBSH (18, 53). The CoBS-CF3 enabled 19F-ENDOR studies to be performed, and demonstrated that following CoBS-CF3 addition the remaining MCRred1 species had Ni(I)-19F distances of 6.2 – 7.7 Å. This distance range agreed with a CoBS-CF3 model created using the CoBSH position observed in the MCRox1-silent crystal structure (53). However, in the MCRred2 species the Ni(I)-19F distances had shortened indicating a movement of ~ 2 Å towards the nickel (MCRred2a Ni(I)-19 F, 4.0 – 5.5 Å; MCRred2r Ni(I)-19F, 4.5 – 5.7 Å). In the case of the CoB8SH analogue (the closest equivalent to CoBS- CF3 used in this study) the thiol to Ni(II) distance lies between the distance ranges observed in the CoBS-CF3 studies, and so the fluorine(s) of the CoBS-CF3 in MCRred2 might penetrate a little further into the hydrophobic annulus in the MCRred2 species. As the alkanoyl chain of CoBSH is not fully extended it could easily undergo a similar conformation change to that observed in the MCRred2 state. Reaction of MCR Ni(III)-alkyl species’ with CoB8SH and CoB9SH The two longer CoBXSH analogues have been shown to undergo alkylation when reacted with MCRPS, a [Ni(III)-alkyl ↔ Ni(II)-alkylsulfonate radical] formed from reaction of Ni(I)-MCRred1 with bromopropanesulfonate (BPS) (Supporting Information, Scheme S1) (20, 23, 30, 45, 54). BPS is a substrate of MCRred1 in a reaction that involves a rapid CoBSH-independent nucleophilic attack by Ni(I) on BPS to displace bromide and generate MCRPS at a rate ~60-fold faster than generation of methane from CoBSH and methyl- HSCoM (20, 45). Certain thiols can eliminate the propylsulfonate to yield a thioether product and regenerate MCRred1, although at a rate 1000-fold slower than methane formation (45). Both CoB8SH and CoB9SH can react with MCRPS to regenerate MCRred1, but CoBSH cannot. The overall second-order rate constant for the reactivation of MCR by CoB8SH is 160 M−1s−1, whereas for CoB9SH the reaction is more sluggish (12 M−1s−1). CoB9SH might be expected to be closer to the proximal Ni ligand. It was therefore proposed that this caused steric interference and explained why CoB9SH was a poorer reactivator of MCR than CoB8SH. Our study has shown that the thiols of these two analogues are placed such that they are approximately the same distance (~3.7 Å) from the thiol of HSCoM ligated to the Ni atom (Table 2). The Ni(II)-HSCoM bond is 2.4 Å, whereas an Ni(III)-alkyl bond is expected to be ~2 Å (24, 33), thus indicating that a conformational change is required to effect the nucleophilic attack of the CoB8SH and CoB9SH thiols on an alkyl- bound species. It would thus appear that a conformational change, such as observed in MCRred2, is required for this chemistry also (53). A Ni(III)-alkyl species is akin to the first intermediate in mechanism I of MCR-catalyzed methane formation, Ni(III)-methyl (MCRMe, Supporting Information, Scheme S1, Scheme 2A) (11, 27). MCRMe has been shown to be capable of generating MCRred1 and methyl- SCoM upon addition of HSCoM (which is the reverse of mechanism I, step 1, Scheme 2A); Cedervall et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript similar chemistry to the observed formation of a thioether product from the Ni(III)-alkyl. Further addition of CoBSH following HSCoM treatment of MCRMe led to methane and heterodisulfide formation, the natural products of methanogenesis. Although this lends credence to mechanism I, it should be noted that like MCRPS, MCRMe in these experiments was generated artificially. However, the MCRCoB9SH crystal structure demonstrates that the two additional methylene units in the alkanoyl chain c.f. CoBSH, do not necessarily translate into direct interaction of the thiol with the nickel proximal ligand. However, this could represent the favorable position for a CoBSH thiol interacting with the methyl group of methyl-SCoM. Just as the alkanoyl chain of CoB6SH has a more extended conformation than CoBSH in the substrate channel, CoBSH could also adopt a more extended conformation so that its thiol was in a similar position as the thiol of CoB8SH, priming it for reaction with a nickel bound species. If a significant conformational change is required early in MCR-catalyzed chemistry, which would be a requirement of mechanism I, catalysis may well involve a rearrangement of the aromatic amino acid annulus due to the presence of the methyl of methyl-SCoM, and this might enable deeper penetration of CoBSH into MCR (Figure 6). All the crystal structures in this study, and those solved previously, are of the inactive Ni(II)-MCR, which disfavors close approach to the nickel in the absence of Ni(I)-bound methyl-SCoM, even in the case of CoB9SH. Conclusion The goal of this study was to induce structural changes within the substrate channel and active site of MCR using analogues of coenzyme CoBSH. It was hoped that this would shed light on the nature of conformational changes that have been proposed to occur in MCR catalysis. We have shown that that the CoBXSH analogues do not lead to any significant conformational changes within the context of inactive Ni(II)-MCR. Therefore, it may be that methyl-SCoM is the key coenzyme, in combination with a nickel oxidation state of 1 (and 3), that triggers a conformational change bringing the thiol of CoBSH closer to the nickel. Thus, the crystal structure of the Ni(I)-methyl-SCoM/MCR complex may be required to structurally define the conformational changes required for MCR-mediated chemistry. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments X-ray data were collected at beam-line 14-BM-C and photoreduction studies at 14-ID-B, BioCARS at the Advanced Photon Source (APS), Argonne National Laboratory, Argonne, IL. We thank Vukica Srajer and Yu- Sheng Chen for valuable assistance during data collection. Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Basic Energy Sciences, Office of Science under Contract No. DE- AC02-06CH11357. Use of the BioCARS Sector 14 was supported by the National Institutes of Health, National Center for Research Resources, under grant number RR007707. Computer resources were provided by the Basic Sciences Computing Laboratory of the University of Minnesota Supercomputing Institute, and we thank Can Ergenekan for his support. We also thank Ed Hoeffner at the Kahlert Structural Biology Laboratory (KSBL) at the University of Minnesota. Use of the KSBL was supported by a Minnesota Partnership for Biotechnology and a Medical Genomics Grant SPAP-05-0013-P-FY06. References 1. Thauer RK. Biochemistry of methanogenesis: a tribute to Marjory Stephenson. Microbiology. 1998; 144:2377–2406. [PubMed: 9782487] 2. Thauer RK, Shima S. Methane as fuel for anaerobic microorganisms. Ann N Y Acad Sci. 2008; 1125:158–170. [PubMed: 18096853] Cedervall et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 3. Ellermann J, Hedderich R, Bocher R, Thauer RK. The final step in methane formation. Investigations with highly purified methyl-CoM reductase (component C) from Methanobacterium thermoautotrophicum (strain Marburg). Eur J Biochem. 1988; 172:669–677. [PubMed: 3350018] 4. Ellefson WL, Wolfe RS. Component C of the methylreductase system of Methanobacterium. J Biol Chem. 1981; 256:4259–4262. [PubMed: 6783657] 5. Ermler U, Grabarse W, Shima S, Goubeaud M, Thauer RK. Crystal structure of methyl-coenzyme M reductase: the key enzyme of biological methane formation. Science. 1997; 278:1457–1462. [PubMed: 9367957] 6. Diekert G, Gilles HH, Jaenchen R, Thauer RK. Incorporation of 8 succinate per mol nickel into factors F430 by Methanobacterium thermoautotrophicum. Arch Microbiol. 1980; 128:256–262. [PubMed: 7212929] 7. Diekert G, Jaenchen R, Thauer RK. Biosynthetic evidence for a nickel tetrapyrrole structure of factor F430 from Methanobacterium thermoautotrophicum. FEBS Letters. 1980; 119:118–120. [PubMed: 7428919] 8. Whitman WB, Wolfe RS. Presence of nickel in Factor F430 from Methanobacterium bryantii. Biochem Biophys Res Comm. 1980; 92:1196–1201. [PubMed: 7370029] 9. Albracht SPJ, Ankel-Fuchs D, Böcher R, Ellermann J, Moll J, van der Zwann JW, Thauer RK. Five new EPR signals assigned to nickel in methyl-coenzyme M reductase from Methanobacterium thermoautotrophicum, strain Marburg. Biochim Biophys Acta. 1988; 955:86–102. 10. Dey M, Kunz RC, Lyons DM, Ragsdale SW. Characterization of alkyl-nickel adducts generated by reaction of methyl-coenzyme m reductase with brominated acids. Biochemistry. 2007; 46:11969– 11978. [PubMed: 17902704] 11. Dey M, Telser J, Kunz RC, Lees NS, Ragsdale SW, Hoffman BM. Biochemical and spectroscopic studies of the electronic structure and reactivity of a methyl-Ni species formed on methyl- coenzyme M reductase. J Am Chem Soc. 2007; 129:11030–11032. [PubMed: 17711283] 12. Duin EC, Cosper NJ, Mahlert F, Thauer RK, Scott RA. Coordination and geometry of the nickel atom in active methyl-coenzyme M reductase from Methanothermobacter marburgensis as detected by X-ray absorption spectroscopy. J Biol Inorg Chem. 2003; 8:141–148. [PubMed: 12459909] 13. Duin EC, Signor L, Piskorski R, Mahlert F, Clay MD, Goenrich M, Thauer RK, Jaun B, Johnson MK. Spectroscopic investigation of the nickel-containing porphinoid cofactor F(430). Comparison of the free cofactor in the (+)1, (+)2 and (+)3 oxidation states with the cofactor bound to methyl- coenzyme M reductase in the silent, red and ox forms. J Biol Inorg Chem. 2004; 9:563–576. [PubMed: 15160314] 14. Finazzo C, Harmer J, Bauer C, Jaun B, Duin EC, Mahlert F, Goenrich M, Thauer RK, Van Doorslaer S, Schweiger A. Coenzyme B induced coordination of coenzyme M via its thiol group to Ni(I) of F430 in active methyl-coenzyme M reductase. J Am Chem Soc. 2003; 125:4988–4989. [PubMed: 12708843] 15. Finazzo C, Harmer J, Jaun B, Duin EC, Mahlert F, Thauer RK, Van Doorslaer S, Schweiger A. Characterization of the MCRred2 form of methyl-coenzyme M reductase: a pulse EPR and ENDOR study. J Biol Inorg Chem. 2003; 8:586–593. [PubMed: 12624730] 16. Goubeaud M, Schreiner G, Thauer RK. Purified methyl-coenzyme-M reductase is activated when the enzyme-bound coenzyme F430 is reduced to the nickel(I) oxidation state by titanium(III) citrate. Eur J Biochem. 1997; 243:110–114. [PubMed: 9030728] 17. Harmer J, Finazzo C, Piskorski R, Bauer C, Jaun B, Duin EC, Goenrich M, Thauer RK, Van Doorslaer S, Schweiger A. Spin density and coenzyme M coordination geometry of the ox1 form of methyl-coenzyme M reductase: a pulse EPR study. J Am Chem Soc. 2005; 127:17744–17755. [PubMed: 16351103] 18. Harmer J, Finazzo C, Piskorski R, Ebner S, Duin EC, Goenrich M, Thauer RK, Reiher M, Schweiger A, Hinderberger D, Jaun B. A nickel hydride complex in the active site of methyl- coenzyme m reductase: implications for the catalytic cycle. J Am Chem Soc. 2008; 130:10907– 10920. [PubMed: 18652465] Cedervall et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 19. Hinderberger D, Ebner S, Mayr S, Jaun B, Reiher M, Goenrich M, Thauer RK, Harmer J. Coordination and binding geometry of methyl-coenzyme M in the red1m state of methyl- coenzyme M reductase. J Biol Inorg Chem. 2008; 13:1275–1289. [PubMed: 18712421] 20. Kunz RC, Horng YC, Ragsdale SW. Spectroscopic and kinetic studies of the reaction of bromopropanesulfonate with methyl-coenzyme M reductase. J Biol Chem. 2006; 281:34663– 34676. [PubMed: 16966321] 21. Mahlert F, Bauer C, Jaun B, Thauer RK, Duin EC. The nickel enzyme methyl-coenzyme M reductase from methanogenic archaea: In vitro induction of the nickel-based MCR-ox EPR signals from MCR-red2. J Biol Inorg Chem. 2002; 7:500–513. [PubMed: 11941508] 22. Mahlert F, Grabarse W, Kahnt J, Thauer RK, Duin EC. The nickel enzyme methyl-coenzyme M reductase from methanogenic archaea: in vitro interconversions among the EPR detectable MCR- red1 and MCR-red2 states. J Biol Inorg Chem. 2002; 7:101–112. [PubMed: 11862546] 23. Rospert S, Voges M, Berkessel A, Albracht SP, Thauer RK. Substrate-analogue-induced changes in the nickel-EPR spectrum of active methyl-coenzyme-M reductase from Methanobacterium thermoautotrophicum. Eur J Biochem. 1992; 210:101–107. [PubMed: 1332856] 24. Sarangi R, Dey M, Ragsdale SW. Geometric and electronic structures of the Ni(I) and methyl- Ni(III) intermediates of methyl-coenzyme M reductase. Biochemistry. 2009; 48:3146–3156. [PubMed: 19243132] 25. Tang Q, Carrington PE, Horng YC, Maroney MJ, Ragsdale SW, Bocian DF. X-ray absorption and resonance Raman studies of methyl-coenzyme M reductase indicating that ligand exchange and macrocycle reduction accompany reductive activation. J Am Chem Soc. 2002; 124:13242–13256. [PubMed: 12405853] 26. Telser J, Davydov R, Horng YC, Ragsdale SW, Hoffman BM. Cryoreduction of methyl-coenzyme M reductase: EPR characterization of forms, MCR(ox1) and MCR (red1). J Am Chem Soc. 2001; 123:5853–5860. [PubMed: 11414817] 27. Yang N, Reiher M, Wang M, Harmer J, Duin EC. Formation of a nickel-methyl species in methyl- coenzyme M reductase, an enzyme catalyzing methane formation. J Am Chem Soc. 2007; 129:11028–11029. [PubMed: 17711279] 28. Albracht SPJ, Ankelfuchs D, Vanderzwaan JW, Fontijn RD, Thauer RK. A New Electron- Paramagnetic-Res Signal of Nickel in Methanobacterium-Thermoautotrophicum. Biochim Biophys Acta. 1986; 870:50–57. 29. Telser J, Horng YC, Becker DF, Hoffman BM, Ragsdale SW. On the assignment of nickel oxidation states of the Ox1, Ox2 forms of methyl-coenzyme M reductase. J Am Chem Soc. 2000; 122:182–183. 30. Hinderberger D, Piskorski RR, Goenrich M, Thauer RK, Schweiger A, Harmer J, Jaun B. A nickel- alkyl bond in an inactivated state of the enzyme catalyzing methane formation. Angewandte Chemie-International Ed. 2006; 45:3602–3607. 31. Kern DI, Goenrich M, Jaun B, Thauer RK, Harmer J, Hinderberger D. Two sub-states of the red2 state of methyl-coenzyme M reductase revealed by high-field EPR spectroscopy. J Biol Inorg Chem. 2007; 12:1097–1105. [PubMed: 17690920] 32. Becker DF, Ragsdale SW. Activation of methyl-SCoM reductase to high specific activity after treatment of whole cells with sodium sulfide. Biochemistry. 1998; 37:2639–2647. [PubMed: 9485414] 33. Grabarse W, Mahlert F, Duin EC, Goubeaud M, Shima S, Thauer RK, Lamzin V, Ermler U. On the mechanism of biological methane formation: structural evidence for conformational changes in methyl-coenzyme M reductase upon substrate binding. J Mol Biol. 2001; 309:315–330. [PubMed: 11491299] 34. Grabarse W, Mahlert F, Shima S, Thauer RK, Ermler U. Comparison of three methyl-coenzyme M reductases from phylogenetically distant organisms: unusual amino acid modification, conservation and adaptation. J Mol Biol. 2000; 303:329–344. [PubMed: 11023796] 35. Horng YC, Becker DF, Ragsdale SW. Mechanistic studies of methane biogenesis by methyl- coenzyme M reductase: evidence that coenzyme B participates in cleaving the C-S bond of methyl-coenzyme M. Biochemistry. 2001; 40:12875–12885. [PubMed: 11669624] Cedervall et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 36. Berkessel A. Methyl-Coenzyme-M Reductase - Model Studies on Pentadentate Nickel-Complexes and a Hypothetical Mechanism. Bioorg Chem. 1991; 19:101–115. 37. Jaun B. Coenzyme-F430 from Methanogenic Bacteria - Oxidation of F430 Pentamethyl Ester to the Ni(Iii) Form. Helvetica Chimica Acta. 1990; 73:2209–2217. 38. Signor L, Knuppe C, Hug R, Schweizer B, Pfaltz A, Jaun B. Methane formation by reaction of a methyl thioether with a photo-excited nickel thiolate - A process mimicking methanogenesis in archaea. Chemistry-a European Journal. 2000; 6:3508–3516. 39. Chen SL, Pelmenschikov V, Blomberg MR, Siegbahn PE. Is there a Ni-methyl intermediate in the mechanism of methyl-coenzyme M reductase? J Am Chem Soc. 2009; 131:9912–9913. [PubMed: 19569621] 40. Pelmenschikov V, Blomberg MRA, Siegbahn PEM, Crabtree RH. A mechanism from quantum chemical studies for methane formation in methanogenesis. J Am Chem Soc. 2002; 124:4039– 4049. [PubMed: 11942842] 41. Pelmenschikov V, Siegbahn PE. Catalysis by methyl-coenzyme M reductase: a theoretical study for heterodisulfide product formation. J Biol Inorg Chem. 2003; 8:653–662. [PubMed: 12728361] 42. Duin EC, McKee ML. A new mechanism for methane production from methyl-coenzyme M reductase as derived from density functional calculations. J Phys Chem. 2008; B 112:2466–2482. 43. Bobik TA, Wolfe RS. Physiological importance of the heterodisulfide of coenzyme M and 7- mercaptoheptanoylthreonine phosphate in the reduction of carbon dioxide to methane in Methanobacterium. Proc Natl Acad Sci U S A. 1988; 85:60–63. [PubMed: 3124103] 44. Goenrich M, Duin EC, Mahlert F, Thauer RK. Temperature dependence of methyl-coenzyme M reductase activity and of the formation of the methyl-coenzyme M reductase red2 state induced by coenzyme B. J Biol Inorg Chem. 2005; 10:333–342. [PubMed: 15846525] 45. Kunz RC, Dey M, Ragsdale SW. Characterization of the Thioether Product Formed from the Thiolytic Cleavage of the Alkyl-Nickel Bond in Methyl-Coenzyme M Reductase. Biochemistry. 2008; 47:2661–2667. [PubMed: 18220418] 46. Noll KM, Donnelly MI, Wolfe RS. Synthesis of 7-mercaptoheptanoylthreonine phosphate and its activity in the methylcoenzyme M methylreductase system. J Biol Chem. 1987; 262:513–515. [PubMed: 3100513] 47. Olson KD, McMahon CW, Wolfe RS. Photoactivation of the 2-(methylthio)ethanesulfonic acid reductase from Methanobacterium. Proc Natl Acad Sci U S A. 1991; 88:4099–4103. [PubMed: 1903534] 48. Zehnder AJ, Wuhrmann K. Titanium (III) citrate as a nontoxic oxidation-reduction buffering system for the culture of obligate anaerobes. Science. 1976; 194:1165–1166. [PubMed: 793008] 49. Gunsalus RP, Romesser JA, Wolfe RS. Preparation of coenzyme M analogues and their activity in the methyl coenzyme M reductase system of Methanobacterium thermoautotrophicum. Biochemistry. 1978; 17:2374–2377. [PubMed: 98178] 50. Otwinowski Z, Minor W. Processing of X-ray Diffraction Data Collected in Oscillation Mode. Methods in Enzymology: Macromolecular Crystallography, part A. 1997; 276:307–326. 51. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr. 1994; 50:760–763. [PubMed: 15299374] 52. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] 53. Ebner S, Jaun B, Goenrich M, Thauer RK, Harmer J. Binding of coenzyme B induces a major conformational change in the active site of methyl-coenzyme M reductase. J Am Chem Soc. 2010; 132:567–575. [PubMed: 20014831] 54. Goenrich M, Mahlert F, Duin EC, Bauer C, Jaun B, Thauer RK. Probing the reactivity of Ni in the active site of methyl-coenzyme M reductase with substrate analogues. J Biol Inorg Chem. 2004; 9:691–705. [PubMed: 15365904] Cedervall et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. The active site and substrate channel in the MCRox1-silent crystal structure (PDB code 1hbn) (9). Coenzyme F430, CoBSH and HSCoM are drawn as stick colored by atom (carbon: dark grey). The nickel is displayed as a green sphere, water as a red sphere. Interactions are drawn as dashed lines, and the corresponding distance is indicted in Angstroms (Å). The path of the substrate channel was defined in the absence of F430, CoBSH, HSCoM and water, with the surface closest to the viewer cut away. The figure was generated using PyMOL (http://www.pymol.org). Cedervall et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Drawing of CoBSH analogues; (A) N-5-mercaptopentanoylthreonine phosphate (CoB5SH); (B) N-6-mercaptohexanoylthreonine phosphate (CoB6SH); (C) N-8- mercaptooctanoylthreonine phosphate (CoB8SH); (D) N-9-mercaptononanoylthreonine phosphate (CoB9SH). Cedervall et al. Page 16 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. The active sites and substrate channels of the MCR crystal structures.; (A) MCRCoB5SH; (B) MCRCoB6SH; (C) MCRHSCoM; (D) MCRCoB8SH; (E) MCRCoB9SH. 2Fo-Fc electron density map around the CoBXSH analogues, waters in the CoBSH binding part of the channel and the acetate ion (contoured at 1σ) is shown as a blue mesh. The protein is drawn as cartoon. CoBXSH and acetate are drawn as stick and colored by atom (carbon: CoB5SH orange; CoB6SH pale yellow; CoB8SH light blue; CoB9SH magenta; acetate white). Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere, and waters as red spheres. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 17 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. Overlay of CoBSH (from PDB code 1hbn) and the different CoBSH analogues. CoBXSH are drawn as stick with the thiol represented by a sphere, and colored CoB5SH orange; CoB6SH pale yellow; CoBSH light green; CoB8SH light blue; CoB9SH magenta. The protein is drawn as cartoon. Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 18 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Hydrogen bonding diagram for the water structure modeled in MCRHSCoM. The water molecules are named as in Figure 3C (W1–W9); WA, WB and WC are water molecules that are present in all structures (i.e. in concert with the substrate CoBSH and the CoBSH analogues). Interactions between surrounding residues and the water molecules are drawn as dashed lines, and the corresponding distance is indicated in Angstroms (Å). Cedervall et al. Page 19 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. Stereo image of the annulus of aromatic amino acids distal of coenzyme F430. The protein is drawn as cartoon with the side-chains of the aromatic residues drawn as white stick. CoBSH, (from PDB code 1hbn (9)), CoB8SH and CoB9SH are drawn as stick with the thiols represented by spheres, and colored CoBSH light green; CoB8SH light blue; CoB9SH magenta. Coenzyme F430 and HSCoM are drawn as stick colored by atom (carbon: F430 dark grey; HSCoM medium grey). The nickel is displayed as a green sphere. The figure was generated using PyMOL (http://www.pymol.org/). Cedervall et al. Page 20 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. Reaction catalyzed by methyl-coenzyme M reductase Cedervall et al. Page 21 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 2. Two of the proposed catalytic mechanisms for methyl-coenzyme M reductase; (A) mechanism I; (B) mechanism II. Cedervall et al. Page 22 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 23 Table 1 X-ray Data Collection, Processing and Refinement Statistics Data collection and processing statistics Name of data set MCRCoB5SH MCRCoB6SH MCRHSCoM MCRCoB8SH MCRCoB9SH Measured reflections 1969388 2427498 1440665 1160543 1425506 Unique reflections 553755 446253 405349 211803 401701 Resolution (Å) a 50.0–1.30 (1.35–1.30) 50.0–1.40 (1.45–1.40) 50.0–1.45 (1.50–1.45) 50.0–1.80 (1.86–1.80) 50.0–1.45 (1.50–1.45) Completeness (%) a 97.1 (78.1) 99.9 (100.0) 99.5 (99.7) 99.8 (100.0) 98.1 (95.4) R-sym (%) a,b 5.5 (32.9) 7.3 (44.7) 6.2 (44.0) 8.4 (47.7) 5.6 (42.5) I/σI a 22.3 (3.6) 20.4 (4.0) 20.2 (3.2) 21.8 (3.9) 24.3 (3.2) Space group P21 P21 P21 P21 P21 Refinement and model building statistics Resolution (Å) a 20.49–1.30 (1.33–1.30) 19.89–1.40 (1.44–1.40) 20.15–1.45 (1.49–1.45) 19.93–1.80 (1.84–1.80) 20.07–1.45 (1.48–1.45) No. of reflection in working set a 525817 (30239) 423854 (25833) 384868 (25791) 201128 (11193) 381474 (23611) No. of reflection in test set a 27777 (1576) 22348 (1331) 20362 (1319) 10625 (557) 20163 (1210) R-work (%) c 14.32 13.04 13.47 14.95 13.58 R-free (%) d 16.56 15.53 16.22 19.54 16.44 ESU (Å) R-work/R-free 0.044/0.046 0.049/0.051 0.056/0.059 0.121/0.119 0.057/0.060 No. protein atoms 20087 19960 20265 19750 20036 No. coenzyme atoms 218 220 180 224 272 No. ligand atoms 37 62 52 26 49 No. water molecules 2443 2352 2516 1893 2432 RMS bond lengths (Å) 0.033 0.033 0.032 0.028 0.032 bond angles (deg.) 2.693 2.625 2.468 2.059 2.549 Ramachandran plot (%) favored 97.8 97.5 97.6 97.2 97.7 allowed 2.1 2.4 2.3 2.7 2.1 Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 24 disallowed 0.1 0.1 0.1 0.1 0.1 Average B-factor (Å2) protein 12.42 13.35 12.12 17.22 12.73 coenzymes 8.20 9.24 7.25 11.24 8.27 ligands 31.95 35.48 28.29 33.76 32.92 waters 22.95 24.89 23.85 26.79 24.09 over all 13.54 14.57 13.40 18.02 13.93 Occupancy of HSCoM per active site (%)e 90/90 50/50 100/100 90/90 90/85 Occupancy of CoBSH per active site (%) e 50/50 50/50 30/30 50/50 40/40 CoBSH analogue, occupancy per active site (%) e CoB5SH, 50/50 CoB6SH, 50/50 CoB8SH, 50/50 CoB9SH, 60/60 Other molecule, occupancy per active site (%) e Acetate, 70/70 aValues in brackets correspond to the highest resolution shell. bR-sym = ΣhklΣNj=1|Ihkl−Ihkl(j)|/ΣhklN*Ihkl, sum over all reflections and all observations N, with Ihkl(j) intensity of the jth observation of reflection hkl and Ihkl mean intensity of the reflection hkl. cR-work = Σ||Fo| − |Fc||/Σ|Fo|, where |Fo| = observed structure factor amplitude and |Fc| = calculated structure factor amplitude. dR-free, R-factor based on 5% of the data excluded from refinement. eOccupancy of model in each of the two crystallographically independent active sites in the ASU Biochemistry. Author manuscript; available in PMC 2011 September 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cedervall et al. Page 25 Table 2 Distances from analogue thiols. CoBXS - SCoM distance (Å) CoBXS - Ni distance (Å) CoB5SH 7.11/7.11a 9.30/9.30 CoB6SH 6.26/6.26 8.70/8.70 CoB7SH (substrate) b 6.37/6.39 8.73/8.77 CoB8SH 3.75/3.78 6.16/6.17 CoB9SH 3.71/3.68 5.96/5.91 aDistances in the two crystallographically independent active sites in the ASU bDistances in the 1.16 Å resolution MCRox1-silent structure (PDB code 1hbn)(33) Biochemistry. Author manuscript; available in PMC 2011 September 7.
3M38
The roles of Glutamates and Metal ions in a rationally designed nitric oxide reductase based on myoglobin: I107E FeBMb (No metal ion binding to FeB site)
Roles of glutamates and metal ions in a rationally designed nitric oxide reductase based on myoglobin Ying-Wu Lina,2, Natasha Yeunga, Yi-Gui Gaob, Kyle D. Minerc, Shiliang Tiana, Howard Robinsond, and Yi Lua,c,1 aDepartment of Chemistry, bGeorge L. Clark X-Ray Facility and 3M Materials Laboratory, andcDepartment of Biochemistry, University of Illinois at Urbana-Champaign, Urbana, IL 61801; and dDepartment of Biology, Brookhaven National Laboratory, Upton, NY 11973 Edited by Harry B. Gray, California Institute of Technology, Pasadena, CA, and approved April 1, 2010 (received for review January 14, 2010) A structural and functional model of bacterial nitric oxide reductase (NOR) has been designed by introducing two glutamates (Glu) and three histidines (His) in sperm whale myoglobin. X-ray structural data indicate that the three His and one Glu (V68E) residues bind iron, mimicking the putative FeB site in NOR, while the second Glu (I107E) interacts with a water molecule and forms a hydrogen bond- ing network in the designed protein. Unlike the first Glu (V68E), which lowered the heme reduction potential by ∼110 mV, the second Glu has little effect on the heme potential, suggesting that thenegativelychargedGluhasadifferentrolein redoxtuning.More importantly, introducing the second Glu resulted in a ∼100% in- crease in NOR activity, suggesting the importance of a hydrogen bonding network in facilitating proton delivery during NOR reactiv- ity. In addition, EPR and X-ray structural studies indicate that the designed protein binds iron, copper, or zinc in the FeB site, each with different effects on the structures and NOR activities, suggesting that both redox activity and an intermediate five-coordinate heme-NO species are important for high NOR activity. The designed protein offers an excellent model for NOR and demonstrates the power of using designed proteins as a simpler and more well- defined system to address important chemical and biological issues. biomimetic models ∣heme-copper oxidase ∣metalloprotein ∣protein design ∣ protein engineering R ational design of proteins that mimic both structure and func- tion of more complex native enzymes has been a long sought- after goal, as the process is an ultimate test of our knowledge and an excellent means to develop advanced biocatalysts (1–3). Although designed proteins that model the structure of native enzymes have been known for a while (4–10), successful designs of proteins that mimic both the structure and function of native enzymes have been reported only recently (11–16). While being able to design such functional proteins is laudable, the impact of such an achievement would be greater if the designed proteins can be used to address fundamental issues in chemistry and biol- ogy that are difficult to tackle by other methods. One primary example is the roles of conserved glutamates and metal ions in bacterial nitric oxide reductase (NOR) (17–19). NO is critical for all life (20). Bacterial denitrification is a cru- cial part of the nitrogen cycle in nature that involves a four-step, five-electron reduction of nitrate (NO3 −) to dinitrogen (N2) (17, 19). Bacterial NOR is a membrane-bound protein that catalyzes one step of this process, namely, the two-electron reduc- tion of NO to N2O (17, 19). With no crystal or solution structure available for bacterial NOR to date, sequence alignments and homology modeling (21, 22) have indicated that NOR is structu- rally homologous to the largest subunit (subunit I) of heme- copper oxidases (HCOs) (23), enzymes that catalyze reduction of O2 to water. The active sites of both NOR and HCO contain a proximal histidine-coordinated heme and a distal three histi- dine-coordinated metal center. However, the metal center in HCOs is occupied by a copper (called CuB), whereas a nonheme iron is present in NOR (called FeB) (23, 24). In addition, two conserved glutamates, shown by modeling to be close to the FeB site (21, 22), are found to be essential for NOR activity (24, 25). Some members of HCOs such as cytochrome cbb3 oxidase display NOR activity (26–28), although the activity is ∼50-fold lower than native NOR (26). Therefore, it is important to elucidate the structural features, specifically the roles of the conserved glutamates close to the FeB site and metal ions (copper vs. iron), responsible for the reduction of NO to N2O. To address these issues, biochemical and biophysical studies of native NOR and its variants have been carried out (24, 25, 29–37). For example, Richardson and coworkers investigated the effects of amino acid substitutions of the five conserved glutamates (E122 and E125 presumed to face the periplasm and E198, E202, and E267 located in the interior of the membrane, close to the catalytic site) in the catalytic subunit of Paracoccus deni- trificans, NorB. The E122A, E125A, E198A, and E267A variants were inactive, indicating that these four glutamates are crucial for NOR activity (24, 25, 32, 33). On the other hand, Reimann et al. constructed a 3D model of NorB using homology modeling with the structures of HCOs as templates and suggested a plausible pathway consisting of these conserved glutamates for proton delivery (22). Despite these successes, the roles of the conserved glutamates and metal ions still remain to be fully elucidated, partly because of the difficulty in obtaining native NOR in high yield and the lack of a 3D structure. Even if these problems are resolved, it is still difficult to replace iron in the native FeB site with other metal ions, and spectroscopic studies of native NOR are often complicated by the presence of other metal cofactors (e.g., low-spin heme). To overcome these limitations, a number of synthetic models of NOR using small organic molecules as ligands, have been made in which the nonheme FeB site can be replaced by a copper ion (17, 38–45). In addition, since these model systems lack addi- tional metal-binding sites, spectroscopic studies are often simpli- fied. Therefore, studies of these synthetic models have offered many insights. For example, Collman et al. showed that a fully reduced heme/nonheme FeB compound can react with two equivalents of NO leading to the formation of one equivalent of N2O and a bis-ferric product (41). On the other hand, Karlin and coworkers showed that a small heme/Cu complex can effi- ciently lead to reductive coupling of NO to N2O (43). However, it is also difficult to obtain the synthetic models in high yield due to the multiple steps required in chemical synthesis. Because of this limitation, no synthetic NOR model containing the two key conserved Glu residues (E198 and E267 in NOR) has been Author contributions: Y.-W.L., N.Y., and Y.L. designed research; Y.-W.L., N.Y., Y.-G.G., K.D.M., and S.T. performed research; Y.-W.L., N.Y., Y.-G.G., K.D.M., H.R., and Y.L. analyzed data; and Y.-W.L. and Y.L. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Data deposition: The atomic coordinates and structure factors have been deposited in the Protein Data Bank, www.pdb.org (PDB ID codes 3M38, 3M39, 3M3A, and 3M3B). 1To whom correspondence should be addressed. E-mail: yi-lu@illinois.edu. 2Present address: School of Chemistry and Chemical Engineering, University of South China, Hengyang 421001, China. This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1000526107/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1000526107 PNAS ∣May 11, 2010 ∣vol. 107 ∣no. 19 ∣8581–8586 BIOCHEMISTRY CHEMISTRY reported. It is also difficult to substitute different metal ions in the same metal-binding site without perturbing the site geometry and distances to the heme iron, as most ligands are not as rigid as those in native enzymes and different metal ions have different geometric and ligand donor set preferences. We have recently designed a structural and functional protein model of bacterial NOR by engineering three histidines and one glutamate into the distal pocket of sperm whale myoglobin (swMb, L29H, F43H, H64, and V68E, named FeBMb) (14). Like synthetic models, this “bottom-up” approach complements the “top-down”approachofthestudyofnativeNORinthatitprovides insights into whether certain “necessary” structural elements are enough to impart enzyme function. Thanks in part to recent advances in computational, molecular, and structural biology, the designed myoglobin protein model is much easier to synthesize and to crystallize than either native NOR or synthetic models. Since myoglobin has often been used for the development and calibration of numerous spectroscopic techniques (46–48), it is an ideal choice for spectroscopic studies. More importantly, the rigid protein network allows precise placement of either glutamate or metal ions in myoglobin to address their roles in NOR activity. Toward this goal, we have demonstrated that both the histidines and one of the glutamates are essential for iron binding and NO reduction activity (14). However, the role of the second Glu close to the FeB site and the role of different metal ions in the FeB site have not been addressed. To address these important issues and to design even closer protein models of NOR, we introduced herein the second Glu to the second coordination sphere of the FeB site by mutating an Ile to a Glu (named I107E FeBMb). We show that the second Glu results in a ∼100% increase in NOR activity through hydro- gen bonding interactions and that the two glutamates have dramatically different effects on the heme reduction potential. Additionally, by comparing the EPR, electrochemistry, X-ray structures, and NOR activity of iron, copper, and zinc derivatives of the designed protein, we have obtained deeper insights into the roles of metal ions in NOR. Results Structure and Function of Fe(II)-I107E FeBMb. The X-ray crystal struc- tures of heme-containing I107E FeBMb without metal ion in the FeB site and with Fe2þ in the FeB site are solved at 1.42-Å and 1.65-Å resolution, respectively (Fig. 1 A and B and Table S1). In the absence of metal ions in the FeB site, the structure shows a water molecule in the FeB site, which forms hydrogen bonds with NE2 atoms of all three His residues, both OE1 and OE2 atoms of E68, and the OE2 atom of E107 (Fig. 1A). Upon binding Fe2þ, the Fe(II)-I107E FeBMb structure shows that Fe2þ is coordinated by three His, the OE2 atom of E68, and one water molecule. Notably, a water molecule bridges Fe2þ in the FeB site and the second glutamate (E107) with a distance of 2.32 Å to the OE2 atom of E107 (Fig. 1B). To probe the conformational changes of introducing the second Glu (E107), we performed a structural alignment of Fe (II)-I107E FeBMb and the previously reported FeðIIÞ-FeBMb (14). The comparison, shown in Fig. 2, indicates that both the polypeptide chain and the active site overlap well with each other. In addition, the two nonheme irons are located at similar posi- tions with a 0.36-Å separation from each other. In contrast, E68 underwent a significant conformational rearrangement in the presence of E107. These observations suggest that the active site of FeBMb can be tuned by the formation of an extended hydrogen bonding network, resulting from the introduction of a second glutamate residue. The binding of Fe2þ to deoxy I107E FeBMb was further mon- itored by EPR (Fig. 3A). Since deoxy myoglobin contains Fe(II) heme that exhibits no EPR signals in X-band EPR (14), we added blue copper Cu(II)-azurin (49), a redox partner of native NOR (19), to oxidize both the reduced heme and nonheme irons in Fe (II)-I107E FeBMb to EPR-active Fe(III). Upon addition of Cu (II)-azurin, the oxidation of deoxy I107E FeBMb resulted in EPR signals at g ¼ 6.12 and 5.56, typical of a high-spin heme- Fe(III). Upon addition of Fe2þ, however, a decrease of the heme-Fe(III) EPR signals was observed, indicating that the Fe2þ, when bound to the FeB site and oxidized by Cu(II)-azurin, is spin-coupled to heme-Fe(III). Such a spin coupling mimics that in NOR (35, 50–53), suggesting that I107E FeBMb models NOR closely, at least in this respect. To probe the role of the second Glu (E107) in NO reduction activity, we measured the yield of N2O production by Fe(II)- I107E FeBMb with excess NO under one turnover conditions. We monitored N2O formation in the headspace of the solution using GC/MS and compared this result to that of FeðIIÞ-FeBMb, which lacks the second Glu (Fig. 4). Remarkably, Fe(II)-I107E FeBMb displays higher activity than FeðIIÞ-FeBMb. After ∼20 hr, ∼24% N2O was produced by Fe(II)-I107E FeBMb, in H29 A H29 B H64 E68 E107 3.04 2.81 3.02 2.20 2.24 2.24 2.12 2.21 H64 E68 E107 3.03 H43 2.15 3.41 3.16 2.62 2.32 H43 2.26 H93 H93 H29 C H29 D 2.04 2.03 H64 E68 E107 2.09 2.10 2.91 2.21 H64 E68 E107 2.26 2.29 2.10 2.18 2.10 4.47 H43 2.07 3.04 H43 2.68 H93 H93 Fig. 1. Crystal structures of I107E FeBMb (A) (PDB ID code 3M38), Fe(II)-I107E FeBMb (B) (PDB ID code 3M39), Cu(II)-I107E FeBMb (C) (PDB ID code 3M3A), and Zn(II)-I107E FeBMb (D) (PDB ID code 3M3B). Water molecules, Fe(II), Cu(II), and Zn(II) are represented by red, green, orange, and gray spheres, respectively. Fig. 2. Overlay of Fe(II)-I107E FeBMb (cyan) (PDB ID code 3M39) with FeðIIÞ-FeBMb (orange) (PDB ID code 3K9Z). 8582 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1000526107 Lin et al. contrast to ∼10% yield for FeðIIÞ-FeBMb, strongly indicating that the second Glu plays an important role in NO reduction, likely facilitating proton uptake during NO reduction. Other Metal Ions Binding to I107E FeBMb. To find out if the resting state of the protein, i.e., oxidized or met I107E FeBMb, can bind other metal ions, Cu2þ or Zn2þ was titrated into met I107E FeBMb and monitored by EPR spectroscopy (Fig. 3 B and C). In the absence of metal ions, met I107E FeBMb exhibited high-spin heme signals at g ¼ 6.03, 5.08, and 1.98 (Fig. 3B, black line). Upon addition of 2 eq of Cu2þ, the signals at g ¼ 6.03 and 5.08 decreased and a broad peak around g ¼ 2.95 increased, probably due to spin coupling between heme-Fe(III) and Cu2þ in the FeB site. In contrast, addition of Zn2þ, a metal ion with no unpaired electrons [i.e., incapable of spin coupling to heme- Fe(III)], produced an increase in the high-spin heme signals at g ¼ 5.88 and 5.60 (Fig. 3C), indicating that the interaction be- tween E68 and heme iron was weakened after metal binding. The X-ray crystal structures of I107E FeBMb with Cu2þ or Zn2þ in the FeB site were solved at 1.37-Å and 1.60-Å resolution, respectively (Fig. 1 C and D and Table S1). Compared to Fe(II)- I107E FeBMb (Fig. 1B), a similar binding site was observed for Cu(II)-I107E FeBMb (Fig. 1C), where H29, H43, and H64 coordinate to Cu2þ with distances of 2.09, 2.10, and 2.04 Å, respectively, slightly shorter than the corresponding distances in the Fe2þ structure. In comparison to Fe(II)-I107E FeBMb, the water bridging the Cu2þ and the second Glu (E107) is shifted toward Cu2þ in the FeB site (2.03 Å) with respect to E107 (3.04 Å). Interestingly, this bridging water molecule was not observed in Zn(II)-I107E FeBMb (Fig. 1D), but the two O atoms of E68 coordinate to Zn2þ with similar distances (2.26 Å for OE1 and 2.29 Å for OE2). The longer distance between OE1 of E68 B 6.12 I107E FeBMb + Azurin I107E Fe Mb + 0 5 eq Fe 2+ + Azurin A 5.56 I107E FeBMb + 0.5 eq Fe + Azurin I107E FeBMb + 1.0 eq Fe 2+ + Azurin I107E FeBMb + 2.0 eq Fe 2+ + Azurin 1.98 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5.08 5.38 Magnetic Field (Gauss) 5.88 I107E FeBMb I107E FeBMb + 0.5 eq Zn 2+ I107E Fe Mb 1 0 eq Zn 2+ C 6.03 I107E FeBMb I107E FeBMb + 0.5 eq Cu 2+ I107E Fe Mb + 1 0 eq Cu 2+ I107E FeBMb + 1.0 eq Zn I107E FeBMb + 2.0 eq Zn 2+ I107E FeBMb + 1.0 eq Cu I107E FeBMb + 2.0 eq Cu 2+ 1.98 ~2.95 1.98 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5.60 Magnetic Field (Gauss) 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5.08 Magnetic Field (Gauss) Fig. 3. EPR spectra of deoxy I107E FeBMb (0.5 mM protein in 50 mM Bis-Tris, pH 7.0) with increasing concentrations of Fe2þ in the presence of wild-type Cu(II)-azurin (A), and oxidized I107E FeBMb with Cu2þ (B) or Zn2þ (C). Spectra were collected at 4 K, 5 mW power, and 9.05 GHz. 30 25 Fe(II)-I107E FeBMb Fe(II)-FeBMb 15 20 10 %) oduction (% N2O pro 0 5 0 4 8 12 16 20 Incubation time (hr) Fig. 4. Time-dependent N2O production by Fe(II)-I107E FeBMb (▴) and FeðIIÞ-FeBMb (●) with ∼50 eq. NO under single turnover conditions. The yield was determined by a comparison of the ratio of NO∶N2O peaks from the GC/MS chromatograms. Lin et al. PNAS ∣ May 11, 2010 ∣ vol. 107 ∣ no. 19 ∣ 8583 BIOCHEMISTRY CHEMISTRY and heme iron in the Zn-bound structure (2.68 Å) in comparison to the Cu- and Fe-bound structures, is likely the result of a weaker interaction, which is also supported by an observed increase of the high-spin heme signals in the EPR spectra upon Zn2þ binding (Fig. 3C). These results suggest I107E FeBMb is capable of incor- porating different metal ions into its designed FeB site, offering an excellent opportunity to compare the role of these metal ions in the same protein scaffold. Effect of Glutamates and Metal Ions on the Redox Potential of I107E FeBMb. Since EPR and X-ray structural studies indicate metal binding to I107E FeBMb, we used spectroelectrochemistry to measure the effects of glutamates and metal ions on the heme reduction potential. When there is no metal ion in the FeB site, the I107E FeBMb displays a reduction potential of −134  3 mV vs. the normal hydrogen electrode (NHE) (Fig. S1A), similar to that of FeBMb (−158  4 mV) without the I107E mutation (14). In the presence of Cu2þ, I107E FeBMb has a reduction potential (−137  2 mV) (Fig. S1B) almost identical to that of the same protein in the absence of metal ions in the FeB site, indicating that copper binding to the FeB site has little effect on the reduction potential of the heme iron. This observation is similar to that observed for Cu2þ binding to CuBMb (CuðIIÞ-CuBMb, 80 mV vs. CuBMb, 77 mV) (54). On the other hand, the presence of Fe2þ and Zn2þ increased the reduction potential of I107E FeBMb from −134  3 mV to −64  3 mV vs. NHE (Fig. S1C) and −105  2 mV vs. NHE (Fig. S1D), respectively. The different effects of Cu2þ, Fe2þ, and Zn2þ on the reduction potential of I107E-FeBMb indicate that these metal ions in the FeB site may play different roles through different coordination properties. NOR Activity of I107E FeBMb in the Presence of Different Metal Ions. The NO reduction activity of I107E FeBMb in the presence of Fe2þ, Cuþ, or Zn2þ was monitored by GC/MS under single turn- over conditions. When Fe(II)-I107E FeBMb was exposed to excess NO, N2O could be observed to form with increased yield over time (Fig. S2). Similarly, N2O formation was observed for Cu(I)-I107E FeBMb, indicating that Fe or Cu binding to the FeB site results in comparable NOR activities. It should be noted that because of the high solubility of N2O (∼25 mM in water at room temperature), GC/MS cannot be used to quantify the rates of NO reduction under these conditions. In contrast, no N2O formation was observed with redox inactive Zn2þ, which demon- strates that redox active Fe2þ or Cuþ in the FeB site plays a crucial role in NO reduction. To gain deeper insight into the process of NO reduction, EPR studies were further performed to monitor the initial process of NO reduction. In the absence of metal ions, the EPR spectrum of ferrous I107E FeBMb-NO shows hyperfine splitting resulting from bound NO and the proximal histidine, indicating the forma- tion of a six-coordinate ferrous heme-NO species (Fig. 5, top line). After incubation of Fe(II)-I107E FeBMb with excess NO, a distinct three-line hyperfine structure appears at 15 min (Fig. 5A), suggesting the formation of a five-coordinate ferrous heme-NO species as a result of cleavage of the proximal His-Fe heme bond (55). A three-line hyperfine structure was also observed for Cuþ and Zn2þ, except that the signal intensity is low- er with Cu(I)-I107E FeBMb-NO (Fig. 5B) and more pronounced in Zn(II)-I107E FeBMb-NO (Fig. 5C). The lower intensity of the three-line hyperfine structure for Cu(I)-I107E FeBMb-NO suggests the major species formed is a six-coordinate ferric heme-NO complex, which is EPR silent (41). These differences further suggest that the metal ion in the FeB site plays a key role in formation of the intermediates, thereby tuning NOR activity. Discussion Using Rationally Designed Proteins to Address Important Issues in Chemistry and Biology. Important issues such as the roles of the conserved glutamates and nonheme FeB in NOR have been previously addressed using biochemical and biophysical studies or biomimetic modeling (24, 25, 27–37, 45, 56, 57). As a comple- mentary approach, rational protein design, using small, easy-to- produce and well-characterized proteins such as myoglobin, offers a powerful method with which to gain insights into more complex native enzymes such as NOR (14). Similar to synthetic models (41, 43), the metal ion at the putative FeB site in the protein model can be substituted freely. Better yet, Glu residues can be placed at precise locations in the protein, including the secondary coordination sphere, due to its rigid network. By care- fully choosing a suitable protein template, rational protein design could be generally applied to address other important issues in chemistry and biology. The Roles of Glutamates. Although two conserved glutamates (E198 and E267) are known to be crucial for NOR activity (24, 25), their roles are not well defined (18, 19). In a previous study (14), we demonstrated that one Glu, E68, is important for both iron binding and NOR activity of FeBMb. The crystal struc- tures of both FeðIIÞ-FeBMb and Fe(II)-I107E FeBMb show that one O atom of E68 directly coordinates to FeB (Fig. 2). In syn- thetic models of NOR, it has also been found that the presence of a glutamic acid mimic significantly increases the stability of iron binding to the FeB site (40). Furthermore, a theoretical study by Blomberg et al. (58) showed that a model with an FeB coordi- nated by three histidines, one glutamate, and one water molecule provides an energetically feasible reaction mechanism of NO reduction. However, the structural model of NOR constructed recently by Reimann et al. (22) shows that the closest conserved Glu (E267) still has its carboxylate O atom 7 Å away from FeB, which suggests that Glu may not bind to FeB in native NOR. One interesting finding from our study is that the Glu (E68) under- went a significant conformational rearrangement in the presence of another Glu (E107) (Fig. 2). Therefore, the FeBMb provides a viable model of NOR that is consistent with Blomberg’s model, but cannot rule out Reimann’s model due to possible conforma- tion changes. While the role of the first Glu is still uncertain until a 3D struc- ture of NOR in its active form is available, the role of the second Glu is even less defined. We address this question by introducing a second Glu (E107) to FeBMb. The crystal structures shown in Fig. 1 indicate that E107 interacts with a water molecule and forms a hydrogen bonding network in both Fe(II)-I107E FeBMb and Cu(II)-I107E FeBMb. Interestingly, although a similar water molecule was observed in the active site of FeðIIÞ- FeBMb (Fig. 2), activity assay data indicate that the presence of E107 in Fe(II)-I107E FeBMb increases NOR activity by ∼100% C 5 min N t l A 5 min B 2+ 5 min 1 min No metal 1 min 5 min F 2+ No metal 1 i 5 min + No metal 2+ Zn 2+ 5 min Fe 2+ 5 min 1 min Fe C + 5 min 1 min Cu + Z 2 Zn 2+ 15 min F 2+ Fe 15 min 5 min C + Cu + 15 min 5 min Zn 2+ 15 min Fe2+ 5 Cu + 15 min 3000 3100 3200 3300 3400 Magnetic Field (Gauss) 3000 3100 3200 3300 3400 Magnetic Field (Gauss) 3000 3100 3200 3300 3400 Magnetic Field (Gauss) Fig. 5. EPR spectra of deoxy I107E FeBMb (0.5 mM) with no metal bound in the presence of NO after 5 min (top line), with 2 eq Fe2þ (A), Cuþ (B), or Zn2þ (C) incubated with excess NO (∼200 eq) for 1, 5, and 15 min. Spectra were collected in 50 mM Bis-Tris pH 7.0 at 30 K, 0.2 mW power, and 9.05 GHz. 8584 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1000526107 Lin et al. (Fig. 4), suggesting that the second Glu may potentionally play a role in providing one of the protons during reduction of NO to N2O. Although free Glu outside a protein has a pKa ∼4.3, the stu- dies of native NOR showed that the pKa of its Glu close to the active site has a value of ∼6.6 (22, 25, 33). The hydrogen binding network in our protein models may contribute to the fine-tuning of the Glu pKa to be more neutral, similar to that in NOR. More- over, it is interesting that Cu(I)-I107E FeBMb also shows NOR activity, which provides an interesting protein model of HCOs with NOR function (26–28), even though Glu residues are not conserved in native HCOs. Additionally, spectroelectrochemical studies showed that the reduction potential of I107E FeBMb with no metal ion in the FeB site is similar to that of FeBMb, but much lower than that of CuBMb (77 mV) (54), which contains the same three His, but no Glu in the metal-binding site above the heme. Since both FeBMb and I107E FeBMb contain the V68E mutation that has been shown to decrease the heme reduction potential of native myoglobin from 59 mV to −137 mV (59), it is likely that the introduction of a negatively charged Glu close to the heme group is what is responsible for the dramatically lower heme redox potential. A conserved glutamate, predicted to be located near the catalytic heme b3 in NOR, was proposed to be responsible for a ∼260-mV decrease in reduction potential (60 mV) in comparison to the other two heme centers, heme b (345 mV) and heme c (310 mV) (60). Our FeBMb and I107E FeBMb mod- els mimic this feature of NOR. Notably, although introduction of the first Glu (E68) lowered the heme potential by ∼110 mV (14), introduction of the second Glu via the I107E mutation did not result in a significant difference in the heme reduction potential, suggesting that the effect of the two conserved Glu residues in NOR on heme reduction potential is not additive, with the effects highly dependent on the location of the Glu. The Roles of Metal Ions. The roles of metal ions in NOR are an- other important question as iron is found in the native FeB site and HCO employs copper at the corresponding CuB site. With different metal ions in the FeB site, the crystal structures clearly show the heme and nonheme dinuclear center existing in differ- ent local environments (Fig. 1). Although a similar hydrogen bond network is formed in both Fe(II)-I107E FeBMb and Cu (II)-I107E FeBMb, the conformation of E68 and E107 with re- spect to the nonheme metal center and heme iron is different from each other. Moreover, the coordination geometry differs significantly with Zn2þ in the FeB site. A hydrogen bond is absent from the Zn crystal structure, but both the O atoms of E68 act as metal-binding ligands. These observations demonstrate that the identity of the metal ion in the FeB site can tune the active site through their interactions with the His and Glu ligands, resulting in formation of different coordination geometries with different hydrogen bonds. In addition to structural fine-tuning, the metal ion at the FeB site can also tune the heme iron reduction potential in I107E FeBMb. Spectroelectrochemical studies showed that the binding of Fe2þ or Zn2þ results in an increase in the heme reduction potential by ∼70 mV and ∼30 mV, respectively (Fig. S1). In the case of Cu(II)-I107E FeBMb, the crystal structure shows that OE1 of E68 is closer to the heme iron (2.07 Å) (Fig. 1C) than its metal-free form (2.15 Å) (Fig. 1A). The stronger interaction from the negatively charged E68 could offset the effect of positively charged Cu2þ binding, resulting in similar reduction potentials observed for Cu(II)-I107E FeBMb and I107E FeBMb. In a previous study (61), EPR data showed that during NO reduction, the binding of Cuþ to the CuB site of CuBMb can weaken the proximal heme Fe-His bond, while complete cleavage of the heme Fe-His bond occurred when Zn2þ was bound to CuBMb-NO. In this study, we observed that a five-coordinate heme-NO species was formed with Fe2þ, Cuþ, or Zn2þ bound to the FeB site of I107E FeBMb (Fig. 5). Significantly, a five- coordinate heme-NO species has also been observed for both NOR (30, 31, 35) and the member of the HCO family with the highest NO reduction activity, cytochrome cbb3 oxidases (26, 62). However, this species was not observed for FeðIIÞ-FeBMb-NO and CuðIÞ-FeBMb-NO, which lack the second Glu (E107). In both these cases, the proximal heme Fe-His bond was only weakened, as indicated by a decrease of the nine-line hyperfine splitting signals in the EPR spectra (Fig. S3). These observations suggest that formation of a five-coordinate heme-NO species may play an important role in NOR reactivity. Conclusions We have successfully designed a structural and functional model of NOR, by introducing a second glutamate in the vicinity of the FeB site, named I107E FeBMb. This protein model mimics native NOR more closely by bearing the structural feature of three his- tidines and two glutamates in the FeB site, as predicted for native NOR. We have demonstrated that the two glutamates can play different roles in NO reduction activity; namely, one acts as a li- gand to FeB (E68), and the other acts as a proton transfer group (E107). Furthermore, by substituting different metal ions into the nonheme metal site, we have demonstrated that FeB plays crucial roles in fine-tuning the active site by donating electrons and by mediating the formation of a five-coordinate heme-NO inter- mediate during NO reduction. In the absence of a crystal struc- ture for native NOR, this study offers an ideal protein model and provides valuable structural as well as mechanistic information for native NOR. Materials and Methods Protein Preparation. I107E FeBMb (swMb L29H/F43H/V68E/I107E) was con- structed, expressed, and purified using the procedure described previously (14). The purity and identity were confirmed by SDS-PAGE and electrospray ionized MS: observed: 17; 392  1 Da; calculated: 17,391 Da. EPR Spectroscopy. EPR spectra were recorded on a Bruker ESP 300 equipped with an Oxford liquid helium cryostat and an ITC4 temperature controller. The samples of met I107E FeBMb, Cu(II)-, or Zn(II)-I107E FeBMb were pre- pared as described previously (14). The samples of NO-bound deoxy I107E FeBMb, Cu(I)-, Fe(II)-, or Zn(II)-I107E FeBMb were prepared by inject- ing 1 mL of purified NO gas into the EPR tube containing 300 μL of protein (0.5 mM, 10% glycerol, in 50 mM Bis-Tris pH 7.0). The samples were then flash frozen in liquid N2 after incubation for 1, 5, or 15 min. The molar extinction coefficient of the Soret band of I107E FeBMb at 406 nm (175 mM−1 · cm−1), calculated using the standard hemochromagen method (63), was used to determine protein concentration. The metal sources of Cu(I), Cu(II), Zn(II), and Fe(II) were ½ðCH3CNÞ4CuPF6, CuSO4, ZnSO4 · 7H2O, and FeCl2, respectively. Spectroelectrochemical Measurements. Protein reduction potentials were measured using an optically transparent thin layer electrode as previously described (64). The potential of the working electrode was applied in the positive direction for deoxy I107E FeBMb with Fe2þ and in the negative direction for metal free and with Cu2þ or Zn2þ. Other procedures are the same as described previously (54). X-Ray Crystallographic Studies. Fe(II)-I107E FeBMb was crystallized anaerobi- cally in a glove box at room temperature using the conditions described for FeðIIÞ-FeBMb (14). I107E FeBMb, Cu(II)-, or Zn(II)-I107E FeBMb were crystallized aerobically. Diffraction-quality crystals were soaked in a cryoprotectant solution of 30% PEG 400 and flash frozen in liquid nitrogen. Diffraction data were collected at the Brookhaven National Lab Synchrotron Light Source X12C beamline. The crystal structure was solved using the same method as for FeðIIÞ-FeBMb (14). NOR Activity Assay. NO reduction was monitored by GC/MS. The protein was reduced to the deoxy form by excess dithionite that was removed with a size-exclusion column (PD-10). Then 2 eq metal, Cu(I), Fe(II), or Zn(II), was added to the protein solution (0.6 mM, 3 mL in 50 mM Bis-Tris buffer, pH 7.0). The samples were prepared anaerobically in a glove box. Purified NO Lin et al. PNAS ∣ May 11, 2010 ∣ vol. 107 ∣ no. 19 ∣ 8585 BIOCHEMISTRY CHEMISTRY gas was injected into the head space of the reaction flask with the molar ratio of NO∶protein ≈50∶1. Other procedures are the same as described previously (14, 61). ACKNOWLEDGMENTS. We thank Dr. Mark J. Nilges for help with EPR analysis, and Furong Sun and Beth D. Eves for aiding in GC/MS data collection. This work was supported by NIH Grant GM062211. 1. Lu Y, Yeung N, Sieracki N, Marshall NM (2009) Design of functional metalloproteins. Nature 460:855–862. 2. Nanda V, Koder RL (2010) Designing artificial enzymes by intuition and computation. Nat Chem 2:15–24. 3. Gray HB (2003) Biological inorganic chemistry at the beginning of the 21st century. Proc Natl Acad Sci USA 100:3563–3568. 4. Regan L, DeGrado WF (1988) Characterization of a helical protein designed from first principles. Science 241:976–978. 5. Hecht MH, Richardson JS, Richardson DC, Ogden RC (1990) De novo design, expression, and characterization of Felix: A four-helix bundle protein of native-like sequence. Science 249:884–891. 6. Robertson DE, et al. (1994) Design and synthesis of multi-heme proteins. Nature 368:425–432. 7. Lu Y, Valentine JS (1997) Engineering metal-binding sites in proteins. Curr Opin Struct Biol 7:495–500. 8. Reedy CJ, Gibney BR (2004) Heme protein assemblies. Chem Rev 104:617–649. 9. Watanabe Y, Hayashi T (2005) Functionalization of myoglobin. Prog Inorg Chem 54:449–493. 10. Ghosh D, Pecoraro VL (2005) Probing metal-protein interactions using a de novo design approach. Curr Opin Chem Biol 9:97–103. 11. Kaplan J, DeGrado WF (2004) De novo design of catalytic proteins. Proc Natl Acad Sci USA 101:11566–11570. 12. Kang SG, Saven JG (2007) Computational protein design: Structure, function and combinatorial diversity. Curr Opin Chem Biol 11:329–334. 13. Jiang L, et al. (2008) De novo computational design of retro-aldol enzymes. Science 319:1387–1391. 14. Yeung N, et al. (2009) Rational design of a structural and functional nitric oxide reductase. Nature 462:1079–1082. 15. Koder RL, et al. (2009) Design and engineering of an O2 transport protein. Nature 458:305–309. 16. Heinisch T, Ward TR (2010) Design strategies for the creation of artifical metalloen- zymes. Curr Opin Chem Biol 14:184–199. 17. Wasser IM, de Vries S, Moënne-Loccoz P, Schröder I, Karlin KD (2002) Nitric oxide in biological denitrification: Fe/Cu metalloenzyme and metal complex NOx redox chemistry. Chem Rev 102:1201–1234. 18. Yeung N, Lu Y (2008) One heme, diverse functions: Using biosynthetic myoglobin models to gain insights into heme-copper oxidases and nitric oxide reductases. Chem Biodivers 5:1437–1454. 19. Watmough NJ, Field SJ, Hughes RJL, Richardson DJ (2009) The bacterial respiratory nitric oxide reductase. Biochem Soc Trans 37:392–399. 20. Cary SPL, Winger JA, Derbyshire ER, Marletta MA (2006) Nitric oxide signaling: No longer simply on or off. Trends Biochem Sci 31:231–239. 21. Sakurai N, Sakurai T (1998) Genomic DNA cloning of the region encoding nitric oxide reductase in paracoccus halodenitrificans and a structure model relevant to cytochrome oxidase. Biochem Biophys Res Commun 243:400–406. 22. Reimann J, Flock U, Lepp H, Honigmann A, Ädelroth P (2007) A pathway for protons in nitric oxide reductase from paracoccus denitrificans. Biochim Biophys Acta 1767:362–373. 23. Pereira MM, Sousa FL, Verissimo AF, Teixeira M (2008) Looking for the minimum common denominator in haem-copper oxygen reductases: Towards a unified catalytic mechanism. Biochim Biophys Acta 1777:929–934. 24. Butland G, Spiro S, Watmough NJ, Richardson DJ (2001) Two conserved glutamates in the bacterial nitric oxide reductase are essential for activity but not assembly of the enzyme. J Bacteriol 183:189–199. 25. Flock U, Lachmann P, Reimann J, Watmough NJ, Äedelroth P (2009) Exploring the terminal region of the proton pathway in the bacterial nitric oxide reductase. J Inorg Biochem 103:845–850. 26. Forte E, et al. (2001) The cytochrome cbb3 from pseudomonas stutzeri displays nitric oxide reductase activity. Eur J Biochem 268:6486–6490. 27. Huang Y, Reimann J, Lepp H, Drici N, Ädelroth P (2008) Vectorial proton transfer coupled to reduction of O2 and NO by a heme-copper oxidase. Proc Natl Acad Sci USA 105:20257–20262. 28. Hayashi T, et al. (2009) Accommodation of two diatomic molecules in cytochrome bo3: Insights into NO reductase activity in terminal oxidases. Biochemistry 48:883–890. 29. Moënne-Loccoz P (2007) Spectroscopic characterization of heme iron-nitrosyl species and their role in NO reductase mechanisms in diiron proteins. Nat Prod Rep 24:610–620. 30. Moënne-Loccoz P, de Vries S (1998) Structural characterization of the catalytic high-spin heme b of nitric oxide reductase: A resonance raman study. J Am Chem Soc 120:5147–5152. 31. Sakurai T, Sakurai N, Matsumoto H, Hirota S, Yamauchi O (1998) Roles of four iron centers in paracoccus halodenitrificans nitric oxide reductase. Biochem Biophys Res Commun 251:248–251. 32. Thorndycroft FH, Butland G, Richardson DJ, Watmough NJ (2007) A new assay for nitric oxide reductase reveals two conserved glutamate residues form the entrance to a proton-conducting channel in the bacterial enzyme. Biochem J 401:111–119. 33. Flock U, et al. (2008) Defining the proton entry point in the bacterial respiratory nitric-oxide reductase. J Biol Chem 283:3839–3845. 34. Hendriks JHM, Jasaitis A, Saraste M, Verkhovsky MI (2002) Proton and electron pathways in the bacterial nitric oxide reductase. Biochemistry 41:2331–2340. 35. Kumita H, et al. (2004) NO reduction by nitric-oxide reductase from denitrifying bacterium pseudomonas aeruginosa: Characterization of reaction intermediates that appear in the single turnover cycle. J Biol Chem 279:55247–55254. 36. Pinakoulaki E, Varotsis C (2008) Resonance raman spectroscopy of nitric oxide reductase and cbb3 heme-copper oxidase. J Phys Chem B 112:1851–1857. 37. Kapetanaki SM, et al. (2008) Ultrafast ligand binding dynamics in the active site of native bacterial nitric oxide reductase. Biochim Biophys Acta 1777:919–924. 38. Wasser IM, et al. (2004) Synthesis and spectroscopy of m-oxo (O2-)-bridged Heme/ Non-heme diiron complexes: Models for the active site of nitric oxide reductase. Inorg Chem 43:651–662. 39. Wasser IM, Huang H, Moënne-Loccoz P, Karlin KD (2005) Heme/Non-heme diiron(II) complexes and O2, CO, and NO adducts as reduced and substrate-bound models for the active site of bacterial nitric oxide reductase. J Am Chem Soc 127:3310–3320. 40. Collman JP, Yan Y, Lei J, Dinolfo PH (2006) Active-site models of bacterial nitric oxide reductase featuring tris-histidyl and glutamic acid mimics: Influence of a carboxylate ligand on FeB binding and the heme Fe∕FeB redox potential. Inorg Chem 45:7581–7583. 41. Collman JP, et al. (2008) A functional nitric oxide reductase model. Proc Natl Acad Sci USA 105:15660–15665. 42. Collman JP, et al. (2008) Intermediates involved in the two electron reduction of NO to N2O by a functional synthetic model of heme containing bacterial NO reductase. J Am Chem Soc 130:16498–16499. 43. Wang J, Schopfer MP, Sarjeant Amy AN, Karlin KD (2009) Heme-copper assembly mediated reductive coupling of nitrogen monoxide (*NO). J Am Chem Soc 131:450–451. 44. Liu J, Naruta Y, Tani F (2005) A functional model of the cytochrome c oxidase active site: Unique conversion of a heme-m-peroxo-CuII intermediate into heme-superoxo/ CuI. Angew Chem Int Edit 44:1836–1840. 45. Ghiladi RA, et al. (2005) Heme-copper/dioxygen adduct formation relevant to cytochrome c oxidase: Spectroscopic characterization of ½ð6LÞFeIII-ðO2 2−Þ-CuIIIþ. J Biol Inorg Chem 10:63–77. 46. Sage JT (1997) Myoglobin and CO: Structure, energetics, and disorder. J Biol Inorg Chem 2:537–543. 47. Sigman JA, Kim HK, Zhao X, Carey JR, Lu Y (2003) The role of copper and protons in heme-copper oxidases: Kinetic study of an engineered heme-copper center in myoglobin. Proc Natl Acad Sci USA 100:3629–3634. 48. Davydov R, Hoffman BM (2008) EPR and ENDOR studies of fe(II) hemoproteins reduced and oxidized at 77 K. J Biol Inorg Chem 13:357–369. 49. Marshall NM, et al. (2009) Rationally tuning the reduction potential of a single cupredoxin beyond the natural range. Nature 462:113–116. 50. Girsch P, de Vries S (1997) Purification and initial kinetic and spectroscopic character- ization of NO reductase from paracoccus denitrificans. Biochim Biophys Acta 1318:202–216. 51. Sakurai N, Sakurai T (1997) Isolation and characterization of nitric oxide reductase from paracoccus halodenitrificans. Biochemistry 36:13809–13815. 52. Hendriks J, et al. (1998) The active site of the bacterial nitric oxide reductase is a dinuclear iron center. Biochemistry 37:13102–13109. 53. Moënne-Loccoz P, et al. (2000) Nitric oxide reductase from paracoccus denitrificans contains an oxo-bridged Heme/Non-heme diiron center. J Am Chem Soc 122:9344–9345. 54. Zhao X, Yeung N, Wang Z, Guo Z, Lu Y (2005) Effects of metal ions in the CuB center on the redox properties of heme in heme-copper oxidases: Spectroelectrochemical stu- dies of an engineered heme-copper center in myoglobin. Biochemistry 44:1210–1214. 55. Decatur SM, et al. (1996) Trans effects in nitric oxide binding to myoglobin cavity mutant H93G. Biochemistry 35:4939–4944. 56. Pervitsky D, Immoos C, van der Veer W, Farmer PJ (2007) Photolysis of the HNO adduct of myoglobin: Transient generation of the aminoxyl radical. J Am Chem Soc 129:9590–9591. 57. Berto TC, Praneeth VKK, Goodrich LE, Lehnert N (2009) Iron-porphyrin NO complexes with covalently attached N-donor ligands: Formation of a stable six-coordinate species in solution. J Am Chem Soc 131:17116–17126. 58. Blomberg LM, Blomberg MRA, Siegbahn PEM (2006) Reduction of nitric oxide in bacterial nitric oxide reductase-a theoretical model study. Biochim Biophys Acta 1757:240–252. 59. Varadarajan R, Zewert TE, Gray HB, Boxer SG (1989) Effects of buried ionizable amino acids on the reduction potential of recombinant myoglobin. Science 243:69–72. 60. Grönberg KLC, et al. (1999) A low-redox potential heme in the dinuclear center of bacterial nitric oxide reductase: Implications for the evolution of energy-conserving heme-copper oxidases. Biochemistry 38:13780–13786. 61. Zhao X, Yeung N, Russell BS, Garner DK, Lu Y (2006) Catalytic reduction of NO to N2O by a designed heme copper center in myoglobin: Implications for the role of metal ions. J Am Chem Soc 128:6766–6767. 62. Pinakoulaki E, Stavrakis S, Urbani A, Varotsis C (2002) Resonance raman detection of a ferrous five-coordinate nitrosylheme b3 complex in cytochrome cbb3 oxidase from pseudomonas stutzeri. J Am Chem Soc 124:9378–9379. 63. Morrison M, Horie S (1965) Determination of heme a concentration in cytochrome preparations by hemochromogen method. Anal Biochem 12:77–82. 64. Taboy CH, Bonaventura C, Crumbliss AL (2002) Anaerobic oxidations of myoglobin and hemoglobin by spectroelectrochemistry. Method Enzymol 353:187–209. 8586 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1000526107 Lin et al.
3M39
The roles of glutamates and metal ions in a rationally designed nitric oxide reductase based on myoglobin: Fe(II)-I107E FeBMb (Fe(II) binding to FeB site)
Roles of glutamates and metal ions in a rationally designed nitric oxide reductase based on myoglobin Ying-Wu Lina,2, Natasha Yeunga, Yi-Gui Gaob, Kyle D. Minerc, Shiliang Tiana, Howard Robinsond, and Yi Lua,c,1 aDepartment of Chemistry, bGeorge L. Clark X-Ray Facility and 3M Materials Laboratory, andcDepartment of Biochemistry, University of Illinois at Urbana-Champaign, Urbana, IL 61801; and dDepartment of Biology, Brookhaven National Laboratory, Upton, NY 11973 Edited by Harry B. Gray, California Institute of Technology, Pasadena, CA, and approved April 1, 2010 (received for review January 14, 2010) A structural and functional model of bacterial nitric oxide reductase (NOR) has been designed by introducing two glutamates (Glu) and three histidines (His) in sperm whale myoglobin. X-ray structural data indicate that the three His and one Glu (V68E) residues bind iron, mimicking the putative FeB site in NOR, while the second Glu (I107E) interacts with a water molecule and forms a hydrogen bond- ing network in the designed protein. Unlike the first Glu (V68E), which lowered the heme reduction potential by ∼110 mV, the second Glu has little effect on the heme potential, suggesting that thenegativelychargedGluhasadifferentrolein redoxtuning.More importantly, introducing the second Glu resulted in a ∼100% in- crease in NOR activity, suggesting the importance of a hydrogen bonding network in facilitating proton delivery during NOR reactiv- ity. In addition, EPR and X-ray structural studies indicate that the designed protein binds iron, copper, or zinc in the FeB site, each with different effects on the structures and NOR activities, suggesting that both redox activity and an intermediate five-coordinate heme-NO species are important for high NOR activity. The designed protein offers an excellent model for NOR and demonstrates the power of using designed proteins as a simpler and more well- defined system to address important chemical and biological issues. biomimetic models ∣heme-copper oxidase ∣metalloprotein ∣protein design ∣ protein engineering R ational design of proteins that mimic both structure and func- tion of more complex native enzymes has been a long sought- after goal, as the process is an ultimate test of our knowledge and an excellent means to develop advanced biocatalysts (1–3). Although designed proteins that model the structure of native enzymes have been known for a while (4–10), successful designs of proteins that mimic both the structure and function of native enzymes have been reported only recently (11–16). While being able to design such functional proteins is laudable, the impact of such an achievement would be greater if the designed proteins can be used to address fundamental issues in chemistry and biol- ogy that are difficult to tackle by other methods. One primary example is the roles of conserved glutamates and metal ions in bacterial nitric oxide reductase (NOR) (17–19). NO is critical for all life (20). Bacterial denitrification is a cru- cial part of the nitrogen cycle in nature that involves a four-step, five-electron reduction of nitrate (NO3 −) to dinitrogen (N2) (17, 19). Bacterial NOR is a membrane-bound protein that catalyzes one step of this process, namely, the two-electron reduc- tion of NO to N2O (17, 19). With no crystal or solution structure available for bacterial NOR to date, sequence alignments and homology modeling (21, 22) have indicated that NOR is structu- rally homologous to the largest subunit (subunit I) of heme- copper oxidases (HCOs) (23), enzymes that catalyze reduction of O2 to water. The active sites of both NOR and HCO contain a proximal histidine-coordinated heme and a distal three histi- dine-coordinated metal center. However, the metal center in HCOs is occupied by a copper (called CuB), whereas a nonheme iron is present in NOR (called FeB) (23, 24). In addition, two conserved glutamates, shown by modeling to be close to the FeB site (21, 22), are found to be essential for NOR activity (24, 25). Some members of HCOs such as cytochrome cbb3 oxidase display NOR activity (26–28), although the activity is ∼50-fold lower than native NOR (26). Therefore, it is important to elucidate the structural features, specifically the roles of the conserved glutamates close to the FeB site and metal ions (copper vs. iron), responsible for the reduction of NO to N2O. To address these issues, biochemical and biophysical studies of native NOR and its variants have been carried out (24, 25, 29–37). For example, Richardson and coworkers investigated the effects of amino acid substitutions of the five conserved glutamates (E122 and E125 presumed to face the periplasm and E198, E202, and E267 located in the interior of the membrane, close to the catalytic site) in the catalytic subunit of Paracoccus deni- trificans, NorB. The E122A, E125A, E198A, and E267A variants were inactive, indicating that these four glutamates are crucial for NOR activity (24, 25, 32, 33). On the other hand, Reimann et al. constructed a 3D model of NorB using homology modeling with the structures of HCOs as templates and suggested a plausible pathway consisting of these conserved glutamates for proton delivery (22). Despite these successes, the roles of the conserved glutamates and metal ions still remain to be fully elucidated, partly because of the difficulty in obtaining native NOR in high yield and the lack of a 3D structure. Even if these problems are resolved, it is still difficult to replace iron in the native FeB site with other metal ions, and spectroscopic studies of native NOR are often complicated by the presence of other metal cofactors (e.g., low-spin heme). To overcome these limitations, a number of synthetic models of NOR using small organic molecules as ligands, have been made in which the nonheme FeB site can be replaced by a copper ion (17, 38–45). In addition, since these model systems lack addi- tional metal-binding sites, spectroscopic studies are often simpli- fied. Therefore, studies of these synthetic models have offered many insights. For example, Collman et al. showed that a fully reduced heme/nonheme FeB compound can react with two equivalents of NO leading to the formation of one equivalent of N2O and a bis-ferric product (41). On the other hand, Karlin and coworkers showed that a small heme/Cu complex can effi- ciently lead to reductive coupling of NO to N2O (43). However, it is also difficult to obtain the synthetic models in high yield due to the multiple steps required in chemical synthesis. Because of this limitation, no synthetic NOR model containing the two key conserved Glu residues (E198 and E267 in NOR) has been Author contributions: Y.-W.L., N.Y., and Y.L. designed research; Y.-W.L., N.Y., Y.-G.G., K.D.M., and S.T. performed research; Y.-W.L., N.Y., Y.-G.G., K.D.M., H.R., and Y.L. analyzed data; and Y.-W.L. and Y.L. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Data deposition: The atomic coordinates and structure factors have been deposited in the Protein Data Bank, www.pdb.org (PDB ID codes 3M38, 3M39, 3M3A, and 3M3B). 1To whom correspondence should be addressed. E-mail: yi-lu@illinois.edu. 2Present address: School of Chemistry and Chemical Engineering, University of South China, Hengyang 421001, China. This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1000526107/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1000526107 PNAS ∣May 11, 2010 ∣vol. 107 ∣no. 19 ∣8581–8586 BIOCHEMISTRY CHEMISTRY reported. It is also difficult to substitute different metal ions in the same metal-binding site without perturbing the site geometry and distances to the heme iron, as most ligands are not as rigid as those in native enzymes and different metal ions have different geometric and ligand donor set preferences. We have recently designed a structural and functional protein model of bacterial NOR by engineering three histidines and one glutamate into the distal pocket of sperm whale myoglobin (swMb, L29H, F43H, H64, and V68E, named FeBMb) (14). Like synthetic models, this “bottom-up” approach complements the “top-down”approachofthestudyofnativeNORinthatitprovides insights into whether certain “necessary” structural elements are enough to impart enzyme function. Thanks in part to recent advances in computational, molecular, and structural biology, the designed myoglobin protein model is much easier to synthesize and to crystallize than either native NOR or synthetic models. Since myoglobin has often been used for the development and calibration of numerous spectroscopic techniques (46–48), it is an ideal choice for spectroscopic studies. More importantly, the rigid protein network allows precise placement of either glutamate or metal ions in myoglobin to address their roles in NOR activity. Toward this goal, we have demonstrated that both the histidines and one of the glutamates are essential for iron binding and NO reduction activity (14). However, the role of the second Glu close to the FeB site and the role of different metal ions in the FeB site have not been addressed. To address these important issues and to design even closer protein models of NOR, we introduced herein the second Glu to the second coordination sphere of the FeB site by mutating an Ile to a Glu (named I107E FeBMb). We show that the second Glu results in a ∼100% increase in NOR activity through hydro- gen bonding interactions and that the two glutamates have dramatically different effects on the heme reduction potential. Additionally, by comparing the EPR, electrochemistry, X-ray structures, and NOR activity of iron, copper, and zinc derivatives of the designed protein, we have obtained deeper insights into the roles of metal ions in NOR. Results Structure and Function of Fe(II)-I107E FeBMb. The X-ray crystal struc- tures of heme-containing I107E FeBMb without metal ion in the FeB site and with Fe2þ in the FeB site are solved at 1.42-Å and 1.65-Å resolution, respectively (Fig. 1 A and B and Table S1). In the absence of metal ions in the FeB site, the structure shows a water molecule in the FeB site, which forms hydrogen bonds with NE2 atoms of all three His residues, both OE1 and OE2 atoms of E68, and the OE2 atom of E107 (Fig. 1A). Upon binding Fe2þ, the Fe(II)-I107E FeBMb structure shows that Fe2þ is coordinated by three His, the OE2 atom of E68, and one water molecule. Notably, a water molecule bridges Fe2þ in the FeB site and the second glutamate (E107) with a distance of 2.32 Å to the OE2 atom of E107 (Fig. 1B). To probe the conformational changes of introducing the second Glu (E107), we performed a structural alignment of Fe (II)-I107E FeBMb and the previously reported FeðIIÞ-FeBMb (14). The comparison, shown in Fig. 2, indicates that both the polypeptide chain and the active site overlap well with each other. In addition, the two nonheme irons are located at similar posi- tions with a 0.36-Å separation from each other. In contrast, E68 underwent a significant conformational rearrangement in the presence of E107. These observations suggest that the active site of FeBMb can be tuned by the formation of an extended hydrogen bonding network, resulting from the introduction of a second glutamate residue. The binding of Fe2þ to deoxy I107E FeBMb was further mon- itored by EPR (Fig. 3A). Since deoxy myoglobin contains Fe(II) heme that exhibits no EPR signals in X-band EPR (14), we added blue copper Cu(II)-azurin (49), a redox partner of native NOR (19), to oxidize both the reduced heme and nonheme irons in Fe (II)-I107E FeBMb to EPR-active Fe(III). Upon addition of Cu (II)-azurin, the oxidation of deoxy I107E FeBMb resulted in EPR signals at g ¼ 6.12 and 5.56, typical of a high-spin heme- Fe(III). Upon addition of Fe2þ, however, a decrease of the heme-Fe(III) EPR signals was observed, indicating that the Fe2þ, when bound to the FeB site and oxidized by Cu(II)-azurin, is spin-coupled to heme-Fe(III). Such a spin coupling mimics that in NOR (35, 50–53), suggesting that I107E FeBMb models NOR closely, at least in this respect. To probe the role of the second Glu (E107) in NO reduction activity, we measured the yield of N2O production by Fe(II)- I107E FeBMb with excess NO under one turnover conditions. We monitored N2O formation in the headspace of the solution using GC/MS and compared this result to that of FeðIIÞ-FeBMb, which lacks the second Glu (Fig. 4). Remarkably, Fe(II)-I107E FeBMb displays higher activity than FeðIIÞ-FeBMb. After ∼20 hr, ∼24% N2O was produced by Fe(II)-I107E FeBMb, in H29 A H29 B H64 E68 E107 3.04 2.81 3.02 2.20 2.24 2.24 2.12 2.21 H64 E68 E107 3.03 H43 2.15 3.41 3.16 2.62 2.32 H43 2.26 H93 H93 H29 C H29 D 2.04 2.03 H64 E68 E107 2.09 2.10 2.91 2.21 H64 E68 E107 2.26 2.29 2.10 2.18 2.10 4.47 H43 2.07 3.04 H43 2.68 H93 H93 Fig. 1. Crystal structures of I107E FeBMb (A) (PDB ID code 3M38), Fe(II)-I107E FeBMb (B) (PDB ID code 3M39), Cu(II)-I107E FeBMb (C) (PDB ID code 3M3A), and Zn(II)-I107E FeBMb (D) (PDB ID code 3M3B). Water molecules, Fe(II), Cu(II), and Zn(II) are represented by red, green, orange, and gray spheres, respectively. Fig. 2. Overlay of Fe(II)-I107E FeBMb (cyan) (PDB ID code 3M39) with FeðIIÞ-FeBMb (orange) (PDB ID code 3K9Z). 8582 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1000526107 Lin et al. contrast to ∼10% yield for FeðIIÞ-FeBMb, strongly indicating that the second Glu plays an important role in NO reduction, likely facilitating proton uptake during NO reduction. Other Metal Ions Binding to I107E FeBMb. To find out if the resting state of the protein, i.e., oxidized or met I107E FeBMb, can bind other metal ions, Cu2þ or Zn2þ was titrated into met I107E FeBMb and monitored by EPR spectroscopy (Fig. 3 B and C). In the absence of metal ions, met I107E FeBMb exhibited high-spin heme signals at g ¼ 6.03, 5.08, and 1.98 (Fig. 3B, black line). Upon addition of 2 eq of Cu2þ, the signals at g ¼ 6.03 and 5.08 decreased and a broad peak around g ¼ 2.95 increased, probably due to spin coupling between heme-Fe(III) and Cu2þ in the FeB site. In contrast, addition of Zn2þ, a metal ion with no unpaired electrons [i.e., incapable of spin coupling to heme- Fe(III)], produced an increase in the high-spin heme signals at g ¼ 5.88 and 5.60 (Fig. 3C), indicating that the interaction be- tween E68 and heme iron was weakened after metal binding. The X-ray crystal structures of I107E FeBMb with Cu2þ or Zn2þ in the FeB site were solved at 1.37-Å and 1.60-Å resolution, respectively (Fig. 1 C and D and Table S1). Compared to Fe(II)- I107E FeBMb (Fig. 1B), a similar binding site was observed for Cu(II)-I107E FeBMb (Fig. 1C), where H29, H43, and H64 coordinate to Cu2þ with distances of 2.09, 2.10, and 2.04 Å, respectively, slightly shorter than the corresponding distances in the Fe2þ structure. In comparison to Fe(II)-I107E FeBMb, the water bridging the Cu2þ and the second Glu (E107) is shifted toward Cu2þ in the FeB site (2.03 Å) with respect to E107 (3.04 Å). Interestingly, this bridging water molecule was not observed in Zn(II)-I107E FeBMb (Fig. 1D), but the two O atoms of E68 coordinate to Zn2þ with similar distances (2.26 Å for OE1 and 2.29 Å for OE2). The longer distance between OE1 of E68 B 6.12 I107E FeBMb + Azurin I107E Fe Mb + 0 5 eq Fe 2+ + Azurin A 5.56 I107E FeBMb + 0.5 eq Fe + Azurin I107E FeBMb + 1.0 eq Fe 2+ + Azurin I107E FeBMb + 2.0 eq Fe 2+ + Azurin 1.98 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5.08 5.38 Magnetic Field (Gauss) 5.88 I107E FeBMb I107E FeBMb + 0.5 eq Zn 2+ I107E Fe Mb 1 0 eq Zn 2+ C 6.03 I107E FeBMb I107E FeBMb + 0.5 eq Cu 2+ I107E Fe Mb + 1 0 eq Cu 2+ I107E FeBMb + 1.0 eq Zn I107E FeBMb + 2.0 eq Zn 2+ I107E FeBMb + 1.0 eq Cu I107E FeBMb + 2.0 eq Cu 2+ 1.98 ~2.95 1.98 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5.60 Magnetic Field (Gauss) 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5.08 Magnetic Field (Gauss) Fig. 3. EPR spectra of deoxy I107E FeBMb (0.5 mM protein in 50 mM Bis-Tris, pH 7.0) with increasing concentrations of Fe2þ in the presence of wild-type Cu(II)-azurin (A), and oxidized I107E FeBMb with Cu2þ (B) or Zn2þ (C). Spectra were collected at 4 K, 5 mW power, and 9.05 GHz. 30 25 Fe(II)-I107E FeBMb Fe(II)-FeBMb 15 20 10 %) oduction (% N2O pro 0 5 0 4 8 12 16 20 Incubation time (hr) Fig. 4. Time-dependent N2O production by Fe(II)-I107E FeBMb (▴) and FeðIIÞ-FeBMb (●) with ∼50 eq. NO under single turnover conditions. The yield was determined by a comparison of the ratio of NO∶N2O peaks from the GC/MS chromatograms. Lin et al. PNAS ∣ May 11, 2010 ∣ vol. 107 ∣ no. 19 ∣ 8583 BIOCHEMISTRY CHEMISTRY and heme iron in the Zn-bound structure (2.68 Å) in comparison to the Cu- and Fe-bound structures, is likely the result of a weaker interaction, which is also supported by an observed increase of the high-spin heme signals in the EPR spectra upon Zn2þ binding (Fig. 3C). These results suggest I107E FeBMb is capable of incor- porating different metal ions into its designed FeB site, offering an excellent opportunity to compare the role of these metal ions in the same protein scaffold. Effect of Glutamates and Metal Ions on the Redox Potential of I107E FeBMb. Since EPR and X-ray structural studies indicate metal binding to I107E FeBMb, we used spectroelectrochemistry to measure the effects of glutamates and metal ions on the heme reduction potential. When there is no metal ion in the FeB site, the I107E FeBMb displays a reduction potential of −134  3 mV vs. the normal hydrogen electrode (NHE) (Fig. S1A), similar to that of FeBMb (−158  4 mV) without the I107E mutation (14). In the presence of Cu2þ, I107E FeBMb has a reduction potential (−137  2 mV) (Fig. S1B) almost identical to that of the same protein in the absence of metal ions in the FeB site, indicating that copper binding to the FeB site has little effect on the reduction potential of the heme iron. This observation is similar to that observed for Cu2þ binding to CuBMb (CuðIIÞ-CuBMb, 80 mV vs. CuBMb, 77 mV) (54). On the other hand, the presence of Fe2þ and Zn2þ increased the reduction potential of I107E FeBMb from −134  3 mV to −64  3 mV vs. NHE (Fig. S1C) and −105  2 mV vs. NHE (Fig. S1D), respectively. The different effects of Cu2þ, Fe2þ, and Zn2þ on the reduction potential of I107E-FeBMb indicate that these metal ions in the FeB site may play different roles through different coordination properties. NOR Activity of I107E FeBMb in the Presence of Different Metal Ions. The NO reduction activity of I107E FeBMb in the presence of Fe2þ, Cuþ, or Zn2þ was monitored by GC/MS under single turn- over conditions. When Fe(II)-I107E FeBMb was exposed to excess NO, N2O could be observed to form with increased yield over time (Fig. S2). Similarly, N2O formation was observed for Cu(I)-I107E FeBMb, indicating that Fe or Cu binding to the FeB site results in comparable NOR activities. It should be noted that because of the high solubility of N2O (∼25 mM in water at room temperature), GC/MS cannot be used to quantify the rates of NO reduction under these conditions. In contrast, no N2O formation was observed with redox inactive Zn2þ, which demon- strates that redox active Fe2þ or Cuþ in the FeB site plays a crucial role in NO reduction. To gain deeper insight into the process of NO reduction, EPR studies were further performed to monitor the initial process of NO reduction. In the absence of metal ions, the EPR spectrum of ferrous I107E FeBMb-NO shows hyperfine splitting resulting from bound NO and the proximal histidine, indicating the forma- tion of a six-coordinate ferrous heme-NO species (Fig. 5, top line). After incubation of Fe(II)-I107E FeBMb with excess NO, a distinct three-line hyperfine structure appears at 15 min (Fig. 5A), suggesting the formation of a five-coordinate ferrous heme-NO species as a result of cleavage of the proximal His-Fe heme bond (55). A three-line hyperfine structure was also observed for Cuþ and Zn2þ, except that the signal intensity is low- er with Cu(I)-I107E FeBMb-NO (Fig. 5B) and more pronounced in Zn(II)-I107E FeBMb-NO (Fig. 5C). The lower intensity of the three-line hyperfine structure for Cu(I)-I107E FeBMb-NO suggests the major species formed is a six-coordinate ferric heme-NO complex, which is EPR silent (41). These differences further suggest that the metal ion in the FeB site plays a key role in formation of the intermediates, thereby tuning NOR activity. Discussion Using Rationally Designed Proteins to Address Important Issues in Chemistry and Biology. Important issues such as the roles of the conserved glutamates and nonheme FeB in NOR have been previously addressed using biochemical and biophysical studies or biomimetic modeling (24, 25, 27–37, 45, 56, 57). As a comple- mentary approach, rational protein design, using small, easy-to- produce and well-characterized proteins such as myoglobin, offers a powerful method with which to gain insights into more complex native enzymes such as NOR (14). Similar to synthetic models (41, 43), the metal ion at the putative FeB site in the protein model can be substituted freely. Better yet, Glu residues can be placed at precise locations in the protein, including the secondary coordination sphere, due to its rigid network. By care- fully choosing a suitable protein template, rational protein design could be generally applied to address other important issues in chemistry and biology. The Roles of Glutamates. Although two conserved glutamates (E198 and E267) are known to be crucial for NOR activity (24, 25), their roles are not well defined (18, 19). In a previous study (14), we demonstrated that one Glu, E68, is important for both iron binding and NOR activity of FeBMb. The crystal struc- tures of both FeðIIÞ-FeBMb and Fe(II)-I107E FeBMb show that one O atom of E68 directly coordinates to FeB (Fig. 2). In syn- thetic models of NOR, it has also been found that the presence of a glutamic acid mimic significantly increases the stability of iron binding to the FeB site (40). Furthermore, a theoretical study by Blomberg et al. (58) showed that a model with an FeB coordi- nated by three histidines, one glutamate, and one water molecule provides an energetically feasible reaction mechanism of NO reduction. However, the structural model of NOR constructed recently by Reimann et al. (22) shows that the closest conserved Glu (E267) still has its carboxylate O atom 7 Å away from FeB, which suggests that Glu may not bind to FeB in native NOR. One interesting finding from our study is that the Glu (E68) under- went a significant conformational rearrangement in the presence of another Glu (E107) (Fig. 2). Therefore, the FeBMb provides a viable model of NOR that is consistent with Blomberg’s model, but cannot rule out Reimann’s model due to possible conforma- tion changes. While the role of the first Glu is still uncertain until a 3D struc- ture of NOR in its active form is available, the role of the second Glu is even less defined. We address this question by introducing a second Glu (E107) to FeBMb. The crystal structures shown in Fig. 1 indicate that E107 interacts with a water molecule and forms a hydrogen bonding network in both Fe(II)-I107E FeBMb and Cu(II)-I107E FeBMb. Interestingly, although a similar water molecule was observed in the active site of FeðIIÞ- FeBMb (Fig. 2), activity assay data indicate that the presence of E107 in Fe(II)-I107E FeBMb increases NOR activity by ∼100% C 5 min N t l A 5 min B 2+ 5 min 1 min No metal 1 min 5 min F 2+ No metal 1 i 5 min + No metal 2+ Zn 2+ 5 min Fe 2+ 5 min 1 min Fe C + 5 min 1 min Cu + Z 2 Zn 2+ 15 min F 2+ Fe 15 min 5 min C + Cu + 15 min 5 min Zn 2+ 15 min Fe2+ 5 Cu + 15 min 3000 3100 3200 3300 3400 Magnetic Field (Gauss) 3000 3100 3200 3300 3400 Magnetic Field (Gauss) 3000 3100 3200 3300 3400 Magnetic Field (Gauss) Fig. 5. EPR spectra of deoxy I107E FeBMb (0.5 mM) with no metal bound in the presence of NO after 5 min (top line), with 2 eq Fe2þ (A), Cuþ (B), or Zn2þ (C) incubated with excess NO (∼200 eq) for 1, 5, and 15 min. Spectra were collected in 50 mM Bis-Tris pH 7.0 at 30 K, 0.2 mW power, and 9.05 GHz. 8584 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1000526107 Lin et al. (Fig. 4), suggesting that the second Glu may potentionally play a role in providing one of the protons during reduction of NO to N2O. Although free Glu outside a protein has a pKa ∼4.3, the stu- dies of native NOR showed that the pKa of its Glu close to the active site has a value of ∼6.6 (22, 25, 33). The hydrogen binding network in our protein models may contribute to the fine-tuning of the Glu pKa to be more neutral, similar to that in NOR. More- over, it is interesting that Cu(I)-I107E FeBMb also shows NOR activity, which provides an interesting protein model of HCOs with NOR function (26–28), even though Glu residues are not conserved in native HCOs. Additionally, spectroelectrochemical studies showed that the reduction potential of I107E FeBMb with no metal ion in the FeB site is similar to that of FeBMb, but much lower than that of CuBMb (77 mV) (54), which contains the same three His, but no Glu in the metal-binding site above the heme. Since both FeBMb and I107E FeBMb contain the V68E mutation that has been shown to decrease the heme reduction potential of native myoglobin from 59 mV to −137 mV (59), it is likely that the introduction of a negatively charged Glu close to the heme group is what is responsible for the dramatically lower heme redox potential. A conserved glutamate, predicted to be located near the catalytic heme b3 in NOR, was proposed to be responsible for a ∼260-mV decrease in reduction potential (60 mV) in comparison to the other two heme centers, heme b (345 mV) and heme c (310 mV) (60). Our FeBMb and I107E FeBMb mod- els mimic this feature of NOR. Notably, although introduction of the first Glu (E68) lowered the heme potential by ∼110 mV (14), introduction of the second Glu via the I107E mutation did not result in a significant difference in the heme reduction potential, suggesting that the effect of the two conserved Glu residues in NOR on heme reduction potential is not additive, with the effects highly dependent on the location of the Glu. The Roles of Metal Ions. The roles of metal ions in NOR are an- other important question as iron is found in the native FeB site and HCO employs copper at the corresponding CuB site. With different metal ions in the FeB site, the crystal structures clearly show the heme and nonheme dinuclear center existing in differ- ent local environments (Fig. 1). Although a similar hydrogen bond network is formed in both Fe(II)-I107E FeBMb and Cu (II)-I107E FeBMb, the conformation of E68 and E107 with re- spect to the nonheme metal center and heme iron is different from each other. Moreover, the coordination geometry differs significantly with Zn2þ in the FeB site. A hydrogen bond is absent from the Zn crystal structure, but both the O atoms of E68 act as metal-binding ligands. These observations demonstrate that the identity of the metal ion in the FeB site can tune the active site through their interactions with the His and Glu ligands, resulting in formation of different coordination geometries with different hydrogen bonds. In addition to structural fine-tuning, the metal ion at the FeB site can also tune the heme iron reduction potential in I107E FeBMb. Spectroelectrochemical studies showed that the binding of Fe2þ or Zn2þ results in an increase in the heme reduction potential by ∼70 mV and ∼30 mV, respectively (Fig. S1). In the case of Cu(II)-I107E FeBMb, the crystal structure shows that OE1 of E68 is closer to the heme iron (2.07 Å) (Fig. 1C) than its metal-free form (2.15 Å) (Fig. 1A). The stronger interaction from the negatively charged E68 could offset the effect of positively charged Cu2þ binding, resulting in similar reduction potentials observed for Cu(II)-I107E FeBMb and I107E FeBMb. In a previous study (61), EPR data showed that during NO reduction, the binding of Cuþ to the CuB site of CuBMb can weaken the proximal heme Fe-His bond, while complete cleavage of the heme Fe-His bond occurred when Zn2þ was bound to CuBMb-NO. In this study, we observed that a five-coordinate heme-NO species was formed with Fe2þ, Cuþ, or Zn2þ bound to the FeB site of I107E FeBMb (Fig. 5). Significantly, a five- coordinate heme-NO species has also been observed for both NOR (30, 31, 35) and the member of the HCO family with the highest NO reduction activity, cytochrome cbb3 oxidases (26, 62). However, this species was not observed for FeðIIÞ-FeBMb-NO and CuðIÞ-FeBMb-NO, which lack the second Glu (E107). In both these cases, the proximal heme Fe-His bond was only weakened, as indicated by a decrease of the nine-line hyperfine splitting signals in the EPR spectra (Fig. S3). These observations suggest that formation of a five-coordinate heme-NO species may play an important role in NOR reactivity. Conclusions We have successfully designed a structural and functional model of NOR, by introducing a second glutamate in the vicinity of the FeB site, named I107E FeBMb. This protein model mimics native NOR more closely by bearing the structural feature of three his- tidines and two glutamates in the FeB site, as predicted for native NOR. We have demonstrated that the two glutamates can play different roles in NO reduction activity; namely, one acts as a li- gand to FeB (E68), and the other acts as a proton transfer group (E107). Furthermore, by substituting different metal ions into the nonheme metal site, we have demonstrated that FeB plays crucial roles in fine-tuning the active site by donating electrons and by mediating the formation of a five-coordinate heme-NO inter- mediate during NO reduction. In the absence of a crystal struc- ture for native NOR, this study offers an ideal protein model and provides valuable structural as well as mechanistic information for native NOR. Materials and Methods Protein Preparation. I107E FeBMb (swMb L29H/F43H/V68E/I107E) was con- structed, expressed, and purified using the procedure described previously (14). The purity and identity were confirmed by SDS-PAGE and electrospray ionized MS: observed: 17; 392  1 Da; calculated: 17,391 Da. EPR Spectroscopy. EPR spectra were recorded on a Bruker ESP 300 equipped with an Oxford liquid helium cryostat and an ITC4 temperature controller. The samples of met I107E FeBMb, Cu(II)-, or Zn(II)-I107E FeBMb were pre- pared as described previously (14). The samples of NO-bound deoxy I107E FeBMb, Cu(I)-, Fe(II)-, or Zn(II)-I107E FeBMb were prepared by inject- ing 1 mL of purified NO gas into the EPR tube containing 300 μL of protein (0.5 mM, 10% glycerol, in 50 mM Bis-Tris pH 7.0). The samples were then flash frozen in liquid N2 after incubation for 1, 5, or 15 min. The molar extinction coefficient of the Soret band of I107E FeBMb at 406 nm (175 mM−1 · cm−1), calculated using the standard hemochromagen method (63), was used to determine protein concentration. The metal sources of Cu(I), Cu(II), Zn(II), and Fe(II) were ½ðCH3CNÞ4CuPF6, CuSO4, ZnSO4 · 7H2O, and FeCl2, respectively. Spectroelectrochemical Measurements. Protein reduction potentials were measured using an optically transparent thin layer electrode as previously described (64). The potential of the working electrode was applied in the positive direction for deoxy I107E FeBMb with Fe2þ and in the negative direction for metal free and with Cu2þ or Zn2þ. Other procedures are the same as described previously (54). X-Ray Crystallographic Studies. Fe(II)-I107E FeBMb was crystallized anaerobi- cally in a glove box at room temperature using the conditions described for FeðIIÞ-FeBMb (14). I107E FeBMb, Cu(II)-, or Zn(II)-I107E FeBMb were crystallized aerobically. Diffraction-quality crystals were soaked in a cryoprotectant solution of 30% PEG 400 and flash frozen in liquid nitrogen. Diffraction data were collected at the Brookhaven National Lab Synchrotron Light Source X12C beamline. The crystal structure was solved using the same method as for FeðIIÞ-FeBMb (14). NOR Activity Assay. NO reduction was monitored by GC/MS. The protein was reduced to the deoxy form by excess dithionite that was removed with a size-exclusion column (PD-10). Then 2 eq metal, Cu(I), Fe(II), or Zn(II), was added to the protein solution (0.6 mM, 3 mL in 50 mM Bis-Tris buffer, pH 7.0). The samples were prepared anaerobically in a glove box. Purified NO Lin et al. PNAS ∣ May 11, 2010 ∣ vol. 107 ∣ no. 19 ∣ 8585 BIOCHEMISTRY CHEMISTRY gas was injected into the head space of the reaction flask with the molar ratio of NO∶protein ≈50∶1. Other procedures are the same as described previously (14, 61). ACKNOWLEDGMENTS. We thank Dr. Mark J. Nilges for help with EPR analysis, and Furong Sun and Beth D. Eves for aiding in GC/MS data collection. This work was supported by NIH Grant GM062211. 1. Lu Y, Yeung N, Sieracki N, Marshall NM (2009) Design of functional metalloproteins. Nature 460:855–862. 2. Nanda V, Koder RL (2010) Designing artificial enzymes by intuition and computation. Nat Chem 2:15–24. 3. Gray HB (2003) Biological inorganic chemistry at the beginning of the 21st century. Proc Natl Acad Sci USA 100:3563–3568. 4. Regan L, DeGrado WF (1988) Characterization of a helical protein designed from first principles. Science 241:976–978. 5. Hecht MH, Richardson JS, Richardson DC, Ogden RC (1990) De novo design, expression, and characterization of Felix: A four-helix bundle protein of native-like sequence. Science 249:884–891. 6. Robertson DE, et al. (1994) Design and synthesis of multi-heme proteins. Nature 368:425–432. 7. Lu Y, Valentine JS (1997) Engineering metal-binding sites in proteins. Curr Opin Struct Biol 7:495–500. 8. Reedy CJ, Gibney BR (2004) Heme protein assemblies. Chem Rev 104:617–649. 9. Watanabe Y, Hayashi T (2005) Functionalization of myoglobin. Prog Inorg Chem 54:449–493. 10. Ghosh D, Pecoraro VL (2005) Probing metal-protein interactions using a de novo design approach. Curr Opin Chem Biol 9:97–103. 11. Kaplan J, DeGrado WF (2004) De novo design of catalytic proteins. Proc Natl Acad Sci USA 101:11566–11570. 12. Kang SG, Saven JG (2007) Computational protein design: Structure, function and combinatorial diversity. Curr Opin Chem Biol 11:329–334. 13. Jiang L, et al. (2008) De novo computational design of retro-aldol enzymes. Science 319:1387–1391. 14. Yeung N, et al. (2009) Rational design of a structural and functional nitric oxide reductase. Nature 462:1079–1082. 15. Koder RL, et al. (2009) Design and engineering of an O2 transport protein. Nature 458:305–309. 16. Heinisch T, Ward TR (2010) Design strategies for the creation of artifical metalloen- zymes. Curr Opin Chem Biol 14:184–199. 17. Wasser IM, de Vries S, Moënne-Loccoz P, Schröder I, Karlin KD (2002) Nitric oxide in biological denitrification: Fe/Cu metalloenzyme and metal complex NOx redox chemistry. Chem Rev 102:1201–1234. 18. Yeung N, Lu Y (2008) One heme, diverse functions: Using biosynthetic myoglobin models to gain insights into heme-copper oxidases and nitric oxide reductases. Chem Biodivers 5:1437–1454. 19. Watmough NJ, Field SJ, Hughes RJL, Richardson DJ (2009) The bacterial respiratory nitric oxide reductase. Biochem Soc Trans 37:392–399. 20. Cary SPL, Winger JA, Derbyshire ER, Marletta MA (2006) Nitric oxide signaling: No longer simply on or off. Trends Biochem Sci 31:231–239. 21. Sakurai N, Sakurai T (1998) Genomic DNA cloning of the region encoding nitric oxide reductase in paracoccus halodenitrificans and a structure model relevant to cytochrome oxidase. Biochem Biophys Res Commun 243:400–406. 22. Reimann J, Flock U, Lepp H, Honigmann A, Ädelroth P (2007) A pathway for protons in nitric oxide reductase from paracoccus denitrificans. Biochim Biophys Acta 1767:362–373. 23. Pereira MM, Sousa FL, Verissimo AF, Teixeira M (2008) Looking for the minimum common denominator in haem-copper oxygen reductases: Towards a unified catalytic mechanism. Biochim Biophys Acta 1777:929–934. 24. Butland G, Spiro S, Watmough NJ, Richardson DJ (2001) Two conserved glutamates in the bacterial nitric oxide reductase are essential for activity but not assembly of the enzyme. J Bacteriol 183:189–199. 25. Flock U, Lachmann P, Reimann J, Watmough NJ, Äedelroth P (2009) Exploring the terminal region of the proton pathway in the bacterial nitric oxide reductase. J Inorg Biochem 103:845–850. 26. Forte E, et al. (2001) The cytochrome cbb3 from pseudomonas stutzeri displays nitric oxide reductase activity. Eur J Biochem 268:6486–6490. 27. Huang Y, Reimann J, Lepp H, Drici N, Ädelroth P (2008) Vectorial proton transfer coupled to reduction of O2 and NO by a heme-copper oxidase. Proc Natl Acad Sci USA 105:20257–20262. 28. Hayashi T, et al. (2009) Accommodation of two diatomic molecules in cytochrome bo3: Insights into NO reductase activity in terminal oxidases. Biochemistry 48:883–890. 29. Moënne-Loccoz P (2007) Spectroscopic characterization of heme iron-nitrosyl species and their role in NO reductase mechanisms in diiron proteins. Nat Prod Rep 24:610–620. 30. Moënne-Loccoz P, de Vries S (1998) Structural characterization of the catalytic high-spin heme b of nitric oxide reductase: A resonance raman study. J Am Chem Soc 120:5147–5152. 31. Sakurai T, Sakurai N, Matsumoto H, Hirota S, Yamauchi O (1998) Roles of four iron centers in paracoccus halodenitrificans nitric oxide reductase. Biochem Biophys Res Commun 251:248–251. 32. Thorndycroft FH, Butland G, Richardson DJ, Watmough NJ (2007) A new assay for nitric oxide reductase reveals two conserved glutamate residues form the entrance to a proton-conducting channel in the bacterial enzyme. Biochem J 401:111–119. 33. Flock U, et al. (2008) Defining the proton entry point in the bacterial respiratory nitric-oxide reductase. J Biol Chem 283:3839–3845. 34. Hendriks JHM, Jasaitis A, Saraste M, Verkhovsky MI (2002) Proton and electron pathways in the bacterial nitric oxide reductase. Biochemistry 41:2331–2340. 35. Kumita H, et al. (2004) NO reduction by nitric-oxide reductase from denitrifying bacterium pseudomonas aeruginosa: Characterization of reaction intermediates that appear in the single turnover cycle. J Biol Chem 279:55247–55254. 36. Pinakoulaki E, Varotsis C (2008) Resonance raman spectroscopy of nitric oxide reductase and cbb3 heme-copper oxidase. J Phys Chem B 112:1851–1857. 37. Kapetanaki SM, et al. (2008) Ultrafast ligand binding dynamics in the active site of native bacterial nitric oxide reductase. Biochim Biophys Acta 1777:919–924. 38. Wasser IM, et al. (2004) Synthesis and spectroscopy of m-oxo (O2-)-bridged Heme/ Non-heme diiron complexes: Models for the active site of nitric oxide reductase. Inorg Chem 43:651–662. 39. Wasser IM, Huang H, Moënne-Loccoz P, Karlin KD (2005) Heme/Non-heme diiron(II) complexes and O2, CO, and NO adducts as reduced and substrate-bound models for the active site of bacterial nitric oxide reductase. J Am Chem Soc 127:3310–3320. 40. Collman JP, Yan Y, Lei J, Dinolfo PH (2006) Active-site models of bacterial nitric oxide reductase featuring tris-histidyl and glutamic acid mimics: Influence of a carboxylate ligand on FeB binding and the heme Fe∕FeB redox potential. Inorg Chem 45:7581–7583. 41. Collman JP, et al. (2008) A functional nitric oxide reductase model. Proc Natl Acad Sci USA 105:15660–15665. 42. Collman JP, et al. (2008) Intermediates involved in the two electron reduction of NO to N2O by a functional synthetic model of heme containing bacterial NO reductase. J Am Chem Soc 130:16498–16499. 43. Wang J, Schopfer MP, Sarjeant Amy AN, Karlin KD (2009) Heme-copper assembly mediated reductive coupling of nitrogen monoxide (*NO). J Am Chem Soc 131:450–451. 44. Liu J, Naruta Y, Tani F (2005) A functional model of the cytochrome c oxidase active site: Unique conversion of a heme-m-peroxo-CuII intermediate into heme-superoxo/ CuI. Angew Chem Int Edit 44:1836–1840. 45. Ghiladi RA, et al. (2005) Heme-copper/dioxygen adduct formation relevant to cytochrome c oxidase: Spectroscopic characterization of ½ð6LÞFeIII-ðO2 2−Þ-CuIIIþ. J Biol Inorg Chem 10:63–77. 46. Sage JT (1997) Myoglobin and CO: Structure, energetics, and disorder. J Biol Inorg Chem 2:537–543. 47. Sigman JA, Kim HK, Zhao X, Carey JR, Lu Y (2003) The role of copper and protons in heme-copper oxidases: Kinetic study of an engineered heme-copper center in myoglobin. Proc Natl Acad Sci USA 100:3629–3634. 48. Davydov R, Hoffman BM (2008) EPR and ENDOR studies of fe(II) hemoproteins reduced and oxidized at 77 K. J Biol Inorg Chem 13:357–369. 49. Marshall NM, et al. (2009) Rationally tuning the reduction potential of a single cupredoxin beyond the natural range. Nature 462:113–116. 50. Girsch P, de Vries S (1997) Purification and initial kinetic and spectroscopic character- ization of NO reductase from paracoccus denitrificans. Biochim Biophys Acta 1318:202–216. 51. Sakurai N, Sakurai T (1997) Isolation and characterization of nitric oxide reductase from paracoccus halodenitrificans. Biochemistry 36:13809–13815. 52. Hendriks J, et al. (1998) The active site of the bacterial nitric oxide reductase is a dinuclear iron center. Biochemistry 37:13102–13109. 53. Moënne-Loccoz P, et al. (2000) Nitric oxide reductase from paracoccus denitrificans contains an oxo-bridged Heme/Non-heme diiron center. J Am Chem Soc 122:9344–9345. 54. Zhao X, Yeung N, Wang Z, Guo Z, Lu Y (2005) Effects of metal ions in the CuB center on the redox properties of heme in heme-copper oxidases: Spectroelectrochemical stu- dies of an engineered heme-copper center in myoglobin. Biochemistry 44:1210–1214. 55. Decatur SM, et al. (1996) Trans effects in nitric oxide binding to myoglobin cavity mutant H93G. Biochemistry 35:4939–4944. 56. Pervitsky D, Immoos C, van der Veer W, Farmer PJ (2007) Photolysis of the HNO adduct of myoglobin: Transient generation of the aminoxyl radical. J Am Chem Soc 129:9590–9591. 57. Berto TC, Praneeth VKK, Goodrich LE, Lehnert N (2009) Iron-porphyrin NO complexes with covalently attached N-donor ligands: Formation of a stable six-coordinate species in solution. J Am Chem Soc 131:17116–17126. 58. Blomberg LM, Blomberg MRA, Siegbahn PEM (2006) Reduction of nitric oxide in bacterial nitric oxide reductase-a theoretical model study. Biochim Biophys Acta 1757:240–252. 59. Varadarajan R, Zewert TE, Gray HB, Boxer SG (1989) Effects of buried ionizable amino acids on the reduction potential of recombinant myoglobin. Science 243:69–72. 60. Grönberg KLC, et al. (1999) A low-redox potential heme in the dinuclear center of bacterial nitric oxide reductase: Implications for the evolution of energy-conserving heme-copper oxidases. Biochemistry 38:13780–13786. 61. Zhao X, Yeung N, Russell BS, Garner DK, Lu Y (2006) Catalytic reduction of NO to N2O by a designed heme copper center in myoglobin: Implications for the role of metal ions. J Am Chem Soc 128:6766–6767. 62. Pinakoulaki E, Stavrakis S, Urbani A, Varotsis C (2002) Resonance raman detection of a ferrous five-coordinate nitrosylheme b3 complex in cytochrome cbb3 oxidase from pseudomonas stutzeri. J Am Chem Soc 124:9378–9379. 63. Morrison M, Horie S (1965) Determination of heme a concentration in cytochrome preparations by hemochromogen method. Anal Biochem 12:77–82. 64. Taboy CH, Bonaventura C, Crumbliss AL (2002) Anaerobic oxidations of myoglobin and hemoglobin by spectroelectrochemistry. Method Enzymol 353:187–209. 8586 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1000526107 Lin et al.
3M3A
The roles of glutamates and metal ions in a rationally designed nitric oxide reductase based on myoglobin: Cu(II)-I107E FeBMb (Cu(II) binding to FeB site)
Roles of glutamates and metal ions in a rationally designed nitric oxide reductase based on myoglobin Ying-Wu Lina,2, Natasha Yeunga, Yi-Gui Gaob, Kyle D. Minerc, Shiliang Tiana, Howard Robinsond, and Yi Lua,c,1 aDepartment of Chemistry, bGeorge L. Clark X-Ray Facility and 3M Materials Laboratory, andcDepartment of Biochemistry, University of Illinois at Urbana-Champaign, Urbana, IL 61801; and dDepartment of Biology, Brookhaven National Laboratory, Upton, NY 11973 Edited by Harry B. Gray, California Institute of Technology, Pasadena, CA, and approved April 1, 2010 (received for review January 14, 2010) A structural and functional model of bacterial nitric oxide reductase (NOR) has been designed by introducing two glutamates (Glu) and three histidines (His) in sperm whale myoglobin. X-ray structural data indicate that the three His and one Glu (V68E) residues bind iron, mimicking the putative FeB site in NOR, while the second Glu (I107E) interacts with a water molecule and forms a hydrogen bond- ing network in the designed protein. Unlike the first Glu (V68E), which lowered the heme reduction potential by ∼110 mV, the second Glu has little effect on the heme potential, suggesting that thenegativelychargedGluhasadifferentrolein redoxtuning.More importantly, introducing the second Glu resulted in a ∼100% in- crease in NOR activity, suggesting the importance of a hydrogen bonding network in facilitating proton delivery during NOR reactiv- ity. In addition, EPR and X-ray structural studies indicate that the designed protein binds iron, copper, or zinc in the FeB site, each with different effects on the structures and NOR activities, suggesting that both redox activity and an intermediate five-coordinate heme-NO species are important for high NOR activity. The designed protein offers an excellent model for NOR and demonstrates the power of using designed proteins as a simpler and more well- defined system to address important chemical and biological issues. biomimetic models ∣heme-copper oxidase ∣metalloprotein ∣protein design ∣ protein engineering R ational design of proteins that mimic both structure and func- tion of more complex native enzymes has been a long sought- after goal, as the process is an ultimate test of our knowledge and an excellent means to develop advanced biocatalysts (1–3). Although designed proteins that model the structure of native enzymes have been known for a while (4–10), successful designs of proteins that mimic both the structure and function of native enzymes have been reported only recently (11–16). While being able to design such functional proteins is laudable, the impact of such an achievement would be greater if the designed proteins can be used to address fundamental issues in chemistry and biol- ogy that are difficult to tackle by other methods. One primary example is the roles of conserved glutamates and metal ions in bacterial nitric oxide reductase (NOR) (17–19). NO is critical for all life (20). Bacterial denitrification is a cru- cial part of the nitrogen cycle in nature that involves a four-step, five-electron reduction of nitrate (NO3 −) to dinitrogen (N2) (17, 19). Bacterial NOR is a membrane-bound protein that catalyzes one step of this process, namely, the two-electron reduc- tion of NO to N2O (17, 19). With no crystal or solution structure available for bacterial NOR to date, sequence alignments and homology modeling (21, 22) have indicated that NOR is structu- rally homologous to the largest subunit (subunit I) of heme- copper oxidases (HCOs) (23), enzymes that catalyze reduction of O2 to water. The active sites of both NOR and HCO contain a proximal histidine-coordinated heme and a distal three histi- dine-coordinated metal center. However, the metal center in HCOs is occupied by a copper (called CuB), whereas a nonheme iron is present in NOR (called FeB) (23, 24). In addition, two conserved glutamates, shown by modeling to be close to the FeB site (21, 22), are found to be essential for NOR activity (24, 25). Some members of HCOs such as cytochrome cbb3 oxidase display NOR activity (26–28), although the activity is ∼50-fold lower than native NOR (26). Therefore, it is important to elucidate the structural features, specifically the roles of the conserved glutamates close to the FeB site and metal ions (copper vs. iron), responsible for the reduction of NO to N2O. To address these issues, biochemical and biophysical studies of native NOR and its variants have been carried out (24, 25, 29–37). For example, Richardson and coworkers investigated the effects of amino acid substitutions of the five conserved glutamates (E122 and E125 presumed to face the periplasm and E198, E202, and E267 located in the interior of the membrane, close to the catalytic site) in the catalytic subunit of Paracoccus deni- trificans, NorB. The E122A, E125A, E198A, and E267A variants were inactive, indicating that these four glutamates are crucial for NOR activity (24, 25, 32, 33). On the other hand, Reimann et al. constructed a 3D model of NorB using homology modeling with the structures of HCOs as templates and suggested a plausible pathway consisting of these conserved glutamates for proton delivery (22). Despite these successes, the roles of the conserved glutamates and metal ions still remain to be fully elucidated, partly because of the difficulty in obtaining native NOR in high yield and the lack of a 3D structure. Even if these problems are resolved, it is still difficult to replace iron in the native FeB site with other metal ions, and spectroscopic studies of native NOR are often complicated by the presence of other metal cofactors (e.g., low-spin heme). To overcome these limitations, a number of synthetic models of NOR using small organic molecules as ligands, have been made in which the nonheme FeB site can be replaced by a copper ion (17, 38–45). In addition, since these model systems lack addi- tional metal-binding sites, spectroscopic studies are often simpli- fied. Therefore, studies of these synthetic models have offered many insights. For example, Collman et al. showed that a fully reduced heme/nonheme FeB compound can react with two equivalents of NO leading to the formation of one equivalent of N2O and a bis-ferric product (41). On the other hand, Karlin and coworkers showed that a small heme/Cu complex can effi- ciently lead to reductive coupling of NO to N2O (43). However, it is also difficult to obtain the synthetic models in high yield due to the multiple steps required in chemical synthesis. Because of this limitation, no synthetic NOR model containing the two key conserved Glu residues (E198 and E267 in NOR) has been Author contributions: Y.-W.L., N.Y., and Y.L. designed research; Y.-W.L., N.Y., Y.-G.G., K.D.M., and S.T. performed research; Y.-W.L., N.Y., Y.-G.G., K.D.M., H.R., and Y.L. analyzed data; and Y.-W.L. and Y.L. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Data deposition: The atomic coordinates and structure factors have been deposited in the Protein Data Bank, www.pdb.org (PDB ID codes 3M38, 3M39, 3M3A, and 3M3B). 1To whom correspondence should be addressed. E-mail: yi-lu@illinois.edu. 2Present address: School of Chemistry and Chemical Engineering, University of South China, Hengyang 421001, China. This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1000526107/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1000526107 PNAS ∣May 11, 2010 ∣vol. 107 ∣no. 19 ∣8581–8586 BIOCHEMISTRY CHEMISTRY reported. It is also difficult to substitute different metal ions in the same metal-binding site without perturbing the site geometry and distances to the heme iron, as most ligands are not as rigid as those in native enzymes and different metal ions have different geometric and ligand donor set preferences. We have recently designed a structural and functional protein model of bacterial NOR by engineering three histidines and one glutamate into the distal pocket of sperm whale myoglobin (swMb, L29H, F43H, H64, and V68E, named FeBMb) (14). Like synthetic models, this “bottom-up” approach complements the “top-down”approachofthestudyofnativeNORinthatitprovides insights into whether certain “necessary” structural elements are enough to impart enzyme function. Thanks in part to recent advances in computational, molecular, and structural biology, the designed myoglobin protein model is much easier to synthesize and to crystallize than either native NOR or synthetic models. Since myoglobin has often been used for the development and calibration of numerous spectroscopic techniques (46–48), it is an ideal choice for spectroscopic studies. More importantly, the rigid protein network allows precise placement of either glutamate or metal ions in myoglobin to address their roles in NOR activity. Toward this goal, we have demonstrated that both the histidines and one of the glutamates are essential for iron binding and NO reduction activity (14). However, the role of the second Glu close to the FeB site and the role of different metal ions in the FeB site have not been addressed. To address these important issues and to design even closer protein models of NOR, we introduced herein the second Glu to the second coordination sphere of the FeB site by mutating an Ile to a Glu (named I107E FeBMb). We show that the second Glu results in a ∼100% increase in NOR activity through hydro- gen bonding interactions and that the two glutamates have dramatically different effects on the heme reduction potential. Additionally, by comparing the EPR, electrochemistry, X-ray structures, and NOR activity of iron, copper, and zinc derivatives of the designed protein, we have obtained deeper insights into the roles of metal ions in NOR. Results Structure and Function of Fe(II)-I107E FeBMb. The X-ray crystal struc- tures of heme-containing I107E FeBMb without metal ion in the FeB site and with Fe2þ in the FeB site are solved at 1.42-Å and 1.65-Å resolution, respectively (Fig. 1 A and B and Table S1). In the absence of metal ions in the FeB site, the structure shows a water molecule in the FeB site, which forms hydrogen bonds with NE2 atoms of all three His residues, both OE1 and OE2 atoms of E68, and the OE2 atom of E107 (Fig. 1A). Upon binding Fe2þ, the Fe(II)-I107E FeBMb structure shows that Fe2þ is coordinated by three His, the OE2 atom of E68, and one water molecule. Notably, a water molecule bridges Fe2þ in the FeB site and the second glutamate (E107) with a distance of 2.32 Å to the OE2 atom of E107 (Fig. 1B). To probe the conformational changes of introducing the second Glu (E107), we performed a structural alignment of Fe (II)-I107E FeBMb and the previously reported FeðIIÞ-FeBMb (14). The comparison, shown in Fig. 2, indicates that both the polypeptide chain and the active site overlap well with each other. In addition, the two nonheme irons are located at similar posi- tions with a 0.36-Å separation from each other. In contrast, E68 underwent a significant conformational rearrangement in the presence of E107. These observations suggest that the active site of FeBMb can be tuned by the formation of an extended hydrogen bonding network, resulting from the introduction of a second glutamate residue. The binding of Fe2þ to deoxy I107E FeBMb was further mon- itored by EPR (Fig. 3A). Since deoxy myoglobin contains Fe(II) heme that exhibits no EPR signals in X-band EPR (14), we added blue copper Cu(II)-azurin (49), a redox partner of native NOR (19), to oxidize both the reduced heme and nonheme irons in Fe (II)-I107E FeBMb to EPR-active Fe(III). Upon addition of Cu (II)-azurin, the oxidation of deoxy I107E FeBMb resulted in EPR signals at g ¼ 6.12 and 5.56, typical of a high-spin heme- Fe(III). Upon addition of Fe2þ, however, a decrease of the heme-Fe(III) EPR signals was observed, indicating that the Fe2þ, when bound to the FeB site and oxidized by Cu(II)-azurin, is spin-coupled to heme-Fe(III). Such a spin coupling mimics that in NOR (35, 50–53), suggesting that I107E FeBMb models NOR closely, at least in this respect. To probe the role of the second Glu (E107) in NO reduction activity, we measured the yield of N2O production by Fe(II)- I107E FeBMb with excess NO under one turnover conditions. We monitored N2O formation in the headspace of the solution using GC/MS and compared this result to that of FeðIIÞ-FeBMb, which lacks the second Glu (Fig. 4). Remarkably, Fe(II)-I107E FeBMb displays higher activity than FeðIIÞ-FeBMb. After ∼20 hr, ∼24% N2O was produced by Fe(II)-I107E FeBMb, in H29 A H29 B H64 E68 E107 3.04 2.81 3.02 2.20 2.24 2.24 2.12 2.21 H64 E68 E107 3.03 H43 2.15 3.41 3.16 2.62 2.32 H43 2.26 H93 H93 H29 C H29 D 2.04 2.03 H64 E68 E107 2.09 2.10 2.91 2.21 H64 E68 E107 2.26 2.29 2.10 2.18 2.10 4.47 H43 2.07 3.04 H43 2.68 H93 H93 Fig. 1. Crystal structures of I107E FeBMb (A) (PDB ID code 3M38), Fe(II)-I107E FeBMb (B) (PDB ID code 3M39), Cu(II)-I107E FeBMb (C) (PDB ID code 3M3A), and Zn(II)-I107E FeBMb (D) (PDB ID code 3M3B). Water molecules, Fe(II), Cu(II), and Zn(II) are represented by red, green, orange, and gray spheres, respectively. Fig. 2. Overlay of Fe(II)-I107E FeBMb (cyan) (PDB ID code 3M39) with FeðIIÞ-FeBMb (orange) (PDB ID code 3K9Z). 8582 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1000526107 Lin et al. contrast to ∼10% yield for FeðIIÞ-FeBMb, strongly indicating that the second Glu plays an important role in NO reduction, likely facilitating proton uptake during NO reduction. Other Metal Ions Binding to I107E FeBMb. To find out if the resting state of the protein, i.e., oxidized or met I107E FeBMb, can bind other metal ions, Cu2þ or Zn2þ was titrated into met I107E FeBMb and monitored by EPR spectroscopy (Fig. 3 B and C). In the absence of metal ions, met I107E FeBMb exhibited high-spin heme signals at g ¼ 6.03, 5.08, and 1.98 (Fig. 3B, black line). Upon addition of 2 eq of Cu2þ, the signals at g ¼ 6.03 and 5.08 decreased and a broad peak around g ¼ 2.95 increased, probably due to spin coupling between heme-Fe(III) and Cu2þ in the FeB site. In contrast, addition of Zn2þ, a metal ion with no unpaired electrons [i.e., incapable of spin coupling to heme- Fe(III)], produced an increase in the high-spin heme signals at g ¼ 5.88 and 5.60 (Fig. 3C), indicating that the interaction be- tween E68 and heme iron was weakened after metal binding. The X-ray crystal structures of I107E FeBMb with Cu2þ or Zn2þ in the FeB site were solved at 1.37-Å and 1.60-Å resolution, respectively (Fig. 1 C and D and Table S1). Compared to Fe(II)- I107E FeBMb (Fig. 1B), a similar binding site was observed for Cu(II)-I107E FeBMb (Fig. 1C), where H29, H43, and H64 coordinate to Cu2þ with distances of 2.09, 2.10, and 2.04 Å, respectively, slightly shorter than the corresponding distances in the Fe2þ structure. In comparison to Fe(II)-I107E FeBMb, the water bridging the Cu2þ and the second Glu (E107) is shifted toward Cu2þ in the FeB site (2.03 Å) with respect to E107 (3.04 Å). Interestingly, this bridging water molecule was not observed in Zn(II)-I107E FeBMb (Fig. 1D), but the two O atoms of E68 coordinate to Zn2þ with similar distances (2.26 Å for OE1 and 2.29 Å for OE2). The longer distance between OE1 of E68 B 6.12 I107E FeBMb + Azurin I107E Fe Mb + 0 5 eq Fe 2+ + Azurin A 5.56 I107E FeBMb + 0.5 eq Fe + Azurin I107E FeBMb + 1.0 eq Fe 2+ + Azurin I107E FeBMb + 2.0 eq Fe 2+ + Azurin 1.98 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5.08 5.38 Magnetic Field (Gauss) 5.88 I107E FeBMb I107E FeBMb + 0.5 eq Zn 2+ I107E Fe Mb 1 0 eq Zn 2+ C 6.03 I107E FeBMb I107E FeBMb + 0.5 eq Cu 2+ I107E Fe Mb + 1 0 eq Cu 2+ I107E FeBMb + 1.0 eq Zn I107E FeBMb + 2.0 eq Zn 2+ I107E FeBMb + 1.0 eq Cu I107E FeBMb + 2.0 eq Cu 2+ 1.98 ~2.95 1.98 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5.60 Magnetic Field (Gauss) 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5.08 Magnetic Field (Gauss) Fig. 3. EPR spectra of deoxy I107E FeBMb (0.5 mM protein in 50 mM Bis-Tris, pH 7.0) with increasing concentrations of Fe2þ in the presence of wild-type Cu(II)-azurin (A), and oxidized I107E FeBMb with Cu2þ (B) or Zn2þ (C). Spectra were collected at 4 K, 5 mW power, and 9.05 GHz. 30 25 Fe(II)-I107E FeBMb Fe(II)-FeBMb 15 20 10 %) oduction (% N2O pro 0 5 0 4 8 12 16 20 Incubation time (hr) Fig. 4. Time-dependent N2O production by Fe(II)-I107E FeBMb (▴) and FeðIIÞ-FeBMb (●) with ∼50 eq. NO under single turnover conditions. The yield was determined by a comparison of the ratio of NO∶N2O peaks from the GC/MS chromatograms. Lin et al. PNAS ∣ May 11, 2010 ∣ vol. 107 ∣ no. 19 ∣ 8583 BIOCHEMISTRY CHEMISTRY and heme iron in the Zn-bound structure (2.68 Å) in comparison to the Cu- and Fe-bound structures, is likely the result of a weaker interaction, which is also supported by an observed increase of the high-spin heme signals in the EPR spectra upon Zn2þ binding (Fig. 3C). These results suggest I107E FeBMb is capable of incor- porating different metal ions into its designed FeB site, offering an excellent opportunity to compare the role of these metal ions in the same protein scaffold. Effect of Glutamates and Metal Ions on the Redox Potential of I107E FeBMb. Since EPR and X-ray structural studies indicate metal binding to I107E FeBMb, we used spectroelectrochemistry to measure the effects of glutamates and metal ions on the heme reduction potential. When there is no metal ion in the FeB site, the I107E FeBMb displays a reduction potential of −134  3 mV vs. the normal hydrogen electrode (NHE) (Fig. S1A), similar to that of FeBMb (−158  4 mV) without the I107E mutation (14). In the presence of Cu2þ, I107E FeBMb has a reduction potential (−137  2 mV) (Fig. S1B) almost identical to that of the same protein in the absence of metal ions in the FeB site, indicating that copper binding to the FeB site has little effect on the reduction potential of the heme iron. This observation is similar to that observed for Cu2þ binding to CuBMb (CuðIIÞ-CuBMb, 80 mV vs. CuBMb, 77 mV) (54). On the other hand, the presence of Fe2þ and Zn2þ increased the reduction potential of I107E FeBMb from −134  3 mV to −64  3 mV vs. NHE (Fig. S1C) and −105  2 mV vs. NHE (Fig. S1D), respectively. The different effects of Cu2þ, Fe2þ, and Zn2þ on the reduction potential of I107E-FeBMb indicate that these metal ions in the FeB site may play different roles through different coordination properties. NOR Activity of I107E FeBMb in the Presence of Different Metal Ions. The NO reduction activity of I107E FeBMb in the presence of Fe2þ, Cuþ, or Zn2þ was monitored by GC/MS under single turn- over conditions. When Fe(II)-I107E FeBMb was exposed to excess NO, N2O could be observed to form with increased yield over time (Fig. S2). Similarly, N2O formation was observed for Cu(I)-I107E FeBMb, indicating that Fe or Cu binding to the FeB site results in comparable NOR activities. It should be noted that because of the high solubility of N2O (∼25 mM in water at room temperature), GC/MS cannot be used to quantify the rates of NO reduction under these conditions. In contrast, no N2O formation was observed with redox inactive Zn2þ, which demon- strates that redox active Fe2þ or Cuþ in the FeB site plays a crucial role in NO reduction. To gain deeper insight into the process of NO reduction, EPR studies were further performed to monitor the initial process of NO reduction. In the absence of metal ions, the EPR spectrum of ferrous I107E FeBMb-NO shows hyperfine splitting resulting from bound NO and the proximal histidine, indicating the forma- tion of a six-coordinate ferrous heme-NO species (Fig. 5, top line). After incubation of Fe(II)-I107E FeBMb with excess NO, a distinct three-line hyperfine structure appears at 15 min (Fig. 5A), suggesting the formation of a five-coordinate ferrous heme-NO species as a result of cleavage of the proximal His-Fe heme bond (55). A three-line hyperfine structure was also observed for Cuþ and Zn2þ, except that the signal intensity is low- er with Cu(I)-I107E FeBMb-NO (Fig. 5B) and more pronounced in Zn(II)-I107E FeBMb-NO (Fig. 5C). The lower intensity of the three-line hyperfine structure for Cu(I)-I107E FeBMb-NO suggests the major species formed is a six-coordinate ferric heme-NO complex, which is EPR silent (41). These differences further suggest that the metal ion in the FeB site plays a key role in formation of the intermediates, thereby tuning NOR activity. Discussion Using Rationally Designed Proteins to Address Important Issues in Chemistry and Biology. Important issues such as the roles of the conserved glutamates and nonheme FeB in NOR have been previously addressed using biochemical and biophysical studies or biomimetic modeling (24, 25, 27–37, 45, 56, 57). As a comple- mentary approach, rational protein design, using small, easy-to- produce and well-characterized proteins such as myoglobin, offers a powerful method with which to gain insights into more complex native enzymes such as NOR (14). Similar to synthetic models (41, 43), the metal ion at the putative FeB site in the protein model can be substituted freely. Better yet, Glu residues can be placed at precise locations in the protein, including the secondary coordination sphere, due to its rigid network. By care- fully choosing a suitable protein template, rational protein design could be generally applied to address other important issues in chemistry and biology. The Roles of Glutamates. Although two conserved glutamates (E198 and E267) are known to be crucial for NOR activity (24, 25), their roles are not well defined (18, 19). In a previous study (14), we demonstrated that one Glu, E68, is important for both iron binding and NOR activity of FeBMb. The crystal struc- tures of both FeðIIÞ-FeBMb and Fe(II)-I107E FeBMb show that one O atom of E68 directly coordinates to FeB (Fig. 2). In syn- thetic models of NOR, it has also been found that the presence of a glutamic acid mimic significantly increases the stability of iron binding to the FeB site (40). Furthermore, a theoretical study by Blomberg et al. (58) showed that a model with an FeB coordi- nated by three histidines, one glutamate, and one water molecule provides an energetically feasible reaction mechanism of NO reduction. However, the structural model of NOR constructed recently by Reimann et al. (22) shows that the closest conserved Glu (E267) still has its carboxylate O atom 7 Å away from FeB, which suggests that Glu may not bind to FeB in native NOR. One interesting finding from our study is that the Glu (E68) under- went a significant conformational rearrangement in the presence of another Glu (E107) (Fig. 2). Therefore, the FeBMb provides a viable model of NOR that is consistent with Blomberg’s model, but cannot rule out Reimann’s model due to possible conforma- tion changes. While the role of the first Glu is still uncertain until a 3D struc- ture of NOR in its active form is available, the role of the second Glu is even less defined. We address this question by introducing a second Glu (E107) to FeBMb. The crystal structures shown in Fig. 1 indicate that E107 interacts with a water molecule and forms a hydrogen bonding network in both Fe(II)-I107E FeBMb and Cu(II)-I107E FeBMb. Interestingly, although a similar water molecule was observed in the active site of FeðIIÞ- FeBMb (Fig. 2), activity assay data indicate that the presence of E107 in Fe(II)-I107E FeBMb increases NOR activity by ∼100% C 5 min N t l A 5 min B 2+ 5 min 1 min No metal 1 min 5 min F 2+ No metal 1 i 5 min + No metal 2+ Zn 2+ 5 min Fe 2+ 5 min 1 min Fe C + 5 min 1 min Cu + Z 2 Zn 2+ 15 min F 2+ Fe 15 min 5 min C + Cu + 15 min 5 min Zn 2+ 15 min Fe2+ 5 Cu + 15 min 3000 3100 3200 3300 3400 Magnetic Field (Gauss) 3000 3100 3200 3300 3400 Magnetic Field (Gauss) 3000 3100 3200 3300 3400 Magnetic Field (Gauss) Fig. 5. EPR spectra of deoxy I107E FeBMb (0.5 mM) with no metal bound in the presence of NO after 5 min (top line), with 2 eq Fe2þ (A), Cuþ (B), or Zn2þ (C) incubated with excess NO (∼200 eq) for 1, 5, and 15 min. Spectra were collected in 50 mM Bis-Tris pH 7.0 at 30 K, 0.2 mW power, and 9.05 GHz. 8584 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1000526107 Lin et al. (Fig. 4), suggesting that the second Glu may potentionally play a role in providing one of the protons during reduction of NO to N2O. Although free Glu outside a protein has a pKa ∼4.3, the stu- dies of native NOR showed that the pKa of its Glu close to the active site has a value of ∼6.6 (22, 25, 33). The hydrogen binding network in our protein models may contribute to the fine-tuning of the Glu pKa to be more neutral, similar to that in NOR. More- over, it is interesting that Cu(I)-I107E FeBMb also shows NOR activity, which provides an interesting protein model of HCOs with NOR function (26–28), even though Glu residues are not conserved in native HCOs. Additionally, spectroelectrochemical studies showed that the reduction potential of I107E FeBMb with no metal ion in the FeB site is similar to that of FeBMb, but much lower than that of CuBMb (77 mV) (54), which contains the same three His, but no Glu in the metal-binding site above the heme. Since both FeBMb and I107E FeBMb contain the V68E mutation that has been shown to decrease the heme reduction potential of native myoglobin from 59 mV to −137 mV (59), it is likely that the introduction of a negatively charged Glu close to the heme group is what is responsible for the dramatically lower heme redox potential. A conserved glutamate, predicted to be located near the catalytic heme b3 in NOR, was proposed to be responsible for a ∼260-mV decrease in reduction potential (60 mV) in comparison to the other two heme centers, heme b (345 mV) and heme c (310 mV) (60). Our FeBMb and I107E FeBMb mod- els mimic this feature of NOR. Notably, although introduction of the first Glu (E68) lowered the heme potential by ∼110 mV (14), introduction of the second Glu via the I107E mutation did not result in a significant difference in the heme reduction potential, suggesting that the effect of the two conserved Glu residues in NOR on heme reduction potential is not additive, with the effects highly dependent on the location of the Glu. The Roles of Metal Ions. The roles of metal ions in NOR are an- other important question as iron is found in the native FeB site and HCO employs copper at the corresponding CuB site. With different metal ions in the FeB site, the crystal structures clearly show the heme and nonheme dinuclear center existing in differ- ent local environments (Fig. 1). Although a similar hydrogen bond network is formed in both Fe(II)-I107E FeBMb and Cu (II)-I107E FeBMb, the conformation of E68 and E107 with re- spect to the nonheme metal center and heme iron is different from each other. Moreover, the coordination geometry differs significantly with Zn2þ in the FeB site. A hydrogen bond is absent from the Zn crystal structure, but both the O atoms of E68 act as metal-binding ligands. These observations demonstrate that the identity of the metal ion in the FeB site can tune the active site through their interactions with the His and Glu ligands, resulting in formation of different coordination geometries with different hydrogen bonds. In addition to structural fine-tuning, the metal ion at the FeB site can also tune the heme iron reduction potential in I107E FeBMb. Spectroelectrochemical studies showed that the binding of Fe2þ or Zn2þ results in an increase in the heme reduction potential by ∼70 mV and ∼30 mV, respectively (Fig. S1). In the case of Cu(II)-I107E FeBMb, the crystal structure shows that OE1 of E68 is closer to the heme iron (2.07 Å) (Fig. 1C) than its metal-free form (2.15 Å) (Fig. 1A). The stronger interaction from the negatively charged E68 could offset the effect of positively charged Cu2þ binding, resulting in similar reduction potentials observed for Cu(II)-I107E FeBMb and I107E FeBMb. In a previous study (61), EPR data showed that during NO reduction, the binding of Cuþ to the CuB site of CuBMb can weaken the proximal heme Fe-His bond, while complete cleavage of the heme Fe-His bond occurred when Zn2þ was bound to CuBMb-NO. In this study, we observed that a five-coordinate heme-NO species was formed with Fe2þ, Cuþ, or Zn2þ bound to the FeB site of I107E FeBMb (Fig. 5). Significantly, a five- coordinate heme-NO species has also been observed for both NOR (30, 31, 35) and the member of the HCO family with the highest NO reduction activity, cytochrome cbb3 oxidases (26, 62). However, this species was not observed for FeðIIÞ-FeBMb-NO and CuðIÞ-FeBMb-NO, which lack the second Glu (E107). In both these cases, the proximal heme Fe-His bond was only weakened, as indicated by a decrease of the nine-line hyperfine splitting signals in the EPR spectra (Fig. S3). These observations suggest that formation of a five-coordinate heme-NO species may play an important role in NOR reactivity. Conclusions We have successfully designed a structural and functional model of NOR, by introducing a second glutamate in the vicinity of the FeB site, named I107E FeBMb. This protein model mimics native NOR more closely by bearing the structural feature of three his- tidines and two glutamates in the FeB site, as predicted for native NOR. We have demonstrated that the two glutamates can play different roles in NO reduction activity; namely, one acts as a li- gand to FeB (E68), and the other acts as a proton transfer group (E107). Furthermore, by substituting different metal ions into the nonheme metal site, we have demonstrated that FeB plays crucial roles in fine-tuning the active site by donating electrons and by mediating the formation of a five-coordinate heme-NO inter- mediate during NO reduction. In the absence of a crystal struc- ture for native NOR, this study offers an ideal protein model and provides valuable structural as well as mechanistic information for native NOR. Materials and Methods Protein Preparation. I107E FeBMb (swMb L29H/F43H/V68E/I107E) was con- structed, expressed, and purified using the procedure described previously (14). The purity and identity were confirmed by SDS-PAGE and electrospray ionized MS: observed: 17; 392  1 Da; calculated: 17,391 Da. EPR Spectroscopy. EPR spectra were recorded on a Bruker ESP 300 equipped with an Oxford liquid helium cryostat and an ITC4 temperature controller. The samples of met I107E FeBMb, Cu(II)-, or Zn(II)-I107E FeBMb were pre- pared as described previously (14). The samples of NO-bound deoxy I107E FeBMb, Cu(I)-, Fe(II)-, or Zn(II)-I107E FeBMb were prepared by inject- ing 1 mL of purified NO gas into the EPR tube containing 300 μL of protein (0.5 mM, 10% glycerol, in 50 mM Bis-Tris pH 7.0). The samples were then flash frozen in liquid N2 after incubation for 1, 5, or 15 min. The molar extinction coefficient of the Soret band of I107E FeBMb at 406 nm (175 mM−1 · cm−1), calculated using the standard hemochromagen method (63), was used to determine protein concentration. The metal sources of Cu(I), Cu(II), Zn(II), and Fe(II) were ½ðCH3CNÞ4CuPF6, CuSO4, ZnSO4 · 7H2O, and FeCl2, respectively. Spectroelectrochemical Measurements. Protein reduction potentials were measured using an optically transparent thin layer electrode as previously described (64). The potential of the working electrode was applied in the positive direction for deoxy I107E FeBMb with Fe2þ and in the negative direction for metal free and with Cu2þ or Zn2þ. Other procedures are the same as described previously (54). X-Ray Crystallographic Studies. Fe(II)-I107E FeBMb was crystallized anaerobi- cally in a glove box at room temperature using the conditions described for FeðIIÞ-FeBMb (14). I107E FeBMb, Cu(II)-, or Zn(II)-I107E FeBMb were crystallized aerobically. Diffraction-quality crystals were soaked in a cryoprotectant solution of 30% PEG 400 and flash frozen in liquid nitrogen. Diffraction data were collected at the Brookhaven National Lab Synchrotron Light Source X12C beamline. The crystal structure was solved using the same method as for FeðIIÞ-FeBMb (14). NOR Activity Assay. NO reduction was monitored by GC/MS. The protein was reduced to the deoxy form by excess dithionite that was removed with a size-exclusion column (PD-10). Then 2 eq metal, Cu(I), Fe(II), or Zn(II), was added to the protein solution (0.6 mM, 3 mL in 50 mM Bis-Tris buffer, pH 7.0). The samples were prepared anaerobically in a glove box. Purified NO Lin et al. PNAS ∣ May 11, 2010 ∣ vol. 107 ∣ no. 19 ∣ 8585 BIOCHEMISTRY CHEMISTRY gas was injected into the head space of the reaction flask with the molar ratio of NO∶protein ≈50∶1. Other procedures are the same as described previously (14, 61). ACKNOWLEDGMENTS. We thank Dr. Mark J. Nilges for help with EPR analysis, and Furong Sun and Beth D. Eves for aiding in GC/MS data collection. This work was supported by NIH Grant GM062211. 1. Lu Y, Yeung N, Sieracki N, Marshall NM (2009) Design of functional metalloproteins. Nature 460:855–862. 2. Nanda V, Koder RL (2010) Designing artificial enzymes by intuition and computation. Nat Chem 2:15–24. 3. Gray HB (2003) Biological inorganic chemistry at the beginning of the 21st century. Proc Natl Acad Sci USA 100:3563–3568. 4. Regan L, DeGrado WF (1988) Characterization of a helical protein designed from first principles. Science 241:976–978. 5. Hecht MH, Richardson JS, Richardson DC, Ogden RC (1990) De novo design, expression, and characterization of Felix: A four-helix bundle protein of native-like sequence. Science 249:884–891. 6. Robertson DE, et al. (1994) Design and synthesis of multi-heme proteins. Nature 368:425–432. 7. Lu Y, Valentine JS (1997) Engineering metal-binding sites in proteins. Curr Opin Struct Biol 7:495–500. 8. Reedy CJ, Gibney BR (2004) Heme protein assemblies. Chem Rev 104:617–649. 9. Watanabe Y, Hayashi T (2005) Functionalization of myoglobin. Prog Inorg Chem 54:449–493. 10. Ghosh D, Pecoraro VL (2005) Probing metal-protein interactions using a de novo design approach. Curr Opin Chem Biol 9:97–103. 11. Kaplan J, DeGrado WF (2004) De novo design of catalytic proteins. Proc Natl Acad Sci USA 101:11566–11570. 12. Kang SG, Saven JG (2007) Computational protein design: Structure, function and combinatorial diversity. Curr Opin Chem Biol 11:329–334. 13. Jiang L, et al. (2008) De novo computational design of retro-aldol enzymes. Science 319:1387–1391. 14. Yeung N, et al. (2009) Rational design of a structural and functional nitric oxide reductase. Nature 462:1079–1082. 15. Koder RL, et al. (2009) Design and engineering of an O2 transport protein. Nature 458:305–309. 16. Heinisch T, Ward TR (2010) Design strategies for the creation of artifical metalloen- zymes. Curr Opin Chem Biol 14:184–199. 17. Wasser IM, de Vries S, Moënne-Loccoz P, Schröder I, Karlin KD (2002) Nitric oxide in biological denitrification: Fe/Cu metalloenzyme and metal complex NOx redox chemistry. Chem Rev 102:1201–1234. 18. Yeung N, Lu Y (2008) One heme, diverse functions: Using biosynthetic myoglobin models to gain insights into heme-copper oxidases and nitric oxide reductases. Chem Biodivers 5:1437–1454. 19. Watmough NJ, Field SJ, Hughes RJL, Richardson DJ (2009) The bacterial respiratory nitric oxide reductase. Biochem Soc Trans 37:392–399. 20. Cary SPL, Winger JA, Derbyshire ER, Marletta MA (2006) Nitric oxide signaling: No longer simply on or off. Trends Biochem Sci 31:231–239. 21. Sakurai N, Sakurai T (1998) Genomic DNA cloning of the region encoding nitric oxide reductase in paracoccus halodenitrificans and a structure model relevant to cytochrome oxidase. Biochem Biophys Res Commun 243:400–406. 22. Reimann J, Flock U, Lepp H, Honigmann A, Ädelroth P (2007) A pathway for protons in nitric oxide reductase from paracoccus denitrificans. Biochim Biophys Acta 1767:362–373. 23. Pereira MM, Sousa FL, Verissimo AF, Teixeira M (2008) Looking for the minimum common denominator in haem-copper oxygen reductases: Towards a unified catalytic mechanism. Biochim Biophys Acta 1777:929–934. 24. Butland G, Spiro S, Watmough NJ, Richardson DJ (2001) Two conserved glutamates in the bacterial nitric oxide reductase are essential for activity but not assembly of the enzyme. J Bacteriol 183:189–199. 25. Flock U, Lachmann P, Reimann J, Watmough NJ, Äedelroth P (2009) Exploring the terminal region of the proton pathway in the bacterial nitric oxide reductase. J Inorg Biochem 103:845–850. 26. Forte E, et al. (2001) The cytochrome cbb3 from pseudomonas stutzeri displays nitric oxide reductase activity. Eur J Biochem 268:6486–6490. 27. Huang Y, Reimann J, Lepp H, Drici N, Ädelroth P (2008) Vectorial proton transfer coupled to reduction of O2 and NO by a heme-copper oxidase. Proc Natl Acad Sci USA 105:20257–20262. 28. Hayashi T, et al. (2009) Accommodation of two diatomic molecules in cytochrome bo3: Insights into NO reductase activity in terminal oxidases. Biochemistry 48:883–890. 29. Moënne-Loccoz P (2007) Spectroscopic characterization of heme iron-nitrosyl species and their role in NO reductase mechanisms in diiron proteins. Nat Prod Rep 24:610–620. 30. Moënne-Loccoz P, de Vries S (1998) Structural characterization of the catalytic high-spin heme b of nitric oxide reductase: A resonance raman study. J Am Chem Soc 120:5147–5152. 31. Sakurai T, Sakurai N, Matsumoto H, Hirota S, Yamauchi O (1998) Roles of four iron centers in paracoccus halodenitrificans nitric oxide reductase. Biochem Biophys Res Commun 251:248–251. 32. Thorndycroft FH, Butland G, Richardson DJ, Watmough NJ (2007) A new assay for nitric oxide reductase reveals two conserved glutamate residues form the entrance to a proton-conducting channel in the bacterial enzyme. Biochem J 401:111–119. 33. Flock U, et al. (2008) Defining the proton entry point in the bacterial respiratory nitric-oxide reductase. J Biol Chem 283:3839–3845. 34. Hendriks JHM, Jasaitis A, Saraste M, Verkhovsky MI (2002) Proton and electron pathways in the bacterial nitric oxide reductase. Biochemistry 41:2331–2340. 35. Kumita H, et al. (2004) NO reduction by nitric-oxide reductase from denitrifying bacterium pseudomonas aeruginosa: Characterization of reaction intermediates that appear in the single turnover cycle. J Biol Chem 279:55247–55254. 36. Pinakoulaki E, Varotsis C (2008) Resonance raman spectroscopy of nitric oxide reductase and cbb3 heme-copper oxidase. J Phys Chem B 112:1851–1857. 37. Kapetanaki SM, et al. (2008) Ultrafast ligand binding dynamics in the active site of native bacterial nitric oxide reductase. Biochim Biophys Acta 1777:919–924. 38. Wasser IM, et al. (2004) Synthesis and spectroscopy of m-oxo (O2-)-bridged Heme/ Non-heme diiron complexes: Models for the active site of nitric oxide reductase. Inorg Chem 43:651–662. 39. Wasser IM, Huang H, Moënne-Loccoz P, Karlin KD (2005) Heme/Non-heme diiron(II) complexes and O2, CO, and NO adducts as reduced and substrate-bound models for the active site of bacterial nitric oxide reductase. J Am Chem Soc 127:3310–3320. 40. Collman JP, Yan Y, Lei J, Dinolfo PH (2006) Active-site models of bacterial nitric oxide reductase featuring tris-histidyl and glutamic acid mimics: Influence of a carboxylate ligand on FeB binding and the heme Fe∕FeB redox potential. Inorg Chem 45:7581–7583. 41. Collman JP, et al. (2008) A functional nitric oxide reductase model. Proc Natl Acad Sci USA 105:15660–15665. 42. Collman JP, et al. (2008) Intermediates involved in the two electron reduction of NO to N2O by a functional synthetic model of heme containing bacterial NO reductase. J Am Chem Soc 130:16498–16499. 43. Wang J, Schopfer MP, Sarjeant Amy AN, Karlin KD (2009) Heme-copper assembly mediated reductive coupling of nitrogen monoxide (*NO). J Am Chem Soc 131:450–451. 44. Liu J, Naruta Y, Tani F (2005) A functional model of the cytochrome c oxidase active site: Unique conversion of a heme-m-peroxo-CuII intermediate into heme-superoxo/ CuI. Angew Chem Int Edit 44:1836–1840. 45. Ghiladi RA, et al. (2005) Heme-copper/dioxygen adduct formation relevant to cytochrome c oxidase: Spectroscopic characterization of ½ð6LÞFeIII-ðO2 2−Þ-CuIIIþ. J Biol Inorg Chem 10:63–77. 46. Sage JT (1997) Myoglobin and CO: Structure, energetics, and disorder. J Biol Inorg Chem 2:537–543. 47. Sigman JA, Kim HK, Zhao X, Carey JR, Lu Y (2003) The role of copper and protons in heme-copper oxidases: Kinetic study of an engineered heme-copper center in myoglobin. Proc Natl Acad Sci USA 100:3629–3634. 48. Davydov R, Hoffman BM (2008) EPR and ENDOR studies of fe(II) hemoproteins reduced and oxidized at 77 K. J Biol Inorg Chem 13:357–369. 49. Marshall NM, et al. (2009) Rationally tuning the reduction potential of a single cupredoxin beyond the natural range. Nature 462:113–116. 50. Girsch P, de Vries S (1997) Purification and initial kinetic and spectroscopic character- ization of NO reductase from paracoccus denitrificans. Biochim Biophys Acta 1318:202–216. 51. Sakurai N, Sakurai T (1997) Isolation and characterization of nitric oxide reductase from paracoccus halodenitrificans. Biochemistry 36:13809–13815. 52. Hendriks J, et al. (1998) The active site of the bacterial nitric oxide reductase is a dinuclear iron center. Biochemistry 37:13102–13109. 53. Moënne-Loccoz P, et al. (2000) Nitric oxide reductase from paracoccus denitrificans contains an oxo-bridged Heme/Non-heme diiron center. J Am Chem Soc 122:9344–9345. 54. Zhao X, Yeung N, Wang Z, Guo Z, Lu Y (2005) Effects of metal ions in the CuB center on the redox properties of heme in heme-copper oxidases: Spectroelectrochemical stu- dies of an engineered heme-copper center in myoglobin. Biochemistry 44:1210–1214. 55. Decatur SM, et al. (1996) Trans effects in nitric oxide binding to myoglobin cavity mutant H93G. Biochemistry 35:4939–4944. 56. Pervitsky D, Immoos C, van der Veer W, Farmer PJ (2007) Photolysis of the HNO adduct of myoglobin: Transient generation of the aminoxyl radical. J Am Chem Soc 129:9590–9591. 57. Berto TC, Praneeth VKK, Goodrich LE, Lehnert N (2009) Iron-porphyrin NO complexes with covalently attached N-donor ligands: Formation of a stable six-coordinate species in solution. J Am Chem Soc 131:17116–17126. 58. Blomberg LM, Blomberg MRA, Siegbahn PEM (2006) Reduction of nitric oxide in bacterial nitric oxide reductase-a theoretical model study. Biochim Biophys Acta 1757:240–252. 59. Varadarajan R, Zewert TE, Gray HB, Boxer SG (1989) Effects of buried ionizable amino acids on the reduction potential of recombinant myoglobin. Science 243:69–72. 60. Grönberg KLC, et al. (1999) A low-redox potential heme in the dinuclear center of bacterial nitric oxide reductase: Implications for the evolution of energy-conserving heme-copper oxidases. Biochemistry 38:13780–13786. 61. Zhao X, Yeung N, Russell BS, Garner DK, Lu Y (2006) Catalytic reduction of NO to N2O by a designed heme copper center in myoglobin: Implications for the role of metal ions. J Am Chem Soc 128:6766–6767. 62. Pinakoulaki E, Stavrakis S, Urbani A, Varotsis C (2002) Resonance raman detection of a ferrous five-coordinate nitrosylheme b3 complex in cytochrome cbb3 oxidase from pseudomonas stutzeri. J Am Chem Soc 124:9378–9379. 63. Morrison M, Horie S (1965) Determination of heme a concentration in cytochrome preparations by hemochromogen method. Anal Biochem 12:77–82. 64. Taboy CH, Bonaventura C, Crumbliss AL (2002) Anaerobic oxidations of myoglobin and hemoglobin by spectroelectrochemistry. Method Enzymol 353:187–209. 8586 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1000526107 Lin et al.
3M3B
The roles of glutamates and metal ions in a rationally designed nitric oxide reductase based on myoglobin: Zn(II)-I107E FeBMb (Zn(II) binding to FeB site)
Roles of glutamates and metal ions in a rationally designed nitric oxide reductase based on myoglobin Ying-Wu Lina,2, Natasha Yeunga, Yi-Gui Gaob, Kyle D. Minerc, Shiliang Tiana, Howard Robinsond, and Yi Lua,c,1 aDepartment of Chemistry, bGeorge L. Clark X-Ray Facility and 3M Materials Laboratory, andcDepartment of Biochemistry, University of Illinois at Urbana-Champaign, Urbana, IL 61801; and dDepartment of Biology, Brookhaven National Laboratory, Upton, NY 11973 Edited by Harry B. Gray, California Institute of Technology, Pasadena, CA, and approved April 1, 2010 (received for review January 14, 2010) A structural and functional model of bacterial nitric oxide reductase (NOR) has been designed by introducing two glutamates (Glu) and three histidines (His) in sperm whale myoglobin. X-ray structural data indicate that the three His and one Glu (V68E) residues bind iron, mimicking the putative FeB site in NOR, while the second Glu (I107E) interacts with a water molecule and forms a hydrogen bond- ing network in the designed protein. Unlike the first Glu (V68E), which lowered the heme reduction potential by ∼110 mV, the second Glu has little effect on the heme potential, suggesting that thenegativelychargedGluhasadifferentrolein redoxtuning.More importantly, introducing the second Glu resulted in a ∼100% in- crease in NOR activity, suggesting the importance of a hydrogen bonding network in facilitating proton delivery during NOR reactiv- ity. In addition, EPR and X-ray structural studies indicate that the designed protein binds iron, copper, or zinc in the FeB site, each with different effects on the structures and NOR activities, suggesting that both redox activity and an intermediate five-coordinate heme-NO species are important for high NOR activity. The designed protein offers an excellent model for NOR and demonstrates the power of using designed proteins as a simpler and more well- defined system to address important chemical and biological issues. biomimetic models ∣heme-copper oxidase ∣metalloprotein ∣protein design ∣ protein engineering R ational design of proteins that mimic both structure and func- tion of more complex native enzymes has been a long sought- after goal, as the process is an ultimate test of our knowledge and an excellent means to develop advanced biocatalysts (1–3). Although designed proteins that model the structure of native enzymes have been known for a while (4–10), successful designs of proteins that mimic both the structure and function of native enzymes have been reported only recently (11–16). While being able to design such functional proteins is laudable, the impact of such an achievement would be greater if the designed proteins can be used to address fundamental issues in chemistry and biol- ogy that are difficult to tackle by other methods. One primary example is the roles of conserved glutamates and metal ions in bacterial nitric oxide reductase (NOR) (17–19). NO is critical for all life (20). Bacterial denitrification is a cru- cial part of the nitrogen cycle in nature that involves a four-step, five-electron reduction of nitrate (NO3 −) to dinitrogen (N2) (17, 19). Bacterial NOR is a membrane-bound protein that catalyzes one step of this process, namely, the two-electron reduc- tion of NO to N2O (17, 19). With no crystal or solution structure available for bacterial NOR to date, sequence alignments and homology modeling (21, 22) have indicated that NOR is structu- rally homologous to the largest subunit (subunit I) of heme- copper oxidases (HCOs) (23), enzymes that catalyze reduction of O2 to water. The active sites of both NOR and HCO contain a proximal histidine-coordinated heme and a distal three histi- dine-coordinated metal center. However, the metal center in HCOs is occupied by a copper (called CuB), whereas a nonheme iron is present in NOR (called FeB) (23, 24). In addition, two conserved glutamates, shown by modeling to be close to the FeB site (21, 22), are found to be essential for NOR activity (24, 25). Some members of HCOs such as cytochrome cbb3 oxidase display NOR activity (26–28), although the activity is ∼50-fold lower than native NOR (26). Therefore, it is important to elucidate the structural features, specifically the roles of the conserved glutamates close to the FeB site and metal ions (copper vs. iron), responsible for the reduction of NO to N2O. To address these issues, biochemical and biophysical studies of native NOR and its variants have been carried out (24, 25, 29–37). For example, Richardson and coworkers investigated the effects of amino acid substitutions of the five conserved glutamates (E122 and E125 presumed to face the periplasm and E198, E202, and E267 located in the interior of the membrane, close to the catalytic site) in the catalytic subunit of Paracoccus deni- trificans, NorB. The E122A, E125A, E198A, and E267A variants were inactive, indicating that these four glutamates are crucial for NOR activity (24, 25, 32, 33). On the other hand, Reimann et al. constructed a 3D model of NorB using homology modeling with the structures of HCOs as templates and suggested a plausible pathway consisting of these conserved glutamates for proton delivery (22). Despite these successes, the roles of the conserved glutamates and metal ions still remain to be fully elucidated, partly because of the difficulty in obtaining native NOR in high yield and the lack of a 3D structure. Even if these problems are resolved, it is still difficult to replace iron in the native FeB site with other metal ions, and spectroscopic studies of native NOR are often complicated by the presence of other metal cofactors (e.g., low-spin heme). To overcome these limitations, a number of synthetic models of NOR using small organic molecules as ligands, have been made in which the nonheme FeB site can be replaced by a copper ion (17, 38–45). In addition, since these model systems lack addi- tional metal-binding sites, spectroscopic studies are often simpli- fied. Therefore, studies of these synthetic models have offered many insights. For example, Collman et al. showed that a fully reduced heme/nonheme FeB compound can react with two equivalents of NO leading to the formation of one equivalent of N2O and a bis-ferric product (41). On the other hand, Karlin and coworkers showed that a small heme/Cu complex can effi- ciently lead to reductive coupling of NO to N2O (43). However, it is also difficult to obtain the synthetic models in high yield due to the multiple steps required in chemical synthesis. Because of this limitation, no synthetic NOR model containing the two key conserved Glu residues (E198 and E267 in NOR) has been Author contributions: Y.-W.L., N.Y., and Y.L. designed research; Y.-W.L., N.Y., Y.-G.G., K.D.M., and S.T. performed research; Y.-W.L., N.Y., Y.-G.G., K.D.M., H.R., and Y.L. analyzed data; and Y.-W.L. and Y.L. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Data deposition: The atomic coordinates and structure factors have been deposited in the Protein Data Bank, www.pdb.org (PDB ID codes 3M38, 3M39, 3M3A, and 3M3B). 1To whom correspondence should be addressed. E-mail: yi-lu@illinois.edu. 2Present address: School of Chemistry and Chemical Engineering, University of South China, Hengyang 421001, China. This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1000526107/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1000526107 PNAS ∣May 11, 2010 ∣vol. 107 ∣no. 19 ∣8581–8586 BIOCHEMISTRY CHEMISTRY reported. It is also difficult to substitute different metal ions in the same metal-binding site without perturbing the site geometry and distances to the heme iron, as most ligands are not as rigid as those in native enzymes and different metal ions have different geometric and ligand donor set preferences. We have recently designed a structural and functional protein model of bacterial NOR by engineering three histidines and one glutamate into the distal pocket of sperm whale myoglobin (swMb, L29H, F43H, H64, and V68E, named FeBMb) (14). Like synthetic models, this “bottom-up” approach complements the “top-down”approachofthestudyofnativeNORinthatitprovides insights into whether certain “necessary” structural elements are enough to impart enzyme function. Thanks in part to recent advances in computational, molecular, and structural biology, the designed myoglobin protein model is much easier to synthesize and to crystallize than either native NOR or synthetic models. Since myoglobin has often been used for the development and calibration of numerous spectroscopic techniques (46–48), it is an ideal choice for spectroscopic studies. More importantly, the rigid protein network allows precise placement of either glutamate or metal ions in myoglobin to address their roles in NOR activity. Toward this goal, we have demonstrated that both the histidines and one of the glutamates are essential for iron binding and NO reduction activity (14). However, the role of the second Glu close to the FeB site and the role of different metal ions in the FeB site have not been addressed. To address these important issues and to design even closer protein models of NOR, we introduced herein the second Glu to the second coordination sphere of the FeB site by mutating an Ile to a Glu (named I107E FeBMb). We show that the second Glu results in a ∼100% increase in NOR activity through hydro- gen bonding interactions and that the two glutamates have dramatically different effects on the heme reduction potential. Additionally, by comparing the EPR, electrochemistry, X-ray structures, and NOR activity of iron, copper, and zinc derivatives of the designed protein, we have obtained deeper insights into the roles of metal ions in NOR. Results Structure and Function of Fe(II)-I107E FeBMb. The X-ray crystal struc- tures of heme-containing I107E FeBMb without metal ion in the FeB site and with Fe2þ in the FeB site are solved at 1.42-Å and 1.65-Å resolution, respectively (Fig. 1 A and B and Table S1). In the absence of metal ions in the FeB site, the structure shows a water molecule in the FeB site, which forms hydrogen bonds with NE2 atoms of all three His residues, both OE1 and OE2 atoms of E68, and the OE2 atom of E107 (Fig. 1A). Upon binding Fe2þ, the Fe(II)-I107E FeBMb structure shows that Fe2þ is coordinated by three His, the OE2 atom of E68, and one water molecule. Notably, a water molecule bridges Fe2þ in the FeB site and the second glutamate (E107) with a distance of 2.32 Å to the OE2 atom of E107 (Fig. 1B). To probe the conformational changes of introducing the second Glu (E107), we performed a structural alignment of Fe (II)-I107E FeBMb and the previously reported FeðIIÞ-FeBMb (14). The comparison, shown in Fig. 2, indicates that both the polypeptide chain and the active site overlap well with each other. In addition, the two nonheme irons are located at similar posi- tions with a 0.36-Å separation from each other. In contrast, E68 underwent a significant conformational rearrangement in the presence of E107. These observations suggest that the active site of FeBMb can be tuned by the formation of an extended hydrogen bonding network, resulting from the introduction of a second glutamate residue. The binding of Fe2þ to deoxy I107E FeBMb was further mon- itored by EPR (Fig. 3A). Since deoxy myoglobin contains Fe(II) heme that exhibits no EPR signals in X-band EPR (14), we added blue copper Cu(II)-azurin (49), a redox partner of native NOR (19), to oxidize both the reduced heme and nonheme irons in Fe (II)-I107E FeBMb to EPR-active Fe(III). Upon addition of Cu (II)-azurin, the oxidation of deoxy I107E FeBMb resulted in EPR signals at g ¼ 6.12 and 5.56, typical of a high-spin heme- Fe(III). Upon addition of Fe2þ, however, a decrease of the heme-Fe(III) EPR signals was observed, indicating that the Fe2þ, when bound to the FeB site and oxidized by Cu(II)-azurin, is spin-coupled to heme-Fe(III). Such a spin coupling mimics that in NOR (35, 50–53), suggesting that I107E FeBMb models NOR closely, at least in this respect. To probe the role of the second Glu (E107) in NO reduction activity, we measured the yield of N2O production by Fe(II)- I107E FeBMb with excess NO under one turnover conditions. We monitored N2O formation in the headspace of the solution using GC/MS and compared this result to that of FeðIIÞ-FeBMb, which lacks the second Glu (Fig. 4). Remarkably, Fe(II)-I107E FeBMb displays higher activity than FeðIIÞ-FeBMb. After ∼20 hr, ∼24% N2O was produced by Fe(II)-I107E FeBMb, in H29 A H29 B H64 E68 E107 3.04 2.81 3.02 2.20 2.24 2.24 2.12 2.21 H64 E68 E107 3.03 H43 2.15 3.41 3.16 2.62 2.32 H43 2.26 H93 H93 H29 C H29 D 2.04 2.03 H64 E68 E107 2.09 2.10 2.91 2.21 H64 E68 E107 2.26 2.29 2.10 2.18 2.10 4.47 H43 2.07 3.04 H43 2.68 H93 H93 Fig. 1. Crystal structures of I107E FeBMb (A) (PDB ID code 3M38), Fe(II)-I107E FeBMb (B) (PDB ID code 3M39), Cu(II)-I107E FeBMb (C) (PDB ID code 3M3A), and Zn(II)-I107E FeBMb (D) (PDB ID code 3M3B). Water molecules, Fe(II), Cu(II), and Zn(II) are represented by red, green, orange, and gray spheres, respectively. Fig. 2. Overlay of Fe(II)-I107E FeBMb (cyan) (PDB ID code 3M39) with FeðIIÞ-FeBMb (orange) (PDB ID code 3K9Z). 8582 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1000526107 Lin et al. contrast to ∼10% yield for FeðIIÞ-FeBMb, strongly indicating that the second Glu plays an important role in NO reduction, likely facilitating proton uptake during NO reduction. Other Metal Ions Binding to I107E FeBMb. To find out if the resting state of the protein, i.e., oxidized or met I107E FeBMb, can bind other metal ions, Cu2þ or Zn2þ was titrated into met I107E FeBMb and monitored by EPR spectroscopy (Fig. 3 B and C). In the absence of metal ions, met I107E FeBMb exhibited high-spin heme signals at g ¼ 6.03, 5.08, and 1.98 (Fig. 3B, black line). Upon addition of 2 eq of Cu2þ, the signals at g ¼ 6.03 and 5.08 decreased and a broad peak around g ¼ 2.95 increased, probably due to spin coupling between heme-Fe(III) and Cu2þ in the FeB site. In contrast, addition of Zn2þ, a metal ion with no unpaired electrons [i.e., incapable of spin coupling to heme- Fe(III)], produced an increase in the high-spin heme signals at g ¼ 5.88 and 5.60 (Fig. 3C), indicating that the interaction be- tween E68 and heme iron was weakened after metal binding. The X-ray crystal structures of I107E FeBMb with Cu2þ or Zn2þ in the FeB site were solved at 1.37-Å and 1.60-Å resolution, respectively (Fig. 1 C and D and Table S1). Compared to Fe(II)- I107E FeBMb (Fig. 1B), a similar binding site was observed for Cu(II)-I107E FeBMb (Fig. 1C), where H29, H43, and H64 coordinate to Cu2þ with distances of 2.09, 2.10, and 2.04 Å, respectively, slightly shorter than the corresponding distances in the Fe2þ structure. In comparison to Fe(II)-I107E FeBMb, the water bridging the Cu2þ and the second Glu (E107) is shifted toward Cu2þ in the FeB site (2.03 Å) with respect to E107 (3.04 Å). Interestingly, this bridging water molecule was not observed in Zn(II)-I107E FeBMb (Fig. 1D), but the two O atoms of E68 coordinate to Zn2þ with similar distances (2.26 Å for OE1 and 2.29 Å for OE2). The longer distance between OE1 of E68 B 6.12 I107E FeBMb + Azurin I107E Fe Mb + 0 5 eq Fe 2+ + Azurin A 5.56 I107E FeBMb + 0.5 eq Fe + Azurin I107E FeBMb + 1.0 eq Fe 2+ + Azurin I107E FeBMb + 2.0 eq Fe 2+ + Azurin 1.98 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5.08 5.38 Magnetic Field (Gauss) 5.88 I107E FeBMb I107E FeBMb + 0.5 eq Zn 2+ I107E Fe Mb 1 0 eq Zn 2+ C 6.03 I107E FeBMb I107E FeBMb + 0.5 eq Cu 2+ I107E Fe Mb + 1 0 eq Cu 2+ I107E FeBMb + 1.0 eq Zn I107E FeBMb + 2.0 eq Zn 2+ I107E FeBMb + 1.0 eq Cu I107E FeBMb + 2.0 eq Cu 2+ 1.98 ~2.95 1.98 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5.60 Magnetic Field (Gauss) 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5.08 Magnetic Field (Gauss) Fig. 3. EPR spectra of deoxy I107E FeBMb (0.5 mM protein in 50 mM Bis-Tris, pH 7.0) with increasing concentrations of Fe2þ in the presence of wild-type Cu(II)-azurin (A), and oxidized I107E FeBMb with Cu2þ (B) or Zn2þ (C). Spectra were collected at 4 K, 5 mW power, and 9.05 GHz. 30 25 Fe(II)-I107E FeBMb Fe(II)-FeBMb 15 20 10 %) oduction (% N2O pro 0 5 0 4 8 12 16 20 Incubation time (hr) Fig. 4. Time-dependent N2O production by Fe(II)-I107E FeBMb (▴) and FeðIIÞ-FeBMb (●) with ∼50 eq. NO under single turnover conditions. The yield was determined by a comparison of the ratio of NO∶N2O peaks from the GC/MS chromatograms. Lin et al. PNAS ∣ May 11, 2010 ∣ vol. 107 ∣ no. 19 ∣ 8583 BIOCHEMISTRY CHEMISTRY and heme iron in the Zn-bound structure (2.68 Å) in comparison to the Cu- and Fe-bound structures, is likely the result of a weaker interaction, which is also supported by an observed increase of the high-spin heme signals in the EPR spectra upon Zn2þ binding (Fig. 3C). These results suggest I107E FeBMb is capable of incor- porating different metal ions into its designed FeB site, offering an excellent opportunity to compare the role of these metal ions in the same protein scaffold. Effect of Glutamates and Metal Ions on the Redox Potential of I107E FeBMb. Since EPR and X-ray structural studies indicate metal binding to I107E FeBMb, we used spectroelectrochemistry to measure the effects of glutamates and metal ions on the heme reduction potential. When there is no metal ion in the FeB site, the I107E FeBMb displays a reduction potential of −134  3 mV vs. the normal hydrogen electrode (NHE) (Fig. S1A), similar to that of FeBMb (−158  4 mV) without the I107E mutation (14). In the presence of Cu2þ, I107E FeBMb has a reduction potential (−137  2 mV) (Fig. S1B) almost identical to that of the same protein in the absence of metal ions in the FeB site, indicating that copper binding to the FeB site has little effect on the reduction potential of the heme iron. This observation is similar to that observed for Cu2þ binding to CuBMb (CuðIIÞ-CuBMb, 80 mV vs. CuBMb, 77 mV) (54). On the other hand, the presence of Fe2þ and Zn2þ increased the reduction potential of I107E FeBMb from −134  3 mV to −64  3 mV vs. NHE (Fig. S1C) and −105  2 mV vs. NHE (Fig. S1D), respectively. The different effects of Cu2þ, Fe2þ, and Zn2þ on the reduction potential of I107E-FeBMb indicate that these metal ions in the FeB site may play different roles through different coordination properties. NOR Activity of I107E FeBMb in the Presence of Different Metal Ions. The NO reduction activity of I107E FeBMb in the presence of Fe2þ, Cuþ, or Zn2þ was monitored by GC/MS under single turn- over conditions. When Fe(II)-I107E FeBMb was exposed to excess NO, N2O could be observed to form with increased yield over time (Fig. S2). Similarly, N2O formation was observed for Cu(I)-I107E FeBMb, indicating that Fe or Cu binding to the FeB site results in comparable NOR activities. It should be noted that because of the high solubility of N2O (∼25 mM in water at room temperature), GC/MS cannot be used to quantify the rates of NO reduction under these conditions. In contrast, no N2O formation was observed with redox inactive Zn2þ, which demon- strates that redox active Fe2þ or Cuþ in the FeB site plays a crucial role in NO reduction. To gain deeper insight into the process of NO reduction, EPR studies were further performed to monitor the initial process of NO reduction. In the absence of metal ions, the EPR spectrum of ferrous I107E FeBMb-NO shows hyperfine splitting resulting from bound NO and the proximal histidine, indicating the forma- tion of a six-coordinate ferrous heme-NO species (Fig. 5, top line). After incubation of Fe(II)-I107E FeBMb with excess NO, a distinct three-line hyperfine structure appears at 15 min (Fig. 5A), suggesting the formation of a five-coordinate ferrous heme-NO species as a result of cleavage of the proximal His-Fe heme bond (55). A three-line hyperfine structure was also observed for Cuþ and Zn2þ, except that the signal intensity is low- er with Cu(I)-I107E FeBMb-NO (Fig. 5B) and more pronounced in Zn(II)-I107E FeBMb-NO (Fig. 5C). The lower intensity of the three-line hyperfine structure for Cu(I)-I107E FeBMb-NO suggests the major species formed is a six-coordinate ferric heme-NO complex, which is EPR silent (41). These differences further suggest that the metal ion in the FeB site plays a key role in formation of the intermediates, thereby tuning NOR activity. Discussion Using Rationally Designed Proteins to Address Important Issues in Chemistry and Biology. Important issues such as the roles of the conserved glutamates and nonheme FeB in NOR have been previously addressed using biochemical and biophysical studies or biomimetic modeling (24, 25, 27–37, 45, 56, 57). As a comple- mentary approach, rational protein design, using small, easy-to- produce and well-characterized proteins such as myoglobin, offers a powerful method with which to gain insights into more complex native enzymes such as NOR (14). Similar to synthetic models (41, 43), the metal ion at the putative FeB site in the protein model can be substituted freely. Better yet, Glu residues can be placed at precise locations in the protein, including the secondary coordination sphere, due to its rigid network. By care- fully choosing a suitable protein template, rational protein design could be generally applied to address other important issues in chemistry and biology. The Roles of Glutamates. Although two conserved glutamates (E198 and E267) are known to be crucial for NOR activity (24, 25), their roles are not well defined (18, 19). In a previous study (14), we demonstrated that one Glu, E68, is important for both iron binding and NOR activity of FeBMb. The crystal struc- tures of both FeðIIÞ-FeBMb and Fe(II)-I107E FeBMb show that one O atom of E68 directly coordinates to FeB (Fig. 2). In syn- thetic models of NOR, it has also been found that the presence of a glutamic acid mimic significantly increases the stability of iron binding to the FeB site (40). Furthermore, a theoretical study by Blomberg et al. (58) showed that a model with an FeB coordi- nated by three histidines, one glutamate, and one water molecule provides an energetically feasible reaction mechanism of NO reduction. However, the structural model of NOR constructed recently by Reimann et al. (22) shows that the closest conserved Glu (E267) still has its carboxylate O atom 7 Å away from FeB, which suggests that Glu may not bind to FeB in native NOR. One interesting finding from our study is that the Glu (E68) under- went a significant conformational rearrangement in the presence of another Glu (E107) (Fig. 2). Therefore, the FeBMb provides a viable model of NOR that is consistent with Blomberg’s model, but cannot rule out Reimann’s model due to possible conforma- tion changes. While the role of the first Glu is still uncertain until a 3D struc- ture of NOR in its active form is available, the role of the second Glu is even less defined. We address this question by introducing a second Glu (E107) to FeBMb. The crystal structures shown in Fig. 1 indicate that E107 interacts with a water molecule and forms a hydrogen bonding network in both Fe(II)-I107E FeBMb and Cu(II)-I107E FeBMb. Interestingly, although a similar water molecule was observed in the active site of FeðIIÞ- FeBMb (Fig. 2), activity assay data indicate that the presence of E107 in Fe(II)-I107E FeBMb increases NOR activity by ∼100% C 5 min N t l A 5 min B 2+ 5 min 1 min No metal 1 min 5 min F 2+ No metal 1 i 5 min + No metal 2+ Zn 2+ 5 min Fe 2+ 5 min 1 min Fe C + 5 min 1 min Cu + Z 2 Zn 2+ 15 min F 2+ Fe 15 min 5 min C + Cu + 15 min 5 min Zn 2+ 15 min Fe2+ 5 Cu + 15 min 3000 3100 3200 3300 3400 Magnetic Field (Gauss) 3000 3100 3200 3300 3400 Magnetic Field (Gauss) 3000 3100 3200 3300 3400 Magnetic Field (Gauss) Fig. 5. EPR spectra of deoxy I107E FeBMb (0.5 mM) with no metal bound in the presence of NO after 5 min (top line), with 2 eq Fe2þ (A), Cuþ (B), or Zn2þ (C) incubated with excess NO (∼200 eq) for 1, 5, and 15 min. Spectra were collected in 50 mM Bis-Tris pH 7.0 at 30 K, 0.2 mW power, and 9.05 GHz. 8584 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1000526107 Lin et al. (Fig. 4), suggesting that the second Glu may potentionally play a role in providing one of the protons during reduction of NO to N2O. Although free Glu outside a protein has a pKa ∼4.3, the stu- dies of native NOR showed that the pKa of its Glu close to the active site has a value of ∼6.6 (22, 25, 33). The hydrogen binding network in our protein models may contribute to the fine-tuning of the Glu pKa to be more neutral, similar to that in NOR. More- over, it is interesting that Cu(I)-I107E FeBMb also shows NOR activity, which provides an interesting protein model of HCOs with NOR function (26–28), even though Glu residues are not conserved in native HCOs. Additionally, spectroelectrochemical studies showed that the reduction potential of I107E FeBMb with no metal ion in the FeB site is similar to that of FeBMb, but much lower than that of CuBMb (77 mV) (54), which contains the same three His, but no Glu in the metal-binding site above the heme. Since both FeBMb and I107E FeBMb contain the V68E mutation that has been shown to decrease the heme reduction potential of native myoglobin from 59 mV to −137 mV (59), it is likely that the introduction of a negatively charged Glu close to the heme group is what is responsible for the dramatically lower heme redox potential. A conserved glutamate, predicted to be located near the catalytic heme b3 in NOR, was proposed to be responsible for a ∼260-mV decrease in reduction potential (60 mV) in comparison to the other two heme centers, heme b (345 mV) and heme c (310 mV) (60). Our FeBMb and I107E FeBMb mod- els mimic this feature of NOR. Notably, although introduction of the first Glu (E68) lowered the heme potential by ∼110 mV (14), introduction of the second Glu via the I107E mutation did not result in a significant difference in the heme reduction potential, suggesting that the effect of the two conserved Glu residues in NOR on heme reduction potential is not additive, with the effects highly dependent on the location of the Glu. The Roles of Metal Ions. The roles of metal ions in NOR are an- other important question as iron is found in the native FeB site and HCO employs copper at the corresponding CuB site. With different metal ions in the FeB site, the crystal structures clearly show the heme and nonheme dinuclear center existing in differ- ent local environments (Fig. 1). Although a similar hydrogen bond network is formed in both Fe(II)-I107E FeBMb and Cu (II)-I107E FeBMb, the conformation of E68 and E107 with re- spect to the nonheme metal center and heme iron is different from each other. Moreover, the coordination geometry differs significantly with Zn2þ in the FeB site. A hydrogen bond is absent from the Zn crystal structure, but both the O atoms of E68 act as metal-binding ligands. These observations demonstrate that the identity of the metal ion in the FeB site can tune the active site through their interactions with the His and Glu ligands, resulting in formation of different coordination geometries with different hydrogen bonds. In addition to structural fine-tuning, the metal ion at the FeB site can also tune the heme iron reduction potential in I107E FeBMb. Spectroelectrochemical studies showed that the binding of Fe2þ or Zn2þ results in an increase in the heme reduction potential by ∼70 mV and ∼30 mV, respectively (Fig. S1). In the case of Cu(II)-I107E FeBMb, the crystal structure shows that OE1 of E68 is closer to the heme iron (2.07 Å) (Fig. 1C) than its metal-free form (2.15 Å) (Fig. 1A). The stronger interaction from the negatively charged E68 could offset the effect of positively charged Cu2þ binding, resulting in similar reduction potentials observed for Cu(II)-I107E FeBMb and I107E FeBMb. In a previous study (61), EPR data showed that during NO reduction, the binding of Cuþ to the CuB site of CuBMb can weaken the proximal heme Fe-His bond, while complete cleavage of the heme Fe-His bond occurred when Zn2þ was bound to CuBMb-NO. In this study, we observed that a five-coordinate heme-NO species was formed with Fe2þ, Cuþ, or Zn2þ bound to the FeB site of I107E FeBMb (Fig. 5). Significantly, a five- coordinate heme-NO species has also been observed for both NOR (30, 31, 35) and the member of the HCO family with the highest NO reduction activity, cytochrome cbb3 oxidases (26, 62). However, this species was not observed for FeðIIÞ-FeBMb-NO and CuðIÞ-FeBMb-NO, which lack the second Glu (E107). In both these cases, the proximal heme Fe-His bond was only weakened, as indicated by a decrease of the nine-line hyperfine splitting signals in the EPR spectra (Fig. S3). These observations suggest that formation of a five-coordinate heme-NO species may play an important role in NOR reactivity. Conclusions We have successfully designed a structural and functional model of NOR, by introducing a second glutamate in the vicinity of the FeB site, named I107E FeBMb. This protein model mimics native NOR more closely by bearing the structural feature of three his- tidines and two glutamates in the FeB site, as predicted for native NOR. We have demonstrated that the two glutamates can play different roles in NO reduction activity; namely, one acts as a li- gand to FeB (E68), and the other acts as a proton transfer group (E107). Furthermore, by substituting different metal ions into the nonheme metal site, we have demonstrated that FeB plays crucial roles in fine-tuning the active site by donating electrons and by mediating the formation of a five-coordinate heme-NO inter- mediate during NO reduction. In the absence of a crystal struc- ture for native NOR, this study offers an ideal protein model and provides valuable structural as well as mechanistic information for native NOR. Materials and Methods Protein Preparation. I107E FeBMb (swMb L29H/F43H/V68E/I107E) was con- structed, expressed, and purified using the procedure described previously (14). The purity and identity were confirmed by SDS-PAGE and electrospray ionized MS: observed: 17; 392  1 Da; calculated: 17,391 Da. EPR Spectroscopy. EPR spectra were recorded on a Bruker ESP 300 equipped with an Oxford liquid helium cryostat and an ITC4 temperature controller. The samples of met I107E FeBMb, Cu(II)-, or Zn(II)-I107E FeBMb were pre- pared as described previously (14). The samples of NO-bound deoxy I107E FeBMb, Cu(I)-, Fe(II)-, or Zn(II)-I107E FeBMb were prepared by inject- ing 1 mL of purified NO gas into the EPR tube containing 300 μL of protein (0.5 mM, 10% glycerol, in 50 mM Bis-Tris pH 7.0). The samples were then flash frozen in liquid N2 after incubation for 1, 5, or 15 min. The molar extinction coefficient of the Soret band of I107E FeBMb at 406 nm (175 mM−1 · cm−1), calculated using the standard hemochromagen method (63), was used to determine protein concentration. The metal sources of Cu(I), Cu(II), Zn(II), and Fe(II) were ½ðCH3CNÞ4CuPF6, CuSO4, ZnSO4 · 7H2O, and FeCl2, respectively. Spectroelectrochemical Measurements. Protein reduction potentials were measured using an optically transparent thin layer electrode as previously described (64). The potential of the working electrode was applied in the positive direction for deoxy I107E FeBMb with Fe2þ and in the negative direction for metal free and with Cu2þ or Zn2þ. Other procedures are the same as described previously (54). X-Ray Crystallographic Studies. Fe(II)-I107E FeBMb was crystallized anaerobi- cally in a glove box at room temperature using the conditions described for FeðIIÞ-FeBMb (14). I107E FeBMb, Cu(II)-, or Zn(II)-I107E FeBMb were crystallized aerobically. Diffraction-quality crystals were soaked in a cryoprotectant solution of 30% PEG 400 and flash frozen in liquid nitrogen. Diffraction data were collected at the Brookhaven National Lab Synchrotron Light Source X12C beamline. The crystal structure was solved using the same method as for FeðIIÞ-FeBMb (14). NOR Activity Assay. NO reduction was monitored by GC/MS. The protein was reduced to the deoxy form by excess dithionite that was removed with a size-exclusion column (PD-10). Then 2 eq metal, Cu(I), Fe(II), or Zn(II), was added to the protein solution (0.6 mM, 3 mL in 50 mM Bis-Tris buffer, pH 7.0). The samples were prepared anaerobically in a glove box. Purified NO Lin et al. PNAS ∣ May 11, 2010 ∣ vol. 107 ∣ no. 19 ∣ 8585 BIOCHEMISTRY CHEMISTRY gas was injected into the head space of the reaction flask with the molar ratio of NO∶protein ≈50∶1. Other procedures are the same as described previously (14, 61). ACKNOWLEDGMENTS. We thank Dr. Mark J. Nilges for help with EPR analysis, and Furong Sun and Beth D. Eves for aiding in GC/MS data collection. This work was supported by NIH Grant GM062211. 1. Lu Y, Yeung N, Sieracki N, Marshall NM (2009) Design of functional metalloproteins. Nature 460:855–862. 2. Nanda V, Koder RL (2010) Designing artificial enzymes by intuition and computation. Nat Chem 2:15–24. 3. Gray HB (2003) Biological inorganic chemistry at the beginning of the 21st century. Proc Natl Acad Sci USA 100:3563–3568. 4. Regan L, DeGrado WF (1988) Characterization of a helical protein designed from first principles. Science 241:976–978. 5. Hecht MH, Richardson JS, Richardson DC, Ogden RC (1990) De novo design, expression, and characterization of Felix: A four-helix bundle protein of native-like sequence. Science 249:884–891. 6. Robertson DE, et al. (1994) Design and synthesis of multi-heme proteins. Nature 368:425–432. 7. Lu Y, Valentine JS (1997) Engineering metal-binding sites in proteins. Curr Opin Struct Biol 7:495–500. 8. Reedy CJ, Gibney BR (2004) Heme protein assemblies. Chem Rev 104:617–649. 9. Watanabe Y, Hayashi T (2005) Functionalization of myoglobin. Prog Inorg Chem 54:449–493. 10. Ghosh D, Pecoraro VL (2005) Probing metal-protein interactions using a de novo design approach. Curr Opin Chem Biol 9:97–103. 11. Kaplan J, DeGrado WF (2004) De novo design of catalytic proteins. Proc Natl Acad Sci USA 101:11566–11570. 12. Kang SG, Saven JG (2007) Computational protein design: Structure, function and combinatorial diversity. Curr Opin Chem Biol 11:329–334. 13. Jiang L, et al. (2008) De novo computational design of retro-aldol enzymes. Science 319:1387–1391. 14. Yeung N, et al. (2009) Rational design of a structural and functional nitric oxide reductase. Nature 462:1079–1082. 15. Koder RL, et al. (2009) Design and engineering of an O2 transport protein. Nature 458:305–309. 16. Heinisch T, Ward TR (2010) Design strategies for the creation of artifical metalloen- zymes. Curr Opin Chem Biol 14:184–199. 17. Wasser IM, de Vries S, Moënne-Loccoz P, Schröder I, Karlin KD (2002) Nitric oxide in biological denitrification: Fe/Cu metalloenzyme and metal complex NOx redox chemistry. Chem Rev 102:1201–1234. 18. Yeung N, Lu Y (2008) One heme, diverse functions: Using biosynthetic myoglobin models to gain insights into heme-copper oxidases and nitric oxide reductases. Chem Biodivers 5:1437–1454. 19. Watmough NJ, Field SJ, Hughes RJL, Richardson DJ (2009) The bacterial respiratory nitric oxide reductase. Biochem Soc Trans 37:392–399. 20. Cary SPL, Winger JA, Derbyshire ER, Marletta MA (2006) Nitric oxide signaling: No longer simply on or off. Trends Biochem Sci 31:231–239. 21. Sakurai N, Sakurai T (1998) Genomic DNA cloning of the region encoding nitric oxide reductase in paracoccus halodenitrificans and a structure model relevant to cytochrome oxidase. Biochem Biophys Res Commun 243:400–406. 22. Reimann J, Flock U, Lepp H, Honigmann A, Ädelroth P (2007) A pathway for protons in nitric oxide reductase from paracoccus denitrificans. Biochim Biophys Acta 1767:362–373. 23. Pereira MM, Sousa FL, Verissimo AF, Teixeira M (2008) Looking for the minimum common denominator in haem-copper oxygen reductases: Towards a unified catalytic mechanism. Biochim Biophys Acta 1777:929–934. 24. Butland G, Spiro S, Watmough NJ, Richardson DJ (2001) Two conserved glutamates in the bacterial nitric oxide reductase are essential for activity but not assembly of the enzyme. J Bacteriol 183:189–199. 25. Flock U, Lachmann P, Reimann J, Watmough NJ, Äedelroth P (2009) Exploring the terminal region of the proton pathway in the bacterial nitric oxide reductase. J Inorg Biochem 103:845–850. 26. Forte E, et al. (2001) The cytochrome cbb3 from pseudomonas stutzeri displays nitric oxide reductase activity. Eur J Biochem 268:6486–6490. 27. Huang Y, Reimann J, Lepp H, Drici N, Ädelroth P (2008) Vectorial proton transfer coupled to reduction of O2 and NO by a heme-copper oxidase. Proc Natl Acad Sci USA 105:20257–20262. 28. Hayashi T, et al. (2009) Accommodation of two diatomic molecules in cytochrome bo3: Insights into NO reductase activity in terminal oxidases. Biochemistry 48:883–890. 29. Moënne-Loccoz P (2007) Spectroscopic characterization of heme iron-nitrosyl species and their role in NO reductase mechanisms in diiron proteins. Nat Prod Rep 24:610–620. 30. Moënne-Loccoz P, de Vries S (1998) Structural characterization of the catalytic high-spin heme b of nitric oxide reductase: A resonance raman study. J Am Chem Soc 120:5147–5152. 31. Sakurai T, Sakurai N, Matsumoto H, Hirota S, Yamauchi O (1998) Roles of four iron centers in paracoccus halodenitrificans nitric oxide reductase. Biochem Biophys Res Commun 251:248–251. 32. Thorndycroft FH, Butland G, Richardson DJ, Watmough NJ (2007) A new assay for nitric oxide reductase reveals two conserved glutamate residues form the entrance to a proton-conducting channel in the bacterial enzyme. Biochem J 401:111–119. 33. Flock U, et al. (2008) Defining the proton entry point in the bacterial respiratory nitric-oxide reductase. J Biol Chem 283:3839–3845. 34. Hendriks JHM, Jasaitis A, Saraste M, Verkhovsky MI (2002) Proton and electron pathways in the bacterial nitric oxide reductase. Biochemistry 41:2331–2340. 35. Kumita H, et al. (2004) NO reduction by nitric-oxide reductase from denitrifying bacterium pseudomonas aeruginosa: Characterization of reaction intermediates that appear in the single turnover cycle. J Biol Chem 279:55247–55254. 36. Pinakoulaki E, Varotsis C (2008) Resonance raman spectroscopy of nitric oxide reductase and cbb3 heme-copper oxidase. J Phys Chem B 112:1851–1857. 37. Kapetanaki SM, et al. (2008) Ultrafast ligand binding dynamics in the active site of native bacterial nitric oxide reductase. Biochim Biophys Acta 1777:919–924. 38. Wasser IM, et al. (2004) Synthesis and spectroscopy of m-oxo (O2-)-bridged Heme/ Non-heme diiron complexes: Models for the active site of nitric oxide reductase. Inorg Chem 43:651–662. 39. Wasser IM, Huang H, Moënne-Loccoz P, Karlin KD (2005) Heme/Non-heme diiron(II) complexes and O2, CO, and NO adducts as reduced and substrate-bound models for the active site of bacterial nitric oxide reductase. J Am Chem Soc 127:3310–3320. 40. Collman JP, Yan Y, Lei J, Dinolfo PH (2006) Active-site models of bacterial nitric oxide reductase featuring tris-histidyl and glutamic acid mimics: Influence of a carboxylate ligand on FeB binding and the heme Fe∕FeB redox potential. Inorg Chem 45:7581–7583. 41. Collman JP, et al. (2008) A functional nitric oxide reductase model. Proc Natl Acad Sci USA 105:15660–15665. 42. Collman JP, et al. (2008) Intermediates involved in the two electron reduction of NO to N2O by a functional synthetic model of heme containing bacterial NO reductase. J Am Chem Soc 130:16498–16499. 43. Wang J, Schopfer MP, Sarjeant Amy AN, Karlin KD (2009) Heme-copper assembly mediated reductive coupling of nitrogen monoxide (*NO). J Am Chem Soc 131:450–451. 44. Liu J, Naruta Y, Tani F (2005) A functional model of the cytochrome c oxidase active site: Unique conversion of a heme-m-peroxo-CuII intermediate into heme-superoxo/ CuI. Angew Chem Int Edit 44:1836–1840. 45. Ghiladi RA, et al. (2005) Heme-copper/dioxygen adduct formation relevant to cytochrome c oxidase: Spectroscopic characterization of ½ð6LÞFeIII-ðO2 2−Þ-CuIIIþ. J Biol Inorg Chem 10:63–77. 46. Sage JT (1997) Myoglobin and CO: Structure, energetics, and disorder. J Biol Inorg Chem 2:537–543. 47. Sigman JA, Kim HK, Zhao X, Carey JR, Lu Y (2003) The role of copper and protons in heme-copper oxidases: Kinetic study of an engineered heme-copper center in myoglobin. Proc Natl Acad Sci USA 100:3629–3634. 48. Davydov R, Hoffman BM (2008) EPR and ENDOR studies of fe(II) hemoproteins reduced and oxidized at 77 K. J Biol Inorg Chem 13:357–369. 49. Marshall NM, et al. (2009) Rationally tuning the reduction potential of a single cupredoxin beyond the natural range. Nature 462:113–116. 50. Girsch P, de Vries S (1997) Purification and initial kinetic and spectroscopic character- ization of NO reductase from paracoccus denitrificans. Biochim Biophys Acta 1318:202–216. 51. Sakurai N, Sakurai T (1997) Isolation and characterization of nitric oxide reductase from paracoccus halodenitrificans. Biochemistry 36:13809–13815. 52. Hendriks J, et al. (1998) The active site of the bacterial nitric oxide reductase is a dinuclear iron center. Biochemistry 37:13102–13109. 53. Moënne-Loccoz P, et al. (2000) Nitric oxide reductase from paracoccus denitrificans contains an oxo-bridged Heme/Non-heme diiron center. J Am Chem Soc 122:9344–9345. 54. Zhao X, Yeung N, Wang Z, Guo Z, Lu Y (2005) Effects of metal ions in the CuB center on the redox properties of heme in heme-copper oxidases: Spectroelectrochemical stu- dies of an engineered heme-copper center in myoglobin. Biochemistry 44:1210–1214. 55. Decatur SM, et al. (1996) Trans effects in nitric oxide binding to myoglobin cavity mutant H93G. Biochemistry 35:4939–4944. 56. Pervitsky D, Immoos C, van der Veer W, Farmer PJ (2007) Photolysis of the HNO adduct of myoglobin: Transient generation of the aminoxyl radical. J Am Chem Soc 129:9590–9591. 57. Berto TC, Praneeth VKK, Goodrich LE, Lehnert N (2009) Iron-porphyrin NO complexes with covalently attached N-donor ligands: Formation of a stable six-coordinate species in solution. J Am Chem Soc 131:17116–17126. 58. Blomberg LM, Blomberg MRA, Siegbahn PEM (2006) Reduction of nitric oxide in bacterial nitric oxide reductase-a theoretical model study. Biochim Biophys Acta 1757:240–252. 59. Varadarajan R, Zewert TE, Gray HB, Boxer SG (1989) Effects of buried ionizable amino acids on the reduction potential of recombinant myoglobin. Science 243:69–72. 60. Grönberg KLC, et al. (1999) A low-redox potential heme in the dinuclear center of bacterial nitric oxide reductase: Implications for the evolution of energy-conserving heme-copper oxidases. Biochemistry 38:13780–13786. 61. Zhao X, Yeung N, Russell BS, Garner DK, Lu Y (2006) Catalytic reduction of NO to N2O by a designed heme copper center in myoglobin: Implications for the role of metal ions. J Am Chem Soc 128:6766–6767. 62. Pinakoulaki E, Stavrakis S, Urbani A, Varotsis C (2002) Resonance raman detection of a ferrous five-coordinate nitrosylheme b3 complex in cytochrome cbb3 oxidase from pseudomonas stutzeri. J Am Chem Soc 124:9378–9379. 63. Morrison M, Horie S (1965) Determination of heme a concentration in cytochrome preparations by hemochromogen method. Anal Biochem 12:77–82. 64. Taboy CH, Bonaventura C, Crumbliss AL (2002) Anaerobic oxidations of myoglobin and hemoglobin by spectroelectrochemistry. Method Enzymol 353:187–209. 8586 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1000526107 Lin et al.
3M3F
PEPA bound to the ligand binding domain of GluA3 (flop form)
The molecular mechanism of flop-selectivity and subsite recognition for an AMPA receptor allosteric modulator: Structures of GluA2 and GluA3 complexed with PEPA Ahmed H. Ahmed§, Christopher P. Ptak§, and Robert E. Oswald* Department of Molecular Medicine, Cornell University, Ithaca, NY 14853 USA Abstract Glutamate receptors are important potential drug targets for cognitive enhancement and the treatment of schizophrenia in part because they are the most prevalent excitatory neurotransmitter receptors in the vertebrate central nervous system. One approach to the application of therapeutic agents to the AMPA subtype of glutamate receptors is the use of allosteric modulators, which promote dimerization by binding to a dimer interface thereby reducing desensitization and deactivation. AMPA receptors exist in two alternatively spliced variants (flip and flop) that differ in desensitization and receptor activation profiles. Most of the structural information on modulators of the AMPA receptor target the flip subtype. We report here the crystal structure of the flop-selective allosteric modulator, PEPA, bound to the binding domains of the GluA2 and GluA3 flop isoforms of AMPA receptors. Specific hydrogen bonding patterns can explain the preference for the flop isoform. This includes a bidentate hydrogen bonding pattern between PEPA and N754 of the flop isoforms of GluA2 and GluA3 (the corresponding position in the flip isoform is S754). Comparison with other allosteric modulators provides a framework for the development of new allosteric modulators with preferences for either the flip or flop isoforms. In addition to interactions with N/S754, specific interactions of the sulfonamide with conserved residues in the binding site are characteristics of a number of allosteric modulators. These, in combination, with variable interactions with five subsites on the binding surface lead to different stoichiometries, orientations within the binding pockets, and functional outcomes. Membrane receptors are the cell's gatekeepers, allowing chemical signals access to the cell's pathways. Through the binding of endogenous ligands, receptors identify relevant environmental cues and facilitate cell-cell communication. The regulation of membrane receptors has become an important goal of drug discovery efforts (1,2). By targeting the physiological (orthosteric) ligand-binding site, agonists and antagonists control the function of membrane receptors. Unfortunately, exogenously induced agonist-activation at the orthosteric site can cause toxic effects from overstimulation. Allosteric modulator binding sites use a distinct avenue for altering the natural response of a receptor. The ability of some allosteric modulators to enhance receptor stimulation, while not actually providing the trigger for stimulation, is a clear advantage that conserves the endogenous signaling pathway. Being important mediators of higher-order processes such as learning and memory, ionotropic glutamate receptors (iGluRs) have attracted a great deal of interest as allosteric modulator targets (3–6). Of clear therapeutic importance, various neurodegenerative disorders such as Parkinson's and Alzheimer's diseases, Huntington's *Corresponding author; telephone: 1-607-253-3877; fax: 1-607-253-3659; email: reo1@cornell.edu. §These authors contributed equally to this work. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 April 7. Published in final edited form as: Biochemistry. 2010 April 6; 49(13): 2843–2850. doi:10.1021/bi1000678. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript chorea, and neurologic disorders including epilepsy and ischemic brain damage have been linked to iGluRs (7). The crystal structure of GluA2 (8) clarifies years of speculation on the complex arrangement of the glutamate receptor's four subunits (9). The GluA2 can be dissected into 3 functionally distinct layers. Farthest from the membrane, the amino terminal domain (ATD) can act as a peripheral regulatory domain but is also involved in assembly and trafficking (10,11). Sandwiched between the ATD and the membrane domain, the ligand-binding domain (LBD) recognizes the neurotransmitter signal and directly regulates receptor activation (12). Structures for both isolated extracellular domains (ATD and LBD) reveal a dimeric organization (13–15). At the membrane interface, two alternative linker conformations transition the 2-fold symmetry, which is adopted by both extracellular domains, into the 4- fold symmetry of a membrane-traversing cation-selective channel (8,16). For iGluRs, the ion channel domain confers functional relevance with its ability to selectively conduct the flow of ions across the cell's membrane. The layers of extracellular domains, each with the potential for multiple control points, allosterically regulate the ion channel domain's function (8). Therefore it is not surprising that the ATD, the LBD, and the LBD-channel linker have all been shown to be effective targets of allosteric modulators (13,17,18). Since the structures of the ATD and the full iGluR channel have only recently been solved, allosteric drug-binding sites external to the LBD have not been fully explored in molecular detail. However, the decade-old LBD structure has proved to be indispensable as a heavily exploited scaffold for understanding agonist, partial agonist, and antagonist binding interactions as well as their ability to regulate channel gating behavior (12,19,20). Although the dimeric organization is consistent across all iGluR subtypes, the molecular details of LBD-agonist specificity define the subtype families into N-methyl-D-aspartic acid (NMDA) receptors (21), α-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid (AMPA) receptors (12), and kainate receptors (22). Because all subtypes are constrained by their conserved sensitivity to glutamate stimulation, diversity at the orthosteric site is evolutionarily limited and most agonists display cross-subtype activity. An allosteric modulator-binding site within the quaternary LBD structure is located along the dimer interface (18) and offers improved discrimination by modulators. Drugs that bind to the allosteric sites on the LBD dimer interface can enhance the activity of iGluRs (23) and increase performance on tests of memory (24). Except for the LBD structures with modulatory ions bound to the dimer interface (25–27), only LBD structures from the AMPA receptor subtype, GluA2, have been reported with bound allosteric modulators (18,28–31). Within the structures, the bound modulatory drugs stabilize the LBD dimer interface, which is required for activation of the ion channel and is dissociated during desensitization (18). Although the residues that line the allosteric modulator-binding pocket do not differ between AMPA receptors subtypes (GluA1–4), the ability of allosteric modulators to stabilize the activated state still varies (32,33). Also, AMPA receptors can be alternatively spliced into what is referred to as flip and flop isoforms (34). Modulator selectivity (23), desensitization (35), and channel closing rates (36) differ between flip and flop. Although several of the amino acid differences between the two forms are located in or near the allosteric modulator-binding site, the difference at position 754 (serine in flip, asparagine in flop) seems to be entirely responsible for the functional differences between allosteric modulator regulation of the flip and flop variants (23,28,32). Cyclothiazide (CTZ) and some other thiazide derivatives have improved binding to the flip form due to a hydrogen bond between S754 and the NH of the fused thiazide ring (28). In the case of the flop form, the alternatively spliced sequence places an asparagine in the 754 position, which is not optimally positioned to form a hydrogen bond. Sekiguchi et al. (33) introduced an allosteric modulator of AMPA receptors (4-[2-(phenylsulphonylamino)ethylthio]-2,6,- Ahmed et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript difluorophenoxyacetamide, PEPA) with a preference for the flop form. In fact, the relative sensitivity of CTZ to PEPA has been used as a diagnostic for the prevalence of flip vs. flop versions of AMPA receptor in particular cell types (37). PEPA shows potential in treatment of post ischemic memory impairment (38) and contextual fear (39) but despite PEPA's unique flop sensitivity, the modulator has not yet been used as a lead compound in SAR studies. For drug discovery to be guided by structures, understanding the possible molecular interactions between modulators and the dimer interface is essential. We have shown previously (31) that changes in the structures of CTZ derivatives can reorient the modulator within the binding site. Subsequently, we proposed that the allosteric modulator site is comprised of 5 subsites (Figure 1C). In the present study, we determine the three dimensional structures of PEPA bound to the GluA2o and GluA3o LBDs (flop forms), and use PEPA's binding interactions to further characterize the subsite specific binding properties displayed by allosteric modulators. The amide group of PEPA makes a direct hydrogen bond to N754, explaining the preferential action of PEPA on the flop form of AMPA receptors. Another key structural element, the sulfonamide group of PEPA, is conserved with the biarylsulfonamide class of allosteric modulators (6) and interacts with the same residues of the dimer interface (8,30). Although previously classified as unrelated, PEPA and the large group of biarylsulfonamide have similarities, which suggest that specific PEPA groups (particularly the unique flop-interacting amide) can be strategically integrated into biarylsulfonamide SAR studies. Experimental Procedures Materials PEPA was purchased from Tocris (Ellisville, MO). The GluA2 S1S2J construct was obtained from Eric Gouaux (Vollum Institute; 12). Protein Preparation and Purification GluA2 S1S2 consists of residues N392 - K506 and P632 - S775 of the full rat GluA2o subunit (40), a `GA' segment at the N-terminus, and a `GT' linker connecting K506 and P632 (12). A similar construct of GluA3 S1S2 was prepared as described previously (41). pET-22b(+) plasmids were transformed in E. coli strain Origami B (DE3) cells and were grown at 37°C to OD600 of 0.9 to 1.0 in LB medium supplemented with the antibiotics (ampicillin and kanamycin). The cultures were cooled to 20°C for 20 min. and isopropyl-β- D-thiogalactoside (IPTG) was added to a final concentration of 0.5 mM. Cultures were allowed to grow at 20°C for 20 h. The cells were then pelleted and the S1S2 protein purified using a Ni-NTA column, followed by a sizing column (Superose 12, XK 26/100), and finally an HT-SP-ion exchange-Sepharose column (Amersham Pharmacia). Glutamate (1 mM) was maintained in all buffers throughout purification. After the last column, the protein was concentrated and stored in 20 mM sodium acetate, 1 mM sodium azide, and 10 mM glutamate at pH 5.5. Crystallography For crystallization trials, the protein was concentrated to 0.2 – 0.5 mM in 10 mM glutamate using a Centricon 10 centrifugal filter (Millipore, Bedford, MA). For the PEPA-bound structures, PEPA was added to 5 mM. The final protein concentration was 0.2 to 0.3 mM. Crystals were grown at 4°C using the hanging drop technique, and the drops contained a 1:1 (v/v) ratio of protein solution to reservoir solution. The reservoir solution contained 14–15% PEG 8K, 0.1 M sodium cacodylate, 0.1–0.15 M zinc acetate, and 0.25 M ammonium sulfate, pH 6.5. Ahmed et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Data were collected at the Cornell High Energy Synchrotron Source beam line A1 using a Quantum-210 Area Detector Systems charge-coupled device detector. Data sets were indexed and scaled with HKL-2000 (42). Structures were solved with molecular replacement using Phenix (43). Refinement was performed with Phenix (43), and Coot 0.5 (44) was used for model building. Results Structure of PEPA bound to GluA2 S1S2 flop The structure of glutamate bound to GluA2o S1S2 (3dp6; 41) was used as the initial search probe for the molecular replacement solution of PEPA bound to GluA2o S1S2 with glutamate in the agonist-binding site. PEPA was then modeled into two symmetrical positions within the density found at the dimer interface, and the structure was optimized using Phenix (43). The refinement statistics are given in Table 1. The resolution is 1.85 Å, and three unique copies are found in the unit cell. The overall structure of the S1S2 domain is very similar to the structure in the absence of PEPA, with contacts between glutamate and the protein unchanged. However, PEPA clearly binds within the dimer interface, making contacts with both monomers within the dimer. As shown in Figure 1, one PEPA molecule binds per dimer interface. However because the dimer interface is symmetrical, two equivalent orientations (related by a 180° rotation) are possible. Electron density for both is seen in the crystal structure, although the intensity of one orientation is greater than the other. The binding of PEPA to the dimer interface increases the distance between the two monomers that form the dimer by approximately 1.5 Å. This allows the relatively large PEPA molecule to fit within the interface, but also increases the separation between the linkers to the ion channel (the distance increases from 39.4 Å to 41 Å; Figure 1A). Relative to the core of Lobe 1, both the J/K helices and one β strand (P105-G110) connecting the two lobes are displaced slightly away from the dimer interface (Figure 1B). In addition, Lobe 2 is slightly twisted relative to glutamate-bound S1S2 in the absence of PEPA (3dp6; 41). PEPA binds at the bottom of a water-filled, inverted U-shaped cleft with five subsites (A, B/ B′, and C/C′; 31). Upon binding, crystallographic waters are displaced from the central A subsite and more buried C/C′ sites, with the waters in the B/B′ subsite remaining (Figure 1C). This displacement of presumably ordered water would be likely to contribute a favorable entropy component to binding. The sidechains of P494 are at the center of the interface and the edge of the two proline rings from each monomer form the base of the binding site in which the difluorophenyl ring resides (Figure 2A). This is close to the position of the methoxybenzoyl ring of aniracetam in its structure bound to GluA2-S1S2(FW) (29). The other side of the ring is exposed to S497 and S729. The sidechain hydroxyl of S497 is oriented toward the dimer interface in the absence of PEPA, but rotates out toward the solvent to accommodate the difluorophenyl ring of PEPA (Figure 1B). The amide of PEPA is involved in a network of hydrogen bonds with sidechain hydroxyl of Y424, the backbone carbonyl of F495, the sidechain carboxyl of D760, the sidechain amide of N754, and two water molecules (Figure 2A). The most striking of these hydrogen bond pairs is with N754. This represents the only difference between the flip (S754) and flop (N754) isoforms in the PEPA binding site and is almost certainly a major source of the preference for the flop isoform. The phenyl-sulfonylamide side of PEPA inserts into a hydrophobic pocket formed by sidechain methyls of I481 and L751 as well as methylene groups contributed by K493, N754, and E755 (Figure 2B). It is possible that the contribution by methylene group of N754 provides a more hydrophobic pocket than S754 in the flip form, further contributing to the preference for the flop form. Ahmed et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Because the dimer interface is symmetrical, PEPA can bind in two orientations and both are observed in the crystal. For this reason, changes in the protein due to a specific interaction with PEPA can be partially masked because each monomer is a weighted average of two orientations of bound PEPA. However, one orientation has a stronger density than the other, providing some insight into the extent of changes in the dimer interface that are produced by PEPA binding. As shown in Figure 2C, the two monomers comprising the dimer differ more within the PEPA binding site than the corresponding monomers in the absence of PEPA. One turn of helix J (L751 to N754) contains important determinants for both orientations of PEPA. In one orientation the amide group of PEPA interacts with the sidechain of N754, and in the other, the aromatic ring of PEPA inserts between a hydrophobic pocket formed by the sidechain of L751 and the methylene group of N754. In the orientation for which the density of the amide of PEPA is stronger, N754 is better positioned to form an H-bond (Figure 2D); whereas, in the other side of the interface, N754 is oriented to form an H-bond with the carbonyl of S729. This change in orientation facilitates the insertion of the aromatic ring of PEPA into the hydrophobic pocket, which is accompanied by a small shift in the sidechain of L751 to accommodate the aromatic ring (Figure 2D). Since these structures are weighted averages, it is possible that the actual positions of these sidechains involve an even greater movement than is seen from the asymmetry of the crystal. Structure of PEPA bound to GluA3 S1S2 flop In studies of the physiological effects of PEPA, a significant difference between subtypes has been observed, with GluA3 being most susceptible to modulation (33). The structure of GluA3i S1S2 bound (flip form) to glutamate has been reported previously (41). Since PEPA preferentially binds to the flop form, the GluA3o structure was determined bound to glutamate with and without PEPA (Figure 3A). Like GluA2o, in the absence of PEPA, GluA3o has three copies in the asymmetric unit. Comparing lobe closure between GluA3i and GluA3o, the flop form is slightly more closed (1.6° ± 0.7°). In the presence of PEPA, GluA3o was present in one copy in the asymmetric unit, and PEPA was observed with the same density in two symmetrical orientations. Like GluA2o bound to PEPA, the dimer interface (assessed using the symmetrical molecule in the crystal) was displaced relative to the unbound from (Figure 3A) by approximately 2.5 Å at the position of the linker replacing the ion channel domain. Within the binding site, three sidechains exhibited different rotamers compared with the GluA2o structure bound to PEPA (Figure 3B). For PEPA-bound GluA3o, both S497 and S729 assumed rotameric states that differed both from GluA2o bound to PEPA and from GluA2o and GluA3o in the absence of PEPA. In the case of S729, the rotameric state in combination with a slight movement of the amide of PEPA (relative to the GluA2o structure) would make an H-bond with the sidechain of S729 (shown in Figure 2A for GluA2o) unlikely. In the case of N754, the sidechain is displaced relative to the GluA2o-PEPA structure so that only one H-bond is made to the amide of PEPA. This may be a result of averaging of the two orientations of PEPA only one of which forms a bidentate H-bond with N754. Discussion The goal of allosteric modulation, like orthosteric modulation, is often to stabilize a conformational state of a dynamic protein (45). The activated state of iGluRs is naturally unstable allowing the channel to desensitize (46). Disruption of the symmetrical dimer interface between LBDs is thought to initiate desensitization-mediated channel closure (47). By maintaining the LBD dimer, positive allosteric modulators can prevent desensitization and prolong activation (18). Currently, 15 crystal structures of the GluA2 LBD with bound allosteric modulators are deposited in the Protein Data Bank (48). All of these modulators bind to a large crevice with 2-fold symmetry along the symmetric dimer interface (18). The Ahmed et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript large variation in structure among allosteric modulators results in significant variations in binding orientations and interactions. At least four distinct binding modes have been identified: (1) A-subsite class (aniracetam, CX614 (29)), (2) classical thiazide (cyclothiazide (18), TCMZ, ALTZ (31)), (3) the shifted thiazide class (IDRA-21, HCMZ, HFMZ; (31)), and (4) the full spanning class (PEPA (this paper), dimeric biarylpropylsulfonamide (30), LY404187 (8)). Overlaying modulators from these structural classes has led to the proposal that the allosteric modulator site is comprised of a series of subsites (Figure 1C;31). Positioned at the center of the binding-site, the symmetric A subsite is narrow and allows entrance to only one molecule. Two subsites (B and C) lie at each end of the A subsite with the hydrophobic C subsite located more deeply in the pocket effectively defining five subsites (A, B, B′, C, and C′). In the open state, the subsites are filled with water, which may act to weakly stabilize the dimer. Allosteric modulators generate stronger interactions across the subsites thereby increasing the linkages between the monomers. The simplest modulator class, including aniracetam and CX614, fills the A subsite with one molecule but does not enter the peripheral B and C subsites (29). The two classes of thiazide-based modulators account for 10 of the 15 solved allosteric modulator-GluA2 crystal structure complexes (18,28,31). The classical thiazide (CTZ-like) binding class and the shifted thiazide (IDRA-21-like) binding class are positioned respectively in the B and C subsite or mainly the C subsite. Most of the thiazide modulators do not extend across the A subsite and therefore can bind two molecules per dimer. However, a few of the newly described shifted thiazides (HFMZ, HCTZ; 31) enter the A subsite but only enough to impair binding of a second modulator. The dimeric biarylpropylsulfonamide compound ((R,R)-N,N-(2,2'-[Biphenyl-4-4'-Diyl]Bis[Propane-2,1- Diyl]) Dimethanesulfonamide) described by Kaae et al. (30) was the first allosteric modulator shown by crystallography to extend along the entire length of the inner dimer cavity from C to C′ subsites. PEPA also interacts with J helices from both monomers, which cap the ends of the modulator-binding pocket. The density occupied by both symmetrical copies of PEPA overlays the dimeric biarylsulfonamide compound as both modulators represent the full spanning class (Figure 4B). The GluA flip and flop splice variants differ by only a few residues along the J helix in the LBD; however, residue 754 (Asn in flop and Ser in flip) is positioned between the B and C subsites. For thiazides, a clear preference in binding to the flip-form is mediated by a hydrogen bond between the hydrobenzothiadiazide ring and S754 (28). In contrast, PEPA is flop-selective and the PEPA-bound structure provides the first structure containing a direct interaction between a modulator and the flop form's N754. The amide of PEPA extends straight out from the A subsite and across the B and C subsite interface to make an amide- amide hydrogen bond with N754 (Figure 2A). Unlike most other AMPA modulators, PEPA fills neither the B nor the C subsites but interacts directly with the J helix. A similar interaction is seen with LY404187 (49) bound to GluA2i (8). Strong hydrogen bonding can occur between two amides (50) and has been shown to be responsible for driving oligomerization of transmembrane leucine zippers (51). The distances between the interacting amides in the PEPA-bound structure support a bidentate hydrogen-bonding pattern, which is much stronger and more specific than a typical hydrogen bond. While PEPA is selective for the AMPA receptor's flop form, a weaker but still existent potentiation of the flip form has been observed (33,52). Replacing N754 (flop) with S754 (flip) would not prevent PEPA from binding; however, serine would provide only one hydrogen-bonding partner for PEPA's amide with an extended interaction distance. In contrast, LY404187 displays a preference for the flip isoform (53), and its cyano group extends out to interact directly with S754. The cyano-S754 interaction is a clear flip analog of the flop-selective PEPA amide-N754 interaction (Figure 4A). Ahmed et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Opposite to the amide on the PEPA molecule, a sulfonamide is tethered to the difluorophenyl ring (Figure 4A). Within the dimer interface, the sulfonamide is positioned so the nitrogen can hydrogen bond directly with the carbonyl of P494 (Figure 4C). A sulfonamide oxygen points toward the amide nitrogen of G731. The angle of the peptide plane is perpendicular to the sulfonamide oxygen, making a hydrogen bonding interaction unlikely (Figure 4D). Instead, a dipole-dipole or charge-dipole interaction may occur. The amide nitrogen of a polypeptide supports at least a partial positive charge (54), which would interact with the strongly electronegative sulfonamide oxygen (55). Interestingly, both the dimeric biarylsulfonamide (30) and LY404187 (8), other members of the full spanning modulator class, also have a sulfonamide that interacts with the same backbone atoms of P494 and G731 as PEPA (Figure 4C). A large number of biarylsulfonamides have been identified that modulate AMPA receptors and are being evaluated for therapeutic use in the treatment of depression and Parkinson's disease (56). The conserved sulfonamide reveals a previously unidentified relationship between PEPA and the biarylsulfonamide modulators. When the perpendicular peptide bond plane including G731 is fixed, the sulfonamide on three overlayed modulators varies by 1.2 Å along the length of the interface with the PEPA sulfonamide being positioned closer to the A subsite (Figure 4D). A shift of the sulfonamide also results in a shift in the corresponding P494 across the interface presumably to maintain the hydrogen bond with the modulator's amine. The sulfonamide forms an important bridge between the two dimer halves. For PEPA, a phenyl-sulfonamide replaces the methyl-sulfonamide in the dimeric biarylsulfonamide and fits snuggly against L751. Based on the orientation-induced asymmetry within the GluA2-complex structure, the phenyl pushes the J helix away from PEPA thereby affecting the C subsite (Figure 2C and D). Residues lining the C subsite are on the same beta strand as G731, which must shift if the C subsite is to remain together and presumably explain the 1.2 Å shift relative to the dimeric biarylsulfonamide. In fact, the same phenyl-sulfonamide group substitution in a biarylpropylsulfonamide decreases the modulatory effect of the derivative in SAR studies (57). For biarylpropylsulfonamides, the optimal sulfonamide substitution was found to be either an ethyl or an iso-propyl group, which should both fit without significantly disrupting the J helix or C subsite (57). The PEPA-bound crystal structure from AMPA receptor subtypes, GluA2 and GluA3, do not display major differences in binding interactions even though PEPA exhibits a stronger effect on GluA3 (33). For GluA2, an asymmetry in the receptor-binding pocket was observed while no significant difference in PEPA density was seen for the each orientation within the GluA3 crystal structure. In addition, a number of side chains exhibit different rotameric states between the two structures, although it is unlikely that these small changes significantly impact the differential effects on the two subtypes. Although no structural differences have been identified between GluA2 and GluA3 that would obviously impact PEPA affinity, the possibility exists that subtle differences arising from the sequence differences peripheral to the binding site may be important as has been described in the case of the agonist binding site of GluA4 (58). We have explored how PEPA (this paper) and other allosteric modulators (31) interact with the GluA interface in the context of drug design. Together the identification of a conserved group between PEPA (this paper) and biarylpropylsulfonamides (8,30) and the regional nature of various subsite-functional group interactions provide a backdrop to extend biarylpropylsulfonamide SAR studies (57) to include PEPA and biarylpropylsulfonamide chimeras. Although optimizing the stability of the dimer interface provides a starting point for SAR studies, additional constraints should be considered including the ability of the modulator to enter the cavity, the dynamic structure of the dimer interface during closed, open, and desensitized state transitions, and the ability of the modulator to cross the blood- Ahmed et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript brain barrier before being metabolized. This definition of the allosteric modulator binding- site should provide guidance in glutamate receptor allosteric modulator pharmacology. Acknowledgments We thank Prof. Eric Gouaux (Vollum Institute) for the GluA2 S1S2J construct, and Prof. Linda Nowak (Cornell) for the full-length GluA3 construct. This work was supported by a grants from the National Institutes of Health (R01-GM068935, R01 NS049223, and R21 NS067562). This work is based upon research conducted at the Cornell High Energy Synchrotron Source (CHESS), which is supported by the National Science Foundation under award DMR 0225180, using the Macromolecular Diffraction at the CHESS (MacCHESS) facility, which is supported by award RR-01646 from the National Institutes of Health, through its National Center for Research Resources. Abbreviations ALTZ althiazide AMPA α-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid CLTZ chlorothiazide CX614 pyrrolidino-1,3-oxazino benzo-1,4-dioxan-10-one CTZ cyclothiazide FW (S)-5-fluorowillardiine flip and flop alternatively spliced versions of AMPA receptors that vary in rates of desensitization and sensitivity to allosteric modulators iGluR ionotropic glutamate receptor GluA1-4 four subtypes of AMPA receptor HCTZ hydrochlorothiazide HFMZ hydroflumethiazide IDRA-21 7-chloro-3-methyl-3,4-dihydro-2H-benzo[e][1,2,4]thiadiazine 1,1-dioxide IPTG isopropyl-β-D-thiogalactoside LY404187 N-[2-(4′-cyanobiphenyl-4-yl)propyl]propane-2-sulfamide PEPA 4-[2-(phenylsulphonylamino)ethylthio]-2,6,-difluorophenoxy acetamide NMDA N-methyl-D-aspartic acid S1S2 extracellular ligand-binding domain of GluA2 and GluA3 SAR structure-activity relationships TCMZ trichlormethiazide References 1. Christopoulos A. Allosteric binding sites on cell-surface receptors: novel targets for drug discovery. Nat Rev Drug Discov. 2002; 1:198–210. [PubMed: 12120504] 2. Changeux JP, Taly A. Nicotinic receptors, allosteric proteins and medicine. Trends Mol Med. 2008; 14:93–102. [PubMed: 18262468] 3. Bowie D. Ionotropic glutamate receptors & CNS disorders. CNS Neurol Disord Drug Targets. 2008; 7:129–143. [PubMed: 18537642] 4. Dingledine R, Borges K, Bowie D, Traynelis S. The glutamate receptor ion channels. Pharmacol. Rev. 1999; 51:7–61. [PubMed: 10049997] Ahmed et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 5. Oswald RE, Ahmed A, Fenwick MK, Loh AP. Structure of glutamate receptors. Current drug targets. 2007; 8:573–582. [PubMed: 17504102] 6. Grigoriev VV, Proshin AN, Kinzirsky AS, Bachurin SO. Modern approaches to the design of memory and cognitive enhancers based on AMPA receptor ligands. Russian Chemical Reviews. 2009; 78:485–494. 7. Black MD. Therapeutic potential of positive AMPA modulators and their relationship to AMPA receptor subunits. A review of preclinical data. Psychopharmacology (Berl). 2005; 179:154–163. [PubMed: 15672275] 8. Sobolevsky AI, Rosconi MP, Gouaux E. X-ray structure, symmetry and mechanism of an AMPA- subtype glutamate receptor. Nature. 2009; 462:745–756. [PubMed: 19946266] 9. Wo ZG, Oswald RE. Unraveling the modular design of glutamate-gated ion channels. Trends Neurosciences. 1995; 18:161–168. 10. Gielen M, Siegler Retchless B, Mony L, Johnson JW, Paoletti P. Mechanism of differential control of NMDA receptor activity by NR2 subunits. Nature. 2009; 459:703–707. [PubMed: 19404260] 11. Greger IH, Ziff EB, Penn AC. Molecular determinants of AMPA receptor subunit assembly. Trends Neurosci. 2007; 30:407–416. [PubMed: 17629578] 12. Armstrong N, Gouaux E. Mechanisms for activation and antagonism of an AMPA-sensitive glutamate receptor: crystal structures of the GluR2 ligand binding core. Neuron. 2000; 28:165– 181. [PubMed: 11086992] 13. Clayton A, Siebold C, Gilbert RJ, Sutton GC, Harlos K, McIlhinney RA, Jones EY, Aricescu AR. Crystal structure of the GluR2 amino-terminal domain provides insights into the architecture and assembly of ionotropic glutamate receptors. J Mol Biol. 2009; 392:1125–1132. [PubMed: 19651138] 14. Jin R, Singh SK, Gu S, Furukawa H, Sobolevsky AI, Zhou J, Jin Y, Gouaux E. Crystal structure and association behaviour of the GluR2 amino-terminal domain. EMBO J. 2009; 28:1812–1823. [PubMed: 19461580] 15. Kumar J, Schuck P, Jin R, Mayer ML. The N-terminal domain of GluR6-subtype glutamate receptor ion channels. Nat Struct Mol Biol. 2009; 16:631–638. [PubMed: 19465914] 16. Doyle DA, Cabral JM, Pfuetzner RA, Kuo AL, Gulbis JM, Cohen SL, Chait BT, MacKinnon R. The structure of a potassium channel: molecular basis of K+ conduction and selectivity. Science. 1998; 280:69–77. [PubMed: 9525859] 17. Balannik V, Menniti FS, Paternain AV, Lerma J, Stern-Bach Y. Molecular mechanism of AMPA receptor noncompetitive antagonism. Neuron. 2005; 48:279–288. [PubMed: 16242408] 18. Sun Y, Olson R, Horning M, Armstrong N, Mayer M, Gouaux E. Mechanism of glutamate receptor desensitization. Nature. 2002; 417:245–253. [PubMed: 12015593] 19. Ahmed A, Thompson M, Fenwick M, Romero B, Loh A, Jane D, Sondermann H, Oswald R. Mechanisms of antagonism of the GluR2 AMPA receptor: Structure and dynamics of the complex of two willardiine antagonists with the glutamate binding domain. Biochemistry. 2009; 48:3894– 3903. [PubMed: 19284741] 20. Jin R, Banke TG, Mayer ML, Traynelis SF, Gouaux E. Structural basis for partial agonist action at ionotropic glutamate receptors. Nat Neurosci. 2003; 6:803–810. [PubMed: 12872125] 21. Furukawa H, Gouaux E. Mechanisms of activation, inhibition and specificity: crystal structures of the NMDA receptor NR1 ligand-binding core. EMBO J. 2003; 22:2873–2885. [PubMed: 12805203] 22. Mayer ML. Crystal structures of the GluR5 and GluR6 ligand binding cores: Molecular mechanisms underlying kainate receptor selectivity. Neuron. 2005; 45:539–552. [PubMed: 15721240] 23. Partin KM, Fleck MW, Mayer ML. AMPA receptor flip/flop mutants affecting deactivation, desensitization, and modulation by cyclothiazide, aniracetam, and thiocyanate. J Neurosci. 1996; 16:6634–6647. [PubMed: 8824304] 24. Martin JR, Cumin R, Aschwanden W, Moreau JL, Jenck F, Haefely WE. Aniracetam improves radial maze performance in rats. Neuroreport. 1992; 3:81–83. [PubMed: 1611039] Ahmed et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 25. Naur P, Vestergaard B, Skov LK, Egebjerg J, Gajhede M, Kastrup JS. Crystal structure of the kainate receptor GluR5 ligand-binding core in complex with (S)-glutamate. FEBS Lett. 2005; 579:1154–1160. [PubMed: 15710405] 26. Plested AJ, Mayer ML. Structure and mechanism of kainate receptor modulation by anions. Neuron. 2007; 53:829–841. [PubMed: 17359918] 27. Plested AJ, Vijayan R, Biggin PC, Mayer ML. Molecular basis of kainate receptor modulation by sodium. Neuron. 2008; 58:720–735. [PubMed: 18549784] 28. Hald H, Ahring PK, Timmermann DB, Liljefors T, Gajhede M, Kastrup JS. Distinct Structural Features of Cyclothiazide are Responsible for Effects on Peak Current Amplitude and Desensitization Kinetics at iGluR2. J Mol Biol. 2009; 391:906–917. [PubMed: 19591837] 29. Jin R, Clark S, Weeks AM, Dudman JT, Gouaux E, Partin KM. Mechanism of positive allosteric modulators acting on AMPA receptors. J Neurosci. 2005; 25:9027–9036. [PubMed: 16192394] 30. Kaae BH, Harpsoe K, Kastrup JS, Sanz AC, Pickering DS, Metzler B, Clausen RP, Gajhede M, Sauerberg P, Liljefors T, Madsen U. Structural proof of a dimeric positive modulator bridging two identical AMPA receptor-binding sites. Chemistry & biology. 2007; 14:1294–1303. [PubMed: 18022568] 31. Ptak CP, Ahmed AH, Oswald RE. Probing the allosteric modulator binding site of GluR2 with thiazide derivatives. Biochemistry. 2009; 48:8594–8602. [PubMed: 19673491] 32. Partin KM, Bowie D, Mayer ML. Structural determinants of allosteric regulation in alternatively spliced AMPA receptors. Neuron. 1995; 14:833–843. [PubMed: 7718245] 33. Sekiguchi M, Fleck MW, Mayer ML, Takeo J, Chiba Y, Yamashita S, Wada K. A novel allosteric potentiator of AMPA receptors: 4-[2-(phenylsulfonylamino)ethylthio]-2,6-difluoro- phenoxyacetamide. J Neurosci. 1997; 17:5760–5771. [PubMed: 9221774] 34. Sommer B, Keinänen K, Verdoorn TA, Wisden W, Burnashev N, Herb A, Köhler M, Takagi T, Sakmann G, Seeburg PH. Flip and flop: A cell-specific functional switch in glutamate-operated channels of the CNS. Science. 1990; 249:1580–1584. [PubMed: 1699275] 35. Mosbacher J, Schoepfer R, Monyer H, Burnashev N, Seeburg PH, Ruppersberg JP. A molecular determinant for submillisecond desensitization in glutamate receptors. Science. 1994; 266:1059– 1062. [PubMed: 7973663] 36. Pei W, Huang Z, Niu L. GluR3 flip and flop: differences in channel opening kinetics. Biochemistry. 2007; 46:2027–2036. [PubMed: 17256974] 37. Sekiguchi M, Takeo J, Harada T, Morimoto T, Kudo Y, Yamashita S, Kohsaka S, Wada K. Pharmacological detection of AMPA receptor heterogeneity by use of two allosteric potentiators in rat hippocampal cultures. Br J Pharmacol. 1998; 123:1294–1303. [PubMed: 9579722] 38. Sekiguchi M, Yamada K, Jin J, Hachitanda M, Murata Y, Namura S, Kamichi S, Kimura I, Wada K. The AMPA receptor allosteric potentiator PEPA ameliorates post-ischemic memory impairment. Neuroreport. 2001; 12:2947–2950. [PubMed: 11588608] 39. Zushida K, Sakurai M, Wada K, Sekiguchi M. Facilitation of extinction learning for contextual fear memory by PEPA: a potentiator of AMPA receptors. J Neurosci. 2007; 27:158–166. [PubMed: 17202483] 40. Hollmann M, Heinemann S. Cloned glutamate receptors. Annu Rev Neurosci. 1994; 17:31–108. [PubMed: 8210177] 41. Ahmed AH, Wang Q, Sondermann H, Oswald RE. Structure of the S1S2 glutamate binding domain of GluR3. Proteins: Structure, Function, and Bioinformatics. 2009; 75:628–637. 42. Otwinowski, Z.; Minor, W. Processing of X-ray diffraction data collected in oscillation mode. In: Carter, CW.; Sweet, RM., editors. Methods in Enzymology, Vol. 276, Macromolecular Crystallography, part A. Academic Press; New York: 1997. p. 307-326. 43. Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Terwilliger TC. PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr D Biol Crystallogr. 2002; 58:1948– 1954. [PubMed: 12393927] 44. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] Ahmed et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 45. Gunasekaran K, Ma B, Nussinov R. Is allostery an intrinsic property of all dynamic proteins? Proteins. 2004; 57:433–443. [PubMed: 15382234] 46. Trussell LO, Fischbach GD. Glutamate receptor desensitization and its role in synaptic transmission. Neuron. 1989; 3:209–218. [PubMed: 2576213] 47. Armstrong N, Jasti J, Beich-Frandsen M, Gouaux E. Measurement of conformational changes accompanying desensitization in an ionotropic glutamate receptor. Cell. 2006; 127:85–97. [PubMed: 17018279] 48. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, Shindyalov IN, Bourne PE. The Protein Data Bank. Nucleic Acids Res. 2000; 28:235–242. [PubMed: 10592235] 49. Miu P, Jarvie KR, Radhakrishnan V, Gates MR, Ogden A, Ornstein PL, Zarrinmayeh H, Ho K, Peters D, Grabell J, Gupta A, Zimmerman DM, Bleakman D. Novel AMPA receptor potentiators LY392098 and LY404187: effects on recombinant human AMPA receptors in vitro. Neuropharmacology. 2001; 40:976–983. [PubMed: 11406188] 50. Shimoni L, Glusker JP. Hydrogen bonding motifs of protein side chains: descriptions of binding of arginine and amide groups. Protein Sci. 1995; 4:65–74. [PubMed: 7773178] 51. Zhou FX, Cocco MJ, Russ WP, Brunger AT, Engelman DM. Interhelical hydrogen bonding drives strong interactions in membrane proteins. Nat Struct Biol. 2000; 7:154–160. [PubMed: 10655619] 52. Sekiguchi M, Nishikawa K, Aoki S, Wada K. A desensitization-selective potentiator of AMPA- type glutamate receptors. Br J Pharmacol. 2002; 136:1033–1041. [PubMed: 12145103] 53. Quirk JC, Nisenbaum ES. Multiple molecular determinants for allosteric modulation of alternatively spliced AMPA receptors. J Neurosci. 2003; 23:10953–10962. [PubMed: 14645491] 54. Kemnitz CR, Loewen MJ. “Amide resonance” correlates with a breadth of C-N rotation barriers. J Am Chem Soc. 2007; 129:2521–2528. [PubMed: 17295481] 55. Cazenave-Gassiot A, Boughtflower R, Caldwell J, Coxhead R, Hitzel L, Lane S, Oakley P, Holyoak C, Pullen F, Langley GJ. Prediction of retention for sulfonamides in supercritical fluid chromatography. J Chromatogr A. 2008; 1189:254–265. [PubMed: 17977551] 56. O'Neill MJ, Witkin JM. AMPA receptor potentiators: application for depression and Parkinson's disease. Current drug targets. 2007; 8:603–620. [PubMed: 17504104] 57. Ornstein PL, Zimmerman DM, Arnold MB, Bleisch TJ, Cantrell B, Simon R, Zarrinmayeh H, Baker SR, Gates M, Tizzano JP, Bleakman D, Mandelzys A, Jarvie KR, Ho K, Deverill M, Kamboj RK. Biarylpropylsulfonamides as novel, potent potentiators of 2-amino-3- (5-methyl-3- hydroxyisoxazol-4-yl)- propanoic acid (AMPA) receptors. J Med Chem. 2000; 43:4354–4358. [PubMed: 11087558] 58. Gill A, Birdsey-Benson A, Jones BL, Henderson LP, Madden DR. Correlating AMPA receptor activation and cleft closure across subunits: crystal structures of the GluR4 ligand-binding domain in complex with full and partial agonists. Biochemistry. 2008; 47:13831–13841. [PubMed: 19102704] Ahmed et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. (A) Comparison of glutamate-bound GluA2o S1S2 in the presence (blue) and the absence of PEPA (green) in two orientations. Both orientations of PEPA are shown. Note that the binding of PEPA results in a separation of the two components of the dimer (distance between the Cα atoms of the threonine in the linker) by approximately 1.5 Å. (B) One monomer of GluA2o S1S2 in the presence (blue) and the absence of PEPA (green) with one orientation of PEPA shown. Both the J/K helices and the strand near S497 are displaced upon binding PEPA. Also, the sidechains of S497 and S729 change rotameric states. (C) Comparison of the water molecules at the dimer interface in the presence (tan spheres) and the absence of PEPA (red spheres). PEPA is shown in both orientations. Despite the greater separation of the dimer interface, a number of the ordered water molecules found in the absence of PEPA are displaced by PEPA. The black circles delineate subsites of the allosteric modulator binding site as described previously (31). Ahmed et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. The PEPA binding site, emphasizing the important interactions, shown in two orientations. (A) A view of the amide side of PEPA bound to GluA2 S1S2. The hydrogen bonding network with the amide of PEPA is shown as dotted lines. The H-bond with the sidechain of S729 is difficult to display in the orientation used in the figure. (B) A view of the phenyl group of PEPA inserted into a hydrophobic pocket in GluA2 S1S2. (C) RMS plot showing more variability in the J/K helices for the PEPA-bound structure than the unbound structure. (D) J/K helix showing where differences in the two orientations were analyzed. The amide of PEPA-N754 interaction (blue) maintains the position of the J helix in the absence of PEPA (green) The J helix is displaced on the phenyl side of PEPA (red). Ahmed et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. (A) Comparison of glutamate-bound GluA3o S1S2 in the presence (blue) and the absence of PEPA (green) in two orientations. Both orientations of PEPA are shown. Note that the binding of PEPA results in a separation of the two components of the dimer (distance between the Cα atoms of the threonine in the linker) by approximately 2.5 Å. (B) One monomer of GluA3o S1S2 in the presence (blue) and the absence of PEPA (green) with one orientation of PEPA shown. Shown for comparison is the PEPA-bound form of GluA2o (red). Both the J/K helices and the strand near S497 are displaced upon binding PEPA for both GluA2o and GluA3o. Also, the sidechains of S497 and S729 are in different rotameric states for GluA3o bound to PEPA compared with GluA3o in the absence of PEPA and GluA2o bound to PEPA. Also, N754 is displaced in PEPA-bound GluA3o, such that only one H-bond is possible with the amide of PEPA. Ahmed et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. (A) Members of the full spanning class of allosteric modulators. The shape-highlighted regions of the modulators illuminate key contact points to the specific binding pocket residues and subsites (as labeled for PEPA). (B) Overlay of the full spanning modulator structures. The structures were aligned at both sets of P494 and G731 residues. PEPA (gray) occupies a similar arrangement of subsites as the dimeric biarylsulfonamide (PDB entry 3bbr, cyan, 30) and LY404187 (PDB entry 3kgc, magenta, 8). (C) The sulfonamide bridges the two monomers in both PEPA and the dimeric biarylsulfonamide with the same interactions to P494 and G731. (D) The hydrogen bond between the carbonyl of P494 and the sulfonamide is maintained when the modulator is in a shifted position relative to the peptide plane of K730 and G731 (green disk) located on the opposite monomer. Ahmed et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Ahmed et al. Page 16 Table 1 Structural Statistics GluA2o (PEPA) GluA3o (PEPA) GluA3o Space Group P22121 P222 P222 Unit Cell (Å) a=47.13 b=113.92 c=164.81 a=46.95 b=52.26 c=115.98 a=46.03 b=110.33 c=161.192 X-ray source CHESS (A1) CHESS (A1) CHESS (A1) Wavelength (Å) 0.977 0.977 0.977 Resolution (Å) 50–2.0 (2.03–2.00) 50–2.5 (2.54–2.00) 50–1.85 (1.88–1.85) Measured reflections (#) 817961 62549 344340 Unique reflections (#) 70175 9141 69379 Data redundancy 6.9 (7.1) 6.0 (4.2) 4.7 (3.0) Completeness (%) 99.9 (100.0) 99.5 (89.8) 96.5 (73.5) Rsym (%) 11.4 (34.2) 13.0 (45.5) 7.5 (24.6) I/σi 33.3 (7.1) 19.2 (2.5) 34.3 (3.3) PDB ID * * * Current Model Refinement Statistics Phasing MR MR MR Molecules/AU 3 (no NCS applied) 1 3 (no NCS applied) Rwork/Rfree (%) 18.7/24.2 18.9/28.5 20.1/23.4 Free R test set size (#/%) 2000 (2.85) 914 (10.0) 2000 (2.88) Number of protein atoms 5979 2030 6091 Number of heteroatoms 111 62 30 Rmsd bond length (Å) 0.011 0.015 0.009 Rmsd bond angles (°) 1.3 1.8 1.3 *to be submitted to RCSB Protein Data Bank Biochemistry. Author manuscript; available in PMC 2011 April 7.
3M3J
A new crystal form of Lys48-linked diubiquitin
structural communications 994 doi:10.1107/S1744309110027600 Acta Cryst. (2010). F66, 994–998 Acta Crystallographica Section F Structural Biology and Crystallization Communications ISSN 1744-3091 A new crystal form of Lys48-linked diubiquitin Jean-Franc¸ois Trempe,a* Nicholas R. Brown,b Martin E. M. Nobleb and Jane A. Endicottb aDepartment of Biochemistry, McGill University, 3649 Promenade Sir William Osler, Montreal, Que´bec H3G 0B1, Canada, and bLaboratory of Molecular Biophysics, Department of Biochemistry, South Parks Road, Oxford OX1 3QU, England Correspondence e-mail: jean.trempe@mail.mcgill.ca Received 10 March 2010 Accepted 12 July 2010 PDB Reference: Lys48-linked diubiquitin, 3m3j. Lys48-linked polyubiquitin chains are recognized by the proteasome as a tag for the degradation of the attached substrates. Here, a new crystal form of Lys48- linked diubiquitin (Ub2) was obtained and the crystal structure was refined to 1.6 A˚ resolution. The structure reveals an ordered isopeptide bond in a trans configuration. All three molecules in the asymmetric unit were in the same closed conformation, in which the hydrophobic patches of both the distal and the proximal moieties interact with each other. Despite the different crystal- lization conditions and different crystal packing, the new crystal structure of Ub2 is similar to the previously published structure of diubiquitin, but differences are observed in the conformation of the flexible isopeptide linkage. 1. Introduction The ubiquitin–proteasome pathway is a fundamental cellular process in eukaryotes that controls protein degradation. Substrates are tagged with ubiquitin through a cascade of enzymatic reactions that is initiated by the activation of ubiquitin by the E1 enzyme, followed by ubiquitin conjugation to E2 and finally transfer of the activated ubiquitin from E2 to a specific substrate via an E3 ligase (Hershko & Ciechanover, 1998). Ubiquitin molecules are assembled through the formation of an isopeptide bond between the carboxyl-terminal group of ubiquitin and the side-chain "-amino group of a lysine in another ubiquitin molecule (termed the distal and proximal moieties, respectively) or on the substrate. The 26S proteasome is able to recognize and degrade substrates tagged with a Lys48-linked poly- ubiquitin chain (Finley, 2009). Several proteasomal ubiquitin receptors have been described, including the 19S regulatory particle base subunits S5a/Rpn10 (Deveraux et al., 1994) and Rpn13 (Husnjak et al., 2008), as well as the UBL-UBA-containing proteins HHR23/Rad23, Dsk2/Dph1 and Ddi1/Mud1 (Bertolaet et al., 2001; Wilkinson et al., 2001). The inter- actions of ubiquitin receptors with Lys48-linked polyubiquitin have been characterized at the structural level (Schreiner et al., 2008; Trempe et al., 2005; Varadan et al., 2005; Zhang, Chen et al., 2009; Zhang, Wang et al., 2009), but as yet a crystal structure of a Lys48- linked polyubiquitin chain bound to its receptor has not been reported. In an attempt to obtain the structure of Lys48-linked di- ubiquitin (Ub2) bound to the Mud1 UBA domain (Trempe et al., 2005), cocrystallization trials were performed. Diffracting crystals were obtained, but subsequent structure determination revealed that the crystals were solely composed of Ub2. The Ub2 subunits in the new crystal structure adopt the closed conformation, as observed in the previous crystal structure (Cook et al., 1992) and in solution (Varadan et al., 2002). The packing in the new crystal form differs from that in the previous crystal structure and the structure reveals differences in the conformation of the isopeptide linkage and the loop connecting 1 and 2. 2. Materials and methods 2.1. Purification and crystallization Ub2 was synthesized in vitro as described previously (Piotrowski et al., 1997; Trempe et al., 2005). Briefly, the reaction mixture contained 50 mM Tris–HCl pH 8.0, 2 mM ATP, 5 mM MgSO4, 0.5 mM bovine ubiquitin, 0.5 mM recombinant human His6-E1 and 50 mM recombi- nant budding yeast His10-Cdc34. The synthesis reaction was per- formed at 310 K overnight. Bovine ubiquitin was purchased as a lyophilized powder (Sigma–Aldrich), His6-E1 ubiquitin-conjugating enzyme was expressed from a recombinant baculovirus in Sf9 insect cells and recombinant His10-Cdc34 was expressed in BL21 (DE3) Escherichia coli cells from a pET16 expression plasmid. Both His- tagged proteins were purified using Ni–NTA agarose resin (Qiagen). The amino-acid sequence of bovine ubiquitin is identical to that of human ubiquitin and yeast Cdc34 has previously been shown to synthesize Lys48-linked polyubiquitin chains in vitro with human E1 (Wu et al., 2002). The Ub2 purification method was a modification of a previously published protocol (Chen & Pickart, 1990). After completion, the synthesis reaction mixture was dialysed against 50 mM ammonium acetate pH 4.5. The mixture was filtered and loaded at 1.0 ml min1 onto a Mono-S cation-exchange chromatography column (HR 5/5, GE Healthcare). The polyubiquitin chains were then eluted with a linear gradient of 0–0.4 M KCl over 60 ml. Elution fractions were collected and further purified by size-exclusion chromatography on a Superdex 75 16/60 column (GE Healthcare) equilibrated in crystal- lization buffer (20 mM Tris–HCl pH 8.0, 50 mM NaCl, 0.01% NaN3). The purity of the different polyubiquitin chains (Ub1, Ub2, Ub3 and Ub4) was assessed by SDS–PAGE. The Ub2 concentration was determined using UVabsorbance at 276 nm. The Mud1 UBA domain (residues 293–332) was expressed and purified as described previously (Trempe et al., 2005) and dialyzed against crystallization buffer. Cocrystallization trials of Mud1 UBA with Ub2 were performed at a final concentration of 0.5 mM Ub2 and 0.5–0.75 mM Mud1 UBA using Structure Screens 1 and 2 (Molecular Dimensions). Crystals were grown at 295 K by vapour diffusion using the sitting-drop method (1.0 ml drops). Thin rectangular plate-shaped crystals (300  100  30 mm) were grown in 30% PEG 4000, 0.2 M Li2SO4, 0.1 M Tris–HCl pH 8.5 from a 1.5:1 molar ratio of UBA:Ub2. Conditions with less or no Mud1 UBAyielded smaller crystals of poor diffraction quality. 2.2. Data collection and processing A crystal was cryoprotected using mother liquor supplemented with 15% ethylene glycol and frozen in liquid nitrogen. Data were collected at 100 K on beamline ID-29 at ESRF, Grenoble. Data- collection statistics are shown in Table 1. Reflections were indexed and integrated using the program MOSFLM (Leslie, 2006) and the intensities were scaled and merged using SCALA (Evans, 2006). 2.3. Structure solution and refinement The phase problem was solved by molecular replacement using the program Phaser (McCoy et al., 2007). The crystal structure of monoubiquitin (PDB code 1ubq; Vijay-Kumar et al., 1987) was used as a search model, excluding the flexible residues 73–76. Six copies of ubiquitin were found, giving a solvent content of 41%. After rigid- body refinement in REFMAC5 (Murshudov et al., 1997), no addi- tional density was observed that could accommodate the UBA domain. Water molecules were added automatically using ARP/ wARP (Perrakis et al., 1997). The model was then adjusted in the electron-density map using the program Coot (Emsley & Cowtan, 2004). The bulk solvent was modelled using the Babinet method with a mask. After a few cycles of restrained refinement in REFMAC5 and model building, a final model was obtained with good overall geometry and a satisfactory fit to the experimental amplitudes (Table 1). The distal moieties of the three Ub2 molecules in the asymmetric unit were named A, C and E and their respective cova- lently bound proximal moieties were named B, D and F. The co- ordinates and structure factors were deposited in the Protein Data Bank under accession code 3m3j. 3. Results and discussion The asymmetric unit of the new crystal form contained three Ub2 molecules, which all adopt the same conformation in which the hydrophobic patches of the proximal and distal ubiquitin moieties, centred around Ile44, interact with each other (Fig. 1a). Most ubiquitin-binding domains interact with the hydrophobic patch of ubiquitin (Hicke et al., 2005) and thus the conformation in which the patch is buried will be referred to as the closed conformation. More specifically, the side chains of Leu8, Ile44, His68 and Val70 of one moiety fit snugly onto a surface formed by the same amino acids on the other moiety (Fig. 1b). Moreover, the same seven hydrogen bonds were found in each of the three distal–proximal pairs, notably between the carbonyl O atoms of Gly47 and Leu71 and the backbone amides of Leu71 and Gln49, respectively. The overall arrangement of the distal and proximal moieties is thus remarkably similar among the three Ub2 molecules in the asymmetric unit (Fig. 1c), with C root- structural communications Acta Cryst. (2010). F66, 994–998 Trempe et al.  Lys48-linked diubiquitin 995 Table 1 X-ray data-collection and refinement statistics for Ub2. Values in parentheses are for the last shell. X-ray source ESRF ID29 Wavelength (A˚ ) 0.97625 Space group C2 Unit-cell parameters (A˚ , ) a = 58.7, b = 78.7, c = 93.1,  =  = 90,  = 97.9 Mosaicity () 0.30 Images 180 Oscillation angle () 1.0 Resolution (A˚ ) 39.90–1.60 (1.69–1.60) Unique reflections 54118 (7792) Completeness (%) 97.9 (96.8) Multiplicity 3.8 (3.8) hIi/h(I)i 16.1 (3.2) Rmerge† 0.057 (0.432) Solvent content (%) 41 No. of reflections in Rfree set (5%) 2738 Rwork 0.183 Rfree 0.229 FOM 0.851 R.m.s. deviations from ideal values‡ Bond lengths (A˚ ) 0.012 Bond angles () 1.5 Torsion angles () 6.1 Protein atoms 3962 Water atoms 360 Ligand atoms (1 ethylene glycol, 3 sulfate ions) 19 Disordered residues (not modelled) Chain B, 76; chains D, F, 74, 75, 76§ Average B factors (A˚ 2) Protein main chain 19 Protein side chain 21 Water 32 Ethylene glycol 28 Sulfate ions 58 Ramachandran outliers} 1 [Gln62 in chain D] Estimated coordinate error†† (A˚ ) 0.18 PDB code 3m3j † P hkl P i jIiðhklÞ  hIðhklÞij=P hkl P i IiðhklÞ, where Ii(hkl) is the intensity of the ith measurement of reflection hkl and hI(hkl)i is the mean value for all i measure- ments. ‡ Ideal values as reported in Engh & Huber (2001). § These residues correspond to the C-termini of proximal ubiquitin moieties. } Residues for which the backbone torsion angles are outside the core region of the Ramachandran plot (Kleywegt & Jones, 1996). †† Coordinate error estimated from a Luzzati plot (R/Rfree versus resolution) as reported by SFCHECK (Vaguine et al., 1999). mean-square deviation (r.m.s.d.) values that are between 0.39 and 0.53 A˚ . A previously reported crystal structure of Ub2 (Cook et al., 1992) has a single molecule in the asymmetric unit, which also adopts the closed conformation (Fig. 1c). C r.m.s.d. values of 0.68–0.89 A˚ were calculated between the previous structure (PDB code 1aar; Cook et al., 1992) and each of the Ub2 subunits in the new crystal structure. The previous crystal form was obtained by crystallizing Ub2 in the presence of 2-methyl-2,4-pentanediol (MPD) and sodium citrate at pH 5.0, instead of PEG 4000, Li2SO4 and Tris at pH 8.5 as used in the current study. Despite these different conditions, the same set of hydrophobic interactions and hydrogen bonds were found as in the previous Ub2 crystal structure. The closed conformation was also observed in one of the crystal forms of Ub4 (Phillips et al., 2001) but not in the other (Cook et al., 1994). Similar to the case reported here, the more recent Ub4 crystal structure was obtained from a crystal grown in the presence of a peptide derived from a ubiquitin-binding protein (S5a), which was not incorporated into the crystal but yielded Ub4 crystals in a different space group (Phillips et al., 2001). NMR residual dipolar couplings and relaxation-anisotropy studies have shown that the closed conformation of Ub2 predominates in solution at pH values above 6.8 and is in rapid equilibrium with an open form (Varadan et al., 2002). The solution structure of the closed confor- mation, which was determined by a docking approach using chemical shift perturbation data and residual dipolar coupling restraints (PDB code 2bgf; van Dijk et al., 2005), superposes with an average C r.m.s.d. of 1.5 A˚ with the three Ub2 conjugates observed in the present crystal structure. This shows that the overall arrangement of the Ub2 conjugate in the crystal is similar to that observed in solution. Although Ub2 adopts the closed conformation in both crystal forms (this study and Cook et al., 1992), differences are observed in the configuration of the isopeptide linkage. Well defined electron density was observed for the isopeptide linkage in the new crystal structure (Fig. 2a), with B factors near main-chain levels for the atoms involved (between 15 and 25 A˚ 2, compared with 10–20 A˚ 2 for main- chain atoms). This contrasts with the previously published Ub2 crystal structure, which showed slight disorder for these residues (B factors of >30 A˚ 2, compared with 10–20 A˚ 2 for main-chain atoms), although electron density was also visible for the isopeptide bond (Cook et al., 1992). The crystal packing probably induces this order in the new crystal form, since isopeptide linkages from molecules within or between different asymmetric units make a number of reciprocal interactions (Fig. 2b). The "-amide group of Lys48 in the distal subunit (involved in the isopeptide bond) makes a hydrogen bond to the backbone carbonyl O atom of Ala46 in a neighbouring subunit and the side chain of Leu73 in the proximal subunit intercalates between Leu71 and Leu73 in the neighbouring subunit (Fig. 2c). These interactions were not observed in the previous structure owing to different crystal packing. A network of intramolecular hydrogen bonds and water molecules that were not observed in the previous crystal structure further stabilizes the isopeptide-linkage conforma- tion. A water molecule makes hydrogen bonds to the carbonyl O atoms of Gly76 and Gln49 in the distal and proximal moieties, respectively, and another water molecule bridges the side chain of Glu51 with the carbonyl O atom of Gly76 (Fig. 2c). Finally, the carbonyl O atom of Leu73 makes a hydrogen bond to the amide group of Gly76 in the distal moiety. These interactions were observed in all three isopeptide linkages in the asymmetric unit, which thus adopt nearly identical conformations with residues 73–76 (distal) and Lys48 (proximal) forming a long U-shaped loop (Fig. 1c). The con- formation of the isopeptide linkage in the previous structure is similar, but shows significant differences in the backbone torsion structural communications 996 Trempe et al.  Lys48-linked diubiquitin Acta Cryst. (2010). F66, 994–998 Figure 1 Crystal structure of Lys48-linked Ub2. (a) Cartoon representation of a Ub2 molecule in the crystal structure. The proximal and distal moieties are coloured magenta and cyan, respectively. The atoms forming the isopeptide bond as well as the interface residues Ile44 and Val70 are shown as sticks. Residues labelled with primes belong to the distal moiety. (b) Close-up view of the residues forming the interface between the distal and proximal subunits. The molecular surface of the proximal subunit is displayed in transparent white. (c) Cross-eye stereoview ribbon display of the overlaid Ub2 crystal structures. The three chains in the new crystal structure are shaded yellow, blue and red for A–B, C–D and E–F, respectively. The previously reported crystal structure of Ub2 is shaded in magenta (PDB code 1aar; Cook et al., 1992). Residues that have different conformations in different subunits are labelled. The disordered C-termini of the proximal moieties are labelled ‘C’. angles for residues 73–76 (Fig. 3a). The isopeptide bond is in a trans configuration in both crystal structures, but the carbonyl O atom of Gly76 points in opposite directions, which imposes a reconfiguration of Gly75 and Gly76. This emphasizes the flexibility of the isopeptide linkage, which is essential for the function of Ub2 because ubiquitin- binding domains need to access the hydrophobic patches of ubiquitin that are occluded in the closed conformation (Fig. 1a). Solution NMR dynamics studies have indeed shown that the closed conformation of Ub2 experiences fast interdomain motion on a 10 ns timescale (Ryabov & Fushman, 2006). Additional differences are found in the backbones of different Ub2 subunits, notably at the free C-termini of the proximal moieties (B, D and F), which show variable levels of disorder for residues Arg74– Gly76 (Fig. 1c and Table 1). The loop residues Thr9 and Gly10, which are located between the 1 and 2 strands, also adopt a different conformation in chain B compared with the other chains (Figs. 1c and 3b) and the electron density around these residues is weaker in chain B in comparison with the other chains. In the previous crystal structure this loop adopts the conformation observed in chains A, C, D, E and F in the new crystal structure. Interestingly, the chemical environment around Thr9 and Gly10 is nearly identical for all chains, including chain B, with Thr9 being in proximity to Ala46/Gly47 and Ser57/Asp58 in two different neighbouring subunits (not shown). This suggests that the two conformations observed have similar potential energy, with the most frequent being slightly more stable. This loop shows significant backbone dynamics in solution (Lakomek et al., 2006), which is consistent with the variability observed here. structural communications Acta Cryst. (2010). F66, 994–998 Trempe et al.  Lys48-linked diubiquitin 997 Figure 2 Conformation of the isopeptide bond in the crystal structure of Ub2. (a) Cross-eye stereoview of the A-weighted 2Fo  Fc electron-density map at the isopeptide linkage contoured in blue at 0.35 e A˚ 3. The atomic model is drawn as sticks. Water molecules are drawn as red spheres. (b) The three Lys48-linked Ub2 molecules in one asymmetric unit are coloured yellow for chains A–B, blue for chains C–D and red for chains E–F. Distal (A, C and E) and proximal (B, D and F) ubiquitin moieties are distinguished by pale and dark shades, respectively. Chains C0 and D0 are from an adjacent asymmetric unit and are labelled in pale and dark cyan, respectively. The isopeptide linkages are shown as spheres coloured by atom type (white, carbon; blue, nitrogen; red, oxygen). (c) Cross-eye stereoview of the isopeptide bond and its interactions. Residues labelled with primes belong to a distal moiety. Hydrogen bonds are shown as dashed lines. C atoms of chains A–B and E–F are shown in yellow and salmon red, respectively. 4. Conclusions A new crystal form of Lys48-linked Ub2 was obtained and its struc- ture was determined by X-ray crystallography to 1.6 A˚ resolution. The asymmetric unit is composed of three Ub2 molecules that all adopt the closed conformation, as observed in solution (Varadan et al., 2002) and in the previous crystal structure (Cook et al., 1992), despite the different crystallization conditions and crystal packing. The new crystal form reveals a new conformation for the isopeptide linkage, which interacts with other isopeptide linkages in the other subunits. A new conformation was also observed for the loop between the 1 and 2 strands. These local differences emphasize the flexibility of the isopeptide linkage and the 1–2 loop. We would like to thank Professor Kazuhiro Iwai for providing the recombinant baculovirus to express mouse E1, Dr Randy Poon for constructs for the expression of human E2 and the staff at ESRF beamline ID29 for providing excellent facilities for data collection. This work was supported by an MRC grant to JAE and Wellcome Trust and British Overseas Research Studentships to JFT. References Bertolaet, B. L., Clarke, D. J., Wolff, M., Watson, M. H., Henze, M., Divita, G. & Reed, S. I. (2001). Nature Struct. Biol. 8, 417–422. Chen, Z. & Pickart, C. M. (1990). J. Biol. Chem. 265, 21835–21842. Cook, W. J., Jeffrey, L. C., Carson, M., Chen, Z. & Pickart, C. M. (1992). J. Biol. Chem. 267, 16467–16471. Cook, W. J., Jeffrey, L. C., Kasperek, E. & Pickart, C. M. (1994). J. Biol. Chem. 236, 601–609. Deveraux, Q., Ustrell, V., Pickart, C. & Rechsteiner, M. (1994). J. Biol. Chem. 269, 7059–7061. Dijk, A. D. van, Fushman, D. & Bonvin, A. M. (2005). Proteins, 60, 367–381. Emsley, P. & Cowtan, K. (2004). Acta Cryst. D60, 2126–2132. Engh, R. A. & Huber, R. (2001). International Tables for Crystallography, Vol. F, edited by M. G. Rossmann & E. Arnold, pp. 382–392. Dordrecht: Kluwer Academic Publishers. Evans, P. (2006). Acta Cryst. D62, 72–82. Finley, D. (2009). Annu. Rev. Biochem. 78, 477–513. Hershko, A. & Ciechanover, A. (1998). Annu. Rev. Biochem. 67, 425–479. Hicke, L., Schubert, H. L. & Hill, C. P. (2005). Nature Rev. Mol. Cell Biol. 6, 610–621. Husnjak, K., Elsasser, S., Zhang, N., Chen, X., Randles, L., Shi, Y., Hofmann, K., Walters, K. J., Finley, D. & Dikic, I. (2008). Nature (London), 453, 481–488. Kleywegt, G. J. & Jones, T. A. (1996). Structure, 4, 1395–1400. Lakomek, N. A., Carlomagno, T., Becker, S., Griesinger, C. & Meiler, J. (2006). J. Biomol. NMR, 34, 101–115. Leslie, A. G. W. (2006). Acta Cryst. D62, 48–57. McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Storoni, L. C. & Read, R. J. (2007). J. Appl. Cryst. 40, 658–674. Murshudov, G. N., Vagin, A. A. & Dodson, E. J. (1997). Acta Cryst. D53, 240–255. Perrakis, A., Sixma, T. K., Wilson, K. S. & Lamzin, V. S. (1997). Acta Cryst. D53, 448–455. Phillips, C. L., Thrower, J., Pickart, C. M. & Hill, C. P. (2001). Acta Cryst. D57, 341–344. Piotrowski, J., Beal, R., Hoffman, L., Wilkinson, K. D., Cohen, R. E. & Pickart, C. M. (1997). J. Biol. Chem. 272, 23712–23721. Ryabov, Y. & Fushman, D. (2006). Proteins, 63, 787–796. Schreiner, P., Chen, X., Husnjak, K., Randles, L., Zhang, N., Elsasser, S., Finley, D., Dikic, I., Walters, K. J. & Groll, M. (2008). Nature (London), 453, 548–552. Trempe, J. F., Brown, N. R., Lowe, E. D., Gordon, C., Campbell, I. D., Noble, M. E. & Endicott, J. A. (2005). EMBO J. 24, 3178–3189. Vaguine, A. A., Richelle, J. & Wodak, S. J. (1999). Acta Cryst. D55, 191–205. Varadan, R., Assfalg, M., Raasi, S., Pickart, C. & Fushman, D. (2005). Mol. Cell, 18, 687–698. Varadan, R., Walker, O., Pickart, C. & Fushman, D. (2002). J. Mol. Biol. 324, 637–647. Vijay-Kumar, S., Bugg, C. E. & Cook, W. J. (1987). J. Mol. Biol. 194, 531– 544. Wilkinson, C. R., Seeger, M., Hartmann-Petersen, R., Stone, M., Wallace, M., Semple, C. & Gordon, C. (2001). Nature Cell Biol. 3, 939–943. Wu, K., Chen, A., Tan, P. & Pan, Z.-Q. (2002). J. Biol. Chem. 277, 516–527. Zhang, D., Chen, T., Ziv, I., Rosenzweig, R., Matiuhin, Y., Bronner, V., Glickman, M. H. & Fushman, D. (2009). Mol. Cell, 36, 1018–1033. Zhang, N., Wang, Q., Ehlinger, A., Randles, L., Lary, J. W., Kang, Y., Haririnia, A., Storaska, A. J., Cole, J. L., Fushman, D. & Walter, K. J. (2009). Mol. Cell, 35, 280–290. structural communications 998 Trempe et al.  Lys48-linked diubiquitin Acta Cryst. (2010). F66, 994–998 Figure 3 Comparison of loop conformations in different Ub2 crystal structures. (a) Comparison of the isopeptide-bond conformation in the two Ub2 crystal structures. Chains A–B of the new crystal structure are coloured yellow and the previous structure (PDB code 1aar; Cook et al., 1992) is coloured magenta. Residues labelled with primes belong to a distal moiety. The conformation of the isopeptide bond in chains C–D and E–F is similar to that in chains A–B. (b) Comparison of the 1–2 loop conformation in chain B (yellow) and the previous crystal structure (magenta). The conformation of this loop in chains C–D and E–F of the new structure is similar to that shown in magenta.
3M3K
Ligand binding domain (S1S2) of GluA3 (flop)
The molecular mechanism of flop-selectivity and subsite recognition for an AMPA receptor allosteric modulator: Structures of GluA2 and GluA3 complexed with PEPA Ahmed H. Ahmed§, Christopher P. Ptak§, and Robert E. Oswald* Department of Molecular Medicine, Cornell University, Ithaca, NY 14853 USA Abstract Glutamate receptors are important potential drug targets for cognitive enhancement and the treatment of schizophrenia in part because they are the most prevalent excitatory neurotransmitter receptors in the vertebrate central nervous system. One approach to the application of therapeutic agents to the AMPA subtype of glutamate receptors is the use of allosteric modulators, which promote dimerization by binding to a dimer interface thereby reducing desensitization and deactivation. AMPA receptors exist in two alternatively spliced variants (flip and flop) that differ in desensitization and receptor activation profiles. Most of the structural information on modulators of the AMPA receptor target the flip subtype. We report here the crystal structure of the flop-selective allosteric modulator, PEPA, bound to the binding domains of the GluA2 and GluA3 flop isoforms of AMPA receptors. Specific hydrogen bonding patterns can explain the preference for the flop isoform. This includes a bidentate hydrogen bonding pattern between PEPA and N754 of the flop isoforms of GluA2 and GluA3 (the corresponding position in the flip isoform is S754). Comparison with other allosteric modulators provides a framework for the development of new allosteric modulators with preferences for either the flip or flop isoforms. In addition to interactions with N/S754, specific interactions of the sulfonamide with conserved residues in the binding site are characteristics of a number of allosteric modulators. These, in combination, with variable interactions with five subsites on the binding surface lead to different stoichiometries, orientations within the binding pockets, and functional outcomes. Membrane receptors are the cell's gatekeepers, allowing chemical signals access to the cell's pathways. Through the binding of endogenous ligands, receptors identify relevant environmental cues and facilitate cell-cell communication. The regulation of membrane receptors has become an important goal of drug discovery efforts (1,2). By targeting the physiological (orthosteric) ligand-binding site, agonists and antagonists control the function of membrane receptors. Unfortunately, exogenously induced agonist-activation at the orthosteric site can cause toxic effects from overstimulation. Allosteric modulator binding sites use a distinct avenue for altering the natural response of a receptor. The ability of some allosteric modulators to enhance receptor stimulation, while not actually providing the trigger for stimulation, is a clear advantage that conserves the endogenous signaling pathway. Being important mediators of higher-order processes such as learning and memory, ionotropic glutamate receptors (iGluRs) have attracted a great deal of interest as allosteric modulator targets (3–6). Of clear therapeutic importance, various neurodegenerative disorders such as Parkinson's and Alzheimer's diseases, Huntington's *Corresponding author; telephone: 1-607-253-3877; fax: 1-607-253-3659; email: reo1@cornell.edu. §These authors contributed equally to this work. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 April 7. Published in final edited form as: Biochemistry. 2010 April 6; 49(13): 2843–2850. doi:10.1021/bi1000678. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript chorea, and neurologic disorders including epilepsy and ischemic brain damage have been linked to iGluRs (7). The crystal structure of GluA2 (8) clarifies years of speculation on the complex arrangement of the glutamate receptor's four subunits (9). The GluA2 can be dissected into 3 functionally distinct layers. Farthest from the membrane, the amino terminal domain (ATD) can act as a peripheral regulatory domain but is also involved in assembly and trafficking (10,11). Sandwiched between the ATD and the membrane domain, the ligand-binding domain (LBD) recognizes the neurotransmitter signal and directly regulates receptor activation (12). Structures for both isolated extracellular domains (ATD and LBD) reveal a dimeric organization (13–15). At the membrane interface, two alternative linker conformations transition the 2-fold symmetry, which is adopted by both extracellular domains, into the 4- fold symmetry of a membrane-traversing cation-selective channel (8,16). For iGluRs, the ion channel domain confers functional relevance with its ability to selectively conduct the flow of ions across the cell's membrane. The layers of extracellular domains, each with the potential for multiple control points, allosterically regulate the ion channel domain's function (8). Therefore it is not surprising that the ATD, the LBD, and the LBD-channel linker have all been shown to be effective targets of allosteric modulators (13,17,18). Since the structures of the ATD and the full iGluR channel have only recently been solved, allosteric drug-binding sites external to the LBD have not been fully explored in molecular detail. However, the decade-old LBD structure has proved to be indispensable as a heavily exploited scaffold for understanding agonist, partial agonist, and antagonist binding interactions as well as their ability to regulate channel gating behavior (12,19,20). Although the dimeric organization is consistent across all iGluR subtypes, the molecular details of LBD-agonist specificity define the subtype families into N-methyl-D-aspartic acid (NMDA) receptors (21), α-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid (AMPA) receptors (12), and kainate receptors (22). Because all subtypes are constrained by their conserved sensitivity to glutamate stimulation, diversity at the orthosteric site is evolutionarily limited and most agonists display cross-subtype activity. An allosteric modulator-binding site within the quaternary LBD structure is located along the dimer interface (18) and offers improved discrimination by modulators. Drugs that bind to the allosteric sites on the LBD dimer interface can enhance the activity of iGluRs (23) and increase performance on tests of memory (24). Except for the LBD structures with modulatory ions bound to the dimer interface (25–27), only LBD structures from the AMPA receptor subtype, GluA2, have been reported with bound allosteric modulators (18,28–31). Within the structures, the bound modulatory drugs stabilize the LBD dimer interface, which is required for activation of the ion channel and is dissociated during desensitization (18). Although the residues that line the allosteric modulator-binding pocket do not differ between AMPA receptors subtypes (GluA1–4), the ability of allosteric modulators to stabilize the activated state still varies (32,33). Also, AMPA receptors can be alternatively spliced into what is referred to as flip and flop isoforms (34). Modulator selectivity (23), desensitization (35), and channel closing rates (36) differ between flip and flop. Although several of the amino acid differences between the two forms are located in or near the allosteric modulator-binding site, the difference at position 754 (serine in flip, asparagine in flop) seems to be entirely responsible for the functional differences between allosteric modulator regulation of the flip and flop variants (23,28,32). Cyclothiazide (CTZ) and some other thiazide derivatives have improved binding to the flip form due to a hydrogen bond between S754 and the NH of the fused thiazide ring (28). In the case of the flop form, the alternatively spliced sequence places an asparagine in the 754 position, which is not optimally positioned to form a hydrogen bond. Sekiguchi et al. (33) introduced an allosteric modulator of AMPA receptors (4-[2-(phenylsulphonylamino)ethylthio]-2,6,- Ahmed et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript difluorophenoxyacetamide, PEPA) with a preference for the flop form. In fact, the relative sensitivity of CTZ to PEPA has been used as a diagnostic for the prevalence of flip vs. flop versions of AMPA receptor in particular cell types (37). PEPA shows potential in treatment of post ischemic memory impairment (38) and contextual fear (39) but despite PEPA's unique flop sensitivity, the modulator has not yet been used as a lead compound in SAR studies. For drug discovery to be guided by structures, understanding the possible molecular interactions between modulators and the dimer interface is essential. We have shown previously (31) that changes in the structures of CTZ derivatives can reorient the modulator within the binding site. Subsequently, we proposed that the allosteric modulator site is comprised of 5 subsites (Figure 1C). In the present study, we determine the three dimensional structures of PEPA bound to the GluA2o and GluA3o LBDs (flop forms), and use PEPA's binding interactions to further characterize the subsite specific binding properties displayed by allosteric modulators. The amide group of PEPA makes a direct hydrogen bond to N754, explaining the preferential action of PEPA on the flop form of AMPA receptors. Another key structural element, the sulfonamide group of PEPA, is conserved with the biarylsulfonamide class of allosteric modulators (6) and interacts with the same residues of the dimer interface (8,30). Although previously classified as unrelated, PEPA and the large group of biarylsulfonamide have similarities, which suggest that specific PEPA groups (particularly the unique flop-interacting amide) can be strategically integrated into biarylsulfonamide SAR studies. Experimental Procedures Materials PEPA was purchased from Tocris (Ellisville, MO). The GluA2 S1S2J construct was obtained from Eric Gouaux (Vollum Institute; 12). Protein Preparation and Purification GluA2 S1S2 consists of residues N392 - K506 and P632 - S775 of the full rat GluA2o subunit (40), a `GA' segment at the N-terminus, and a `GT' linker connecting K506 and P632 (12). A similar construct of GluA3 S1S2 was prepared as described previously (41). pET-22b(+) plasmids were transformed in E. coli strain Origami B (DE3) cells and were grown at 37°C to OD600 of 0.9 to 1.0 in LB medium supplemented with the antibiotics (ampicillin and kanamycin). The cultures were cooled to 20°C for 20 min. and isopropyl-β- D-thiogalactoside (IPTG) was added to a final concentration of 0.5 mM. Cultures were allowed to grow at 20°C for 20 h. The cells were then pelleted and the S1S2 protein purified using a Ni-NTA column, followed by a sizing column (Superose 12, XK 26/100), and finally an HT-SP-ion exchange-Sepharose column (Amersham Pharmacia). Glutamate (1 mM) was maintained in all buffers throughout purification. After the last column, the protein was concentrated and stored in 20 mM sodium acetate, 1 mM sodium azide, and 10 mM glutamate at pH 5.5. Crystallography For crystallization trials, the protein was concentrated to 0.2 – 0.5 mM in 10 mM glutamate using a Centricon 10 centrifugal filter (Millipore, Bedford, MA). For the PEPA-bound structures, PEPA was added to 5 mM. The final protein concentration was 0.2 to 0.3 mM. Crystals were grown at 4°C using the hanging drop technique, and the drops contained a 1:1 (v/v) ratio of protein solution to reservoir solution. The reservoir solution contained 14–15% PEG 8K, 0.1 M sodium cacodylate, 0.1–0.15 M zinc acetate, and 0.25 M ammonium sulfate, pH 6.5. Ahmed et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Data were collected at the Cornell High Energy Synchrotron Source beam line A1 using a Quantum-210 Area Detector Systems charge-coupled device detector. Data sets were indexed and scaled with HKL-2000 (42). Structures were solved with molecular replacement using Phenix (43). Refinement was performed with Phenix (43), and Coot 0.5 (44) was used for model building. Results Structure of PEPA bound to GluA2 S1S2 flop The structure of glutamate bound to GluA2o S1S2 (3dp6; 41) was used as the initial search probe for the molecular replacement solution of PEPA bound to GluA2o S1S2 with glutamate in the agonist-binding site. PEPA was then modeled into two symmetrical positions within the density found at the dimer interface, and the structure was optimized using Phenix (43). The refinement statistics are given in Table 1. The resolution is 1.85 Å, and three unique copies are found in the unit cell. The overall structure of the S1S2 domain is very similar to the structure in the absence of PEPA, with contacts between glutamate and the protein unchanged. However, PEPA clearly binds within the dimer interface, making contacts with both monomers within the dimer. As shown in Figure 1, one PEPA molecule binds per dimer interface. However because the dimer interface is symmetrical, two equivalent orientations (related by a 180° rotation) are possible. Electron density for both is seen in the crystal structure, although the intensity of one orientation is greater than the other. The binding of PEPA to the dimer interface increases the distance between the two monomers that form the dimer by approximately 1.5 Å. This allows the relatively large PEPA molecule to fit within the interface, but also increases the separation between the linkers to the ion channel (the distance increases from 39.4 Å to 41 Å; Figure 1A). Relative to the core of Lobe 1, both the J/K helices and one β strand (P105-G110) connecting the two lobes are displaced slightly away from the dimer interface (Figure 1B). In addition, Lobe 2 is slightly twisted relative to glutamate-bound S1S2 in the absence of PEPA (3dp6; 41). PEPA binds at the bottom of a water-filled, inverted U-shaped cleft with five subsites (A, B/ B′, and C/C′; 31). Upon binding, crystallographic waters are displaced from the central A subsite and more buried C/C′ sites, with the waters in the B/B′ subsite remaining (Figure 1C). This displacement of presumably ordered water would be likely to contribute a favorable entropy component to binding. The sidechains of P494 are at the center of the interface and the edge of the two proline rings from each monomer form the base of the binding site in which the difluorophenyl ring resides (Figure 2A). This is close to the position of the methoxybenzoyl ring of aniracetam in its structure bound to GluA2-S1S2(FW) (29). The other side of the ring is exposed to S497 and S729. The sidechain hydroxyl of S497 is oriented toward the dimer interface in the absence of PEPA, but rotates out toward the solvent to accommodate the difluorophenyl ring of PEPA (Figure 1B). The amide of PEPA is involved in a network of hydrogen bonds with sidechain hydroxyl of Y424, the backbone carbonyl of F495, the sidechain carboxyl of D760, the sidechain amide of N754, and two water molecules (Figure 2A). The most striking of these hydrogen bond pairs is with N754. This represents the only difference between the flip (S754) and flop (N754) isoforms in the PEPA binding site and is almost certainly a major source of the preference for the flop isoform. The phenyl-sulfonylamide side of PEPA inserts into a hydrophobic pocket formed by sidechain methyls of I481 and L751 as well as methylene groups contributed by K493, N754, and E755 (Figure 2B). It is possible that the contribution by methylene group of N754 provides a more hydrophobic pocket than S754 in the flip form, further contributing to the preference for the flop form. Ahmed et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Because the dimer interface is symmetrical, PEPA can bind in two orientations and both are observed in the crystal. For this reason, changes in the protein due to a specific interaction with PEPA can be partially masked because each monomer is a weighted average of two orientations of bound PEPA. However, one orientation has a stronger density than the other, providing some insight into the extent of changes in the dimer interface that are produced by PEPA binding. As shown in Figure 2C, the two monomers comprising the dimer differ more within the PEPA binding site than the corresponding monomers in the absence of PEPA. One turn of helix J (L751 to N754) contains important determinants for both orientations of PEPA. In one orientation the amide group of PEPA interacts with the sidechain of N754, and in the other, the aromatic ring of PEPA inserts between a hydrophobic pocket formed by the sidechain of L751 and the methylene group of N754. In the orientation for which the density of the amide of PEPA is stronger, N754 is better positioned to form an H-bond (Figure 2D); whereas, in the other side of the interface, N754 is oriented to form an H-bond with the carbonyl of S729. This change in orientation facilitates the insertion of the aromatic ring of PEPA into the hydrophobic pocket, which is accompanied by a small shift in the sidechain of L751 to accommodate the aromatic ring (Figure 2D). Since these structures are weighted averages, it is possible that the actual positions of these sidechains involve an even greater movement than is seen from the asymmetry of the crystal. Structure of PEPA bound to GluA3 S1S2 flop In studies of the physiological effects of PEPA, a significant difference between subtypes has been observed, with GluA3 being most susceptible to modulation (33). The structure of GluA3i S1S2 bound (flip form) to glutamate has been reported previously (41). Since PEPA preferentially binds to the flop form, the GluA3o structure was determined bound to glutamate with and without PEPA (Figure 3A). Like GluA2o, in the absence of PEPA, GluA3o has three copies in the asymmetric unit. Comparing lobe closure between GluA3i and GluA3o, the flop form is slightly more closed (1.6° ± 0.7°). In the presence of PEPA, GluA3o was present in one copy in the asymmetric unit, and PEPA was observed with the same density in two symmetrical orientations. Like GluA2o bound to PEPA, the dimer interface (assessed using the symmetrical molecule in the crystal) was displaced relative to the unbound from (Figure 3A) by approximately 2.5 Å at the position of the linker replacing the ion channel domain. Within the binding site, three sidechains exhibited different rotamers compared with the GluA2o structure bound to PEPA (Figure 3B). For PEPA-bound GluA3o, both S497 and S729 assumed rotameric states that differed both from GluA2o bound to PEPA and from GluA2o and GluA3o in the absence of PEPA. In the case of S729, the rotameric state in combination with a slight movement of the amide of PEPA (relative to the GluA2o structure) would make an H-bond with the sidechain of S729 (shown in Figure 2A for GluA2o) unlikely. In the case of N754, the sidechain is displaced relative to the GluA2o-PEPA structure so that only one H-bond is made to the amide of PEPA. This may be a result of averaging of the two orientations of PEPA only one of which forms a bidentate H-bond with N754. Discussion The goal of allosteric modulation, like orthosteric modulation, is often to stabilize a conformational state of a dynamic protein (45). The activated state of iGluRs is naturally unstable allowing the channel to desensitize (46). Disruption of the symmetrical dimer interface between LBDs is thought to initiate desensitization-mediated channel closure (47). By maintaining the LBD dimer, positive allosteric modulators can prevent desensitization and prolong activation (18). Currently, 15 crystal structures of the GluA2 LBD with bound allosteric modulators are deposited in the Protein Data Bank (48). All of these modulators bind to a large crevice with 2-fold symmetry along the symmetric dimer interface (18). The Ahmed et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript large variation in structure among allosteric modulators results in significant variations in binding orientations and interactions. At least four distinct binding modes have been identified: (1) A-subsite class (aniracetam, CX614 (29)), (2) classical thiazide (cyclothiazide (18), TCMZ, ALTZ (31)), (3) the shifted thiazide class (IDRA-21, HCMZ, HFMZ; (31)), and (4) the full spanning class (PEPA (this paper), dimeric biarylpropylsulfonamide (30), LY404187 (8)). Overlaying modulators from these structural classes has led to the proposal that the allosteric modulator site is comprised of a series of subsites (Figure 1C;31). Positioned at the center of the binding-site, the symmetric A subsite is narrow and allows entrance to only one molecule. Two subsites (B and C) lie at each end of the A subsite with the hydrophobic C subsite located more deeply in the pocket effectively defining five subsites (A, B, B′, C, and C′). In the open state, the subsites are filled with water, which may act to weakly stabilize the dimer. Allosteric modulators generate stronger interactions across the subsites thereby increasing the linkages between the monomers. The simplest modulator class, including aniracetam and CX614, fills the A subsite with one molecule but does not enter the peripheral B and C subsites (29). The two classes of thiazide-based modulators account for 10 of the 15 solved allosteric modulator-GluA2 crystal structure complexes (18,28,31). The classical thiazide (CTZ-like) binding class and the shifted thiazide (IDRA-21-like) binding class are positioned respectively in the B and C subsite or mainly the C subsite. Most of the thiazide modulators do not extend across the A subsite and therefore can bind two molecules per dimer. However, a few of the newly described shifted thiazides (HFMZ, HCTZ; 31) enter the A subsite but only enough to impair binding of a second modulator. The dimeric biarylpropylsulfonamide compound ((R,R)-N,N-(2,2'-[Biphenyl-4-4'-Diyl]Bis[Propane-2,1- Diyl]) Dimethanesulfonamide) described by Kaae et al. (30) was the first allosteric modulator shown by crystallography to extend along the entire length of the inner dimer cavity from C to C′ subsites. PEPA also interacts with J helices from both monomers, which cap the ends of the modulator-binding pocket. The density occupied by both symmetrical copies of PEPA overlays the dimeric biarylsulfonamide compound as both modulators represent the full spanning class (Figure 4B). The GluA flip and flop splice variants differ by only a few residues along the J helix in the LBD; however, residue 754 (Asn in flop and Ser in flip) is positioned between the B and C subsites. For thiazides, a clear preference in binding to the flip-form is mediated by a hydrogen bond between the hydrobenzothiadiazide ring and S754 (28). In contrast, PEPA is flop-selective and the PEPA-bound structure provides the first structure containing a direct interaction between a modulator and the flop form's N754. The amide of PEPA extends straight out from the A subsite and across the B and C subsite interface to make an amide- amide hydrogen bond with N754 (Figure 2A). Unlike most other AMPA modulators, PEPA fills neither the B nor the C subsites but interacts directly with the J helix. A similar interaction is seen with LY404187 (49) bound to GluA2i (8). Strong hydrogen bonding can occur between two amides (50) and has been shown to be responsible for driving oligomerization of transmembrane leucine zippers (51). The distances between the interacting amides in the PEPA-bound structure support a bidentate hydrogen-bonding pattern, which is much stronger and more specific than a typical hydrogen bond. While PEPA is selective for the AMPA receptor's flop form, a weaker but still existent potentiation of the flip form has been observed (33,52). Replacing N754 (flop) with S754 (flip) would not prevent PEPA from binding; however, serine would provide only one hydrogen-bonding partner for PEPA's amide with an extended interaction distance. In contrast, LY404187 displays a preference for the flip isoform (53), and its cyano group extends out to interact directly with S754. The cyano-S754 interaction is a clear flip analog of the flop-selective PEPA amide-N754 interaction (Figure 4A). Ahmed et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Opposite to the amide on the PEPA molecule, a sulfonamide is tethered to the difluorophenyl ring (Figure 4A). Within the dimer interface, the sulfonamide is positioned so the nitrogen can hydrogen bond directly with the carbonyl of P494 (Figure 4C). A sulfonamide oxygen points toward the amide nitrogen of G731. The angle of the peptide plane is perpendicular to the sulfonamide oxygen, making a hydrogen bonding interaction unlikely (Figure 4D). Instead, a dipole-dipole or charge-dipole interaction may occur. The amide nitrogen of a polypeptide supports at least a partial positive charge (54), which would interact with the strongly electronegative sulfonamide oxygen (55). Interestingly, both the dimeric biarylsulfonamide (30) and LY404187 (8), other members of the full spanning modulator class, also have a sulfonamide that interacts with the same backbone atoms of P494 and G731 as PEPA (Figure 4C). A large number of biarylsulfonamides have been identified that modulate AMPA receptors and are being evaluated for therapeutic use in the treatment of depression and Parkinson's disease (56). The conserved sulfonamide reveals a previously unidentified relationship between PEPA and the biarylsulfonamide modulators. When the perpendicular peptide bond plane including G731 is fixed, the sulfonamide on three overlayed modulators varies by 1.2 Å along the length of the interface with the PEPA sulfonamide being positioned closer to the A subsite (Figure 4D). A shift of the sulfonamide also results in a shift in the corresponding P494 across the interface presumably to maintain the hydrogen bond with the modulator's amine. The sulfonamide forms an important bridge between the two dimer halves. For PEPA, a phenyl-sulfonamide replaces the methyl-sulfonamide in the dimeric biarylsulfonamide and fits snuggly against L751. Based on the orientation-induced asymmetry within the GluA2-complex structure, the phenyl pushes the J helix away from PEPA thereby affecting the C subsite (Figure 2C and D). Residues lining the C subsite are on the same beta strand as G731, which must shift if the C subsite is to remain together and presumably explain the 1.2 Å shift relative to the dimeric biarylsulfonamide. In fact, the same phenyl-sulfonamide group substitution in a biarylpropylsulfonamide decreases the modulatory effect of the derivative in SAR studies (57). For biarylpropylsulfonamides, the optimal sulfonamide substitution was found to be either an ethyl or an iso-propyl group, which should both fit without significantly disrupting the J helix or C subsite (57). The PEPA-bound crystal structure from AMPA receptor subtypes, GluA2 and GluA3, do not display major differences in binding interactions even though PEPA exhibits a stronger effect on GluA3 (33). For GluA2, an asymmetry in the receptor-binding pocket was observed while no significant difference in PEPA density was seen for the each orientation within the GluA3 crystal structure. In addition, a number of side chains exhibit different rotameric states between the two structures, although it is unlikely that these small changes significantly impact the differential effects on the two subtypes. Although no structural differences have been identified between GluA2 and GluA3 that would obviously impact PEPA affinity, the possibility exists that subtle differences arising from the sequence differences peripheral to the binding site may be important as has been described in the case of the agonist binding site of GluA4 (58). We have explored how PEPA (this paper) and other allosteric modulators (31) interact with the GluA interface in the context of drug design. Together the identification of a conserved group between PEPA (this paper) and biarylpropylsulfonamides (8,30) and the regional nature of various subsite-functional group interactions provide a backdrop to extend biarylpropylsulfonamide SAR studies (57) to include PEPA and biarylpropylsulfonamide chimeras. Although optimizing the stability of the dimer interface provides a starting point for SAR studies, additional constraints should be considered including the ability of the modulator to enter the cavity, the dynamic structure of the dimer interface during closed, open, and desensitized state transitions, and the ability of the modulator to cross the blood- Ahmed et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript brain barrier before being metabolized. This definition of the allosteric modulator binding- site should provide guidance in glutamate receptor allosteric modulator pharmacology. Acknowledgments We thank Prof. Eric Gouaux (Vollum Institute) for the GluA2 S1S2J construct, and Prof. Linda Nowak (Cornell) for the full-length GluA3 construct. This work was supported by a grants from the National Institutes of Health (R01-GM068935, R01 NS049223, and R21 NS067562). This work is based upon research conducted at the Cornell High Energy Synchrotron Source (CHESS), which is supported by the National Science Foundation under award DMR 0225180, using the Macromolecular Diffraction at the CHESS (MacCHESS) facility, which is supported by award RR-01646 from the National Institutes of Health, through its National Center for Research Resources. Abbreviations ALTZ althiazide AMPA α-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid CLTZ chlorothiazide CX614 pyrrolidino-1,3-oxazino benzo-1,4-dioxan-10-one CTZ cyclothiazide FW (S)-5-fluorowillardiine flip and flop alternatively spliced versions of AMPA receptors that vary in rates of desensitization and sensitivity to allosteric modulators iGluR ionotropic glutamate receptor GluA1-4 four subtypes of AMPA receptor HCTZ hydrochlorothiazide HFMZ hydroflumethiazide IDRA-21 7-chloro-3-methyl-3,4-dihydro-2H-benzo[e][1,2,4]thiadiazine 1,1-dioxide IPTG isopropyl-β-D-thiogalactoside LY404187 N-[2-(4′-cyanobiphenyl-4-yl)propyl]propane-2-sulfamide PEPA 4-[2-(phenylsulphonylamino)ethylthio]-2,6,-difluorophenoxy acetamide NMDA N-methyl-D-aspartic acid S1S2 extracellular ligand-binding domain of GluA2 and GluA3 SAR structure-activity relationships TCMZ trichlormethiazide References 1. Christopoulos A. Allosteric binding sites on cell-surface receptors: novel targets for drug discovery. Nat Rev Drug Discov. 2002; 1:198–210. [PubMed: 12120504] 2. Changeux JP, Taly A. Nicotinic receptors, allosteric proteins and medicine. Trends Mol Med. 2008; 14:93–102. [PubMed: 18262468] 3. Bowie D. Ionotropic glutamate receptors & CNS disorders. CNS Neurol Disord Drug Targets. 2008; 7:129–143. [PubMed: 18537642] 4. Dingledine R, Borges K, Bowie D, Traynelis S. The glutamate receptor ion channels. Pharmacol. Rev. 1999; 51:7–61. [PubMed: 10049997] Ahmed et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 5. Oswald RE, Ahmed A, Fenwick MK, Loh AP. Structure of glutamate receptors. Current drug targets. 2007; 8:573–582. [PubMed: 17504102] 6. Grigoriev VV, Proshin AN, Kinzirsky AS, Bachurin SO. Modern approaches to the design of memory and cognitive enhancers based on AMPA receptor ligands. Russian Chemical Reviews. 2009; 78:485–494. 7. Black MD. Therapeutic potential of positive AMPA modulators and their relationship to AMPA receptor subunits. A review of preclinical data. Psychopharmacology (Berl). 2005; 179:154–163. [PubMed: 15672275] 8. Sobolevsky AI, Rosconi MP, Gouaux E. X-ray structure, symmetry and mechanism of an AMPA- subtype glutamate receptor. Nature. 2009; 462:745–756. [PubMed: 19946266] 9. Wo ZG, Oswald RE. Unraveling the modular design of glutamate-gated ion channels. Trends Neurosciences. 1995; 18:161–168. 10. Gielen M, Siegler Retchless B, Mony L, Johnson JW, Paoletti P. Mechanism of differential control of NMDA receptor activity by NR2 subunits. Nature. 2009; 459:703–707. [PubMed: 19404260] 11. Greger IH, Ziff EB, Penn AC. Molecular determinants of AMPA receptor subunit assembly. Trends Neurosci. 2007; 30:407–416. [PubMed: 17629578] 12. Armstrong N, Gouaux E. Mechanisms for activation and antagonism of an AMPA-sensitive glutamate receptor: crystal structures of the GluR2 ligand binding core. Neuron. 2000; 28:165– 181. [PubMed: 11086992] 13. Clayton A, Siebold C, Gilbert RJ, Sutton GC, Harlos K, McIlhinney RA, Jones EY, Aricescu AR. Crystal structure of the GluR2 amino-terminal domain provides insights into the architecture and assembly of ionotropic glutamate receptors. J Mol Biol. 2009; 392:1125–1132. [PubMed: 19651138] 14. Jin R, Singh SK, Gu S, Furukawa H, Sobolevsky AI, Zhou J, Jin Y, Gouaux E. Crystal structure and association behaviour of the GluR2 amino-terminal domain. EMBO J. 2009; 28:1812–1823. [PubMed: 19461580] 15. Kumar J, Schuck P, Jin R, Mayer ML. The N-terminal domain of GluR6-subtype glutamate receptor ion channels. Nat Struct Mol Biol. 2009; 16:631–638. [PubMed: 19465914] 16. Doyle DA, Cabral JM, Pfuetzner RA, Kuo AL, Gulbis JM, Cohen SL, Chait BT, MacKinnon R. The structure of a potassium channel: molecular basis of K+ conduction and selectivity. Science. 1998; 280:69–77. [PubMed: 9525859] 17. Balannik V, Menniti FS, Paternain AV, Lerma J, Stern-Bach Y. Molecular mechanism of AMPA receptor noncompetitive antagonism. Neuron. 2005; 48:279–288. [PubMed: 16242408] 18. Sun Y, Olson R, Horning M, Armstrong N, Mayer M, Gouaux E. Mechanism of glutamate receptor desensitization. Nature. 2002; 417:245–253. [PubMed: 12015593] 19. Ahmed A, Thompson M, Fenwick M, Romero B, Loh A, Jane D, Sondermann H, Oswald R. Mechanisms of antagonism of the GluR2 AMPA receptor: Structure and dynamics of the complex of two willardiine antagonists with the glutamate binding domain. Biochemistry. 2009; 48:3894– 3903. [PubMed: 19284741] 20. Jin R, Banke TG, Mayer ML, Traynelis SF, Gouaux E. Structural basis for partial agonist action at ionotropic glutamate receptors. Nat Neurosci. 2003; 6:803–810. [PubMed: 12872125] 21. Furukawa H, Gouaux E. Mechanisms of activation, inhibition and specificity: crystal structures of the NMDA receptor NR1 ligand-binding core. EMBO J. 2003; 22:2873–2885. [PubMed: 12805203] 22. Mayer ML. Crystal structures of the GluR5 and GluR6 ligand binding cores: Molecular mechanisms underlying kainate receptor selectivity. Neuron. 2005; 45:539–552. [PubMed: 15721240] 23. Partin KM, Fleck MW, Mayer ML. AMPA receptor flip/flop mutants affecting deactivation, desensitization, and modulation by cyclothiazide, aniracetam, and thiocyanate. J Neurosci. 1996; 16:6634–6647. [PubMed: 8824304] 24. Martin JR, Cumin R, Aschwanden W, Moreau JL, Jenck F, Haefely WE. Aniracetam improves radial maze performance in rats. Neuroreport. 1992; 3:81–83. [PubMed: 1611039] Ahmed et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 25. Naur P, Vestergaard B, Skov LK, Egebjerg J, Gajhede M, Kastrup JS. Crystal structure of the kainate receptor GluR5 ligand-binding core in complex with (S)-glutamate. FEBS Lett. 2005; 579:1154–1160. [PubMed: 15710405] 26. Plested AJ, Mayer ML. Structure and mechanism of kainate receptor modulation by anions. Neuron. 2007; 53:829–841. [PubMed: 17359918] 27. Plested AJ, Vijayan R, Biggin PC, Mayer ML. Molecular basis of kainate receptor modulation by sodium. Neuron. 2008; 58:720–735. [PubMed: 18549784] 28. Hald H, Ahring PK, Timmermann DB, Liljefors T, Gajhede M, Kastrup JS. Distinct Structural Features of Cyclothiazide are Responsible for Effects on Peak Current Amplitude and Desensitization Kinetics at iGluR2. J Mol Biol. 2009; 391:906–917. [PubMed: 19591837] 29. Jin R, Clark S, Weeks AM, Dudman JT, Gouaux E, Partin KM. Mechanism of positive allosteric modulators acting on AMPA receptors. J Neurosci. 2005; 25:9027–9036. [PubMed: 16192394] 30. Kaae BH, Harpsoe K, Kastrup JS, Sanz AC, Pickering DS, Metzler B, Clausen RP, Gajhede M, Sauerberg P, Liljefors T, Madsen U. Structural proof of a dimeric positive modulator bridging two identical AMPA receptor-binding sites. Chemistry & biology. 2007; 14:1294–1303. [PubMed: 18022568] 31. Ptak CP, Ahmed AH, Oswald RE. Probing the allosteric modulator binding site of GluR2 with thiazide derivatives. Biochemistry. 2009; 48:8594–8602. [PubMed: 19673491] 32. Partin KM, Bowie D, Mayer ML. Structural determinants of allosteric regulation in alternatively spliced AMPA receptors. Neuron. 1995; 14:833–843. [PubMed: 7718245] 33. Sekiguchi M, Fleck MW, Mayer ML, Takeo J, Chiba Y, Yamashita S, Wada K. A novel allosteric potentiator of AMPA receptors: 4-[2-(phenylsulfonylamino)ethylthio]-2,6-difluoro- phenoxyacetamide. J Neurosci. 1997; 17:5760–5771. [PubMed: 9221774] 34. Sommer B, Keinänen K, Verdoorn TA, Wisden W, Burnashev N, Herb A, Köhler M, Takagi T, Sakmann G, Seeburg PH. Flip and flop: A cell-specific functional switch in glutamate-operated channels of the CNS. Science. 1990; 249:1580–1584. [PubMed: 1699275] 35. Mosbacher J, Schoepfer R, Monyer H, Burnashev N, Seeburg PH, Ruppersberg JP. A molecular determinant for submillisecond desensitization in glutamate receptors. Science. 1994; 266:1059– 1062. [PubMed: 7973663] 36. Pei W, Huang Z, Niu L. GluR3 flip and flop: differences in channel opening kinetics. Biochemistry. 2007; 46:2027–2036. [PubMed: 17256974] 37. Sekiguchi M, Takeo J, Harada T, Morimoto T, Kudo Y, Yamashita S, Kohsaka S, Wada K. Pharmacological detection of AMPA receptor heterogeneity by use of two allosteric potentiators in rat hippocampal cultures. Br J Pharmacol. 1998; 123:1294–1303. [PubMed: 9579722] 38. Sekiguchi M, Yamada K, Jin J, Hachitanda M, Murata Y, Namura S, Kamichi S, Kimura I, Wada K. The AMPA receptor allosteric potentiator PEPA ameliorates post-ischemic memory impairment. Neuroreport. 2001; 12:2947–2950. [PubMed: 11588608] 39. Zushida K, Sakurai M, Wada K, Sekiguchi M. Facilitation of extinction learning for contextual fear memory by PEPA: a potentiator of AMPA receptors. J Neurosci. 2007; 27:158–166. [PubMed: 17202483] 40. Hollmann M, Heinemann S. Cloned glutamate receptors. Annu Rev Neurosci. 1994; 17:31–108. [PubMed: 8210177] 41. Ahmed AH, Wang Q, Sondermann H, Oswald RE. Structure of the S1S2 glutamate binding domain of GluR3. Proteins: Structure, Function, and Bioinformatics. 2009; 75:628–637. 42. Otwinowski, Z.; Minor, W. Processing of X-ray diffraction data collected in oscillation mode. In: Carter, CW.; Sweet, RM., editors. Methods in Enzymology, Vol. 276, Macromolecular Crystallography, part A. Academic Press; New York: 1997. p. 307-326. 43. Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Terwilliger TC. PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr D Biol Crystallogr. 2002; 58:1948– 1954. [PubMed: 12393927] 44. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] Ahmed et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 45. Gunasekaran K, Ma B, Nussinov R. Is allostery an intrinsic property of all dynamic proteins? Proteins. 2004; 57:433–443. [PubMed: 15382234] 46. Trussell LO, Fischbach GD. Glutamate receptor desensitization and its role in synaptic transmission. Neuron. 1989; 3:209–218. [PubMed: 2576213] 47. Armstrong N, Jasti J, Beich-Frandsen M, Gouaux E. Measurement of conformational changes accompanying desensitization in an ionotropic glutamate receptor. Cell. 2006; 127:85–97. [PubMed: 17018279] 48. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, Shindyalov IN, Bourne PE. The Protein Data Bank. Nucleic Acids Res. 2000; 28:235–242. [PubMed: 10592235] 49. Miu P, Jarvie KR, Radhakrishnan V, Gates MR, Ogden A, Ornstein PL, Zarrinmayeh H, Ho K, Peters D, Grabell J, Gupta A, Zimmerman DM, Bleakman D. Novel AMPA receptor potentiators LY392098 and LY404187: effects on recombinant human AMPA receptors in vitro. Neuropharmacology. 2001; 40:976–983. [PubMed: 11406188] 50. Shimoni L, Glusker JP. Hydrogen bonding motifs of protein side chains: descriptions of binding of arginine and amide groups. Protein Sci. 1995; 4:65–74. [PubMed: 7773178] 51. Zhou FX, Cocco MJ, Russ WP, Brunger AT, Engelman DM. Interhelical hydrogen bonding drives strong interactions in membrane proteins. Nat Struct Biol. 2000; 7:154–160. [PubMed: 10655619] 52. Sekiguchi M, Nishikawa K, Aoki S, Wada K. A desensitization-selective potentiator of AMPA- type glutamate receptors. Br J Pharmacol. 2002; 136:1033–1041. [PubMed: 12145103] 53. Quirk JC, Nisenbaum ES. Multiple molecular determinants for allosteric modulation of alternatively spliced AMPA receptors. J Neurosci. 2003; 23:10953–10962. [PubMed: 14645491] 54. Kemnitz CR, Loewen MJ. “Amide resonance” correlates with a breadth of C-N rotation barriers. J Am Chem Soc. 2007; 129:2521–2528. [PubMed: 17295481] 55. Cazenave-Gassiot A, Boughtflower R, Caldwell J, Coxhead R, Hitzel L, Lane S, Oakley P, Holyoak C, Pullen F, Langley GJ. Prediction of retention for sulfonamides in supercritical fluid chromatography. J Chromatogr A. 2008; 1189:254–265. [PubMed: 17977551] 56. O'Neill MJ, Witkin JM. AMPA receptor potentiators: application for depression and Parkinson's disease. Current drug targets. 2007; 8:603–620. [PubMed: 17504104] 57. Ornstein PL, Zimmerman DM, Arnold MB, Bleisch TJ, Cantrell B, Simon R, Zarrinmayeh H, Baker SR, Gates M, Tizzano JP, Bleakman D, Mandelzys A, Jarvie KR, Ho K, Deverill M, Kamboj RK. Biarylpropylsulfonamides as novel, potent potentiators of 2-amino-3- (5-methyl-3- hydroxyisoxazol-4-yl)- propanoic acid (AMPA) receptors. J Med Chem. 2000; 43:4354–4358. [PubMed: 11087558] 58. Gill A, Birdsey-Benson A, Jones BL, Henderson LP, Madden DR. Correlating AMPA receptor activation and cleft closure across subunits: crystal structures of the GluR4 ligand-binding domain in complex with full and partial agonists. Biochemistry. 2008; 47:13831–13841. [PubMed: 19102704] Ahmed et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. (A) Comparison of glutamate-bound GluA2o S1S2 in the presence (blue) and the absence of PEPA (green) in two orientations. Both orientations of PEPA are shown. Note that the binding of PEPA results in a separation of the two components of the dimer (distance between the Cα atoms of the threonine in the linker) by approximately 1.5 Å. (B) One monomer of GluA2o S1S2 in the presence (blue) and the absence of PEPA (green) with one orientation of PEPA shown. Both the J/K helices and the strand near S497 are displaced upon binding PEPA. Also, the sidechains of S497 and S729 change rotameric states. (C) Comparison of the water molecules at the dimer interface in the presence (tan spheres) and the absence of PEPA (red spheres). PEPA is shown in both orientations. Despite the greater separation of the dimer interface, a number of the ordered water molecules found in the absence of PEPA are displaced by PEPA. The black circles delineate subsites of the allosteric modulator binding site as described previously (31). Ahmed et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. The PEPA binding site, emphasizing the important interactions, shown in two orientations. (A) A view of the amide side of PEPA bound to GluA2 S1S2. The hydrogen bonding network with the amide of PEPA is shown as dotted lines. The H-bond with the sidechain of S729 is difficult to display in the orientation used in the figure. (B) A view of the phenyl group of PEPA inserted into a hydrophobic pocket in GluA2 S1S2. (C) RMS plot showing more variability in the J/K helices for the PEPA-bound structure than the unbound structure. (D) J/K helix showing where differences in the two orientations were analyzed. The amide of PEPA-N754 interaction (blue) maintains the position of the J helix in the absence of PEPA (green) The J helix is displaced on the phenyl side of PEPA (red). Ahmed et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. (A) Comparison of glutamate-bound GluA3o S1S2 in the presence (blue) and the absence of PEPA (green) in two orientations. Both orientations of PEPA are shown. Note that the binding of PEPA results in a separation of the two components of the dimer (distance between the Cα atoms of the threonine in the linker) by approximately 2.5 Å. (B) One monomer of GluA3o S1S2 in the presence (blue) and the absence of PEPA (green) with one orientation of PEPA shown. Shown for comparison is the PEPA-bound form of GluA2o (red). Both the J/K helices and the strand near S497 are displaced upon binding PEPA for both GluA2o and GluA3o. Also, the sidechains of S497 and S729 are in different rotameric states for GluA3o bound to PEPA compared with GluA3o in the absence of PEPA and GluA2o bound to PEPA. Also, N754 is displaced in PEPA-bound GluA3o, such that only one H-bond is possible with the amide of PEPA. Ahmed et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. (A) Members of the full spanning class of allosteric modulators. The shape-highlighted regions of the modulators illuminate key contact points to the specific binding pocket residues and subsites (as labeled for PEPA). (B) Overlay of the full spanning modulator structures. The structures were aligned at both sets of P494 and G731 residues. PEPA (gray) occupies a similar arrangement of subsites as the dimeric biarylsulfonamide (PDB entry 3bbr, cyan, 30) and LY404187 (PDB entry 3kgc, magenta, 8). (C) The sulfonamide bridges the two monomers in both PEPA and the dimeric biarylsulfonamide with the same interactions to P494 and G731. (D) The hydrogen bond between the carbonyl of P494 and the sulfonamide is maintained when the modulator is in a shifted position relative to the peptide plane of K730 and G731 (green disk) located on the opposite monomer. Ahmed et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Ahmed et al. Page 16 Table 1 Structural Statistics GluA2o (PEPA) GluA3o (PEPA) GluA3o Space Group P22121 P222 P222 Unit Cell (Å) a=47.13 b=113.92 c=164.81 a=46.95 b=52.26 c=115.98 a=46.03 b=110.33 c=161.192 X-ray source CHESS (A1) CHESS (A1) CHESS (A1) Wavelength (Å) 0.977 0.977 0.977 Resolution (Å) 50–2.0 (2.03–2.00) 50–2.5 (2.54–2.00) 50–1.85 (1.88–1.85) Measured reflections (#) 817961 62549 344340 Unique reflections (#) 70175 9141 69379 Data redundancy 6.9 (7.1) 6.0 (4.2) 4.7 (3.0) Completeness (%) 99.9 (100.0) 99.5 (89.8) 96.5 (73.5) Rsym (%) 11.4 (34.2) 13.0 (45.5) 7.5 (24.6) I/σi 33.3 (7.1) 19.2 (2.5) 34.3 (3.3) PDB ID * * * Current Model Refinement Statistics Phasing MR MR MR Molecules/AU 3 (no NCS applied) 1 3 (no NCS applied) Rwork/Rfree (%) 18.7/24.2 18.9/28.5 20.1/23.4 Free R test set size (#/%) 2000 (2.85) 914 (10.0) 2000 (2.88) Number of protein atoms 5979 2030 6091 Number of heteroatoms 111 62 30 Rmsd bond length (Å) 0.011 0.015 0.009 Rmsd bond angles (°) 1.3 1.8 1.3 *to be submitted to RCSB Protein Data Bank Biochemistry. Author manuscript; available in PMC 2011 April 7.
3M3L
PEPA bound to the ligand binding domain of GluA2 (flop form)
The molecular mechanism of flop-selectivity and subsite recognition for an AMPA receptor allosteric modulator: Structures of GluA2 and GluA3 complexed with PEPA Ahmed H. Ahmed§, Christopher P. Ptak§, and Robert E. Oswald* Department of Molecular Medicine, Cornell University, Ithaca, NY 14853 USA Abstract Glutamate receptors are important potential drug targets for cognitive enhancement and the treatment of schizophrenia in part because they are the most prevalent excitatory neurotransmitter receptors in the vertebrate central nervous system. One approach to the application of therapeutic agents to the AMPA subtype of glutamate receptors is the use of allosteric modulators, which promote dimerization by binding to a dimer interface thereby reducing desensitization and deactivation. AMPA receptors exist in two alternatively spliced variants (flip and flop) that differ in desensitization and receptor activation profiles. Most of the structural information on modulators of the AMPA receptor target the flip subtype. We report here the crystal structure of the flop-selective allosteric modulator, PEPA, bound to the binding domains of the GluA2 and GluA3 flop isoforms of AMPA receptors. Specific hydrogen bonding patterns can explain the preference for the flop isoform. This includes a bidentate hydrogen bonding pattern between PEPA and N754 of the flop isoforms of GluA2 and GluA3 (the corresponding position in the flip isoform is S754). Comparison with other allosteric modulators provides a framework for the development of new allosteric modulators with preferences for either the flip or flop isoforms. In addition to interactions with N/S754, specific interactions of the sulfonamide with conserved residues in the binding site are characteristics of a number of allosteric modulators. These, in combination, with variable interactions with five subsites on the binding surface lead to different stoichiometries, orientations within the binding pockets, and functional outcomes. Membrane receptors are the cell's gatekeepers, allowing chemical signals access to the cell's pathways. Through the binding of endogenous ligands, receptors identify relevant environmental cues and facilitate cell-cell communication. The regulation of membrane receptors has become an important goal of drug discovery efforts (1,2). By targeting the physiological (orthosteric) ligand-binding site, agonists and antagonists control the function of membrane receptors. Unfortunately, exogenously induced agonist-activation at the orthosteric site can cause toxic effects from overstimulation. Allosteric modulator binding sites use a distinct avenue for altering the natural response of a receptor. The ability of some allosteric modulators to enhance receptor stimulation, while not actually providing the trigger for stimulation, is a clear advantage that conserves the endogenous signaling pathway. Being important mediators of higher-order processes such as learning and memory, ionotropic glutamate receptors (iGluRs) have attracted a great deal of interest as allosteric modulator targets (3–6). Of clear therapeutic importance, various neurodegenerative disorders such as Parkinson's and Alzheimer's diseases, Huntington's *Corresponding author; telephone: 1-607-253-3877; fax: 1-607-253-3659; email: reo1@cornell.edu. §These authors contributed equally to this work. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 April 7. Published in final edited form as: Biochemistry. 2010 April 6; 49(13): 2843–2850. doi:10.1021/bi1000678. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript chorea, and neurologic disorders including epilepsy and ischemic brain damage have been linked to iGluRs (7). The crystal structure of GluA2 (8) clarifies years of speculation on the complex arrangement of the glutamate receptor's four subunits (9). The GluA2 can be dissected into 3 functionally distinct layers. Farthest from the membrane, the amino terminal domain (ATD) can act as a peripheral regulatory domain but is also involved in assembly and trafficking (10,11). Sandwiched between the ATD and the membrane domain, the ligand-binding domain (LBD) recognizes the neurotransmitter signal and directly regulates receptor activation (12). Structures for both isolated extracellular domains (ATD and LBD) reveal a dimeric organization (13–15). At the membrane interface, two alternative linker conformations transition the 2-fold symmetry, which is adopted by both extracellular domains, into the 4- fold symmetry of a membrane-traversing cation-selective channel (8,16). For iGluRs, the ion channel domain confers functional relevance with its ability to selectively conduct the flow of ions across the cell's membrane. The layers of extracellular domains, each with the potential for multiple control points, allosterically regulate the ion channel domain's function (8). Therefore it is not surprising that the ATD, the LBD, and the LBD-channel linker have all been shown to be effective targets of allosteric modulators (13,17,18). Since the structures of the ATD and the full iGluR channel have only recently been solved, allosteric drug-binding sites external to the LBD have not been fully explored in molecular detail. However, the decade-old LBD structure has proved to be indispensable as a heavily exploited scaffold for understanding agonist, partial agonist, and antagonist binding interactions as well as their ability to regulate channel gating behavior (12,19,20). Although the dimeric organization is consistent across all iGluR subtypes, the molecular details of LBD-agonist specificity define the subtype families into N-methyl-D-aspartic acid (NMDA) receptors (21), α-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid (AMPA) receptors (12), and kainate receptors (22). Because all subtypes are constrained by their conserved sensitivity to glutamate stimulation, diversity at the orthosteric site is evolutionarily limited and most agonists display cross-subtype activity. An allosteric modulator-binding site within the quaternary LBD structure is located along the dimer interface (18) and offers improved discrimination by modulators. Drugs that bind to the allosteric sites on the LBD dimer interface can enhance the activity of iGluRs (23) and increase performance on tests of memory (24). Except for the LBD structures with modulatory ions bound to the dimer interface (25–27), only LBD structures from the AMPA receptor subtype, GluA2, have been reported with bound allosteric modulators (18,28–31). Within the structures, the bound modulatory drugs stabilize the LBD dimer interface, which is required for activation of the ion channel and is dissociated during desensitization (18). Although the residues that line the allosteric modulator-binding pocket do not differ between AMPA receptors subtypes (GluA1–4), the ability of allosteric modulators to stabilize the activated state still varies (32,33). Also, AMPA receptors can be alternatively spliced into what is referred to as flip and flop isoforms (34). Modulator selectivity (23), desensitization (35), and channel closing rates (36) differ between flip and flop. Although several of the amino acid differences between the two forms are located in or near the allosteric modulator-binding site, the difference at position 754 (serine in flip, asparagine in flop) seems to be entirely responsible for the functional differences between allosteric modulator regulation of the flip and flop variants (23,28,32). Cyclothiazide (CTZ) and some other thiazide derivatives have improved binding to the flip form due to a hydrogen bond between S754 and the NH of the fused thiazide ring (28). In the case of the flop form, the alternatively spliced sequence places an asparagine in the 754 position, which is not optimally positioned to form a hydrogen bond. Sekiguchi et al. (33) introduced an allosteric modulator of AMPA receptors (4-[2-(phenylsulphonylamino)ethylthio]-2,6,- Ahmed et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript difluorophenoxyacetamide, PEPA) with a preference for the flop form. In fact, the relative sensitivity of CTZ to PEPA has been used as a diagnostic for the prevalence of flip vs. flop versions of AMPA receptor in particular cell types (37). PEPA shows potential in treatment of post ischemic memory impairment (38) and contextual fear (39) but despite PEPA's unique flop sensitivity, the modulator has not yet been used as a lead compound in SAR studies. For drug discovery to be guided by structures, understanding the possible molecular interactions between modulators and the dimer interface is essential. We have shown previously (31) that changes in the structures of CTZ derivatives can reorient the modulator within the binding site. Subsequently, we proposed that the allosteric modulator site is comprised of 5 subsites (Figure 1C). In the present study, we determine the three dimensional structures of PEPA bound to the GluA2o and GluA3o LBDs (flop forms), and use PEPA's binding interactions to further characterize the subsite specific binding properties displayed by allosteric modulators. The amide group of PEPA makes a direct hydrogen bond to N754, explaining the preferential action of PEPA on the flop form of AMPA receptors. Another key structural element, the sulfonamide group of PEPA, is conserved with the biarylsulfonamide class of allosteric modulators (6) and interacts with the same residues of the dimer interface (8,30). Although previously classified as unrelated, PEPA and the large group of biarylsulfonamide have similarities, which suggest that specific PEPA groups (particularly the unique flop-interacting amide) can be strategically integrated into biarylsulfonamide SAR studies. Experimental Procedures Materials PEPA was purchased from Tocris (Ellisville, MO). The GluA2 S1S2J construct was obtained from Eric Gouaux (Vollum Institute; 12). Protein Preparation and Purification GluA2 S1S2 consists of residues N392 - K506 and P632 - S775 of the full rat GluA2o subunit (40), a `GA' segment at the N-terminus, and a `GT' linker connecting K506 and P632 (12). A similar construct of GluA3 S1S2 was prepared as described previously (41). pET-22b(+) plasmids were transformed in E. coli strain Origami B (DE3) cells and were grown at 37°C to OD600 of 0.9 to 1.0 in LB medium supplemented with the antibiotics (ampicillin and kanamycin). The cultures were cooled to 20°C for 20 min. and isopropyl-β- D-thiogalactoside (IPTG) was added to a final concentration of 0.5 mM. Cultures were allowed to grow at 20°C for 20 h. The cells were then pelleted and the S1S2 protein purified using a Ni-NTA column, followed by a sizing column (Superose 12, XK 26/100), and finally an HT-SP-ion exchange-Sepharose column (Amersham Pharmacia). Glutamate (1 mM) was maintained in all buffers throughout purification. After the last column, the protein was concentrated and stored in 20 mM sodium acetate, 1 mM sodium azide, and 10 mM glutamate at pH 5.5. Crystallography For crystallization trials, the protein was concentrated to 0.2 – 0.5 mM in 10 mM glutamate using a Centricon 10 centrifugal filter (Millipore, Bedford, MA). For the PEPA-bound structures, PEPA was added to 5 mM. The final protein concentration was 0.2 to 0.3 mM. Crystals were grown at 4°C using the hanging drop technique, and the drops contained a 1:1 (v/v) ratio of protein solution to reservoir solution. The reservoir solution contained 14–15% PEG 8K, 0.1 M sodium cacodylate, 0.1–0.15 M zinc acetate, and 0.25 M ammonium sulfate, pH 6.5. Ahmed et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Data were collected at the Cornell High Energy Synchrotron Source beam line A1 using a Quantum-210 Area Detector Systems charge-coupled device detector. Data sets were indexed and scaled with HKL-2000 (42). Structures were solved with molecular replacement using Phenix (43). Refinement was performed with Phenix (43), and Coot 0.5 (44) was used for model building. Results Structure of PEPA bound to GluA2 S1S2 flop The structure of glutamate bound to GluA2o S1S2 (3dp6; 41) was used as the initial search probe for the molecular replacement solution of PEPA bound to GluA2o S1S2 with glutamate in the agonist-binding site. PEPA was then modeled into two symmetrical positions within the density found at the dimer interface, and the structure was optimized using Phenix (43). The refinement statistics are given in Table 1. The resolution is 1.85 Å, and three unique copies are found in the unit cell. The overall structure of the S1S2 domain is very similar to the structure in the absence of PEPA, with contacts between glutamate and the protein unchanged. However, PEPA clearly binds within the dimer interface, making contacts with both monomers within the dimer. As shown in Figure 1, one PEPA molecule binds per dimer interface. However because the dimer interface is symmetrical, two equivalent orientations (related by a 180° rotation) are possible. Electron density for both is seen in the crystal structure, although the intensity of one orientation is greater than the other. The binding of PEPA to the dimer interface increases the distance between the two monomers that form the dimer by approximately 1.5 Å. This allows the relatively large PEPA molecule to fit within the interface, but also increases the separation between the linkers to the ion channel (the distance increases from 39.4 Å to 41 Å; Figure 1A). Relative to the core of Lobe 1, both the J/K helices and one β strand (P105-G110) connecting the two lobes are displaced slightly away from the dimer interface (Figure 1B). In addition, Lobe 2 is slightly twisted relative to glutamate-bound S1S2 in the absence of PEPA (3dp6; 41). PEPA binds at the bottom of a water-filled, inverted U-shaped cleft with five subsites (A, B/ B′, and C/C′; 31). Upon binding, crystallographic waters are displaced from the central A subsite and more buried C/C′ sites, with the waters in the B/B′ subsite remaining (Figure 1C). This displacement of presumably ordered water would be likely to contribute a favorable entropy component to binding. The sidechains of P494 are at the center of the interface and the edge of the two proline rings from each monomer form the base of the binding site in which the difluorophenyl ring resides (Figure 2A). This is close to the position of the methoxybenzoyl ring of aniracetam in its structure bound to GluA2-S1S2(FW) (29). The other side of the ring is exposed to S497 and S729. The sidechain hydroxyl of S497 is oriented toward the dimer interface in the absence of PEPA, but rotates out toward the solvent to accommodate the difluorophenyl ring of PEPA (Figure 1B). The amide of PEPA is involved in a network of hydrogen bonds with sidechain hydroxyl of Y424, the backbone carbonyl of F495, the sidechain carboxyl of D760, the sidechain amide of N754, and two water molecules (Figure 2A). The most striking of these hydrogen bond pairs is with N754. This represents the only difference between the flip (S754) and flop (N754) isoforms in the PEPA binding site and is almost certainly a major source of the preference for the flop isoform. The phenyl-sulfonylamide side of PEPA inserts into a hydrophobic pocket formed by sidechain methyls of I481 and L751 as well as methylene groups contributed by K493, N754, and E755 (Figure 2B). It is possible that the contribution by methylene group of N754 provides a more hydrophobic pocket than S754 in the flip form, further contributing to the preference for the flop form. Ahmed et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Because the dimer interface is symmetrical, PEPA can bind in two orientations and both are observed in the crystal. For this reason, changes in the protein due to a specific interaction with PEPA can be partially masked because each monomer is a weighted average of two orientations of bound PEPA. However, one orientation has a stronger density than the other, providing some insight into the extent of changes in the dimer interface that are produced by PEPA binding. As shown in Figure 2C, the two monomers comprising the dimer differ more within the PEPA binding site than the corresponding monomers in the absence of PEPA. One turn of helix J (L751 to N754) contains important determinants for both orientations of PEPA. In one orientation the amide group of PEPA interacts with the sidechain of N754, and in the other, the aromatic ring of PEPA inserts between a hydrophobic pocket formed by the sidechain of L751 and the methylene group of N754. In the orientation for which the density of the amide of PEPA is stronger, N754 is better positioned to form an H-bond (Figure 2D); whereas, in the other side of the interface, N754 is oriented to form an H-bond with the carbonyl of S729. This change in orientation facilitates the insertion of the aromatic ring of PEPA into the hydrophobic pocket, which is accompanied by a small shift in the sidechain of L751 to accommodate the aromatic ring (Figure 2D). Since these structures are weighted averages, it is possible that the actual positions of these sidechains involve an even greater movement than is seen from the asymmetry of the crystal. Structure of PEPA bound to GluA3 S1S2 flop In studies of the physiological effects of PEPA, a significant difference between subtypes has been observed, with GluA3 being most susceptible to modulation (33). The structure of GluA3i S1S2 bound (flip form) to glutamate has been reported previously (41). Since PEPA preferentially binds to the flop form, the GluA3o structure was determined bound to glutamate with and without PEPA (Figure 3A). Like GluA2o, in the absence of PEPA, GluA3o has three copies in the asymmetric unit. Comparing lobe closure between GluA3i and GluA3o, the flop form is slightly more closed (1.6° ± 0.7°). In the presence of PEPA, GluA3o was present in one copy in the asymmetric unit, and PEPA was observed with the same density in two symmetrical orientations. Like GluA2o bound to PEPA, the dimer interface (assessed using the symmetrical molecule in the crystal) was displaced relative to the unbound from (Figure 3A) by approximately 2.5 Å at the position of the linker replacing the ion channel domain. Within the binding site, three sidechains exhibited different rotamers compared with the GluA2o structure bound to PEPA (Figure 3B). For PEPA-bound GluA3o, both S497 and S729 assumed rotameric states that differed both from GluA2o bound to PEPA and from GluA2o and GluA3o in the absence of PEPA. In the case of S729, the rotameric state in combination with a slight movement of the amide of PEPA (relative to the GluA2o structure) would make an H-bond with the sidechain of S729 (shown in Figure 2A for GluA2o) unlikely. In the case of N754, the sidechain is displaced relative to the GluA2o-PEPA structure so that only one H-bond is made to the amide of PEPA. This may be a result of averaging of the two orientations of PEPA only one of which forms a bidentate H-bond with N754. Discussion The goal of allosteric modulation, like orthosteric modulation, is often to stabilize a conformational state of a dynamic protein (45). The activated state of iGluRs is naturally unstable allowing the channel to desensitize (46). Disruption of the symmetrical dimer interface between LBDs is thought to initiate desensitization-mediated channel closure (47). By maintaining the LBD dimer, positive allosteric modulators can prevent desensitization and prolong activation (18). Currently, 15 crystal structures of the GluA2 LBD with bound allosteric modulators are deposited in the Protein Data Bank (48). All of these modulators bind to a large crevice with 2-fold symmetry along the symmetric dimer interface (18). The Ahmed et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript large variation in structure among allosteric modulators results in significant variations in binding orientations and interactions. At least four distinct binding modes have been identified: (1) A-subsite class (aniracetam, CX614 (29)), (2) classical thiazide (cyclothiazide (18), TCMZ, ALTZ (31)), (3) the shifted thiazide class (IDRA-21, HCMZ, HFMZ; (31)), and (4) the full spanning class (PEPA (this paper), dimeric biarylpropylsulfonamide (30), LY404187 (8)). Overlaying modulators from these structural classes has led to the proposal that the allosteric modulator site is comprised of a series of subsites (Figure 1C;31). Positioned at the center of the binding-site, the symmetric A subsite is narrow and allows entrance to only one molecule. Two subsites (B and C) lie at each end of the A subsite with the hydrophobic C subsite located more deeply in the pocket effectively defining five subsites (A, B, B′, C, and C′). In the open state, the subsites are filled with water, which may act to weakly stabilize the dimer. Allosteric modulators generate stronger interactions across the subsites thereby increasing the linkages between the monomers. The simplest modulator class, including aniracetam and CX614, fills the A subsite with one molecule but does not enter the peripheral B and C subsites (29). The two classes of thiazide-based modulators account for 10 of the 15 solved allosteric modulator-GluA2 crystal structure complexes (18,28,31). The classical thiazide (CTZ-like) binding class and the shifted thiazide (IDRA-21-like) binding class are positioned respectively in the B and C subsite or mainly the C subsite. Most of the thiazide modulators do not extend across the A subsite and therefore can bind two molecules per dimer. However, a few of the newly described shifted thiazides (HFMZ, HCTZ; 31) enter the A subsite but only enough to impair binding of a second modulator. The dimeric biarylpropylsulfonamide compound ((R,R)-N,N-(2,2'-[Biphenyl-4-4'-Diyl]Bis[Propane-2,1- Diyl]) Dimethanesulfonamide) described by Kaae et al. (30) was the first allosteric modulator shown by crystallography to extend along the entire length of the inner dimer cavity from C to C′ subsites. PEPA also interacts with J helices from both monomers, which cap the ends of the modulator-binding pocket. The density occupied by both symmetrical copies of PEPA overlays the dimeric biarylsulfonamide compound as both modulators represent the full spanning class (Figure 4B). The GluA flip and flop splice variants differ by only a few residues along the J helix in the LBD; however, residue 754 (Asn in flop and Ser in flip) is positioned between the B and C subsites. For thiazides, a clear preference in binding to the flip-form is mediated by a hydrogen bond between the hydrobenzothiadiazide ring and S754 (28). In contrast, PEPA is flop-selective and the PEPA-bound structure provides the first structure containing a direct interaction between a modulator and the flop form's N754. The amide of PEPA extends straight out from the A subsite and across the B and C subsite interface to make an amide- amide hydrogen bond with N754 (Figure 2A). Unlike most other AMPA modulators, PEPA fills neither the B nor the C subsites but interacts directly with the J helix. A similar interaction is seen with LY404187 (49) bound to GluA2i (8). Strong hydrogen bonding can occur between two amides (50) and has been shown to be responsible for driving oligomerization of transmembrane leucine zippers (51). The distances between the interacting amides in the PEPA-bound structure support a bidentate hydrogen-bonding pattern, which is much stronger and more specific than a typical hydrogen bond. While PEPA is selective for the AMPA receptor's flop form, a weaker but still existent potentiation of the flip form has been observed (33,52). Replacing N754 (flop) with S754 (flip) would not prevent PEPA from binding; however, serine would provide only one hydrogen-bonding partner for PEPA's amide with an extended interaction distance. In contrast, LY404187 displays a preference for the flip isoform (53), and its cyano group extends out to interact directly with S754. The cyano-S754 interaction is a clear flip analog of the flop-selective PEPA amide-N754 interaction (Figure 4A). Ahmed et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Opposite to the amide on the PEPA molecule, a sulfonamide is tethered to the difluorophenyl ring (Figure 4A). Within the dimer interface, the sulfonamide is positioned so the nitrogen can hydrogen bond directly with the carbonyl of P494 (Figure 4C). A sulfonamide oxygen points toward the amide nitrogen of G731. The angle of the peptide plane is perpendicular to the sulfonamide oxygen, making a hydrogen bonding interaction unlikely (Figure 4D). Instead, a dipole-dipole or charge-dipole interaction may occur. The amide nitrogen of a polypeptide supports at least a partial positive charge (54), which would interact with the strongly electronegative sulfonamide oxygen (55). Interestingly, both the dimeric biarylsulfonamide (30) and LY404187 (8), other members of the full spanning modulator class, also have a sulfonamide that interacts with the same backbone atoms of P494 and G731 as PEPA (Figure 4C). A large number of biarylsulfonamides have been identified that modulate AMPA receptors and are being evaluated for therapeutic use in the treatment of depression and Parkinson's disease (56). The conserved sulfonamide reveals a previously unidentified relationship between PEPA and the biarylsulfonamide modulators. When the perpendicular peptide bond plane including G731 is fixed, the sulfonamide on three overlayed modulators varies by 1.2 Å along the length of the interface with the PEPA sulfonamide being positioned closer to the A subsite (Figure 4D). A shift of the sulfonamide also results in a shift in the corresponding P494 across the interface presumably to maintain the hydrogen bond with the modulator's amine. The sulfonamide forms an important bridge between the two dimer halves. For PEPA, a phenyl-sulfonamide replaces the methyl-sulfonamide in the dimeric biarylsulfonamide and fits snuggly against L751. Based on the orientation-induced asymmetry within the GluA2-complex structure, the phenyl pushes the J helix away from PEPA thereby affecting the C subsite (Figure 2C and D). Residues lining the C subsite are on the same beta strand as G731, which must shift if the C subsite is to remain together and presumably explain the 1.2 Å shift relative to the dimeric biarylsulfonamide. In fact, the same phenyl-sulfonamide group substitution in a biarylpropylsulfonamide decreases the modulatory effect of the derivative in SAR studies (57). For biarylpropylsulfonamides, the optimal sulfonamide substitution was found to be either an ethyl or an iso-propyl group, which should both fit without significantly disrupting the J helix or C subsite (57). The PEPA-bound crystal structure from AMPA receptor subtypes, GluA2 and GluA3, do not display major differences in binding interactions even though PEPA exhibits a stronger effect on GluA3 (33). For GluA2, an asymmetry in the receptor-binding pocket was observed while no significant difference in PEPA density was seen for the each orientation within the GluA3 crystal structure. In addition, a number of side chains exhibit different rotameric states between the two structures, although it is unlikely that these small changes significantly impact the differential effects on the two subtypes. Although no structural differences have been identified between GluA2 and GluA3 that would obviously impact PEPA affinity, the possibility exists that subtle differences arising from the sequence differences peripheral to the binding site may be important as has been described in the case of the agonist binding site of GluA4 (58). We have explored how PEPA (this paper) and other allosteric modulators (31) interact with the GluA interface in the context of drug design. Together the identification of a conserved group between PEPA (this paper) and biarylpropylsulfonamides (8,30) and the regional nature of various subsite-functional group interactions provide a backdrop to extend biarylpropylsulfonamide SAR studies (57) to include PEPA and biarylpropylsulfonamide chimeras. Although optimizing the stability of the dimer interface provides a starting point for SAR studies, additional constraints should be considered including the ability of the modulator to enter the cavity, the dynamic structure of the dimer interface during closed, open, and desensitized state transitions, and the ability of the modulator to cross the blood- Ahmed et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript brain barrier before being metabolized. This definition of the allosteric modulator binding- site should provide guidance in glutamate receptor allosteric modulator pharmacology. Acknowledgments We thank Prof. Eric Gouaux (Vollum Institute) for the GluA2 S1S2J construct, and Prof. Linda Nowak (Cornell) for the full-length GluA3 construct. This work was supported by a grants from the National Institutes of Health (R01-GM068935, R01 NS049223, and R21 NS067562). This work is based upon research conducted at the Cornell High Energy Synchrotron Source (CHESS), which is supported by the National Science Foundation under award DMR 0225180, using the Macromolecular Diffraction at the CHESS (MacCHESS) facility, which is supported by award RR-01646 from the National Institutes of Health, through its National Center for Research Resources. Abbreviations ALTZ althiazide AMPA α-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid CLTZ chlorothiazide CX614 pyrrolidino-1,3-oxazino benzo-1,4-dioxan-10-one CTZ cyclothiazide FW (S)-5-fluorowillardiine flip and flop alternatively spliced versions of AMPA receptors that vary in rates of desensitization and sensitivity to allosteric modulators iGluR ionotropic glutamate receptor GluA1-4 four subtypes of AMPA receptor HCTZ hydrochlorothiazide HFMZ hydroflumethiazide IDRA-21 7-chloro-3-methyl-3,4-dihydro-2H-benzo[e][1,2,4]thiadiazine 1,1-dioxide IPTG isopropyl-β-D-thiogalactoside LY404187 N-[2-(4′-cyanobiphenyl-4-yl)propyl]propane-2-sulfamide PEPA 4-[2-(phenylsulphonylamino)ethylthio]-2,6,-difluorophenoxy acetamide NMDA N-methyl-D-aspartic acid S1S2 extracellular ligand-binding domain of GluA2 and GluA3 SAR structure-activity relationships TCMZ trichlormethiazide References 1. Christopoulos A. Allosteric binding sites on cell-surface receptors: novel targets for drug discovery. Nat Rev Drug Discov. 2002; 1:198–210. [PubMed: 12120504] 2. Changeux JP, Taly A. Nicotinic receptors, allosteric proteins and medicine. Trends Mol Med. 2008; 14:93–102. [PubMed: 18262468] 3. Bowie D. Ionotropic glutamate receptors & CNS disorders. CNS Neurol Disord Drug Targets. 2008; 7:129–143. [PubMed: 18537642] 4. Dingledine R, Borges K, Bowie D, Traynelis S. The glutamate receptor ion channels. Pharmacol. Rev. 1999; 51:7–61. [PubMed: 10049997] Ahmed et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 5. Oswald RE, Ahmed A, Fenwick MK, Loh AP. Structure of glutamate receptors. Current drug targets. 2007; 8:573–582. [PubMed: 17504102] 6. Grigoriev VV, Proshin AN, Kinzirsky AS, Bachurin SO. Modern approaches to the design of memory and cognitive enhancers based on AMPA receptor ligands. Russian Chemical Reviews. 2009; 78:485–494. 7. Black MD. Therapeutic potential of positive AMPA modulators and their relationship to AMPA receptor subunits. A review of preclinical data. Psychopharmacology (Berl). 2005; 179:154–163. [PubMed: 15672275] 8. Sobolevsky AI, Rosconi MP, Gouaux E. X-ray structure, symmetry and mechanism of an AMPA- subtype glutamate receptor. Nature. 2009; 462:745–756. [PubMed: 19946266] 9. Wo ZG, Oswald RE. Unraveling the modular design of glutamate-gated ion channels. Trends Neurosciences. 1995; 18:161–168. 10. Gielen M, Siegler Retchless B, Mony L, Johnson JW, Paoletti P. Mechanism of differential control of NMDA receptor activity by NR2 subunits. Nature. 2009; 459:703–707. [PubMed: 19404260] 11. Greger IH, Ziff EB, Penn AC. Molecular determinants of AMPA receptor subunit assembly. Trends Neurosci. 2007; 30:407–416. [PubMed: 17629578] 12. Armstrong N, Gouaux E. Mechanisms for activation and antagonism of an AMPA-sensitive glutamate receptor: crystal structures of the GluR2 ligand binding core. Neuron. 2000; 28:165– 181. [PubMed: 11086992] 13. Clayton A, Siebold C, Gilbert RJ, Sutton GC, Harlos K, McIlhinney RA, Jones EY, Aricescu AR. Crystal structure of the GluR2 amino-terminal domain provides insights into the architecture and assembly of ionotropic glutamate receptors. J Mol Biol. 2009; 392:1125–1132. [PubMed: 19651138] 14. Jin R, Singh SK, Gu S, Furukawa H, Sobolevsky AI, Zhou J, Jin Y, Gouaux E. Crystal structure and association behaviour of the GluR2 amino-terminal domain. EMBO J. 2009; 28:1812–1823. [PubMed: 19461580] 15. Kumar J, Schuck P, Jin R, Mayer ML. The N-terminal domain of GluR6-subtype glutamate receptor ion channels. Nat Struct Mol Biol. 2009; 16:631–638. [PubMed: 19465914] 16. Doyle DA, Cabral JM, Pfuetzner RA, Kuo AL, Gulbis JM, Cohen SL, Chait BT, MacKinnon R. The structure of a potassium channel: molecular basis of K+ conduction and selectivity. Science. 1998; 280:69–77. [PubMed: 9525859] 17. Balannik V, Menniti FS, Paternain AV, Lerma J, Stern-Bach Y. Molecular mechanism of AMPA receptor noncompetitive antagonism. Neuron. 2005; 48:279–288. [PubMed: 16242408] 18. Sun Y, Olson R, Horning M, Armstrong N, Mayer M, Gouaux E. Mechanism of glutamate receptor desensitization. Nature. 2002; 417:245–253. [PubMed: 12015593] 19. Ahmed A, Thompson M, Fenwick M, Romero B, Loh A, Jane D, Sondermann H, Oswald R. Mechanisms of antagonism of the GluR2 AMPA receptor: Structure and dynamics of the complex of two willardiine antagonists with the glutamate binding domain. Biochemistry. 2009; 48:3894– 3903. [PubMed: 19284741] 20. Jin R, Banke TG, Mayer ML, Traynelis SF, Gouaux E. Structural basis for partial agonist action at ionotropic glutamate receptors. Nat Neurosci. 2003; 6:803–810. [PubMed: 12872125] 21. Furukawa H, Gouaux E. Mechanisms of activation, inhibition and specificity: crystal structures of the NMDA receptor NR1 ligand-binding core. EMBO J. 2003; 22:2873–2885. [PubMed: 12805203] 22. Mayer ML. Crystal structures of the GluR5 and GluR6 ligand binding cores: Molecular mechanisms underlying kainate receptor selectivity. Neuron. 2005; 45:539–552. [PubMed: 15721240] 23. Partin KM, Fleck MW, Mayer ML. AMPA receptor flip/flop mutants affecting deactivation, desensitization, and modulation by cyclothiazide, aniracetam, and thiocyanate. J Neurosci. 1996; 16:6634–6647. [PubMed: 8824304] 24. Martin JR, Cumin R, Aschwanden W, Moreau JL, Jenck F, Haefely WE. Aniracetam improves radial maze performance in rats. Neuroreport. 1992; 3:81–83. [PubMed: 1611039] Ahmed et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 25. Naur P, Vestergaard B, Skov LK, Egebjerg J, Gajhede M, Kastrup JS. Crystal structure of the kainate receptor GluR5 ligand-binding core in complex with (S)-glutamate. FEBS Lett. 2005; 579:1154–1160. [PubMed: 15710405] 26. Plested AJ, Mayer ML. Structure and mechanism of kainate receptor modulation by anions. Neuron. 2007; 53:829–841. [PubMed: 17359918] 27. Plested AJ, Vijayan R, Biggin PC, Mayer ML. Molecular basis of kainate receptor modulation by sodium. Neuron. 2008; 58:720–735. [PubMed: 18549784] 28. Hald H, Ahring PK, Timmermann DB, Liljefors T, Gajhede M, Kastrup JS. Distinct Structural Features of Cyclothiazide are Responsible for Effects on Peak Current Amplitude and Desensitization Kinetics at iGluR2. J Mol Biol. 2009; 391:906–917. [PubMed: 19591837] 29. Jin R, Clark S, Weeks AM, Dudman JT, Gouaux E, Partin KM. Mechanism of positive allosteric modulators acting on AMPA receptors. J Neurosci. 2005; 25:9027–9036. [PubMed: 16192394] 30. Kaae BH, Harpsoe K, Kastrup JS, Sanz AC, Pickering DS, Metzler B, Clausen RP, Gajhede M, Sauerberg P, Liljefors T, Madsen U. Structural proof of a dimeric positive modulator bridging two identical AMPA receptor-binding sites. Chemistry & biology. 2007; 14:1294–1303. [PubMed: 18022568] 31. Ptak CP, Ahmed AH, Oswald RE. Probing the allosteric modulator binding site of GluR2 with thiazide derivatives. Biochemistry. 2009; 48:8594–8602. [PubMed: 19673491] 32. Partin KM, Bowie D, Mayer ML. Structural determinants of allosteric regulation in alternatively spliced AMPA receptors. Neuron. 1995; 14:833–843. [PubMed: 7718245] 33. Sekiguchi M, Fleck MW, Mayer ML, Takeo J, Chiba Y, Yamashita S, Wada K. A novel allosteric potentiator of AMPA receptors: 4-[2-(phenylsulfonylamino)ethylthio]-2,6-difluoro- phenoxyacetamide. J Neurosci. 1997; 17:5760–5771. [PubMed: 9221774] 34. Sommer B, Keinänen K, Verdoorn TA, Wisden W, Burnashev N, Herb A, Köhler M, Takagi T, Sakmann G, Seeburg PH. Flip and flop: A cell-specific functional switch in glutamate-operated channels of the CNS. Science. 1990; 249:1580–1584. [PubMed: 1699275] 35. Mosbacher J, Schoepfer R, Monyer H, Burnashev N, Seeburg PH, Ruppersberg JP. A molecular determinant for submillisecond desensitization in glutamate receptors. Science. 1994; 266:1059– 1062. [PubMed: 7973663] 36. Pei W, Huang Z, Niu L. GluR3 flip and flop: differences in channel opening kinetics. Biochemistry. 2007; 46:2027–2036. [PubMed: 17256974] 37. Sekiguchi M, Takeo J, Harada T, Morimoto T, Kudo Y, Yamashita S, Kohsaka S, Wada K. Pharmacological detection of AMPA receptor heterogeneity by use of two allosteric potentiators in rat hippocampal cultures. Br J Pharmacol. 1998; 123:1294–1303. [PubMed: 9579722] 38. Sekiguchi M, Yamada K, Jin J, Hachitanda M, Murata Y, Namura S, Kamichi S, Kimura I, Wada K. The AMPA receptor allosteric potentiator PEPA ameliorates post-ischemic memory impairment. Neuroreport. 2001; 12:2947–2950. [PubMed: 11588608] 39. Zushida K, Sakurai M, Wada K, Sekiguchi M. Facilitation of extinction learning for contextual fear memory by PEPA: a potentiator of AMPA receptors. J Neurosci. 2007; 27:158–166. [PubMed: 17202483] 40. Hollmann M, Heinemann S. Cloned glutamate receptors. Annu Rev Neurosci. 1994; 17:31–108. [PubMed: 8210177] 41. Ahmed AH, Wang Q, Sondermann H, Oswald RE. Structure of the S1S2 glutamate binding domain of GluR3. Proteins: Structure, Function, and Bioinformatics. 2009; 75:628–637. 42. Otwinowski, Z.; Minor, W. Processing of X-ray diffraction data collected in oscillation mode. In: Carter, CW.; Sweet, RM., editors. Methods in Enzymology, Vol. 276, Macromolecular Crystallography, part A. Academic Press; New York: 1997. p. 307-326. 43. Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Terwilliger TC. PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr D Biol Crystallogr. 2002; 58:1948– 1954. [PubMed: 12393927] 44. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] Ahmed et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 45. Gunasekaran K, Ma B, Nussinov R. Is allostery an intrinsic property of all dynamic proteins? Proteins. 2004; 57:433–443. [PubMed: 15382234] 46. Trussell LO, Fischbach GD. Glutamate receptor desensitization and its role in synaptic transmission. Neuron. 1989; 3:209–218. [PubMed: 2576213] 47. Armstrong N, Jasti J, Beich-Frandsen M, Gouaux E. Measurement of conformational changes accompanying desensitization in an ionotropic glutamate receptor. Cell. 2006; 127:85–97. [PubMed: 17018279] 48. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, Shindyalov IN, Bourne PE. The Protein Data Bank. Nucleic Acids Res. 2000; 28:235–242. [PubMed: 10592235] 49. Miu P, Jarvie KR, Radhakrishnan V, Gates MR, Ogden A, Ornstein PL, Zarrinmayeh H, Ho K, Peters D, Grabell J, Gupta A, Zimmerman DM, Bleakman D. Novel AMPA receptor potentiators LY392098 and LY404187: effects on recombinant human AMPA receptors in vitro. Neuropharmacology. 2001; 40:976–983. [PubMed: 11406188] 50. Shimoni L, Glusker JP. Hydrogen bonding motifs of protein side chains: descriptions of binding of arginine and amide groups. Protein Sci. 1995; 4:65–74. [PubMed: 7773178] 51. Zhou FX, Cocco MJ, Russ WP, Brunger AT, Engelman DM. Interhelical hydrogen bonding drives strong interactions in membrane proteins. Nat Struct Biol. 2000; 7:154–160. [PubMed: 10655619] 52. Sekiguchi M, Nishikawa K, Aoki S, Wada K. A desensitization-selective potentiator of AMPA- type glutamate receptors. Br J Pharmacol. 2002; 136:1033–1041. [PubMed: 12145103] 53. Quirk JC, Nisenbaum ES. Multiple molecular determinants for allosteric modulation of alternatively spliced AMPA receptors. J Neurosci. 2003; 23:10953–10962. [PubMed: 14645491] 54. Kemnitz CR, Loewen MJ. “Amide resonance” correlates with a breadth of C-N rotation barriers. J Am Chem Soc. 2007; 129:2521–2528. [PubMed: 17295481] 55. Cazenave-Gassiot A, Boughtflower R, Caldwell J, Coxhead R, Hitzel L, Lane S, Oakley P, Holyoak C, Pullen F, Langley GJ. Prediction of retention for sulfonamides in supercritical fluid chromatography. J Chromatogr A. 2008; 1189:254–265. [PubMed: 17977551] 56. O'Neill MJ, Witkin JM. AMPA receptor potentiators: application for depression and Parkinson's disease. Current drug targets. 2007; 8:603–620. [PubMed: 17504104] 57. Ornstein PL, Zimmerman DM, Arnold MB, Bleisch TJ, Cantrell B, Simon R, Zarrinmayeh H, Baker SR, Gates M, Tizzano JP, Bleakman D, Mandelzys A, Jarvie KR, Ho K, Deverill M, Kamboj RK. Biarylpropylsulfonamides as novel, potent potentiators of 2-amino-3- (5-methyl-3- hydroxyisoxazol-4-yl)- propanoic acid (AMPA) receptors. J Med Chem. 2000; 43:4354–4358. [PubMed: 11087558] 58. Gill A, Birdsey-Benson A, Jones BL, Henderson LP, Madden DR. Correlating AMPA receptor activation and cleft closure across subunits: crystal structures of the GluR4 ligand-binding domain in complex with full and partial agonists. Biochemistry. 2008; 47:13831–13841. [PubMed: 19102704] Ahmed et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. (A) Comparison of glutamate-bound GluA2o S1S2 in the presence (blue) and the absence of PEPA (green) in two orientations. Both orientations of PEPA are shown. Note that the binding of PEPA results in a separation of the two components of the dimer (distance between the Cα atoms of the threonine in the linker) by approximately 1.5 Å. (B) One monomer of GluA2o S1S2 in the presence (blue) and the absence of PEPA (green) with one orientation of PEPA shown. Both the J/K helices and the strand near S497 are displaced upon binding PEPA. Also, the sidechains of S497 and S729 change rotameric states. (C) Comparison of the water molecules at the dimer interface in the presence (tan spheres) and the absence of PEPA (red spheres). PEPA is shown in both orientations. Despite the greater separation of the dimer interface, a number of the ordered water molecules found in the absence of PEPA are displaced by PEPA. The black circles delineate subsites of the allosteric modulator binding site as described previously (31). Ahmed et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. The PEPA binding site, emphasizing the important interactions, shown in two orientations. (A) A view of the amide side of PEPA bound to GluA2 S1S2. The hydrogen bonding network with the amide of PEPA is shown as dotted lines. The H-bond with the sidechain of S729 is difficult to display in the orientation used in the figure. (B) A view of the phenyl group of PEPA inserted into a hydrophobic pocket in GluA2 S1S2. (C) RMS plot showing more variability in the J/K helices for the PEPA-bound structure than the unbound structure. (D) J/K helix showing where differences in the two orientations were analyzed. The amide of PEPA-N754 interaction (blue) maintains the position of the J helix in the absence of PEPA (green) The J helix is displaced on the phenyl side of PEPA (red). Ahmed et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. (A) Comparison of glutamate-bound GluA3o S1S2 in the presence (blue) and the absence of PEPA (green) in two orientations. Both orientations of PEPA are shown. Note that the binding of PEPA results in a separation of the two components of the dimer (distance between the Cα atoms of the threonine in the linker) by approximately 2.5 Å. (B) One monomer of GluA3o S1S2 in the presence (blue) and the absence of PEPA (green) with one orientation of PEPA shown. Shown for comparison is the PEPA-bound form of GluA2o (red). Both the J/K helices and the strand near S497 are displaced upon binding PEPA for both GluA2o and GluA3o. Also, the sidechains of S497 and S729 are in different rotameric states for GluA3o bound to PEPA compared with GluA3o in the absence of PEPA and GluA2o bound to PEPA. Also, N754 is displaced in PEPA-bound GluA3o, such that only one H-bond is possible with the amide of PEPA. Ahmed et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. (A) Members of the full spanning class of allosteric modulators. The shape-highlighted regions of the modulators illuminate key contact points to the specific binding pocket residues and subsites (as labeled for PEPA). (B) Overlay of the full spanning modulator structures. The structures were aligned at both sets of P494 and G731 residues. PEPA (gray) occupies a similar arrangement of subsites as the dimeric biarylsulfonamide (PDB entry 3bbr, cyan, 30) and LY404187 (PDB entry 3kgc, magenta, 8). (C) The sulfonamide bridges the two monomers in both PEPA and the dimeric biarylsulfonamide with the same interactions to P494 and G731. (D) The hydrogen bond between the carbonyl of P494 and the sulfonamide is maintained when the modulator is in a shifted position relative to the peptide plane of K730 and G731 (green disk) located on the opposite monomer. Ahmed et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 April 7. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Ahmed et al. Page 16 Table 1 Structural Statistics GluA2o (PEPA) GluA3o (PEPA) GluA3o Space Group P22121 P222 P222 Unit Cell (Å) a=47.13 b=113.92 c=164.81 a=46.95 b=52.26 c=115.98 a=46.03 b=110.33 c=161.192 X-ray source CHESS (A1) CHESS (A1) CHESS (A1) Wavelength (Å) 0.977 0.977 0.977 Resolution (Å) 50–2.0 (2.03–2.00) 50–2.5 (2.54–2.00) 50–1.85 (1.88–1.85) Measured reflections (#) 817961 62549 344340 Unique reflections (#) 70175 9141 69379 Data redundancy 6.9 (7.1) 6.0 (4.2) 4.7 (3.0) Completeness (%) 99.9 (100.0) 99.5 (89.8) 96.5 (73.5) Rsym (%) 11.4 (34.2) 13.0 (45.5) 7.5 (24.6) I/σi 33.3 (7.1) 19.2 (2.5) 34.3 (3.3) PDB ID * * * Current Model Refinement Statistics Phasing MR MR MR Molecules/AU 3 (no NCS applied) 1 3 (no NCS applied) Rwork/Rfree (%) 18.7/24.2 18.9/28.5 20.1/23.4 Free R test set size (#/%) 2000 (2.85) 914 (10.0) 2000 (2.88) Number of protein atoms 5979 2030 6091 Number of heteroatoms 111 62 30 Rmsd bond length (Å) 0.011 0.015 0.009 Rmsd bond angles (°) 1.3 1.8 1.3 *to be submitted to RCSB Protein Data Bank Biochemistry. Author manuscript; available in PMC 2011 April 7.
3M3N
Structure of a Longitudinal Actin Dimer Assembled by Tandem W Domains
Structure of a Longitudinal Actin Dimer Assembled by Tandem W Domains – Implications for Actin Filament Nucleation Grzegorz Rebowski1,§, Suk Namgoong1,§, Malgorzata Boczkowska1, Paul C. Leavis2, Jorge Navaza3, and Roberto Dominguez1,* 1Department of Physiology, 3700 Hamilton Walk, University of Pennsylvania School of Medicine, Philadelphia, PA 19104-6085, USA. 2Boston Biomedical Research Institute, Watertown, MA 02472-2899, USA. 3Institut de Biologie Structurale, F-38027 Grenoble, France Abstract Actin filament nucleators initiate polymerization in cells in a regulated manner. A common architecture among these molecules consists of tandem W domains that recruit three to four actin subunits to form a polymerization nucleus. We describe a low-resolution crystal structure of an actin dimer assembled by tandem W domains, where the first W domain is crosslinked to Cys-374 of the actin subunit bound to it, whereas the last W domain is followed by the C-terminal pointed end-capping helix of Tβ4. While the arrangement of actin subunits in the dimer resembles that of a long-pitch helix of the actin filament, important differences are observed. These differences result from steric hindrance of the W domain with inter-subunit contacts in the actin filament. We also determined the structure of the first W domain of Vibrio parahaemolyticus VopL crosslinked to actin Cys-374, and show it to be nearly identical to non-crosslinked W-actin structures. This result validates the use of crosslinking as a tool for the study of actin nucleation complexes, whose natural tendency to polymerize interferes with most structural methods. Combined with a biochemical analysis of nucleation, the structures may explain why nucleators based on tandem W domains with short inter-W linkers have relatively weak activity, cannot stay bound to filaments after nucleation, and are unlikely to influence filament elongation. The findings may also explain why Nucleation Promoting Factors of the Arp2/3 complex, which are related to tandem W domain nucleators, are ejected from branch junctions after nucleation. We finally show that the simple addition of the C-terminal pointed end-capping helix of Tβ4 to tandem W domains can change their activity from actin filament nucleation to monomer sequestration. Keywords Actin nucleation; W domain; Nucleation Promoting Factors; Tβ4; actin monomer sequestration © 2010 Elsevier Ltd. All rights reserved. *Corresponding author (Phone: 215-573-4559; droberto@mail.med.upenn.edu). §These authors contributed equally to this work Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. PDB ACCESSION NUMBERS Coordinates and structure factors were deposited under PDB ID 3M1F (WxActin) and 3M3N (3W-Actin). SUPPLEMENTARY MATERIAL Supplementary material is available online, including supplementary Material and Methods and Movies S1–S4. NIH Public Access Author Manuscript J Mol Biol. Author manuscript; available in PMC 2011 October 15. Published in final edited form as: J Mol Biol. 2010 October 15; 403(1): 11–23. doi:10.1016/j.jmb.2010.08.040. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript INTRODUCTION The nucleation of actin filaments in cells is kinetically unfavorable because of the instability of polymerization intermediates (dimers, trimers and tetramers) and the actions of actin monomer binding proteins such as profilin and thymosin-β4 (Tβ4) 1; 2. This creates an opportunity for cells to use molecules known as actin filament nucleators to initiate the formation of actin polymerization nuclei in a spatially and temporally controlled manner. The actin filament can be described as either a single left-handed short-pitch helix, where consecutive subunits are staggered with respect to one another by half a monomer length, or two right-handed long-pitch helices of head-to-tail bound actin subunits 3; 4; 5. Different nucleators work by different mechanisms, stabilizing small actin oligomers along either the long- or the short-pitch helices of the actin filament 6; 7. Most actin filament nucleators use the WASP-Homology 2 (WH2 or W) domain for interaction with actin. The W domain has a short length (17–27aa) and is extremely abundant and functionally versatile 7; 8; 9. The N-terminal portion of the W domain forms a helix that binds in the hydrophobic (or target-binding) cleft 10 formed between subdomains 1 and 3 at the barbed end of the actin monomer 11; 12; 13. After this helix, the W domain presents an extended region that is directed towards the pointed end of the actin monomer (formed by subdomains 2 and 4 of actin). This region is variable in length and sequence, but comprises the conserved four residue motif LKKT(V), which is critical for the interaction with actin 11. Filament nucleators are characterized by the presence of multiple actin-binding sites. The simplest and most common architecture consists of tandem repeats of the W domain, occurring in the proteins Spire 14, Cobl 15 and VopL/VopF 16; 17. The W domain also participates in filament nucleation through the Nucleation Promoting Factors (NPFs) of the Arp2/3 complex, which can have between one and three W domains 18; 19; 20. The muscle- specific nucleator Lmod also contains one W domain 21. The nucleation activities of tandem W domain-based nucleators vary widely. At least in part, the reason for these differences may lie in the highly variable linkers between W domains. When the linkers are short, as in the relatively weak nucleator Spire 14, only actin subunits along the long-pitch helix of the actin filament can be connected. In contrast, the brain-enriched protein Cobl is a strong nucleator, featuring three W domains with a long linker between its second and third W domains, and is thought to stabilize a short-pitch actin trimer for nucleation 15. The examples of Cobl, the Arp2/3 complex, and formins suggest that stabilization of a short- pitch actin nucleus is a more effective way to promote polymerization than stabilization of a long-pitch actin nucleus 6; 7. However, the structural bases for this observation are not well understood. In an attempt to understand the nucleation mechanism of tandem W domain-based nucleators, we recently reported a solution study, using Small Angle X-ray Scattering (SAXS), of an actin dimer and a trimer stabilized by tandem W domain constructs 22. These complexes, referred to as 2W-Actin and 3W-Actin, and containing respectively two and three W domains, were capped at the barbed end by structure-based crosslinking of the first W domain to Cys-374 of the first actin subunit and at the pointed end by addition of the C- terminal helix of Tβ4. Constructs 2W and 3W are based on the W domain repeat present in the NPF protein N-WASP, which like Spire presents short inter-W linkers. The SAXS study suggested that the actin subunits in the complexes adopted an elongated conformation similar to that of the long-pitch helix of the actin filament. However, the resolution of this study was insufficient to establish a direct comparison between the longitudinal contacts of actin subunits in the complexes and in the actin filament. Rebowski et al. Page 2 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Here we report the crystal structure of 3W-Actin at 7 Å resolution. Only two actin subunits are present in the structure, indicating that one of the actin subunits is released during crystallization. Despite its low resolution, this structure, obtained by fitting high-resolution structures of W-Actin complexes into the low-resolution data, offers a clearer picture of the relative disposition of actin subunits bound to tandem W domains that are separated by Spire-like short inter-W linkers. While the longitudinal arrangement of actin subunits in the structure is somewhat related to that of the long-pitch helix of the actin filament 3; 4, important differences are observed. These differences probably result from steric hindrance of the W domain with inter-subunit contacts in the filament. The determination of the structure of 3W-Actin was aided by determination of the 2.9 Å resolution crystal structure of the first W domains of VopL 16 crosslinked to actin Cys-374 (hereafter referred to as WxActin). The structure of WxActin is nearly undistinguishable from non-crosslinked W- Actin structures determined previously 11; 12; 13, thus validating the use of crosslinking as a tool to stabilize actin polymerization complexes for structural investigation. The structures, and a biochemical analysis of nucleation, reveal important clues about the existing disparities in the nucleation activities of tandem W domain-based nucleators. RESULTS AND DISCUSSION Crystal structure of crosslinked WxActin In two previous studies, we reported low-resolution SAXS structures of actin nucleation complexes formed by the Arp2/3 complex and tandem W domains 22; 23. Barbed end polymerization in these studies was blocked by crosslinking of the W domain to Cys-374 of the actin subunit located at the barbed end of the complexes. This approach was based on analysis of the structures of various W-actin complexes 11; 12; 13, which placed the N- terminus of the W domain within disulfide bond distance to actin Cys-374. In each case, a Cys residue was introduced into the W domain at the most favorable position for crosslinking to actin Cys-374. Here, this approach was used again to obtain the low- resolution crystal structure of 3W-Actin. However, it remained unclear whether the crosslink altered the structure of actin and/or the W domain in a significant way, which prompted us to pursue the determination of a crosslinked WxActin structure. It later became apparent that this structure also provided the best molecular replacement model for determination of the structure of 3W-Actin. After testing crystallization with various W domains, good diffracting crystals were obtained of the crosslinked complex of actin with a synthetic peptide corresponding to the first W domain (amino acids 130–160) of Vibrio parahemolyticus VopL. During synthesis, residue Val-131 of this W domain was replaced by Cys and crosslinked to actin Cys-374 (Materials and Methods). The crystal structure of WxActin was determined by molecular replacement to 2.9 Å resolution (Fig. 1A and Table 1). The structure of WxActin is very similar to those of non-crosslinked W-Actin complexes determined with bound DNase I 11; 13 and that of Drosophila ciboulot bound to actin- latrunculin A 12. Figure 1B shows a comparison of the structure of WxActin with that of the non-crosslinked complex of actin with the W domain WASP (PDB code 2A3Z). The two structures superimpose with r.m.s deviation of 0.66 Å for 358 equivalent Cα atoms. The most important differences occur in regions that were visualized in one of the structures but not the other, including the DNase I-binding loop (D-loop), the C-terminus of actin, and the N-terminus of the W domain. The D-loop is disordered in most structures of actin, as well as in the structure of WxActin described here, but forms an extended β-sheet with β-strands of DNase I in the non-crosslinked structure. The C-terminus of actin is also disordered in most crystal structures, except complexes with profilin, which interacts with the C-terminus of actin 24; 25; 26. In the non-crosslinked W-Actin complex, the last 10 amino acids of actin Rebowski et al. Page 3 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (Gly-366 to Phe-375) are disordered and the W domain is only visualized starting from WASP residue Arg-431 (corresponding to VopL Asn-132). In contrast, in the crosslinked structure only the last amino acid of actin (Phe-375) is unresolved in the electron density map, whereas the W domain of VopL is visualized from residue 130 to 151, i.e. the last nine amino acids of the synthetic peptide were not resolved. The disulfide bond between actin Cys-374 and VopL Cys-131 is visualized in the electron density map (inset in Fig. 1A), although it is poorly defined compared to the rest of the structure. The similarity of the structures suggests that the crosslinking approach used here and in previous studies 22; 23 as a tool to cap the barbed end of actin polymerization nuclei for structural investigation does not introduce significant structural distortions. Furthermore, as we show next, the availability of the structure of WxActin aided the determination of the structure of 3W-Actin. Crystal structure of 3W-Actin The solution SAXS study of 3W-Actin revealed an elongated molecule, consistent with the presence of three actin subunits, somewhat similar to the long-pitch helix of the actin filament 22. However, the nature of actin-actin contacts in the complex could not be determined. We had suggested that subdomain 2 of actin could move slightly, which combined with a helical conformation in the D-loop, would make the binding of tandem W domains fully compatible with intersubunit contacts in the actin filament 3; 4. Other investigators had suggested that the W domain would probably interfere with intersubunit contacts in the filament 27; 28. Knowing which proposal is correct is important, because it may shed light on the mechanism of nucleation, and possibly explain why tandem W domain-based nucleators do not influence elongation the way formins do. It may also answer important questions about differences in the activities of tandem W domain-based nucleators, and the mechanism of action of NPFs of the Arp2/3 complex, which also contain tandem W domains 7. Therefore, we set out to crystallize the complexes of 2W-Actin and 3W-Actin. While both complexes were crystallized readily, the crystals did not diffract the X-rays. Additional search for conditions led to the identification of additives, such as RbCl and polyvinylpyrrolidone K15, which improved diffraction somewhat. After several attempts, the best result consisted of a rather complete and highly redundant X-ray dataset collected from crystals of 3W-Actin to 7 Å resolution. While we initially considered not reporting this structure, we later recognized that significant information could be obtained by positioning high-resolution W-Actin structures into the unit cell of the 3W-Actin crystals by molecular replacement. Because the individual structures are known at high-resolution, this approach overcomes some of the typical limitations of low-resolution structures in which the content of the unit cell is totally unknown. The limitations, however, are that individual atomic positions cannot be refined and the inter-W linkers cannot be visualized. Consistent with the design and mass measurements in solution of the complex of 3W-Actin 22, three copies of the W-Actin basic unit were expected in the asymmetric unit of the crystal. The volume of the asymmetric unit was also compatible with it containing three copies of the W-Actin unit (corresponding to a solvent content of 43%). However, weak diffraction is typically consistent with a higher solvent content. Not surprisingly, the molecular replacement solution, performed independently with the programs Phenix 29; 30 and AMoRe 31, located only two W-Actin complexes in the asymmetric unit, for a solvent content of 62% (see Material and Methods and detailed description in Supplementary Material). We do not understand why one of the actin molecules dissociates during crystallization, although it could simply be that this molecule is bound loosely and is therefore displaced by favorable crystal contacts. Analysis of the crystal packing demonstrates why a third actin molecule was never found. Consecutive actin dimers are stacked head-to-tail, forming a helix along the crystallographic c axis (Movie S1). Two such Rebowski et al. Page 4 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript helices assemble tightly in anti-parallel fashion (see Movie S2). Each anti-parallel pair comprises 24 actin subunits along the length of the c axis, which constitutes the basic building block of the crystal lattice. Adjacent pairs of helices crossover twice in a repeat (or helical turn), corresponding to the length of the c axis (see Movies S3 and S4), thus assuring the connectivity of the crystal lattice and leaving no extra-space for the missing third actin subunit (or rather 12 actin subunits, when the P6522 symmetry of the crystal is taken into consideration). Because of the limited resolution, we could not identify which of the actin subunits is lost during crystallization, or whether the crystals consist solely of the actin subunit crosslinked to the long 3W polypeptide. Note that any non-crosslinked actin dissociated from the complex would be expected to polymerize during crystallization, and would therefore not be present in the crystals. To address this question, a large number of crystals were collected, washed multiple times in the crystallization solution by transferring them with a cryo-loop, and then dissolved in water for analysis by non-reducing gel electrophoresis and mass spectrometry (Materials and Methods). The results clearly illustrate that the crystals consist of a 50/50 mixture of actin crosslinked to construct 3W and non-crosslinked actin (Fig. 2A). Therefore, we conclude that one of the non-crosslinked actins was lost during crystallization which, based on the arrangement of actin subunits in the asymmetric unit, is most likely that bound to the last W domain. The disposition of the actin subunits in the structure of 3W-Actin (Fig. 2B) is somewhat similar to the longitudinal arrangement of actin subunits in the long-pitch helix of the actin filament model 3; 4 (Fig. 2C). However, important differences are observed. To better understand these differences, it is important to discuss what is currently known about longitudinal contacts in the actin filament. Multiple crystal structures of actin show similar longitudinal contacts between actin subunits (including both non-crystallographic dimers and symmetry-related dimers), which are thought to mimic inter-subunit contacts in the actin filament (Table 2). However, because of constraints imposed by crystal symmetry, these dimers are unwound, i.e. they lack the natural twist of the actin filament. A detailed analysis of these structures and their implications for our understanding of the actin filament has been carried out by other investigators 32, and will not be repeated here. However, it is important to compare the structure of 3W-Actin to both the actin filament model 3; 4 and the longitudinal dimers observed in crystal structures, with the understanding that the structure of 3W-Actin does not address the conformation of the actin filament per se, but rather the mechanism of recruitment of actin subunits by tandem W domain proteins. The dimers observed in crystal structures are generally similar and often crystallographically isomorphous. Based on a superimposition of their structures, we have identified three subgroups that diverge more significantly (represented by PDB entries 2FXU, 1Y64 and 2HMP) (Table 2). Compared to a long-pitch dimer of the actin filament model in which consecutive subunits are rotated by ~27° 3; 4, these three subgroups present flat structures, i.e. rotated counterclockwise with respect to the filament dimer by approximately −27° (Fig. 2D) (although the orientation of the axis of rotation is markedly different for entry 2HMP). Remarkably, the longitudinal contacts between subdomains 4 and 3 of neighboring actin subunits are well conserved in the three subgroups (Fig. 3). It thus appears that longitudinal contacts between actin subunits in the filament have a strong tendency to reemerge as crystal contacts in actin structures 32. It is important to note that these structures offer the most accurate view of longitudinal contacts currently available 32, because the resolution of the actin filament model 3; 4 is still insufficient to address specific atomic interactions. Additional longitudinal contacts are thought to involve the D-loop in subdomain 2 33, which is proposed to bind in the hydrophobic cleft between subdomains 1 and 3 of the actin subunit immediately above it 4. However, the D-loop is disordered in all the crystal Rebowski et al. Page 5 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript structures containing longitudinal actin dimers, and its conformation(s) and actual contacts in the filament are unknown. On the other hand, the rotation between the two actin subunits in the structure of 3W-Actin is approximately −33°, i.e. −60° compared with a longitudinal dimer of the actin filament 3; 4 (Fig. 2D). As a result, the longitudinal contacts observed in other crystal dimers are generally broken in the structure of 3W-Actin (Fig. 3), whereas the contacts involving subdomain 2 are unresolved. Therefore, it appears that the presence of the W domain at the interface between actin subunits breaks the natural tendency that actin has to preserve filament-like longitudinal contacts in crystal structures, and induces a rotation between actin subunits that is of similar magnitude but opposite direction to that of the filament (−33° vs 27°). These results are generally consistent with our previous SAXS studies 22, which revealed an extended (pseudo long-pitch) arrangement of the actin subunits stabilized by tandem W domains. However, the SAXS envelope lacked the resolution to distinguish between the dimer observed here in the crystal structure of 3W-actin and a longitudinal dimer of the actin filament model. It is interesting to note that there is also a crystal contact in the structure of 3W-Actin (between two adjacent dimers) that resembles the dimer of the asymmetric unit. This so- called ‘crystal’ dimer differs even more significantly from both the actin filament and the other actin dimers described above, due to an overall translation of ~8 Å between actin subunits compared to the dimer of the asymmetric unit (Fig. 4). In the crystal dimer, the crosslink with construct 3W is at the interface between actin subunits, which may explain the added translation. However, it is significant that the actin subunits of both the non- crystallographic and crystal dimers are rotated counterclockwise by about the same amount compared to all the other actin dimers observed in crystal structures, suggesting that this is a general constraint imposed by the W domain at the interface between actin subunits. We conclude that while Spire-like tandem W domains can bring actin subunits into close proximity for nucleation, the conformation of the polymerization nucleus that they form differs significantly from that of the actin filament. This may explain their weak nucleation activity as analyzed next. Long-pitch nucleation by tandem W domains is suboptimal The structural results prompted us to test the polymerization activity of construct 3W as compared to those of the prototypical tandem W domain nucleator Spire, which stabilizes a long-pitch nucleus, and the Arp2/3 complex, which forms a short-pitch nucleus. The nucleation activity of Drosophila Spire 14 has been mapped to the fragment Spire366–482 comprising the four W domains (Fig. 5A), which was used in the current study. We used the pyrene-actin polymerization assay to study the effect of Spire366–482 on the polymerization of 2 µM actin (6% pyrene labeled) by monitoring the fluorescence increase resulting from the incorporation of labeled actin monomers into the filament (Fig. 5B). At the concentration of 25 nM, Spire366–482 had very little effect on actin polymerization (polymerization rate 1.0±0.2 nM/sec as compared to 0.8±0.1 nM/sec for actin alone), whereas the Arp2/3 complex activated by the WCA fragment of mouse N-WASP showed a major increase in polymerization (polymerization rate 31.5±1 nM/sec). However, the nucleation activity of Spire366–482 increased with concentration, becoming a stronger nucleator at 250 nM (polymerization rate 4.8±0.2 nM/sec). The opposite effect was observed with construct 3W, which had no effect on actin polymerization at the concentration of 25 nM, but inhibited polymerization when used at 250 nM. This could be an indication that construct 3W, like Tβ4, sequesters actin monomers (Fig. 5B). Rebowski et al. Page 6 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Tβ4 is a short 43-aa polypeptide related to the W domain 8; 9, but it contains an additional helix at the C-terminus that binds atop actin subdomains 2 and 4 34 (Fig. 5A). As a result, and despite the apparent simplicity of its helix-loop-helix design, Tβ4 has the ability to block actin monomer addition to both the pointed and barbed ends of the actin filament, making it an extremely effective monomer sequestering protein 35; 36. Some proteins contain tandem repeats of the Tβ4 fold. Examples include, Acanthamoeba castellanii actobindin 37, Drosophila melanogaster ciboulot 12 and Caenorhabditis elegans tetrathymosin 38, which respectively contain two-and-a-half, three and four copies of the Tβ4 fold. Contrary to tandem repeats of the W domain, that frequently mediate filament nucleation 7, tandem Tβ4 proteins are characterized by their ability to sequester actin monomers 37; 38. Therefore, we asked whether 3W, consisting of a tandem repeat of three W domains followed by the C-terminal helix of Tβ4 (Fig. 5A), would sequester actin monomers. A concentration-dependence analysis of steady-state actin polymerization revealed that construct 3W sequesters actin monomers even more effectively than Tβ4 (Fig. 5C). We had previously shown, using analytical ultracentrifugation, light scattering and native gel electrophoresis, that 3W binds three actin monomers in solution 22, which may explain its stronger sequestering activity compared to Tβ4. Therefore, the effect of 3W on actin polymerization is more closely related to that of tetrathymosin, which binds and sequesters multiple actin monomers 38. Although actobindin and ciboulot also sequester actin monomers, perhaps surprisingly they form 1:1 complexes with actin, indicating that only one of their actin-binding sites is fully functional 12; 37. It thus appears that the simple addition of the pointed end capping helix of Tβ4 to tandem W domains changes their activity from nucleation, as in Spire 14, to monomer sequestration as in Tβ4 35; 36. CONCLUSIONS The crystal structure of crosslinked WxActin was found to be nearly undistinguishable from those of non-crosslinked W-Actin complexes. We have used W domain crosslinking in this work, as well as in two previous studies 22; 23. The finding that the structure is not altered in a significant way by the crosslink suggests that this is a structurally sound approach that can be used as a way to stabilize large polymerization complexes, which are intrinsically dynamic, for structural investigation. Various proteins contain tandem repeats of the W domain 7; 8; 9. While the W domain itself presents well-conserved features (N-terminal helix and LKKT(V) motif), the linkers between W domains are highly variable, and no single structure can be fully representative of this large family of proteins. Irrespective of this variability, the inter-W linkers can be sub-divided into two subgroups: short (as in Spire and N-WASP) and long (as in Cobl linker-2). While 3W is a synthetic construct with no natural counterpart, it is based on the tandem W repeat of N-WASP, and it therefore represents the short inter-W linker subgroup. One general implication of the structure of its complex with actin is that the binding of the W domain is intrinsically incompatible with inter-subunit contacts along the long-pitch helix of the actin filament, which is contrary to what we had anticipated 3; 4. The structure of 3W- Actin further suggests that the actin subunits recruited by tandem W domains with short inter-W linkers are positioned in a way that resembles the long-pitch helix of the actin filament, a conformation that would be expected to favor polymerization. However, due to steric hindrance of the W domain, the contacts between actin subunits in these complexes differ significantly from those of the actin filament. This may explain the weak nucleation activity of Spire as compared to the Arp2/3 complex, formins, Cobl and Lmod, all proteins that are thought to stabilize short-pitch actin nuclei to initiate polymerization. The incompatibility of the W domain with longitudinal inter-subunit contacts in the filament also implies that when the actin nucleus transitions into a filament and begins to elongate, tandem W domain nucleators cannot stay bound to newly formed filaments, and would Rebowski et al. Page 7 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript therefore be unlikely to influence elongation. Steric hindrance of the W domain may also be a contributing factor in the release of NPFs of the Arp2/3 complex from branch junctions once the branch filament begins to elongate (conformational changes within the Arp2/3 complex itself could be another factor). We finally found that the simple addition of the C- terminal pointed end-capping helix of Tβ4 to tandem W domains can change their activity from actin filament nucleation to monomer sequestration. MATERIALS AND METHODS Preparation of proteins and protein complexes Detailed descriptions of the preparation and characterization of the complex of 3W-Actin 22 and the purification of the Arp2/3 complex from bovine brain and preparation of the WCA fragment of mouse N-WASP 23 were reported previously. Actin was purified from rabbit skeletal muscle 39. Tβ4 and the first W domain (amino acids Ser-130 to Ser-160) of Vibrio parahemolyticus VopL (UniProt accession code: Q87GE5) were made as synthetic peptides, and purified by reverse-phase chromatography. During peptide synthesis, an amino acid substitution was made (Val-131->Cys) in the first W domain of VopL, a position chosen based on analysis of the various W-Actin structures 11; 12; 13 as the most favorable for crosslinking to actin Cys-374. The crosslinking reaction was performed by activation of the W domain peptide with DTNB (5,5'-dithiobis(2-nitrobenzoic acid)), before mixing with actin at an actin:W peptide ratio of 1:1.2. The crosslinked fraction was then separated by gel filtration on a S200 column (Pfizer-Pharmacia) in 25 mM Tris-HCl pH 7.5, 100 mM NaCl, 0.2 mM CaCl2 and 0.2 mM ATP. The fragment 366–482 of Drosophila Spire (Spire366–482), comprising the four W domains, was amplified by PCR from cDNA purchased from Open Biosystems. The PCR product was cloned between the NdeI and EcoRI sites of vector pTYB12 (New England BioLabs). Protein expression was performed in BL21(DE3) cells (Invitrogen) grown in Terrific Broth medium at 37°C until the OD600 was 1.0–1.2. Expression was induced with addition of 0.5 mM isopropylthio-β-D-galactoside for 5 h at 20°C. Cells were resuspended in chitin-column equilibration buffer [20 mM HEPES (pH 7.5), 500 mM NaCl, 1 mM EDTA, and 100 µM PMSF]. After purification on the chitin affinity column and release of the protein by DTT- induced autocleavage of the intein, Spire366–482 was additionally purified on a reverse-phase C18 column (0.1% trifluoroacetic acid and 0–90% acetonitrile), and then dialyzed extensively against 25 mM Tris-HCl pH 7.5, 100 mM NaCl. Crystallization of the complexes of 3W-Actin and WxActin The complex of 3W-Actin (consisting of a tandem repeat of three W domains with three actin subunits bound 22) was dialyzed against 20 mM HEPES pH 7.5, 100 mM NaCl, 0.2 mM CaCl2 and 0.2 mM ATP and concentrated to 15 mg/ml using an Amicon centrifugal filter (Millipore). Needle-like crystals grow within hours, or even minutes, using the hanging drop vapor diffusion method at 20°C, and from drops consisting of a 1:1 (v/v) mixture of protein solution and a well solution containing 100 mM CAPS pH 10.0, and 24% polyethylene glycol 3350. However, these crystals did not diffract the X-rays. Crystal quality and diffraction were improved with addition of 10–100 mM RbCl or polyvinylpyrrolidone K15 as additives. The crystals were flash-frozen in liquid nitrogen, with addition of 20% glycerol as cryoprotectant. The crosslinked complex WxActin was concentrated to 5 mg/ml, and crystallized using the hanging drop vapor diffusion method at 20°C from a well solution containing 0.2 M LiNO3 and 20% polyethylene glycol 3350. The content of the crystals of 3W-Actin was analyzed by non-reducing gel electrophoresis and mass spectrometry, using a Voyager DE Pro MALDI-TOF Mass Spectrometer (Applied Rebowski et al. Page 8 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Biosystems) and sinapinic acid as a matrix. For this analysis, multiple crystals were collected and washed five times through the crystallization solution by transferring them with a cryo-loop, and then dissolved in water. Data collection and determination of the structures X-ray datasets were collected from crystals of WxActin and 3W-Actin at the beamline 17- BM of the IMCA-CAT facility of the Advance Photon Source (Argonne, IL). Data indexation and scaling were carried out with the program HKL2000 (HKL Research, Inc.). The crystals of 3W-Actin diffracted only to 7.0 Å resolution (Table 1). The data in the last resolution shell (7.25 – 7.0 Å) is weak (I/σ 1.1) and only 34.3% complete. Yet, ~70% of the data was obtained between 7.9 – 7.0 Å, with average I/σ of 2.1 and redundancy of 6. This range includes 449 reflections (~20% of the total). Because of the limited resolution, special emphasis was placed on obtaining a highly redundant dataset (the average redundancy is 20.5 for the entire dataset), which should minimize intensity errors. The structure of WxActin was determined by molecular replacement, using as search model the structure of actin complexed with the W domain of WASP (PDB code: 2A3Z). Molecular replacement and refinement were carried out with the program Phenix 29, and model building was performed with the program Coot 40. The structure of 3W-Actin was determined by molecular replacement, using the stronger data between 15 and 8 Å resolution, and independently with two different programs; Phaser 30 belonging to the Phenix package 29 and AMoRe 31. The two programs gave the same solution. The likelihood-based scoring function (LLG) of the program Phenix is highly sensitive to the quality of the search model 41. Several search models were tested, including monomeric actin 42, complexes of W-Actin determined as ternary complexes with DNase I 11; 13, the complex of ciboulot-actin with bound latrunculin A 12, the structure of actin with the C-terminal portion of Tβ4 34, and the structure of WxActin determined here. The best- contrasted solution was obtained using the structure of WxActin as search model. Two different models were prepared based on this structure, one consisting of the entire crosslinked complex and one lacking the crosslinked portion (i.e. the last 5 amino acids of actin and the first 3 amino acids of the W domain). These two models were positioned independently using a multibody-body search, and clearly defined the locations of the first (crosslinked) and the second (non-crosslinked) actin subunits of the dimer. While Phenix was used in the automated mode, a more exhaustive search was performed with the program AMoRe (details in Supplementary Material). AMoRe’s self-rotation function gave a single prominent peak with correlation coefficient 0.62. Thus, while the volume of the unit cell seemed to be compatible with the presence of three W-Actin complexes in the asymmetric unit (corresponding to a Matthews’ coefficient Vm of 2.15 Å3/ Da and a solvent content of 43%), only two were found (for a Vm of 3.23 Å3/Da and a solvent content of 62%). We tested many possible configurations in which the orientation of one W-Actin complex was constrained with respect to the other according to the non- crystallographic two-fold axis resulting from the self-rotation function. This gave a clearly contrasted solution for two W-Actin complexes, where the correlation between calculated and observed structure factor amplitudes was 0.66 (0.50 for the next peak that was not contrasted above background). Because of the limited resolution, the only refinement performed after molecular replacement was by rigid body, and using all the diffraction data available. Rebowski et al. Page 9 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Actin Polymerization Assay Pyrene-actin polymerization assays were carried out and analyzed as described 43, using a Cary Eclipse fluorescence spectrophotometer (Varian). All the experiments were performed at 20°C. Prior to data acquisition, 2 µM Mg-ATP-actin (6% pyrene-labeled) was mixed with different concentrations of construct 3W (25 nM, 250 nM or 1 µM), Tβ4 (1 µM), and Spire (25 nM, 250 nM) in F-buffer (10 mM Tris-HCl pH 7.5, 1 mM MgCl2, 50 mM KCl, 1 mM EGTA, 0.02 mg/mL BSA, 0.2 mM ATP, 1 mM DTT, 0.1 mM NaN3). Note that the addition of DTT prevents crosslinking of construct 3W to actin Cys-374 during the polymerization assay. Polymerization rates were measured from the slope of the polymerization curve at 50% polymerization and converted to nM/sec assuming that the total concentration of polymerizable actin monomers is 1.9 µM (2 µM – 0.1 µM, i.e. by subtracting the critical concentration for actin monomer addition to the barbed-end from the total concentration of actin) 43. Steady-state experiments with varying Tβ4 or 3W concentrations were carried out under similar condition by allowing actin to polymerize for 16h. Miscellaneous The program DynDom 44 was used to calculate the relative rotation of actin subunits in crystal structures of longitudinal actin dimers. Illustrations of the structures were prepared with the program PyMol (DeLano Scientific LLC). Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments Supported by National Institutes of Health grant P01 HL086655. Use of IMCA-CAT beamline 17-BM was supported by the Industrial Macromolecular Crystallography Association through a contract with Hauptman- Woodward Medical Research Institute. The Advanced Photon Source was supported by Department of Energy Contract W-31-109-Eng-38. REFERENCES 1. Sept D, McCammon JA. Thermodynamics and kinetics of actin filament nucleation. Biophys J. 2001; 81:667–674. [PubMed: 11463615] 2. Pollard TD, Borisy GG. Cellular motility driven by assembly and disassembly of actin filaments. Cell. 2003; 112:453–465. [PubMed: 12600310] 3. Holmes KC, Popp D, Gebhard W, Kabsch W. Atomic model of the actin filament. Nature. 1990; 347:44–49. [PubMed: 2395461] 4. Oda T, Iwasa M, Aihara T, Maeda Y, Narita A. The nature of the globular- to fibrous-actin transition. Nature. 2009; 457:441–445. [PubMed: 19158791] 5. Holmes KC. Structural biology: actin in a twist. Nature. 2009; 457:389–390. [PubMed: 19158779] 6. Dominguez R. Structural insights into de novo actin polymerization. Curr Opin Struct Biol. 2010; 20:217–225. [PubMed: 20096561] 7. Dominguez R. Actin filament nucleation and elongation factors--structure-function relationships. Crit Rev Biochem Mol Biol. 2009; 44:351–366. [PubMed: 19874150] 8. Paunola E, Mattila PK, Lappalainen P. WH2 domain: a small, versatile adapter for actin monomers. FEBS Lett. 2002; 513:92–97. [PubMed: 11911886] 9. Dominguez R. The beta-thymosin/WH2 fold: multifunctionality and structure. Ann N Y Acad Sci. 2007; 1112:86–94. [PubMed: 17468236] 10. Dominguez R. Actin-binding proteins--a unifying hypothesis. Trends Biochem Sci. 2004; 29:572– 578. [PubMed: 15501675] Rebowski et al. Page 10 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 11. Chereau D, Kerff F, Graceffa P, Grabarek Z, Langsetmo K, Dominguez R. Actin-bound structures of Wiskott-Aldrich syndrome protein (WASP)-homology domain 2 and the implications for filament assembly. Proc Natl Acad Sci U S A. 2005; 102:16644–16649. [PubMed: 16275905] 12. Hertzog M, van Heijenoort C, Didry D, Gaudier M, Coutant J, Gigant B, Didelot G, Preat T, Knossow M, Guittet E, Carlier MF. The beta-thymosin/WH2 domain; structural basis for the switch from inhibition to promotion of actin assembly. Cell. 2004; 117:611–623. [PubMed: 15163409] 13. Lee SH, Kerff F, Chereau D, Ferron F, Klug A, Dominguez R. Structural Basis for the Actin- Binding Function of Missing-in-Metastasis. Structure. 2007; 15:145–155. [PubMed: 17292833] 14. Quinlan ME, Heuser JE, Kerkhoff E, Mullins RD. Drosophila Spire is an actin nucleation factor. Nature. 2005; 433:382–388. [PubMed: 15674283] 15. Ahuja R, Pinyol R, Reichenbach N, Custer L, Klingensmith J, Kessels MM, Qualmann B. Cordon- Bleu Is an Actin Nucleation Factor and Controls Neuronal Morphology. Cell. 2007; 131:337–350. [PubMed: 17956734] 16. Liverman AD, Cheng HC, Trosky JE, Leung DW, Yarbrough ML, Burdette DL, Rosen MK, Orth K. Arp2/3-independent assembly of actin by Vibrio type III effector VopL. Proc Natl Acad Sci U S A. 2007; 104:17117–17122. [PubMed: 17942696] 17. Tam VC, Serruto D, Dziejman M, Brieher W, Mekalanos JJ. A Type III Secretion System in Vibrio cholerae Translocates a Formin/Spire Hybrid-like Actin Nucleator to Promote Intestinal Colonization. Cell Host and Microbe. 2007; 1:95–107. [PubMed: 18005688] 18. Pollard TD. Regulation of actin filament assembly by Arp2/3 complex and formins. Annu Rev Biophys Biomol Struct. 2007; 36:451–477. [PubMed: 17477841] 19. Goley ED, Welch MD. The ARP2/3 complex: an actin nucleator comes of age. Nat Rev Mol Cell Biol. 2006; 7:713–726. [PubMed: 16990851] 20. Zuchero JB, Coutts AS, Quinlan ME, Thangue NB, Mullins RD. p53-cofactor JMY is a multifunctional actin nucleation factor. Nat Cell Biol. 2009; 11:451–459. [PubMed: 19287377] 21. Chereau D, Boczkowska M, Skwarek-Maruszewska A, Fujiwara I, Hayes DB, Rebowski G, Lappalainen P, Pollard TD, Dominguez R. Leiomodin is an actin filament nucleator in muscle cells. Science. 2008; 320:239–243. [PubMed: 18403713] 22. Rebowski G, Boczkowska M, Hayes DB, Guo L, Irving TC, Dominguez R. X-ray scattering study of actin polymerization nuclei assembled by tandem W domains. Proc Natl Acad Sci U S A. 2008; 105:10785–10790. [PubMed: 18669664] 23. Boczkowska M, Rebowski G, Petoukhov MV, Hayes DB, Svergun DI, Dominguez R. X-ray scattering study of activated Arp2/3 complex with bound actin-WCA. Structure. 2008; 16:695– 704. [PubMed: 18462674] 24. Schutt CE, Myslik JC, Rozycki MD, Goonesekere NC, Lindberg U. The structure of crystalline profilin-beta-actin. Nature. 1993; 365:810–816. [PubMed: 8413665] 25. Ferron F, Rebowski G, Lee SH, Dominguez R. Structural basis for the recruitment of profilin-actin complexes during filament elongation by Ena/VASP. Embo J. 2007; 26:4597–4606. [PubMed: 17914456] 26. Baek K, Liu X, Ferron F, Shu S, Korn ED, Dominguez R. Modulation of actin structure and function by phosphorylation of Tyr-53 and profilin binding. Proc Natl Acad Sci U S A. 2008; 105:11748–11753. [PubMed: 18689676] 27. Aguda AH, Burtnick LD, Robinson RC. The state of the filament. EMBO Rep. 2005; 6:220–226. [PubMed: 15741975] 28. Renault L, Bugyi B, Carlier MF. Spire and Cordon-bleu: multifunctional regulators of actin dynamics. Trends Cell Biol. 2008; 18:494–504. [PubMed: 18774717] 29. Zwart PH, Afonine PV, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, McKee E, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Storoni LC, Terwilliger TC, Adams PD. Automated structure solution with the PHENIX suite. Methods Mol Biol. 2008; 426:419–435. [PubMed: 18542881] 30. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software. J Appl Crystallogr. 2007; 40:658–674. [PubMed: 19461840] Rebowski et al. Page 11 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 31. Navaza J. Implementation of molecular replacement in AMoRe. Acta Crystallogr D Biol Crystallogr. 2001; 57:1367–1372. [PubMed: 11567147] 32. Sawaya MR, Kudryashov DS, Pashkov I, Adisetiyo H, Reisler E, Yeates TO. Multiple crystal structures of actin dimers and their implications for interactions in the actin filament. Acta Crystallogr D Biol Crystallogr. 2008; 64:454–465. [PubMed: 18391412] 33. Reisler E, Egelman EH. Actin structure and function: what we still do not understand. J Biol Chem. 2007; 282:36133–36137. [PubMed: 17965017] 34. Irobi E, Aguda AH, Larsson M, Guerin C, Yin HL, Burtnick LD, Blanchoin L, Robinson RC. Structural basis of actin sequestration by thymosin-beta4: implications for WH2 proteins. Embo J. 2004; 23:3599–3608. [PubMed: 15329672] 35. Weber A, Nachmias VT, Pennise CR, Pring M, Safer D. Interaction of thymosin beta 4 with muscle and platelet actin: implications for actin sequestration in resting platelets. Biochemistry. 1992; 31:6179–6185. [PubMed: 1627561] 36. Xue B, Aguda AH, Robinson RC. Models of the actin-bound forms of the beta-thymosins. Ann N Y Acad Sci. 2007; 1112:56–66. [PubMed: 17468228] 37. Hertzog M, Yarmola EG, Didry D, Bubb MR, Carlier MF. Control of actin dynamics by proteins made of beta-thymosin repeats: the actobindin family. J Biol Chem. 2002; 277:14786–14792. [PubMed: 11856744] 38. Van Troys M, Ono K, Dewitte D, Jonckheere V, De Ruyck N, Vandekerckhove J, Ono S, Ampe C. TetraThymosinbeta is required for actin dynamics in Caenorhabditis elegans and acts via functionally different actin-binding repeats. Mol Biol Cell. 2004; 15:4735–4748. [PubMed: 15269284] 39. Pardee JD, Spudich JA. Purification of muscle actin. Methods Enzymol. 1982; 85(Pt B):164–181. [PubMed: 7121269] 40. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] 41. McCoy AJ, Grosse-Kunstleve RW, Storoni LC, Read RJ. Likelihood-enhanced fast translation functions. Acta Crystallogr D Biol Crystallogr. 2005; 61:458–464. [PubMed: 15805601] 42. Rould MA, Wan Q, Joel PB, Lowey S, Trybus KM. Crystal structures of expressed non- polymerizable monomeric actin in the ADP and ATP states. J Biol Chem. 2006; 281:31909– 31919. [PubMed: 16920713] 43. Harris ES, Higgs HN. Biochemical analysis of mammalian formin effects on actin dynamics. Methods Enzymol. 2006; 406:190–214. [PubMed: 16472659] 44. Hayward S, Berendsen HJ. Systematic analysis of domain motions in proteins from conformational change: new results on citrate synthase and T4 lysozyme. Proteins. 1998; 30:144–154. [PubMed: 9489922] 45. Allingham JS, Zampella A, D'Auria MV, Rayment I. Structures of microfilament destabilizing toxins bound to actin provide insight into toxin design and activity. Proc Natl Acad Sci U S A. 2005; 102:14527–14532. [PubMed: 16192358] 46. Rizvi SA, Tereshko V, Kossiakoff AA, Kozmin SA. Structure of bistramide A-actin complex at a 1.35 angstroms resolution. J Am Chem Soc. 2006; 128:3882–3883. [PubMed: 16551075] 47. Kudryashov DS, Sawaya MR, Adisetiyo H, Norcross T, Hegyi G, Reisler E, Yeates TO. The crystal structure of a cross-linked actin dimer suggests a detailed molecular interface in F-actin. Proc Natl Acad Sci U S A. 2005; 102:13105–13110. [PubMed: 16141336] 48. Otomo T, Tomchick DR, Otomo C, Panchal SC, Machius M, Rosen MK. Structural basis of actin filament nucleation and processive capping by a formin homology 2 domain. Nature. 2005; 433:488–494. [PubMed: 15635372] 49. Klenchin VA, Khaitlina SY, Rayment I. Crystal structure of polymerization-competent actin. J Mol Biol. 2006; 362:140–150. [PubMed: 16893553] Rebowski et al. Page 12 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript FIG. 1. Structure of WxActin. (A) Two perpendicular views of the structure of WxActin. The inset shows the 2Fo-Fc electron density map (contoured at 1σ) in the region around the crosslink. Although the crosslink was visualized, this is one of less well-defined regions of the map. (B) Superimposition of the structures of WxActin (blue, actin; red, W domain) and the non- crosslinked complex of actin with the W domain of WASP (pink, actin; yellow, W domain), showing the similarity of the structures. (Two Column Figure). Rebowski et al. Page 13 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript FIG. 2. Structure of 3W-Actin. (A) Non-reducing gel electrophoresis and mass spectrometry analysis indicate that the crystals of 3W-Actin consist of a 50/50 mixture of actin crosslinked to construct 3W (expected mass 53,021 Da) and non-crosslinked actin. Actin is also shown in the gel as a control. (B) Illustration of the actin dimer in the structure of 3W- Actin. The linker between W domains was modeled. (C) Illustration of a longitudinal actin dimer from the actin filament model 4. (D) Comparisons of the relative rotations between actin subunits in the actin filament model (gray and magenta) and the structures of 3W- Actin and three representative actin dimers observed in crystal structures (including non- crystallographic and symmetry-related dimers, see also Table 2). For this comparison, the structures were superimposed using as reference the lower actin subunit (gray), which is only shown for the filament model. Note that compared to a long-pitch dimer of the actin filament, in which subunits are rotated by ~27° (magenta arrow), there is a −60° rotation between the two crystallographically independent actin subunits in the structure of 3W- Actin. Other dimers observed in crystal structures tend to be flat due to symmetry constraints and are therefore rotated −27° relative to a longitudinal dimer of the filament. Rebowski et al. Page 14 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The relative rotations between actin subunits were calculated with the program DynDom 44. (Two Column Figure) Rebowski et al. Page 15 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript FIG. 3. Inter-subunit contacts in the structure of 3W-Actin compared to those of crystallographic actin dimers. Insets show that longitudinal contacts between subdomains 4 and 3 of adjacent actin subunits observed in various crystal structures (right) are mostly broken in the structure of 3W-Actin (left). Representative distances between Cβ atoms are shown for reference. (Two Column Figure) Rebowski et al. Page 16 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript FIG. 4. Comparison of the non-crystallographic and crystallographic dimers in the structure of 3W- Actin. (A) Representation of two consecutive dimers related by crystal symmetry. (B) Superimposition of the non-crystallographic (yellow-blue and red W domains) and crystallographic (blue-yellow and magenta W domains) dimers. Note that despite their general similarity, the crystallographic dimer differs more significantly from other actin dimers observed in crystals structures (Table 2) and the actin filament model. While the actin subunits in the crystallographic dimer are rotated ~13° relative to the non- crystallographic dimer, which undoes part of the initial −60° rotation, there is also a translation of ~8 Å, probably imposed by steric hindrance with the crosslinked W domain. It is nonetheless significant that the actin subunits of both the non-crystallographic and crystallographic dimers are rotated counterclockwise by about the same amount compared to all the other dimers observed in crystal structures, which are generally unwound (see Fig. 2), suggesting that this is a general property of the W domain at the interface between actin subunits. (Two Column Figure) Rebowski et al. Page 17 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript FIG. 5. Different effects of Spire, 3W, and Tβ4 on actin polymerization. (A) Schematic diagram of Tβ4, the four W domain region of Drosophila Spire, and construct 3W. Note that construct 3W consists of three W domains (occurring naturally in mouse N-WASP) separated by short linkers (as in Spire) and the pointed end capping helix of Tβ4. This construct also contains a Cys residue at the N-terminus that was crosslinked to actin Cys-374 for crystallization, but the crosslink was reduced with DTT to measure the nucleation activity. (B) Time course of polymerization of 2 µM Mg-ATP-actin (6% pyrene-labeled) alone (black) or in the presence of different concentrations of Spire366–482 (different shades of blue), construct 3W (different shades of green), Tβ4 (pink), and 25 nM Arp2/3 complex with 250 nM mouse N-WASP Rebowski et al. Page 18 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript WCA (red). Each experiment was repeated at least three times. The polymerization rates are: Actin (0.8±0.1 nM/sec), Spire (1.0±0.2 nM/sec at 25nM and 4.8±0.2 nM/sec at 250 nM), Arp2/3 complex (31.5±1 nM/sec). (C) Steady-state concentration-dependence of actin monomer sequestration by Tβ4 and 3W. (One Column Figure) Rebowski et al. Page 19 J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Rebowski et al. Page 20 Table 1 Crystallographic Data and Refinement Statistics WxActin 3W-Actin Diffraction data Wavelength (Å) 1.0 1.0 Space group P 212121 P 65 2 2 Unit cell a/b/c (Å) 66.6 / 76.4 / 86.1 100.7 / 100.7 / 458.8 Unit cell α/β/γ (°) 90.0 / 90.0 / 90.0 90.0 / 90.0 / 120.0 Resolution (Å) 50.0-2.89 (2.99–2.89) 50.0-7.0 (7.9–7.0) Unique reflections 10207 2182 Completeness (%) 99.2 (92.5) 90.0 (70.1) Redundancy 12.9 (6.4) 20.5 (6.0) Rmergea (%) 16.8 (46.1) 8.6 (37.3) I/σ 16.3 (1.8) 16.5 (2.1) Refinement Resolution (Å) 37.51–2.89 Atoms used in refinement 3058 Rfactorb (%) 21.2 Rfreec (%) 26.5 No atomic refinement was performed Rmsd bond lengths (Å) 0.011 Rmsd bond angles (°) 1.910 Average B factors (Å2) All atoms 62.90 Protein atoms 62.92 Solvent 58.57 Residues in Ramachandran plot Most favored regions (%) 90.2 Other allowed regions (%) 9.8 PDB accession code 3M1F 3M3N Values in parentheses correspond to highest resolution shell aRmerge=Σhkl(I-<I>)/ΣI, where I and <I> are the observed and mean intensities of all the observations of reflection hkl, including its symmetry- related equivalents bRfactor=Σhkl∥Fobs| - |Fcalc∥/Σ|Fobs|, where Fobs and Fcalc are the observed and calculated structure factors of reflection hkl cRfree, Rfactor calculated for a randomly selected subset of reflections (5%) that were not used in refinement J Mol Biol. Author manuscript; available in PMC 2011 October 15. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Rebowski et al. Page 21 Table 2 Structures of Longitudinal Actin Dimers. PDB actins per AU Description Symmetry, Resolution (Å) and Cell a, b, c (Å), α, β, γ (°) References 3M3N 2 Dimer stabilized by tandem W domains P6522, 7.0 100.7, 100.7, 458.8, 90.0, 90.0, 120.0 This work 2HF3 2HF4 1 non-polymerizable actin mutant (Ala-204Glu/Pro-243Lys) C2, 1.8 199.7, 54.1, 39.6, 90.0, 93.2, 90.0 42 2ASM 2ASO 1 Complexes with marine macrolides C2, 1.6 171.2, 54.7, 40.7, 90.0, 96.0, 90.00 45; 46 2ASP or 2FXU C2, 1.35 60.1, 56.5, 101.7, 90.0, 94.6, 90.0 2A5X 1 Longitudinally crosslinked actin dimer C2, 2.49 207.4, 54.4, 36.2, 90.0, 98.6, 90.0 47 2Q1N 2Q31 2 Longitudinally crosslinked actin dimer P21, 2.7 108.1, 71.8, 54.8, 90.0, 104.7, 90.0 32 1Y64 1 Complex with formin homology 2 domain C2, 3.05 232.0, 56.2, 100.9, 90.0 107.7, 90.0 48 2HMP 2 Non-polymerizable actin, cleaved between Gly42 and Val43 P212121, 1.9 64, 198, 69.6, 90.0, 90.0, 90.0 49 J Mol Biol. Author manuscript; available in PMC 2011 October 15.
3M3R
Crystal structure of the M113F alpha-hemolysin mutant complexed with beta-cyclodextrin
Molecular bases of cyclodextrin adapter interactions with engineered protein nanopores Arijit Banerjeea, Ellina Mikhailovaa, Stephen Cheleya, Li-Qun Gub,2, Michelle Montoyac,3, Yasuo Nagaokaa,4, Eric Gouauxd, and Hagan Bayleya,1 aDepartment of Chemistry, University of Oxford, Oxford, OX1 3TA, United Kingdom; bDepartment of Medical Biochemistry & Genetics, Texas A&M University System Health Science Center, College Station, TX 77843-1114; cDepartment of Biochemistry and Molecular Biophysics and Howard Hughes Medical Institute, Columbia University, New York, NY 10032; and dVollum Institute and Howard Hughes Medical Institute, Oregon Health and Science University, Portland, OR 97239 Edited by Gregory A. Petsko, Brandeis University, Waltham, MA, and approved March 3, 2010 (received for review December 15, 2009) Engineered protein pores have several potential applications in biotechnology: as sensor elements in stochastic detection and ultrarapid DNA sequencing, as nanoreactors to observe single- molecule chemistry, and in the construction of nano- and micro- devices. One important class of pores contains molecular adapters, which provide internal binding sites for small molecules. Mutants of the α-hemolysin (αHL) pore that bind the adapter β-cyclodextrin (βCD) ∼104 times more tightly than the wild type have been ob- tained. We now use single-channel electrical recording, protein en- gineering including unnatural amino acid mutagenesis, and high- resolution x-ray crystallography to provide definitive structural in- formation on these engineered protein nanopores in unparalleled detail. alpha-hemolysin ∣single molecule ∣stochastic sensing ∣structure ∣ unnatural amino acid M any research groups have used protein engineering to obtain enzymes and antibodies with new properties suited for specific tasks (1–6). Fewer groups have taken on the difficult problem of engineering membrane proteins (7). We have engi- neered the α-hemolysin protein pore, mindful of several potential applications in biotechnology, including its ability to act as a de- tector in stochastic sensing (8) and ultrarapid DNA sequencing (9), to serve as a nanoreactor for the observation of single- molecule chemistry (10) and to act as a component for the con- struction of nano- and microdevices (11). An important breakthrough in this area, which enabled the sto- chastic sensing of organic molecules including the detection of DNA bases in the form of nucleoside monophosphates (12, 13), was the discovery of internal molecular adapters, a form of non- covalent protein modification (14). Most useful have been cyclo- dextrin (CD) adapters, which have until now been used in the absence of detailed structural information about how they work. The present paper is a definitive investigation, which provides such information through the application of a wide variety of technical approaches: single-channel electrical recording, protein engineering including unnatural amino acid mutagenesis, and x-ray crystallography. The studies employing mutagenesis show that the striking interactions seen in the crystal structures also occur in individual pores in lipid bilayers. We reveal that the tight-binding αHL mutants (15) M113N7 and M113F7 bind βCD in different orientations within the hep- tameric pore. In the case of M113N7, the top (primary hydroxyls) of the CD ring faces the trans entrance of the pore. In the case of M113F7, the bottom (secondary hydroxyls) of the CD ring faces the trans entrance, while the top of the ring is bonded to the pore through remarkable CH-π interactions. Another tight-binding mutant, M113V7, can bind the CD in both orientations. These results illustrate the exquisite level of engineering that can be achieved with protein nanopores, which is, for example, far be- yond what is possible with solid-state pores. The work also pro- vides information valuable for the design of new binding sites within the lumen of the αHL pore or within other β-barrel pro- teins. Our results will be of interest to others exploring the inter- actions of CDs with the αHL pore (16, 17), including groups involved in computational studies (18, 19). In addition CDs bind to a variety of other pores, including porins (20, 21) and connex- ins (22), and are being tested in vivo as blockers of the anthrax protective antigen pore (23, 24). The CD adapter concept has also been incorporated into other formats, e.g., with glass nano- pores (25), and artificial pores based on CDs have been made by several groups (26–28). Our work is pertinent to these studies. Results Kinetics and Thermodynamics of the Interactions of βCD with αHL Pores Containing Met, Phe and Asn at Position 113. We showed earlier that position 113 in the αHL pore (Fig. 1A) is critical for the bind- ing of βCD (14). Subsequently, residue 113, which is Met in the WT protein, was changed to each of the remaining 19 naturally occurring amino acids by site-directed mutagenesis (15). We found that 11 of these mutants, expressed as homoheptamers, bound βCD with a similar affinity and with similar kinetics to the WT homoheptamer. Two mutants (P, W) bound βCD about 10 times more strongly than the WT homoheptamer, while six of them (V, H, Y, D, N, F) bound with high affinity, i.e., with a Kd value 103 to 104 times lower than the WT. Remarkably, the side chains of the latter six amino acids bear little resemblance to one another, and this issue is addressed in the present paper. We first examined the two amino acids with the most disparate side chains (Fand N) by making heteromeric pores containing WT (Met-113), M113F, and M113N subunits. Three series of heteroheptamers were produced: WT7−nM113Nn, WT7−nM113Fn, and M113F7−nM113Nn. The heteroheptamers were separated by SDS-polyacrylamide gel electrophoresis aided by an oligoaspartate (D8) tail on the first of the two types of sub- unit (Fig. 1B) (29). All 21 combinations of WT, M113F, and M113N subunits formed αHL pores that interacted with βCD as shown by single-channel current recordings, which revealed the extent of block by βCD (Fig. S1), the association and dissociation Author contributions: A.B., S.C., E.G., and H.B. designed research; A.B., E.M., S.C., L.-Q.G., M.M., and Y.N. performed research; A.B., E.G., and H.B. analyzed data; and A.B. and H.B. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. 1To whom correspondence should be addressed. E-mail: hagan.bayley@chem.ox.ac.uk. 2Present address: Department of Biological Engineering and Dalton Cardiovascular Research Center, University of Missouri, Columbia, MO 65211. 3Present address: Nature Structural & Molecular Biology, 75 Varick Street, 9th Floor, New York NY 10013-1917. 4Present address: Department of Biotechnology, Faculty of Engineering, Kansai University, 3-3-35 Yamate-cho, Suita, Osaka 564-8680, Japan. This article contains supporting information online at www.pnas.org/cgi/content/full/ 0914229107/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.0914229107 PNAS ∣May 4, 2010 ∣vol. 107 ∣no. 18 ∣8165–8170 BIOCHEMISTRY rate constants for βCD (kon and koff), and (from the latter) the equilibrium dissociation constant for βCD (Kd ¼ koff∕kon) (15). The kon values for βCD for the 21 combinations of subunits were all similar at ∼5 × 105 M−1 s−1 (Fig. 1C, Upper). By contrast, the koff values differed widely, ranging from ∼5 × 10−2 s−1 to ∼103 s−1. For WT7−nM113Nn and WT7−nM113Fn, the koff values decreased as M113N or M113F subunits were added. In the case of M113N, there was a steep drop in the value of koff after the fifth subunit had been incorporated. In the case of M113F, the decrease in the value of koff occurred less precipitously as the M113F subunits were added (Fig. 1C, Lower). Intriguingly, with M113F7−nM113Nn, koff first increased as M113N subunits were added to M113F7 until n ¼ 4 (M113F3M113N4) and then de- creased for larger values of n (Fig. 1C, Lower). We recognize that there is more than one permutation of heteromers containing two to five mutant subunits (Fig. 1B), but we have ignored this fact here because no significant differences in the properties of indi- vidual heteromers were observed. For example, 42 recordings were made of WT5M113N2, which has three permutations. Because, kon showed little variation with subunit composition, the variation in Kd was similar to the variation in koff (Fig. 1C). While these studies were in progress, the crystal structures of βCD complexed to M113N7 (Fig. 2B) and M113F7 (Fig. 2C) were solved (Table S1) (30). High-resolution structures could be obtained because the CD and the αHL pore have the same C7 symmetry. In the case of M113N7, βCD is bound with the second- ary hydroxyl face “upward” (Fig. 2B). In each glucose unit of the βCD, the 2-hydroxyl is hydrogen bonded to the side-chain amide of an Asn-113 (the residue introduced by mutagenesis) and the 3-hydroxyl is hydrogen bonded to the ϵ-amino group of Lys-147. In the case of M113F7, two βCDs are bound to the αHL pore (Fig. 2C). It is the top βCD in the structure that concerns us, be- cause it is in contact with the Phe-113 residues introduced by mu- tagenesis. It is immediately apparent that the top βCD in M113F7 is in the opposite orientation to the βCD in M113N7 with each 6-hydroxyl group in a CH-π bonding interaction (31–35) with a Phe-113 side chain. The opposite orientations of the βCDs in M113N7 and M113F7 immediately explain why heteromers formed from similar numbers of M113N and M113F subunits (e.g., M113N4M113F3) bind βCD weakly (see also Discussion). Unnatural Amino Acid Mutagenesis. To further explore the range of noncovalent interactions that are available when βCD binds to the αHL pore, five unnatural amino acids (Fig. 3A and Fig. S2) were incorporated at position 113, by using the in vitro nonsense codon suppression method (36). In particular, we had noted that M113V7 containing the β-branched Val binds βCD tightly (15), and therefore we compared cyclopropylglycine (Cpg) and cyclo- propylalanine (Cpa). We also further examined the means by which M113F7 binds βCD tightly, by comparing the properties of 4-fluorophenylalanine (f1F), pentafluorophenylalanine (f5F), and cyclohexylalanine (Cha) at position 113. The five homomeric pores all produced single-channel cur- rents with unitary conductance values in the range expected for properly assembled heptamers (Fig. S3). All five bound βCD (Fig. 3B, Level 2), either tightly (f1F, Cpg) or weakly (f5F, Cha, Cpa) as described in detail below. During the long βCD binding events, additional current spikes were seen (Fig. 3B). Similar Fig. 1. Binding of βCD by heteromeric αHL pores formed by WT, M113F and M113N subunits. (A) Crystal structure of WT-αHL (61) showing residue 113 (Met, yellow). Left panel, side view and right panel, top view. (B) Separation of 35S-labeled αHL heteroheptamers by SDS-polyacrylamide electrophoresis. The separation of the M113F7-nM113Nn heteromers is shown as detected by autoradiography of a dried gel. The M113F subunits carried a D8 tail. Lane 1, M113N7; lane 2, M113F7−nM113Nn (the heteromers formed from several preparations made with differing ratios of M113F and M113N subunits were mixed to give roughly equal amounts of each subunit combination); lane 3, M113F7. A diagram of the eight different combinations of subunits and their permuta- tions is shown to the right of the autoradiogram. The various permutations are not separated by electrophoresis. (C) Kinetics of the interaction of βCD with single heteromeric αHL pores as determined by bilayer recording. Values of kon were calculated by using kon ¼ 1∕ðτon½βCDÞ, where τon is the mean interevent interval. Values of koff were determined by using koff ¼ 1∕τoff, where τoff is the mean dwell time of βCD in the pore. Values of Kd were calculated by using Kd ¼ koff∕kon. Each point represents the mean  s:d: for three or more determinations. Where they cannot be seen, the s.d. values lie within the symbol. Black squares, WT7−nM113Nn; gray squares, M113F7−nM113Nn; empty squares, M113F7−nWTn. 8166 ∣ www.pnas.org/cgi/doi/10.1073/pnas.0914229107 Banerjee et al. events had been observed previously with certain Met-113 repla- cement mutants and may represent movement of the βCD at its binding site (e.g., rotation about axes perpendicular to the C7 axis) (15). The additional current spikes were more prevalent for M113V7 and M113Cpg7, which may take part in more con- formationally labile interactions with βCD, compared with say M113F7 (Fig. S4). Interactions of βCD with Homoheptamers Bearing Aromatic Residues at Position 113. To further understand the nature of the binding of βCD to aromatic side chains, we examined the kinetics of βCD binding to the homoheptamers containing f1F or f5F at position 113, M113f1F7 and M113f5F7 (Fig. 3C). For both mutants, the value of kon was very similar to that of WT7, but the values of koff and therefore Kd for M113f1F7 differed dramatically from WT7 and were close to the values for the tight-binding mutant M113F7 (Table S2A). By contrast, koff and Kd for M113f5F7 were similar to the values for WT7 (Table S2A). To determine whether M113f1F7 binds βCD in the same orien- tation as M113F7 (Fig. 2C), we made heteromers of the M113f1F subunit with M113N or M113F and examined M113F4M113f1F3 and M113N4M113f1F3. M113F4M113f1F3 binds βCD as strongly as either M113F7 or M113f1F7, but M113N4M113f1F3 binds βCD weakly with a similar affinity to WT7 (Fig. 3D and Table S3). Therefore, it is reasonable to infer that M113F7 and M113f1F7 bind βCD in the same orientation with the 6- hydroxyl groups of the CD in proximity to the aromatic rings on the protein. Cyclohexylalanine (Cha) was used to replace the aromatic side chains with a roughly isosteric hydrophobic group. Again the va- lue of kon for βCD was little changed, but koff for M113Cha7 had an intermediate value of 42  6 s−1. Therefore, M113Cha7 binds βCD more weakly than M113F7 but distinctly more strongly than the WT7 pore (Table S2A and Fig. 3C). Interactions of βCD with Homoheptamers Bearing Hydrophobic Resi- dues at Position 113. M113V7 binds βCD very strongly, and there- fore we compared αHL pores with Cpg or Cpa at position 113. Cpg is roughly isosteric with Val, and like Val has a β-branched side chain. Gratifyingly, M113Cpg7 has a kon value similar to the other αHL pores, and koff and Kd values close to those of M113V7 (Table S2B and Fig. 3E). Cyclopropylalanine (Cpa), with an additional methylene group compared to Cpg, is roughly isosteric with Leu, a weak binder, and M113Cpa7 also binds βCD weakly with kon, koff and Kd values similar to those of WT7 (Table S2B and Fig. 3E). M113I7 and M113T7, which are β-branched, are also weak binders, but Ile and Thr are less closely related to Val than Cpg. To determine whether M113V7 binds βCD in the same orien- tation as M113F7 or M113N7 (Fig. 2), we made heteromers of M113V and the M113N or M113F subunits. M113V3M113F4, M113V4M113F3, M113V3M113N4, and M113V4M113N3 were examined in detail. All four heteroheptamers bound βCD more weakly than M113V7, M113F7 or M113N7 (Fig. 3F and Table S4), suggesting that Val at position 113 interacts with βCD strongly but in a different manner to either Phe or Asn. Each heteromer exhibited a range of Kd values, perhaps reflecting the various pos- sible permutations of the two different subunits around the cen- tral axis of the heptamer, although this heterogeneity was not seen for heteromers made from WT, M113F and M113N (Fig. 1). Discussion Soon after we discovered that βCD binds to the WT-αHL pore for around a millisecond, we found a mutant pore, M113N7, that re- leases βCD ∼104 times more slowly (14). This prompted us to examine all 19 mutants in which residue 113 is replaced by a nat- ural amino acid, with the surprising result that a collection of ami- no acids with structurally unrelated side chains (V, H, Y, D, N, F) are tight binders (15). Here, we have examined the nature of the binding interactions more closely by single-channel electrical re- cording, protein engineering including unnatural amino acid mu- tagenesis, and high-resolution x-ray crystallography, and we provide the first definitive structural information on an engi- neered protein nanopore. We find that βCD can bind tightly to the αHL pore in three different ways depending on the residue at 113, as exemplified by Asn, Phe, and Val. Because Asn and Phe have quite different side chains, we first compared the ability of M113N and M113F subunits to take part in binding the CD. The examination of het- eromeric proteins containing WT (Met-113), M113N and M113F subunits showed that the replacement of WT subunits in WT7 with M113N or M113F subunits led to increased affinity for βCD. The more M113N or M113F subunits that were added, the tighter binding became. By contrast, when subunits in M113N7 were replaced with M113F subunits, binding became weaker, reaching a minimum at three to four M113F subunits, and then increasing in strength with five M113F subunits or more (Fig. 1C). Parallel structural studies (30) revealed the basis of the “oppos- ing” effects of the M113N and M113F subunits. βCD binds to M113N7 in the opposite orientation to that in which it binds to M113F7. In M113N7, the secondary hydroxyls in the βCD ring are hydrogen bonded to Lys-147 and Asn-113 (Fig. 2A). By con- trast, βCD interacts with M113F7 through its primary hydroxyl face (Fig. 2B). It seemed likely that M113V7, bound βCD in yet another way, and this was examined by forming heteromers between M113V and M113N or M113F. The presence of three or four subunits of either M113N or M113F greatly decreases the affinity of the pore for βCD (Fig. 3F), with an average koff of 7.3 × 101 s−1, indicating that a third binding mode is indeed operating Fig. 2. X-ray structures of M113N and M113F homoheptamers with βCD bound. (A) Side view of heptameric αHL. βCD binds in the blue highlighted region. (B) βCD bound to M113N7 (dotted lines indicate hy- drogen bonding). The side chains of Lys-147 are in pale brown and the side chains of Asn- 113 in yellow. (C) βCD bound to M113F7 (dotted lines indicate CH-π bonding). The side chains of Phe-113 are in yellow. The sec- ond βCD in the M113F7 · ðβCDÞ2 structure is hydrogen bonded to the top βCD in a head- to-head arrangement and has no apparent interactions with the protein. For both (B) and (C), four β strands were omitted from the barrel to give a better view. Banerjee et al. PNAS ∣ May 4, 2010 ∣ vol. 107 ∣ no. 18 ∣ 8167 BIOCHEMISTRY (Table S4). In summary, the three groups of tight-binding mutants comprise αHL pores incorporating, at position 113: (i) the hydro- gen-bonding amino acids N, D (the latter would have to be largely in the protonated form), and possibly H; (ii) the aromatics F, Y, f1F, and possibly H, and more weakly W; (iii) the β-branched ami- no acids V, Cpg. There may be yet other means by which CDs can bind to the αHL pore. For example, we earlier found that hepta- 6-sulfato-βCD can bind tightly to αHL pores containing the N139Q mutation (37). Presumably, this CD is bound at a site low- er down in the β barrel in a fashion that includes hydrogen bond- ing to the Gln at position 139. While the various mutants exhibited widely different koff values, the value of kon was almost invariant and averaged ∼2.3 × 105 M−1 s−1 (Table S2) (15). Ap- parently, transport to the binding site is rate limiting, through a route unaffected by mutagenesis. koff increased precipitously with the addition of WTsubunits to M113N7 (Fig. 1C). Crystal structures of M113N7 show that resi- dues 111, 113, and 147 are reorganized by compari- son with WT7 and then undergo a more limited rearrangement when βCD binds (Fig. S5). For example, the side chain of Lys-147 shifts position to form a bifurcated hydrogen bond with a 3-hydroxyl group of βCD and the side chain carbonyl of an Asn- 113 (Fig. S6). Therefore, the side chains of residues 111, 113, and 147 might be in a variety of conformations in WT7−nM113Nn het- eromers and offer less well preorganized binding sites for βCD than they do in M113N7. Further, the intramolecular hydrogen bonds of the secondary hydroxyls in βCD (38) must be disrupted upon binding as both hydroxyls on each glucose ring form hydro- gen bonds to the mutant subunits (Fig. 2B). Because the hydrogen bonds that are broken in βCD are arranged in a circle, the break- age of bonds involving a single glucose (three bonds in all) will be relatively more disruptive than those involving adjoining glucose residues or the entire circle. The overall binding cooperativity in M113N7 could be attributed to enthalpic cooperativity outweigh- ing entropic penalties to binding (39). Positive cooperativity has been observed previously in fairly rigid model systems (40). By contrast with M113N7, there is little movement of side chains in ðM113FÞ7 by comparison with WT7 and little move- ment, including Phe-113, upon binding βCD (Fig. S7A). Further, the crystal structure of M113F7 · βCD suggests that each Phe re- sidue interacts independently with the βCD through what appear to be CH-π interactions (Fig. S7B). These interactions are ex- pected to be weak and not strongly directional and hence offer less enthalpic cooperativity, as supported by the B-factors (crys- tallographic temperatures factors) at the primary βCD binding site, which are between ∼40 and 50. Positive cooperativity is ob- served, but it is less pronounced than in the case of M113N7 (Table S5). In the case of M113N7, the B-factors of the residues that bind βCD are in the 20s implying that the βCD is more rigidly held than it is in M113F7. The binding of sugars to aromatic residues in proteins can in- clude CH-π bonding (41) or OH-π bonding or a finely balanced complement of both (42, 43). However, we have dismissed the possibility of an OH-π interaction between Phe-113 and the 6-hydroxyl groups of βCD as the distance between the center of the phenyl rings to the nearest hydroxyl oxygen is higher (5.2  0.65 Å, n ¼ 7) than that expected for a favorable OH-π interaction (33). While we propose that βCD binds to Phe-113 through a C-6 CH-π interaction (Fig. S7B), the distances between the center of the Phe-113 ring and the nearest C-6 of βCD ob- served in the M113F7 · βCD structure (4.66  0.24 Å, n ¼ 7) are in the upper range of the expected distance for a strong inter- action, which is ∼4.5 Å (33). The angle between the normal to the aromatic rings and the line connecting the C-6 atoms to the aro- matic midpoint is 8.0  5.6°, which is well within the expected range (44). The measurements with M113f5F7 argue against a hydrophobic interaction between Phe residues at position 113 and the βCD ring. In f5F, the hydrophobicity of the phenyl ring is significantly increased (45) yet M113f5F7 binds βCD weakly, like WT7 (Fig. 3C and Table S2A). By contrast with F, f1F, Y and N, homomeric αHL pores with f5F and W at position 113 bound βCD relatively weakly (Fig. 3C and Table S2A). In the case of f5F, the powerful electron with- drawing action of the five fluorine atoms leaves a highly increased positive charge at the center of the ring (46, 47), mitigating against a hydrogen-bonding interaction. The electron-rich Trp Fig. 3. Properties of pores containing natural and unnatural amino acid sub- stitutions at position 113. The data were recorded at þ40 mV in 1.0 M NaCl, 10 mM sodium phosphate, pH 7.5. (A) Unnatural amino acids used in this study: 4-fluorophenylalanine, f1F; pentafluorophenylalanine, f5F; cyclohex- ylalanine, Cha; cyclopropylglycine, Cpg; cyclopropylalanine, Cpa. (B) Repre- sentative current traces from single homoheptameric αHL pores, containing unnatural amino acids at position 113, in the presence of βCD. βCD (40 μM final) was added to the trans chamber. Level 1, open pore current; level 2, pore occupied by βCD. The broken line indicates zero current. (C) In- teraction of βCD with homomeric αHL pores containing aromatic amino acids at position 113. Kd values for the interaction between βCD and the αHL pore were calculated by using Kd ¼ koff∕kon. Each column represents the mean  s:d: for 10 or more determinations: dark gray, natural amino acids; light gray, unnatural amino acids. Data adapted from Gu and colleagues (15) are marked (*). (D) Representative current traces from single-channel recordings of βCD binding to M113F4M113f1F3 and M113N4M113f1F3. βCD (40 μM final) was added to the trans chamber. The broken line indicates zero current. (E) Interaction of βCD with homomeric αHL pores containing hydrophobic amino acids at position 113. Kd values for the interaction between βCD and the αHL pore were calculated by using Kd ¼ koff∕kon. Each column represents the mean  s:d: for ten or more determinations: dark gray, natural amino acids; light gray, unnatural amino acids. Data adapted from Gu and colleagues (15) are marked (*). (F) koff values for βCD from heteroheptamers formed with M113F and M113V subunits and with M113N and M113V subunits. βCD (40 μM final) was added to the trans chamber. The kon values for βCD for all these mutants are similar, at ∼3 × 105 M−1 s−1. Empty square: average koff values for the mutant (bar is s:d). Filled square: M113V3M113F4; filled circle: M113V4M113F3; filled upright triangle: M113V3M113N4; filled in- verted triangle: M113V4M113N3. 8168 ∣ www.pnas.org/cgi/doi/10.1073/pnas.0914229107 Banerjee et al. ring (44, 46, 47) should favor hydrogen bonding, but here we can- not make a direct comparison with the crystal structure of M113F7 as the indole ring is far larger than benzene. It is possible that it cannot become oriented in the same manner and that it is misaligned for hydrogen bonding. Our experiments suggest that M113V7 and M113Cpg7 bind βCD in a third way. In heteromers with M113V, both M113F and M113N reduce the affinity of the pore for βCD suggesting that neither the CH-π interaction with Phe-113 nor the hydrogen- bonding interactions with Asn-113 and Lys-147 are compatible with binding to Val. Close interactions of Val with glucose rings have been noted previously (48). Therefore, we propose that the Val side-chain interacts with the side of the glucose ring. This in- teraction might occur in one or both orientations of the CD ring (Fig. 4). Conclusion We provide structural information on engineered protein nano- pores and describe three distinct ways in which βCD can bind within the lumen of mutant αHL pores in atomic detail. Our re- sults will be useful in several areas of basic science and biotech- nology. By using host molecules lodged within the αHL pore, host-guest interactions can be investigated in fine detail at the single-molecule level (17, 49). The present work will now permit us to examine binding events at a known face of a CD. The work also provides information for designing new binding sites within the lumen of the αHL pore (37) or within other β barrel proteins (21, 50) and for using molecular design to devise ways in which to covalently attach CDs within pores (13, 51). These areas impact practical applications of nanopore technology including stochas- tic sensing (8), single-molecule DNA sequencing (9, 12, 13, 52), the use of nanoreactors for the observation of single-molecule chemistry (10), and the construction of nano- and microdevices (11, 53), as well as the design of CDs as therapeutic agents (23, 24). Methods Full details of the experimental procedures can be found in SI Appendix. Materials L-Amino acids were obtained as follows: 4-fluorophenylalanine (f1F) (Fluka); pentafluorophenylalanine (f5F) (PepTech Corp.); cyclopropylglycine (Cpg) (Ty- ger); cyclopropylalanine (Cpa) (Tyger). 4-N-benzoyl-5′-O-(4,4′-dimethoxytri- tyl)-2′-deoxycytidine-3′-[(2-cyanoethyl)-(N,N-diisopropyl)]-phosphoramidite and bis(2-cyanoethyl)-N,N-diisopropylphosphoramidite for the synthesis of pdCpA were purchased from Glen Research and Toronto Research Chemicals, respectively. Preparation of NVOC-Protected Aminoacyl-pdCpA. NVOC-protected aminoacyl-pdCpAs were prepared as reported previously by reacting the dinucleotide pdCpA with N-protected, carboxylic acid-activated, amino acids (54–56). Preparation of NVOC-Protected Aminoacyl-tRNA. NVOC-protected aminoacyl- pdCpAs were ligated enzymatically with a truncated tRNA, prepared by using methods described elsewhere (57, 58). Genetic Constructs and Mutagenesis. All new αHL constructs were verified by DNA sequencing. Details of each construct can be found in SI Appendix. Synthesis, Assembly, and Purification of Mutant αHL pores. αHL monomers (WT and mutants) were prepared in vitro by coupled transcription and translation (IVTT) and assembled into homoheptamers on rabbit red blood cell membranes followed by purification by SDS–PAGE as described earlier (59). Heteroheptamers were prepared by mixing the two required DNAs (one encoding an αHL with a D8 tail) before IVTT and then oligomerizing the mixed translation products on rabbit red blood cell membranes. Pores with the desired combinations of subunits were purified by SDS–PAGE (59). Synthesis, Assembly, and Purification of αHL Mutants Containing Unnatural Ami- no Acids. αHL polypeptides containing unnatural amino acids were synthe- sized by IVTT in the presence of rabbit red blood cell membranes. The plasmid with a stop codon (TAG) at position 113 was used. Deprotected ami- noacyl-tRNAs (SI Appendix) were added to the IVTT mixtures. For heterohep- tamers with subunits containing unnatural amino acids in combination with M113N or M113F, monomers were first made, which were then coassembled on rabbit red blood cell membranes and subsequently purified by SDS–PAGE. Single-Channel Current Recordings in Planar Lipid Bilayers. (15, 60) Recordings were made with 1.0 M NaCl, 10 mM sodium phosphate, pH 7.5, in both cham- bers, at an applied potential of þ40 mV. Data were recorded at 22  2°C. The bilayer was formed from 1,2-diphytanoyl-sn-glycero-phosphocholine (Avanti Polar Lipids). Proteins were added to the cis chamber, and βCD to the trans chamber. Single-channel currents were recorded with an Axopatch 200B patch-clamp amplifier (Axon Instruments) and filtered at 2 kHz with a built-in 4-pole low-pass Bessel Filter. The data were acquired at a sampling rate of 10 kHz. For mutants that bind βCD strongly, the data were acquired for at least 30 min and for weak-binding mutants for at least 10 min. Kinetic Data Analysis. Current amplitude and dwell-time histograms were made by using ClampFit 9.0. The mean dwell times, τoff, were determined by fitting the dwell-time histograms to single exponentials. Values of kon and koff were obtained by using the mean dwell times and mean interevent intervals, as described previously (15, 60). This analysis assumes a binary in- teraction, which was supported in all cases examined by the finding of only one major blockade level and a single exponential distribution of dwell times (τoff). Fig. 4. Molecular model showing the three classes of interaction between the αHL pore and βCD identified in this work. The model identifies the region of βCD responsible for each interaction (H atoms interacting with Phe-113 or Asn-113 and Lys-147: gray). The first class of interaction is with aromatic residues and involves the seven -CH2OH groups of the βCD. The second class is typified by the interactions with Asn at position 113, which involve hydro- gen-bonds to the secondary 2-hydroxyls of the βCD. Structural studies show that this interaction is supported by hydrogen bonding between Lys-147 and the secondary 3-hydroxyls of the βCD. Structural studies and experiments with heteromers suggest that the βCD in M113F7 is in the opposite orienta- tion to the βCD in M113N7, in support of the model shown here. As the inter- action with Val is hydrophobic, it is not directional and βCD may not bind at the same position inside the β barrel as it does in M113F7 or M113N7. Banerjee et al. PNAS ∣ May 4, 2010 ∣ vol. 107 ∣ no. 18 ∣ 8169 BIOCHEMISTRY Protein Crystallography. Details can be found in SI Appendix. Protein Data Bank: The coordinates and structure factors of the described structures have been deposited with accession codes 3M2L ðM113F7Þ, 3M3R ðM113F7 · βCDÞ, 3M4D ðM113N7Þ, 3M4E ðM113N7 · βCDÞ. ACKNOWLEDGMENTS. We thank Dennis Dougherty for the plasmid pTHG73. This work was funded by a Royal Society Wolfson Research Merit Award (to H.B.), the Medical Research Council (H.B.), the National Institutes of Health (H.B.), and the Howard Hughes Medical Institute (E.G.). 1. Lu Y, Yeung N, Sieracki N, Marshall NM (2009) Design of functional metalloproteins. Nature 460:855–862. 2. Gebauer M, Skerra A (2009) Engineered protein scaffolds as next-generation antibody therapeutics. Curr Opin Chem Biol 13:245–255. 3. Gronwall C, Stahl S (2009) Engineered affinity proteins—Generation and applications. J Biotechnol 140:254–269. 4. Arnold U (2009) Incorporation of non-natural modules into proteins: Structural features beyond the genetic code. Biotechnol Lett 31:1129–1139. 5. Tracewell CA, Arnold FH (2009) Directed enzyme evolution: Climbing fitness peaks one amino acid at a time. Curr Opin Chem Biol 13:3–9. 6. Fruk L, Kuo CH, Torres E, Niemeyer CM (2009) Apoenzyme reconstitution as a chemical tool for structural enzymology and biotechnology. Angew Chem Int Ed Engl 48:1550–1574. 7. Bayley H, Jayasinghe L (2004) Functional engineered channels and pores. Mol Membr Biol 21:209–220. 8. Bayley H, Cremer PS (2001) Stochastic sensors inspired by biology. Nature 413:226–230. 9. Branton D, et al. (2008) The potential and challenges of nanopore sequencing. Nature Biotechnol 26:1146–1153. 10. Bayley H, Luchian T, Shin S-H, Steffensen MB (2008) Single Molecules and Nanotech- nology, eds R Rigler and H Vogel (Springer, Heidelberg), pp 251–277. 11. Maglia G, et al. (2009) Droplet networks with incorporated protein diodes show collective properties. Nat Nanotechnol 4:437–440. 12. Astier Y, Braha O, Bayley H (2006) Toward single molecule DNA sequencing: Direct identification of ribonucleoside and deoxyribonucleoside 5'-monophosphates by using an engineered protein nanopore equipped with a molecular adapter. J Am Chem Soc 128:1705–1710. 13. Clarke J, et al. (2009) Continuous base identification for single-molecule nanopore DNA sequencing. Nature Nanotechnol 4:265–270. 14. Gu L-Q, Braha O, Conlan S, Cheley S, Bayley H (1999) Stochastic sensing of organic analytes by a pore-forming protein containing a molecular adapter. Nature 398:686–690. 15. Gu L-Q, Cheley S, Bayley H (2001) Prolonged residence time of a noncovalent molecular adapter, β-cyclodextrin, within the lumen of mutant α-hemolysin pores. J Gen Physiol 118:481–494. 16. Ervin EN, Kawano R, White RJ, White HS (2008) Simultaneous alternating and direct current readout of protein ion channel blocking events using glass nanopore membranes. Anal Chem 80:2069–2076. 17. Gurnev PA, Harries D, Parsegian VA, Bezrukov SM (2009) The dynamic side of the Hof- meister effect: A single-molecule nanopore study of specific complex formation. ChemPhysChem 10:1445–1449. 18. Mamonova T, Kurnikova M (2006) Structure and energetics of channel-forming protein-polysaccharide complexes inferred via computational statistical thermody- namics. J Phys Chem B 110:25091–25100. 19. Egwolf B, Luo Y, Walters DE, Roux B (2010) Ion selectivity of alpha-hemolysin with beta-cyclodextrin adapter. II. Multi-ion effects studied with grand canonical monte carlo/brownian dynamics simulations. J Phys Chem B 114:2901–2909. 20. Orlik F, et al. (2003) CymA of Klebsiella oxytoca outer membrane: Binding of cyclodex- trins and study of the current noise of the open membrane. Biophys J 85:876–885. 21. Chen M, Khalid S, Sansom MSP, Bayley H (2008) OmpG: Engineering a quiet pore for biosensing. Proc Natl Acad Sci USA 105:6272–6277. 22. Locke D, Koreen IV, Liu JY, Harris AL (2004) Reversible pore block of connexin channels by cyclodextrins. J Biol Chem 279:22883–22892. 23. Karginov VA, Nestorovich EM, Moayeri M, Leppla SH, Bezrukov SM (2005) Blocking anthrax lethal toxin at the protective antigen channel by using structure-inspired drug design. Proc Natl Acad Sci USA 102:15075–15080. 24. Moayeri M, Robinson TM, Leppla SH, Karginov VA (2008) In vivo efficacy of beta- cyclodextrin derivatives against anthrax lethal toxin. Antimicrob Agents Chemother 52:2239–2241. 25. Gao C, Ding S, Tan Q, Gu LQ (2009) Method of creating a nanopore-terminated probe for single-molecule enantiomer discrimination. Anal Chem 81:80–86. 26. Pregel MJ, Jullien L, Lehn J-M (1992) Towards artificial ion channels: Transport of alkali metal ions across liposomal membranes by “bouquet” molecules. Angew Chem Int Edit 31:1637–1639. 27. Bacri L, Benkhaled A, Guegan P, Auvray L (2005) Ionic channel behavior of modified cyclodextrins inserted in lipid membranes. Langmuir 21:5842–5846. 28. Jog PV, Gin MS (2008) A light-gated synthetic ion channel. Org Lett 10:3693–3696. 29. Howorka S, Cheley S, Bayley H (2001) Sequence-specific detection of individual DNA strands using engineered nanopores. Nat Biotechnol 19:636–639. 30. Montoya M (2004) Insights into Membrane Association and Bioengineering of a Pore- Forming Toxin: Structural Studies of Staphylococcal α-Hemolysin (Columbia University, New York). 31. Steiner T (2002) The hydrogen bond in the solid state. Angew Chem Int Ed 41:49–76. 32. Steiner T (2002) Hydrogen bonds from water molecules to aromatic acceptors in very high-resolution protein crystal structures. Biophys Chem 95:195–201. 33. Steiner T, Koellner G (2001) Hydrogen bonds with pi-acceptors in proteins: Frequencies and role in stabilizing local 3D structures. J Mol Biol 305:535–557. 34. Brandl M, Weiss MS, Jabs A, Sühnel J, Hilgenfeld R (2001) CH...π-interactions in proteins. J Mol Biol 307:357–377. 35. Weiss MS, Brandl M, Suhnel J, Pal D, Hilgenfeld R (2001) More hydrogen bonds for the (structural) biologist. Trends Biochem Sci 26:521–523. 36. Wang L, Xie J, Schultz PG (2006) Expanding the genetic code. Annu Rev Biophys Biomol Struct 35:225–249. 37. Gu L-Q, Cheley S, Bayley H (2001) Capture of a single molecule in a nanocavity. Science 291:636–640. 38. Saenger W, et al. (1998) Structures of the common cyclodextrins and their larger analogues—Beyond the doughnut. Chem Rev 98:1787–1802. 39. Hunter CA, Tomas S (2003) Cooperativity, partially bound states, and enthalpy-entropy compensation. Chem Biol 10:1023–1032. 40. Bisson AP, Hunter CA (1996) Cooperativity in the assembly of zipper complexes. Chem Commun 1723–1724. 41. del Carmen Fernandez-Alonso M, Canada FJ, Jimenez-Barbero J, Cuevas G (2005) Molecular recognition of saccharides by proteins. Insights on the origin of the carbohydrate-aromatic interactions. J Am Chem Soc 127:7379–7386. 42. Jimenez-Barbero J, Asensio JL, Canada FJ, Poveda A (1999) Free and protein-bound carbohydrate structures. Curr Opin Struct Biol 9:549–555. 43. Stanca-Kaposta EC, et al. (2007) Carbohydrate molecular recognition: A spectroscopic investigation of carbohydrate-aromatic interactions. Phys Chem Chem Phys 9:4444–4451. 44. Brandl M, Weiss MS, Jabs A, Suhnel J, Hilgenfeld R (2001) CH...π-interactions in proteins. J Mol Biol 307:357–377. 45. Woll MG, Hadley EB, Mecozzi S, Gellman SH (2006) Stabilizing and destabilizing effects of phenylalanine –>F5-phenylalanine mutations on the folding of a small protein. J Am Chem Soc 128:15932–15933. 46. Mecozzi S, West AP, Jr, Dougherty DA (1996) Cation-pi interactions in aromatics of biological and medicinal interest: Electrostatic potential surfaces as a useful qualitative guide. Proc Natl Acad Sci USA 93:10566–10571. 47. Dougherty DA (2008) Cys-loop neuroreceptors: Structure to the rescue?. Chem Rev 108:1642–1653. 48. Hondoh H, et al. (2008) Substrate recognition mechanism of alpha-1,6-glucosidic linkage hydrolyzing enzyme, dextran glucosidase from Streptococcus mutans. J Mol Biol 378:913–922. 49. Kang XF, Cheley S, Guan X, Bayley H (2006) Stochastic detection of enantiomers. J Am Chem Soc 128:10684–10685. 50. Chen M, Li QH, Bayley H (2008) Orientation of the monomeric porin OmpG in planar lipid bilayers. ChemBioChem 9:3029–3036. 51. Wu H-C, Astier Y, Maglia G, Mikhailova E, Bayley H (2007) Protein nanopores with covalently attached molecular adapters. J Am Chem Soc 129:16142–16148. 52. Bayley H (2006) Sequencing single molecules of DNA. Curr Opin Chem Biol 10:628–637. 53. Astier Y, Bayley H, Howorka S (2005) Protein components for nanodevices. Curr Opin Chem Biol 9:576–584. 54. Robertson SA, Noren CJ, Anthony-Cahill SJ, Griffith MC, Schultz PG (1989) The use of 5'-phospho-2 deoxyribocytidylylriboadenosine as a facile route to chemical aminoacy- lation of tRNA. Nucleic Acids Res 17:9649–9660. 55. Ellman J, Mendel D, Anthony-Cahill S, Noren CJ, Schultz PG (1991) Biosynthetic method for introducing unnatural amino acids site-specifically into proteins. Method Enzymol 202:301–336. 56. Kearney PC, et al. (1996) Dose-response relations for unnatural amino acids at the agonist binding site of the nicotinic acetylcholine receptor: Tests with novel side chains and with several agonists. Mol Pharmacol 50:1401–1412. 57. England TE, Bruce AG, Uhlenbeck OC (1980) Specific labeling of 3′ termini of RNA with T4 RNA ligase. Method Enzymol 65:65–74. 58. Nowak MW, et al. (1998) In vivo incorporation of unnatural amino acids into ion channels in xenopus oocyte expression system. Method Enzymol 293:504–529. 59. Cheley S, Braha O, Lu X, Conlan S, Bayley H (1999) A functional protein pore with a “retro” transmembrane domain. Protein Sci 8:1257–1267. 60. Gu L-Q, et al. (2000) Reversal of charge selectivity in transmembrane protein pores by using non-covalent molecular adapters. Proc Natl Acad Sci USA 97:3959–3964. 61. Song L, et al. (1996) Structure of staphylococcal α-hemolysin, a heptameric transmem- brane pore. Science 274:1859–1865. 8170 ∣ www.pnas.org/cgi/doi/10.1073/pnas.0914229107 Banerjee et al.
3M3W
Crystal structure of mouse PACSIN3 BAR domain mutant
Rigidity of Wedge Loop in PACSIN 3 Protein Is a Key Factor in Dictating Diameters of Tubules*□ S Received for publication,March 4, 2012, and in revised form, April 26, 2012 Published, JBC Papers in Press,May 9, 2012, DOI 10.1074/jbc.M112.358960 Xiaoyun Bai‡§1, Geng Meng‡§1, Ming Luo¶, and Xiaofeng Zheng‡§2 From the ‡State Key Laboratory of Protein and Plant Gene Research, §Department of Biochemistry and Molecular Biology, School of Life Sciences, Peking University, Beijing 100871, China and the ¶Department of Microbiology, University of Alabama at Birmingham, Birmingham, Alabama 35294 Background: PACSINs participate in cellular membrane remodeling. Results: PACSIN 3 F-BAR induces different tubules compared with PACSIN 1 and 2. Structures of PACSINs reveal a novel wedge loop-mediated lateral interaction and different packing mode in PACSIN 3. Conclusion: The rigidity of the wedge loop determines the angles between neighboring dimers and further dictates tubule diameters. Significance: The study provides novel insights into F-BAR domain-induced membrane deformation. BAR (Bin/amphiphysin/Rvs) domain-containing proteins participate in cellular membrane remodeling. The F-BAR pro- teins normally generate low curvature tubules. However, in the PACSIN subfamily, the F-BAR domain from PACSIN 1 and 2 can induce both high and low curvature tubules. We found that unlike PACSIN 1 and 2, PACSIN 3 could only induce low curvature tubules. To elucidate the key factors that dictate the tubule curvature, crystal structures of all three PACSIN F-BAR domains were determined. A novel type of lateral interaction mediated by a wedge loop is observed between the F-BAR neighboring dimers. Comparisons of the structures of PACSIN 3 with PACSIN 1 and 2 indicate that the wedge loop of PACSIN 3 is more rigid, which influences the lateral interactions between assembled dimers. We fur- ther identified the residues that affect the rigidity of the loop by mutagenesis and determined the structures of two PAC- SIN 3 wedge loop mutants. Our results suggest that the rigid- ity-mediated conformations of the wedge loop correlate well with the various crystal packing modes and membrane tubu- lations. Thus, the rigidity of the wedge loop is a key factor in dictating tubule diameters. Cellular membrane deformation is important in the process of cargo transportation and cell movement (1–5). Membrane remodeling is induced by the packing of protein oligomers on the negatively charged membrane surface (6–8). The Bin/am- phiphysin/Rvs (BAR)3 domain proteins, including N-BAR (N-terminal amphipathic helix-BAR), EFC/F-BAR (Fes/CIP4 homology-BAR), and IMD/IBAR (inverse-BAR) domain pro- teins, are important for membrane remodeling in vesicle bud- ding, membrane trafficking between intracellular compart- ments, and cell division (6, 9–14). F-BAR domain proteins contain a central -helix bundle that stabilizes the membrane via the positively charged protein surface of homodimer and bends the membrane into tubules with low curvature (8, 15–17). The crystal structures of FBP17 and CIP4 revealed that F-BAR domain proteins form filaments through end-to-end interactions between dimers in the crystal, and the diameters of the induced tubules were proposed to be related to the intrinsic large radial curvature of the F-BAR dimer (16). Cryo-EM stud- ies of CIP4 and FBP17 showed that the F-BAR domain-induced membrane tubulation is caused by the packing of the protein helical lattice on the membrane surface via lateral and tip-to-tip interactions of the F-BAR dimers (18). PACSINs (also named syndapin) constitute a branch of the F-BAR domain protein family, which are cytoplasmic proteins involved in receptor-mediated endocytosis, synaptic vesicle trafficking, and biogenesis of different cellular organelles (11, 19–26). PACSINs contain the N-terminal F-BAR domain and a C-terminal mono-Src homology 3 domain. The PACSIN family contains PACSIN 1, 2, and 3, which differ in their tissue distri- butions. PACSIN 1 is detected primarily in neuron cells, and PACSIN 3 is found in lung and muscle tissues, whereas PACSIN 2 is expressed ubiquitously (21, 27, 28). The F-BAR-mediated membrane deformation of PACSIN 1 is autoinhibited by its Src homology 3 domain (29). Interactions of PACSINs with other proteins such as dynamin and the neutal Wiskott-Aldrich syn- drome protein also play important roles in membrane remod- eling activities (30). In contrast to the typical F-BAR domain proteins that mainly generate low curvature tubules (100 nm in diameter), PACSIN * This work was supported by National High Technology and Development Program of China Grant 2010CB911804, National Science Foundation of China Grant 30930020, and International Centre for Genetic Engineering and Biotechnology Project CRP/CHN09-01. □ S This article contains supplemental Figs. S1–S7. The atomic coordinates and structure factors (codes 3Q84, 3Q0K, 3QE6, 3M3W, and 3SYV) have been deposited in the Protein Data Bank, Research Collabora- tory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). 1 Both authors contributed equally to this work. 2 To whom correspondence should be addressed: School of Life Sciences, Peking University, Beijing, 100871 China. Tel.: 86-10-6275-5712; Fax: 86- 10-6276-5913; E-mail: xiaofengz@pku.edu.cn. 3 The abbreviations used are: BAR protein, Bin/amphiphysin/Rvs domain pro- tein; PACSIN, protein kinase C and casein kinase substrate in neurons pro- tein; PDB, Protein Data Bank. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 287, NO. 26, pp. 22387–22396, June 22, 2012 © 2012 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A. JUNE 22, 2012•VOLUME 287•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 22387 1 and 2 can induce a wide range radius of membrane tubules including not only the typical 100-nm tubule but also smaller tubules with an average diameter of 10 nm. In addition, they also facilitate tubule constrictions (31). Both crystal structures of the F-BAR domain of PACSIN 1 and 2 reveal a unique wedge loop, which was proposed to contribute to membrane insertion and binding (31). A hinge in the distal end of the murine PACSIN 2 F-BAR domain was predicted to contribute to membrane cur- vature sensing (32). However, the detailed mechanism of how the F-BAR dimers of PACSIN induce various curvature tubu- lations remains unclear. To elucidate further the mechanism of membrane remodel- ing by the F-BAR domain of PACSINs, we investigated the membrane deformation by the PACSIN 3 F-BAR domain com- pared with PACSIN 1 and PACSIN 2 F-BAR domains, by neg- ative-stain electron microscopy. We further determined the structures of the F-BAR domains from PACSIN 3, PACSIN 1, PACSIN 2, and two PACSIN 3 loop mutants, E128A and P121Q. We found that different from PACSIN 1 and 2, PACSIN 3 could only induce 100-nm tubules. Based on structure stud- ies, we discovered that the more rigid PACSIN 3 wedge loop, which is in a different structural conformation compared with PACSIN 1 and 2, contributes to the distinct lattice formed by the PACSIN 3 F-BAR domain in the crystal. The parallel lattice of PACSIN 3 reveals a novel type of wedge loop-mediated lat- eral interaction between neighboring dimers, and the lattice is related to protein packing on the tubulated membrane. We fur- ther identified the residues that determine the rigidity of the wedge loop in the PACSIN 3 F-BAR domain. The mutations of these residues change the structural conformation of the wedge loop, which alters the wedge loop-mediated lateral interactions between neighboring dimers and further influences the mem- brane curvature. Therefore, the rigidity of the wedge loop in PACSIN 3 is a key factor in dictating the diameters of tubules. EXPERIMENTAL PROCEDURES Protein Expression and Purification—The F-BAR domain of PACSIN 1 (residues 1–344), PACSIN 1 mutants, the F-BAR domain of PACSIN 2 (residues 1–372) and PACSIN 2 mutant E130A, the F-BAR domain of mouse PACSIN 3 (residues 1–341, which has 94% amino acid sequence identity to human PACSIN 3) and PACSIN 3 mutants, were cloned into pET28a vector. All proteins except PACSIN 1 truncate 1–344 were expressed in Escherichia coli BL21 (DE3) cells and purified on a Ni2-HiTrap affinity column followed by a Superdex-75 col- umn (GE Healthcare) (33). PACSIN 1 truncate 1–344 was expressed in B834 cells which were cultured in M9 minimal medium supplemented with 50 mg/liter kanamycin and 40 mg/liter selenomethionine, and purified as described (33). Protein Crystallization and Structural Determination—Ini- tial crystallization conditions were screened using kits from Hampton Research including PEG, Crystal Screen, Crystal Screen 2, and Index. Crystals were optimized by the hanging- drop vapor diffusion method. Drops were prepared by mixing 2 l of protein solution (8 mg/ml protein in 500 mM NaCl, 10 mM HEPES, pH 7.5) with 2 l of reservoir solution and were equil- ibrated against 500 l of reservoir solution at 293 K. A large crystal of PACSIN 1 F-BAR (600  190  200 m) was obtained within 2 weeks from 200 mM NH4H2PO4, 100 mM HEPES, pH 7.5, 16% PEG 3350 (w/v), and 5% glycerol. The crystal of PACSIN 2 F-BAR was obtained from 100 mM MgCl2, 100 mM sodium cacodylate, pH 6.5, 18% PEG 3350 (w/v) within 2 days. The crystal of PACSIN 3 F-BAR was obtained from 200 mM KSCN, 100 mM HEPES, pH 7.3, 100 mM CaCl2, 20% PEG 3350 (w/v) within 3 days. The crystal of PACSIN 3 E128A was obtained from 200 mM CH3COONH4, 100 mM CaCl2, 16% PEG 3350 (w/v), and PACSIN 3 P121Q from 0.5 M ammonium sul- fate, 0.1 M sodium citrate tribasic dehydrate, pH 5.6, 1.0 M lith- ium sulfate monohydrate within 10 days. 15, 15, 25, 20, and 28% glycerol was used as cryoprotectant for PACSIN 1, PACSIN 2, PACSIN 3, PACSIN 3 E128A, and PACSIN 3 P121Q. For crystal data collection, crystals were flash-cooled with a nitrogen stream at 100 K, and x-ray diffraction data were col- lected on Mar 345 image plate detector at SSRF. Data of PACSIN 1(1–344) were processed with HKL2000 and MLPHARE soft- ware (34). Images of the F-BAR domains of PACSIN 1, PACSIN 2, PACSIN 3, PACSIN 3 mutant E128A, and PACSIN 3 P121Q were integrated with Mosflm (35), and data were carried out by Molecular Replacement by CCP4 (34) with PACSIN 1–344 as the model, and finally refined by COOT (36). Liposome Preparation—Lipids containing 80% DOPC and 20% DOPA (Avanti) were mixed and dissolved in chloroform. The organic solvent was removed by evaporation under a stream of nitrogen gas, followed by incubation for 2 h in a vac- uum to ensure complete removal of solvent. Lipid films were resuspended in HEPES buffer (10 mM HEPES, pH 7.4, 50 mM NaCl, 0.2 mM EDTA) and subjected to 10 freeze-thaw cycles. Large unilamellar vesicles were then formed by extrusion through 100-nm nucleopore polycarbonate membranes. The prepared liposomes were stored at 4 °C. Liposome Sedimentation Assay—Protein (1 mg/ml) was incubated with liposome (1 mg/ml) with 1:1 protein-lipid vol- ume ratio for 20 min at room temperature. The protein-lipid mixture was centrifuged at 140,000  g for 30 min at 4 °C in an ultracentrifuge (Beckman TLA100 rotor). Supernatants were then collected, and pellets were resuspended in 40 l of sample buffer. Proteins in both fractions were subjected to SDS-PAGE, stained with Coomassie Blue, and visualized by a Bio-Rad XRS system. Tubulation Assays—Protein (1 mg/ml) was incubated with liposome (1 mg/ml) with 1:1 protein-lipid volume ratio for 5 min at room temperature. 6-l protein-liposome samples were then spread onto freshly glow-discharged Formvar- and car- bon-coated electron microscopy grids, stained with 2% uranyl acetate for 1 min, and air dried at room temperature. The grid was examined on a transmission electron microscope (FEI 200 kV) with the electron energy set to 120 kV. Nanogold Labeling Tubulation Assays—Protein (1 mg/ml) was incubated with liposome (1 mg/ml) with 1:1 protein-lipid volume ratio for 5 min at room temperature. 6-l protein-lipo- some samples were then spread onto freshly glow-discharged Formvar- and carbon-coated electron microscopy grids and stained with 5 nm Ni-nitrilotriacetic acid-nanogold (Nano- probes) for 1 min at room temperature followed by staining with 2% uranyl acetate for 1 min and air dried at room temper- Key Factor in Dictating Diameters of Tubules 22388 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287•NUMBER 26•JUNE 22, 2012 ature. The grid was examined on a transmission electron microscope (FEI 200 kV) with electron energy set to 120 kV. RESULTS Membrane Tubulation Induced by F-BAR Domain of PAC- SIN 3 Is Different from PACSIN 1 and 2—To investigate the membrane deformation induced by PACSINs, the F-BAR domains from PACSIN 1, 2, and PACSIN 3 were expressed in E. coli and purified to near homogeneity. Liposome tubulation induced by F-BAR domains was examined by negative stain electron microscopy. Unlike other F-BAR domain-containing proteins, tubules induced by PACSIN 1 and 2 have various diameters, which is consistent with previous reports (31). These tubules are mainly classified into three classes: the first class has low curvature with a diameter of 100 nm, the second class has intermediate curvature with a diameter of 50 nm, and the third class has high curvature with a diameter of 10 nm (Fig. 1A). In addition, tubules with other diameters were also observed. In contrast, the PACSIN 3 F-BAR domain could only induce low curvature tubules of 100-nm diameter (Fig. 1A), despite sharing 55 and 60% sequence identity with the F-BAR domain of PACSIN 1 and 2, respectively. Crystal Structure of PACSIN 3 F-BAR Domain Shows Novel Type of Lateral Interaction Mediated by Wedge Loop—To elu- cidate the molecular mechanism to explain the differences in membrane tubulation between PACSIN 3 and PACSIN 1, 2, we crystallized and determined the structures of the F-BAR domain of PACSIN 3 (residues 1–304), as well as that of PAC- SIN 1 (residues 14–308) and PACSIN 2 (residues 1–304) (sup- plemental Fig. 1 and Table 1). The structure of the PACSIN 1 F-BAR domain was determined by single anomalous dispersion phasing, using data collected with a crystal of selenomethio- nine-derivatized protein, and the structures of PACSIN 2 and PACSIN 3 F-BAR domains were solved by molecular replace- ment using the initial structure of the PACSIN 1 F-BAR domain as the searching model. The structures of PACSIN 1, PACSIN 2, and PACSIN 3 F-BAR domains were refined to resolutions of 2.9, 2.7, and 2.6 Å in the space groups C2, P21, and P21, respec- tively (Table 1). The structures showed that PACSINs form a dimer in the crystal and adopt a crescent shape with a six-helix bundle core and two three-helix bundle arms extending from the central body. Each monomer is composed of six -helices that interact with the other monomer in an antiparallel man- ner. Two neighboring helices 1 form the main central dimerization interface that are flanked by two wings composed of helix 3 and the curved region of helix 4. Helix 3 and a longer helix 4 both extend the length of the monomer with bending ends. Helix 2 and the shorter helices 5 and 6 gen- erate the bundle center that is associated with helix 1 (Fig. 1B FIGURE 1. PACSIN-induced liposome tubulation. A, negative stain EM images of liposomes and tubules induced by PACSIN 3 and PACSIN 1, two F-BAR domains, and quantification of liposome tubulation in the same experiment. Liposomes were incubated with the purified F-BAR domain of PACSIN 1 (residues 1–344), PACSIN 2 (residues 1–372), and PACSIN 3 (residues 1–341). Insets show the higher magnifications of the tubules indicated by arrows. Scale bars, 50, 100, and 500 nm, respectively. B, ribbon presentation of the F-BAR dimers of PASCIN 3. The two monomers interact with each other in an antiparallel manner. Each monomer consists of six -helices named 1–6. One monomer is tinted in a different color for the six -helices; the other is colored in green. C, superimposition of the F-BAR domain between PACSIN 3 (Protein Data Bank (PDB) code 3QE6) and FBP17 (PDB code 2EFL). Ribbon representation is shown. PACSIN 3 and FBP17 are colored in cyan and yellow, respectively. The arrow indicates the unique wedge loop in the PACSIN family proteins that is not found in any other F-BAR domain structure. Key Factor in Dictating Diameters of Tubules JUNE 22, 2012•VOLUME 287•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 22389 and supplemental Fig. 1A). The structure-based sequence alignment of the residues on the protein surface of the F-BAR domains of PACSINs showed that the positively charged resi- dues on the concave surface are highly conserved (supplemen- tal Fig. 1, B and C). Consistent with the previously determined structures of the F-BAR domain of PACSIN 1 and 2 (31), the wedge loop between helices 2 and 3 is also present in PAC- SIN 3, which is a distinct feature of the F-BAR domain of PAC- SINs and has not been observed in other BAR domain-contain- ing families (Fig. 1C). Despite the structure similarity between the F-BAR dimer of PACSIN 3 and those of PACSIN 1 and 2, we found that the packing patterns of the neighboring dimers of these proteins are strikingly different in the crystal: PACSIN 3 shows only a parallel packing among all the different crystallization condi- tions that have been examined, whereas PACSIN 1 and 2 show different packing modes (supplemental Fig. 2). We also found that the parallel packing of the PACSIN 3 F-BAR domain is similar to the packing lattice of the CIP4 F-BAR domain observed on the tubules by cryo-EM (18). How- ever, a close-up view of the PACSIN 3 structure reveals a dis- tinct type of lateral interaction between two neighboring dimers (Fig. 2A), in addition to the typical CIP4-like tip-to-tip interaction. In CIP4, the lateral interactions between the neigh- boring dimers involve only the residues of neighboring helices, whereas in PACSIN 3, the lateral interactions are between one of the PACSIN-specific wedge loops with the helix 2 of the neighboring dimer (Fig. 2, A, left, and B, upper). The other wedge loop of the same dimer extends out into the membrane direction (Fig. 2, A, right, and B, lower). This observation sug- gests that the wedge loop might function in two ways: one loop in the dimer is involved in the lateral interaction, and the other loop in the dimer extends out and inserts into the membrane bilayer. Dual Roles of Wedge Loop in Mediating Protein Packing and Membrane Binding—To confirm that the wedge loop is involved in the interactions between the neighboring dimers, we mutated the residues located in the lateral contacting area (Fig. 2C, left). A total of 15 mutants of the five residues on the contacting surface were constructed. The mutants were puri- fied and examined by a liposome tubulation assay (Fig. 3A). The binding of these mutants to liposomes was investigated by a sedimentation assay (Fig. 3B). Mutations in the wedge loop (H119E/A/Q, R127E/A/Q) or in the contacting area (E93R/ A/Q, E97R/A/Q and E100R/A/Q) abolished the liposome tubu- lation activity completely, but retained the liposome binding ability (Fig. 3 and supplemental Fig. 3B). This observation indi- cates that the interaction between the wedge loop and its neigh- boring dimer plays critical roles in protein packing on the mem- brane to induce liposome tubulation. We further examined the residues on the wedge loop that may insert into the membrane bilayer. A series of mutants of the wedge loop, V122L123, V122E/L123E, and V122R/L123R of PACSIN 3 (Fig. 2C, right) were constructed in which the two bulky hydrophobic residues at the base of the wedge loop were either deleted or mutated. The liposome tubulation activities and the binding of these mutants to liposomes were also inves- tigated. Our results showed that the deletion of the wedge loop (V122L123) or mutation of residues Val-122 and Leu-123 at the center of the wedge loop to Glu or Arg (V122E/L123E, V122R/L123R) totally abolished induction of tubules and reduced liposome binding (Fig. 3B), which is consistent with the observation in the site mutation constructs I125E or M126E previously reported in PACSIN 1 (31). These results indicate that the wedge loop binds liposomes through hydro- phobic residues Val-122 and Leu-123 (or Ile-125/Met-126 in PACSIN 1, and Met-124/Met-125 in PACSIN 2, supplemen- tal Fig. 3B). TABLE 1 Data collection and refinement statistics Se-PACSIN (residues 1–307) PACSIN 2 (residues 1–304) PACSIN 3 (residues 1–304) PACSIN3E128A (residues 1–302) PACSIN3P121Q (residues 6–299) Diffraction data Space group C2 P21 P21 P21 P21 Unit cell parameters (a, b, c) (Å) 85.3, 154.3, 215.4 31.58, 86.13, 353.80 46.9, 54.7, 193.7 47.5, 52.3, 196.5 120.570, 108.901, 222.319  90, 90.3, 90 90, 90, 90 90, 96.9, 90 90, 94.8, 90 90.00, 90.05, 90.00 Resolution range (Å) 30–2.8 (2.91–2.8) 30–2.6 (2.75–2.6) 30–2.6 (2.67–2.6) 15–2.6 (2.67–2.6) 50–3.1 (3.18–3.1) Unique reflections 63,353 41,646 28,175 28,283 98,202 Completeness (%) 96.2 (87.43) 99.7 (85.9) 97.8 (90.8) 99.8 (93.7) 98.68 (98.02) Mean I/(I) 16.65 (2.5) 14.5 (2.5) 35.4 (3.4) 14.2 (2.4) 18.3 (3.1) Multiplicity 4.3 3.1 4.1 4 4.3 Refinement Rwork/Rfree (%)a 22.1/29.4 20.77/29.2 22.7/27.4 23.2/28.1 27/33.6 Modeled chain: residues A Non-H atoms (protein/water) 13,874/213 9,683//202 4,648/34 4,587/10 16,535/197 Root mean square deviations Bonds (Å)/Angles (°) 0.017/1.692 0.015/1.593 0.022/2.041 0.011/1.297 0.0213/2.009 Ramachandran analysis Favored 98.61 98.14 93.94 96.22 94.01 Allowed 1.33 1.68 5.81 3.78 5.14 Generously allowed 0.06 0.18 0.24 0 0.86 Disallowed 0 0 0 0 0 a R   Fo  Fc /Fo. The full-length PACSINs proteins were first crystallized but degraded from the C-terminal Src homology 3 domain. Then the truncation of Se- PACSIN (residues 1–305), PACSIN 2 (residues 1–305), and PACSIN 3 (residues 1–302) was designed and crystallized. Key Factor in Dictating Diameters of Tubules 22390 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287•NUMBER 26•JUNE 22, 2012 There is a possibility that the mutant proteins may be dis- torted in structure, thus influencing the membrane tubulation property. We therefore monitored the secondary structure of all of the mutant proteins by a circular dichroism assay. Our results showed that the secondary structures of these mutants are similar to the wild-type protein (supplemental Fig. 3), indi- cating that the lost of the tubulation activity is caused by the elimination of protein-protein interaction, not by destroying the structure of the protein. To characterize further where the PACSIN proteins localize in the tubulation process, 5-nm Ni2-nitrilotriacetic acid- linked nanogold particles were used to label the His-tagged PACSIN F-BAR proteins in tubulation assays (Fig. 3C). The results showed that the nanogold particles were concentrated on the surface of the tubules. This observation indicates that tubules are induced by the PACSIN F-BAR proteins. Rigidity of Wedge Loop Influences Lateral Interactions and Dictates Diameter of Tubules—Superimposition of the wedge loops among PACSIN 3 and PACSIN 1 and 2 revealed that the wedge loop of PACSIN 3 has a different conformation. The wedge loop of PACSIN 3 is wider and bends more toward the inner side of the dimer compared with the relatively flat, outward-splayed loops in PACSIN 1 and 2 (Fig. 4A, left). We therefore hypothesized that the conformational difference of the wedge loops would be an important factor in determination of the interaction pattern between neighboring dimers and would con- tribute further to the variations in membrane tubulation. Structure-based sequence alignments of the wedge loop region in PACSINs revealed that the predominant difference between PACSIN 3 and PACSIN 1 or 2 is residue Pro-121 located in the center of the wedge loop of PACSIN 3, which superimposes onto residue Gln-124 and Gln-123 in PACSIN 1 and 2, respectively (Fig. 4, B and C). The rigidity of the proline residue may restrain the flexibility of the wedge loop in PACSIN 3 and affect the interaction pattern between neighboring dimers and membrane tubulation. To verify the above predictions, we mutated Pro-121 of PACSIN 3 to glutamine, which is found in PACSIN 1 and 2, and substituted the equivalent glutamine of PACSIN 1 and 2 with proline. The effects of the mutant proteins on liposome tubu- lation were investigated by negative stain EM. Consistent with the hypothesis, replacement of proline with glutamine at posi- tion 121 (P121Q) in PACSIN 3 led to the formation of a large number of pseudopod-like small tubules, in addition to the large tubules (Fig. 5A). These pseudopod-like tubules were approximately 10 nm in diameter, and some are budded from the surface of the large tubules. At the same time, replacement of Gln with Pro in PACSIN 1 Q124P and PACSIN 2 Q123P induced only one type of large tubules that are 100-nm in diameter (Fig. 5B). Moreover, we determined the crystal struc- ture of the PACSIN 3 P121Q mutant to 2.9 Å resolution (sup- plemental Fig. 4 and Table 1). The structure revealed that the mutation of proline to glutamine changed the orientation of the wedge loop of PACSIN 3, and the wedge loop in the mutant structure points outward and superimposes very well with the wedge loop of native PACSIN 1 and 2 (Fig. 4A, middle). In addition, the crystal packing of PACSIN 3 P121Q showed an obvious change of the angle between the neighboring dimers, which resembles that found in PACSIN 1 and 2 packing lattices (supplemental Fig. 2). These results demonstrate that the rigid- ity of the wedge loop in the PACSINs F-BAR domain correlates with the conformational change of the wedge loop and further FIGURE 2. Wedge loop plays dual roles in PACSIN 3 packing. A, novel type of loop-mediated lateral interaction found in PACSIN 3 F-BAR packing. The PACSIN 3 F-BAR domain packs in a parallel manner which connected by tip-to-tip interactions and the loop-mediated lateral interactions. B, close-up view of the two wedgeloopsinthesamedimer.Oneofthewedgeloopsinteractswithitsadjacentdimer(upper).Theotherwedgeloopisexposedandextrudesintothetarget membrane direction (lower). The right view is rotated by 90° clockwise relative to the left view. The two adjacent dimers are colored marine and yellow, and presented as ribbon and surface, respectively. C, left image showing residues on the contacting surface between the wedge loop and the neighboring dimer. Indicated residues are shown as side chain sticks. Right image shows the membrane binding wedge loop from residue His-119 to Glu-128 and is presented as side chain sticks. Key Factor in Dictating Diameters of Tubules JUNE 22, 2012•VOLUME 287•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 22391 Key Factor in Dictating Diameters of Tubules 22392 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287•NUMBER 26•JUNE 22, 2012 affects lateral interactions of neighboring dimers and mem- brane tubulation. To investigate whether mutation of Pro-121 in PACSIN 3 to glutamine affects the specific interactions between the wedge loop and membrane, a sedimentation assay was performed. The result showed that PACSIN 3 P121Q did not change the bind- ing of protein to lipid bilayers (supplemental Fig. 3B). Con- versely, the mutations of PACSIN3 Val-122 and Leu-123 to charged residues, or deletion of them, resulted in the complete loss of the membrane binding ability even though Pro-121 is still intact (Fig. 3). These results indicate that Pro-121 is not involved in membrane insertion. Therefore, the reason that P121Q can induce the large type of tubules is not due to the change of the wedge loop insertion. Structure-based sequence alignments also suggest that there is another residue, Glu-128 in PACSIN 3 (Glu-131 in PACSIN 1 FIGURE 3. Tubulation and sedimentation assays of PACSIN 3 F-BAR and mutants. A, negative-stained electron micrographs are shown. Liposomes were incubated with PACSIN 3 and its mutant proteins and examined by EM. V122L123, V122E/L123E, V122R/L123R altered only membrane binding, whereas R127E, H119E altered only the interactions between neighboring dimers. E93R, E97R, and E100R are mutations on the contacting surface of neighboring dimer as shown in Fig. 2C, left. B, liposomes were incubated with PACSIN 3 and its mutant proteins and examined by sedimentation assays. Supernatant (S) and pellet (P) fractions were analyzed by SDS-PAGE. Band in pellet represents the protein bound to liposome. Histograms with means  S.E. (error bars) show the quantified protein in supernatant and pellet with the total protein defined as 100%. C, liposomes were incubated with the His-tagged F-BAR proteins of PACSINs followed by Ni-nitrilotriacetic acid-nanogold particles (5-nm diameter) treatment and then visualized by EM. FIGURE 4. Comparisons of the wedge loop among PACSINs. A, structural alignments of the wedge loop among PACSIN 1, 2, 3 and PACSIN 3 mutants P121Q and E128A. The wedge loop in PACSIN 3 points inward by 45° compared with PACSIN 2 and points inward by 30° compared with PACSIN 1 (left). Compared with the wedge loop in the wild-type PACSIN 3 F-BAR domain, the wedge loop in the PACSIN 3 P121Q mutant swings outward and is superimposable with the wedge loop of PACSIN 1 and 2 (middle); and the wedge loop in the PACSIN3 E128A mutant reoriented significantly by swinging 30° outward to resemble the outward-splayed wedge loop in PACSIN 1 and PACSIN 2 (right). B, close-up view of the wedge loops of PACSIN 1, PACSIN 2, PACSIN 3, and PACSIN 3 E128A mutant. All of the native structures of the PACSIN F-BAR domains include a conserved hydrogen bond between Glu and His at the base of the wedge loop (Glu-131 and His-122 for PACSIN 1, Glu-130 and His-121 for PACSIN 2, and Glu-128 and His-119 for PACSIN 3), whereas this hydrogen bond is absent in the PACSIN 3 E128A mutant. C, sequence alignments of the wedge loop among PACSINs from human and mouse using ClustalX. Instead of the conserved residue Gln in PASCIN 1 and 2, it is a conformationally restrained Pro residue at position 121 in PACSIN 3. Asterisks and colons indicate homology sequence among PACSIN 1 PACSIN 2 and PACSIN 3. Key Factor in Dictating Diameters of Tubules JUNE 22, 2012•VOLUME 287•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 22393 and Glu-130 in PACSIN 2) that forms a hydrogen bond with the mainchainofHis-119(His-122inPACSIN1andHis-121inPAC- SIN2)(Fig.4B),mayalsobeimportantfortheconformationofthe wedge loop. The crystal structure of the PACSIN 3 E128A mutant determined here to 2.8 Å resolution (supplemental Fig. 4 and Table 1) showed that substitution of Glu-128 with alanine elimi- nated the hydrogen bond between Glu-128 and His-119 (Fig. 4B) andresultedina significant conformational change in the wedge loop. The E128A wedge loop of PACSIN 3 was found to resem- ble the wedge loop of PACSIN 1 and 2 in an outward-splayed pattern (Fig. 4A, right). Structural comparisons of the PACSIN 3 E128A loop with that of the wild type suggest that the hydro- gen bond between Glu-128 and His-119 could function as a lock to maintain the loop conformation. The PACSIN 3 E128A was found to induce predominantly tubules with a diameter of 50 nm in addition to the typical 100-nm tubules induced by wild- type PACSIN3 (Fig. 5C), which resemble the intermediate and low curvature tubules induced by PACSIN 1 and 2. Similarly, mutants E131A and E130A of PACSIN 1 and 2 also showed more membrane constrictions budding from the tubules (sup- plemental Fig. 5) and more high curvature tubules compared with the wild-type PACSIN 1 and 2. These data suggest that, in PACSINs, the hydrogen bond between Glu-128 (Glu-131 or Glu-130) and His-119 (His-122 or His-121) helps to reinforce the rigidity and orientation of the wedge loop and therefore affects the curvature of the tubules. DISCUSSION Membrane curving by BAR domain proteins is an important biological process involved in vesicle trafficking and subcellular structure stabilization. Various tubules induced by the F-BAR domain of PACSINs are important in membrane deformation (20–25). It was suggested that the distinct wedge loop of PACSIN 1 and 2 F-BAR domains inserts into lipid bilayers of membrane (31). However, there is no clear interpretation to explain why the F-BAR domains of PACSIN 1 and 2 induce tubules with such diverse diameters. Here, we performed structural and biochemical studies on the F-BAR domains from all three PACSINs and systematically compared the differences of the F-BAR domains between PAC- SIN 3 and 1, or 2 with respect to structure and membrane tubu- lation activities. Our analyses revealed that the more rigid wedge loop in the F-BAR domain of PACSIN 3 is responsible for its unique feature of inducing only low curvature tubules compared FIGURE 5. Key residues in determination of diameter of liposome tubules. Liposomes were incubated with different mutant proteins and examined by EM. The numbers of the tubules were quantified on five independently prepared EM grids. Membrane constrictions are denoted by arrows. A, mutation in the wedge loop of PACSIN 3 F-BAR P121Q has a profound influence on the diameters of the induced liposome tubules. B, substitution of residue Gln to Pro in the wedge loop of PACSIN 1 and 2 only results in low curvature tubules as shown in PACSIN 3 F-BAR in Fig. 1C. C, mutation in the wedge loop of PACSIN 3 E128A induces predominantly tubules with a diameter of 50 nm, in addition to the typical 100-nm tubules as induced by PACSIN 3 F-BAR. Key Factor in Dictating Diameters of Tubules 22394 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287•NUMBER 26•JUNE 22, 2012 withPACSIN1and2.WhenthewedgeloopintheF-BARdomain of PACSIN 1 and 2 was mutated to the more rigid structure as found in PACSIN 3, these mutants induced only low curvature tubules as the wild-type PACSIN 3. Conversely, mutants that modified the PACSIN 3 wedge loop to resemble the wedge loop in PACSIN 1 and 2 induced tubules with dimensions similar to those found in the wild-type PACSIN 1 and 2. The distal end of murine PACSIN 2 was proposed to sense the membrane (32). Comparisons of the distal ends of wild-type PAC- SIN 3 with PACSIN 1 and 2 show no obvious difference in all the structures we determined (supplemental Fig. 6), suggesting that the various phenotypes presented by PACSIN 1 and 2 F-BAR may not be correlated with the intrinsic curvature of the dimer surface. It was previously reported that during tubulation, F-BAR proteins make tip-to-tip interactions and contacts between lat- erally adjacent dimers (18). The diameters of the induced tubules were proposed to be related to the intrinsic large radial curvature of the F-BAR domain, such as CIP4 and FBP17 (16). However, in PACSINs, the intrinsic curvature itself certainly could not result in such diverse tubules with various diameters because PACSIN 1 and 2 induce tubules different from those induced by PACSIN 3 even though all three PACSINs almost have the same intrinsic curvature. According to the crystal packing pattern of PACSIN 3, we propose a potential mem- brane tubulation model for PACSINs (supplemental Fig. 7): the F-BAR domain is connected by tip-to-tip interactions and the wedge loop-mediated lateral interactions to form filaments. The filament of the F-BAR protein bends as a hinge motion at the tip-to-tip interaction to various extents. The extent of the motion is dependent on the angle between two adjacent dimers, and this angle is determined by the conformation of the wedge loop. This results in the filament winding spirally around a cylindrical membrane. The larger the angle is between the two dimers, the more the dimer bends the membrane; the more the membrane is bent, and the smaller the diameter of the tubule is (supplemental Fig. 7E). This model also shed light on under- standing why PACSINs generate different size tubules in so many biological processes, such as trans-Golgi network vesicle formation, filopodia tips, and lamellipodia dynamics, micro- spike formation, and caveola fission (22, 24, 37, 38). In conclusion, the membrane tubulation by PACSIN 3 was shown to be different from PACSIN 1 and 2, and these differ- ences are due to different degrees of rigidity of the wedge loop. We demonstrated that the rigidity of the wedge loop in the PACSIN F-BAR domain is a key factor that determines the dif- ferent angles between two neighboring dimers and dictates the diameters of various tubules. Our study provides new insights for understanding the mechanism of membrane deformation by the PACSIN family proteins. Acknowledgments—We thank Prof. Fuyu Yang and Dr. Kai Zhao from Institute of Biophysics, Chinese Academy of Sciences, for lipid preparation; Dr. Ning Gao at Tsinghua University for the nanogold labeling assay; and Dr. Plomann at Stanford University School of Medicine for providing the PACSIN 3 plasmid. X-ray diffraction data collection was carried out at the Beijing Synchrotron Radiation Lab- oratory and Shanghai Synchrotron Radiation Facility. REFERENCES 1. Merrifield, C. J., Perrais, D., and Zenisek, D. (2005) Coupling between clathrin-coated pit invagination, cortactin recruitment, and membrane scission observed in live cells. Cell 121, 593–606 2. Ford, M. G., Mills, I. G., Peter, B. J., Vallis, Y., Praefcke, G. J., Evans, P. R., and McMahon, H. T. (2002) Curvature of clathrin-coated pits driven by epsin. Nature 419, 361–366 3. Scott, I. C., and Stainier, D. Y. (2003) Developmental biology: twisting the body into shape. Nature 425, 461–463 4. Dvorak, A. M., and Feng, D. (2001) The vesiculo-vacuolar organelle (VVO): a new endothelial cell permeability organelle. J. Histochem. Cy- tochem. 49, 419–432 5. Rippe, B., Rosengren, B. I., Carlsson, O., and Venturoli, D. (2002) Tran- sendothelial transport: the vesicle controversy. J. Vasc. Res. 39, 375–390 6. Gallop, J. L., Jao, C. C., Kent, H. M., Butler, P. J., Evans, P. R., Langen, R., and McMahon, H. T. (2006) Mechanism of endophilin N-BAR domain- mediated membrane curvature. EMBO J. 25, 2898–2910 7. Peter, B. J., Kent, H. M., Mills, I. G., Vallis, Y., Butler, P. J., Evans, P. R., and McMahon, H. T. (2004) BAR domains as sensors of membrane curvature: the amphiphysin BAR structure. Science 303, 495–499 8. Itoh, T., Erdmann, K. S., Roux, A., Habermann, B., Werner, H., and De Camilli, P. (2005) Dynamin and the actin cytoskeleton cooperatively reg- ulate plasma membrane invagination by BAR and F-BAR proteins. Dev. Cell 9, 791–804 9. Bhatia, V. K., Madsen, K. L., Bolinger, P. Y., Kunding, A., Hedegård, P., Gether, U., and Stamou, D. (2009) Amphipathic motifs in BAR domains are essential for membrane curvature sensing. EMBO J. 28, 3303–3314 10. Masuda, M., Takeda, S., Sone, M., Ohki, T., Mori, H., Kamioka, Y., and Mochizuki, N. (2006) Endophilin BAR domain drives membrane curva- ture by two newly identified structure-based mechanisms. EMBO J. 25, 2889–2897 11. Takei, K., Slepnev, V. I., Haucke, V., and De Camilli, P. (1999) Functional partnership between amphiphysin and dynamin in clathrin-mediated en- docytosis. Nat. Cell Biol. 1, 33–39 12. Farsad, K., Ringstad, N., Takei, K., Floyd, S. R., Rose, K., and De Camilli, P. (2001) Generation of high curvature membranes mediated by direct en- dophilin bilayer interactions. J. Cell Biol. 155, 193–200 13. Mattila, P. K., Pyka¨la¨inen, A., Saarikangas, J., Paavilainen, V. O., Vihinen, H., Jokitalo, E., and Lappalainen, P. (2007) Missing-in-metastasis and IRSp53 deform PI(4,5)P2-rich membranes by an inverse BAR domain-like mechanism. J. Cell Biol. 176, 953–964 14. Pyka¨la¨inen, A., Boczkowska, M., Zhao, H., Saarikangas, J., Rebowski, G., Jansen, M., Hakanen, J., Koskela, E. V., Pera¨nen, J., Vihinen, H., Jokitalo, E., Salminen, M., Ikonen, E., Dominguez, R., and Lappalainen, P. (2011) Pink- bar is an epithelial-specific BAR domain protein that generates planar membrane structures. Nat. Struct. Mol. Biol. 18, 902–907 15. Fu¨tterer, K., and Machesky, L. M. (2007) “Wunder” F-BAR domains: going from pits to vesicles. Cell 129, 655–657 16. Shimada, A., Niwa, H., Tsujita, K., Suetsugu, S., Nitta, K., Hanawa-Suet- sugu, K., Akasaka, R., Nishino, Y., Toyama, M., Chen, L., Liu, Z. J., Wang, B. C., Yamamoto, M., Terada, T., Miyazawa, A., Tanaka, A., Sugano, S., Shirouzu, M., Nagayama, K., Takenawa, T., and Yokoyama, S. (2007) Curved EFC/F-BAR-domain dimers are joined end to end into a filament for membrane invagination in endocytosis. Cell 129, 761–772 17. Henne, W. M., Kent, H. M., Ford, M. G., Hegde, B. G., Daumke, O., Butler, P. J., Mittal, R., Langen, R., Evans, P. R., and McMahon, H. T. (2007) Structure and analysis of FCHo2 F-BAR domain: a dimerizing and mem- brane recruitment module that effects membrane curvature. Structure 15, 839–852 18. Frost, A., Perera, R., Roux, A., Spasov, K., Destaing, O., Egelman, E. H., De Camilli, P., and Unger, V. M. (2008) Structural basis of membrane invag- ination by F-BAR domains. Cell 132, 807–817 19. Damke, H., Baba, T., Warnock, D. E., and Schmid, S. L. (1994) Induction of mutant dynamin specifically blocks endocytic coated vesicle formation. J. Cell Biol. 127, 915–934 20. Takei, Y., Harada, A., Takeda, S., Kobayashi, K., Terada, S., Noda, T., Takahashi, T., and Hirokawa, N. (1995) Synapsin I deficiency results in the Key Factor in Dictating Diameters of Tubules JUNE 22, 2012•VOLUME 287•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 22395 structural change in the presynaptic terminals in the murine nervous sys- tem. J. Cell Biol. 131, 1789–1800 21. Modregger, J., Ritter, B., Witter, B., Paulsson, M., and Plomann, M. (2000) All three PACSIN isoforms bind to endocytic proteins and inhibit endo- cytosis. J. Cell Sci. 113, 4511–4521 22. Qualmann, B., and Kelly, R. B. (2000) Syndapin isoforms participate in receptor-mediated endocytosis and actin organization. J. Cell Biol. 148, 1047–1062 23. Braun, A., Pinyol, R., Dahlhaus, R., Koch, D., Fonarev, P., Grant, B. D., Kessels, M. M., and Qualmann, B. (2005) EHD proteins associate with syndapin I and II and such interactions play a crucial role in endosomal recycling. Mol. Biol. Cell 16, 3642–3658 24. Kessels, M. M., Dong, J., Leibig, W., Westermann, P., and Qualmann, B. (2006) Complexes of syndapin II with dynamin II promote vesicle forma- tion at the trans-Golgi network. J. Cell Sci. 119, 1504–1516 25. Grimm-Gu¨nter, E. M., Milbrandt, M., Merkl, B., Paulsson, M., and Plo- mann, M. (2008) PACSIN proteins bind tubulin and promote microtubule assembly. Exp. Cell Res. 314, 1991–2003 26. Halbach, A., Mo¨rgelin, M., Baumgarten, M., Milbrandt, M., Paulsson, M., and Plomann, M. (2007) PACSIN 1 forms tetramers via its N-terminal F-BAR domain. FEBS J. 274, 773–782 27. Plomann, M., Lange, R., Vopper, G., Cremer, H., Heinlein, U. A., Scheff, S., Baldwin, S. A., Leitges, M., Cramer, M., Paulsson, M., and Barthels, D. (1998) PACSIN, a brain protein that is up-regulated upon differentiation into neuronal cells. Eur. J. Biochem. 256, 201–211 28. Ritter, B., Modregger, J., Paulsson, M., and Plomann, M. (1999) PACSIN 2, a novel member of the PACSIN family of cytoplasmic adapter proteins. FEBS Lett. 454, 356–362 29. Rao, Y., Ma, Q., Vahedi-Faridi, A., Sundborger, A., Pechstein, A., Puchkov, D., Luo, L., Shupliakov, O., Saenger, W., and Haucke, V. (2010) Molecular basis for SH3 domain regulation of F-BAR-mediated membrane deforma- tion. Proc. Natl. Acad. Sci. U.S.A. 107, 8213–8218 30. Qualmann, B., Roos, J., DiGregorio, P. J., and Kelly, R. B. (1999) Syndapin I, a synaptic dynamin-binding protein that associates with the neural Wis- kott-Aldrich syndrome protein. Mol. Biol. Cell 10, 501–513 31. Wang, Q., Navarro, M. V., Peng, G., Molinelli, E., Goh, S. L., Judson, B. L., Rajashankar, K. R., and Sondermann, H. (2009) Molecular mechanism of membrane constriction and tubulation mediated by the F-BAR protein PACSIN/syndapin. Proc. Natl. Acad. Sci. U.S.A. 106, 12700–12705 32. Plomann, M., Wittmann, J. G., and Rudolph, M. G. (2010) A hinge in the distal end of the PACSIN 2 F-BAR domain may contribute to membrane- curvature sensing. J. Mol. Biol. 400, 129–136 33. Bai, X., Meng, G., Li, G., Luo, M., and Zheng, X. (2010) Crystallization and preliminary x-ray crystallographic analysis of human PACSIN 1 protein. Acta Crystallogr. Sect. F Struct. Biol. Cryst. Commun. 66, 73–75 34. Collaborative Computational Project, Number 4 (1994) The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D Biol. Crystallogr. 50, 760–763 35. Powell, H. R. (1999) The Rossmann Fourier autoindexing algorithm in MOSFLM. Acta Crystallogr. D Biol. Crystallogr. 55, 1690–1695 36. Emsley, P., and Cowtan, K. (2004) COOT: model-building tools for mo- lecular graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 37. Senju, Y., Itoh, Y., Takano, K., Hamada, S., and Suetsugu, S. (2011) Essen- tial role of PACSIN2/syndapin-II in caveolae membrane sculpting. J. Cell Sci. 124, 2032–2040 38. Shimada, A., Takano, K., Shirouzu, M., Hanawa-Suetsugu, K., Terada, T., Toyooka, K., Umehara, T., Yamamoto, M., Yokoyama, S., and Suetsugu, S. (2010) Mapping of the basic amino acid residues responsible for tubula- tion and cellular protrusion by the EFC/F-BAR domain of PACSIN2/syn- dapin II. FEBS Lett. 584, 1111–1118 Key Factor in Dictating Diameters of Tubules 22396 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287•NUMBER 26•JUNE 22, 2012
3M3Y
RNA polymerase II elongation complex C
X-ray structure and mechanism of RNA polymerase II stalled at an antineoplastic monofunctional platinum-DNA adduct Dong Wanga,b,1, Guangyu Zhuc, Xuhui Huangd, and Stephen J. Lippardc,1 aDepartment of Structural Biology, Stanford University School of Medicine, Stanford, CA 94305; bSkaggs School of Pharmacy and Pharmaceutical Sciences, University of California, San Diego, La Jolla, CA 92093; cDepartment of Chemistry, Massachusetts Institute of Technology, Cambridge, MA 02139; and dDepartment of Chemistry, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, P.R. China Contributed by Stephen J. Lippard, March 3, 2010 (sent for review February 9, 2010) DNA is a major target of anticancer drugs. The resulting adducts interfere with key cellular processes, such as transcription, to trigger downstream events responsible for drug activity. cis- Diammine(pyridine)chloroplatinum(II), cDPCP or pyriplatin, is a monofunctional platinum(II) analogue of the widely used antican- cer drug cisplatin having significant anticancer properties with a different spectrum of activity. Its novel structure-activity properties hold promise for overcoming drug resistance and improving the spectrum of treatable cancers over those responsive to cisplatin. However, the detailed molecular mechanism by which cells process DNA modified by pyriplatin and related monofunctional complexes is not at all understood. Here we report the structure of a transcri- bing RNA polymerase II (pol II) complex stalled at a site-specific monofunctional pyriplatin-DNA adduct in the active site. The re- sults reveal a molecular mechanism of pol II transcription inhibition and drug action that is dramatically different from transcription in- hibition by cisplatin and UV-induced 1,2-intrastrand cross-links. Our findings provide insight into structure-activity relationships that may apply to the entire family of monofunctional DNA-damaging agents and pave the way for rational improvement of monofunc- tional platinum anticancer drugs. anticancer ∣chemotherapy ∣DNA damage ∣pyriplatin ∣transcription T he DNA template for transcription is not only the site of in- born errors of metabolism and of continuous attack by harm- ful environmental agents, but it also represents a major target for cancer therapy. Platinum-based anticancer drugs such as cisplatin, cis-diamminedichloroplatinum(II), are widely used and among the most effective antineoplastic treatments (1, 2). Platinum-based drugs typically form bifunctional intra- or inter- strand DNA cross-links by covalent bonding to the N7 positions of two guanosine residues, triggering a variety of cellular processes, including transcription inhibition with attendant apoptosis (1, 2). However, resistance and side effects can require with- drawal of these drugs before they can effect a cure in certain types of cancer (3). In the effort to find new compounds that circumvent resis- tance to conventional bifunctional platinum-based drugs, a class of monofunctional platinum compounds were synthesized and screened for anticancer activity (4–6). In contrast to other inactive monofunctional platinum(II) compounds such as ½PtðdienÞClþ and ½PtðNH3Þ3Clþ, cis-diammine(pyridine)chloro- platinum(II) [cDPCP or “pyriplatin” (Fig. 1)] and related com- plexes display significant anticancer properties and a different spectrum of activity compared to conventional platinum-based drugs. These features render them attractive candidates for treat- ing cisplatin-refractory patients if the potency could be raised to or beyond the level of that of cisplatin (4, 5, 7). Pyriplatin exhibits unique chemical and biological properties, forming monofunc- tional DNA adducts (Fig. 1 and Fig. S1) that can inhibit transcrip- tion and better elude DNA repair (7). The x-ray crystal structure of pyriplatin bound to a DNA duplex reveals substantially different features than those of DNA adducts formed by conven- tional, bifunctional platinum-based drugs. The overall DNA duplex is much less distorted, with the pyridine ligand of the cis-fPtðNH3Þ2ðpyÞg2þ moiety directed toward the 50-end of the platinated strand. A hydrogen bond forms between the NH3 ligand trans to pyridine and O6 of the platinated guanosine residue (7). The detailed molecular mechanism by which cells process DNA modified by monofunctional complexes such as pyriplatin is not understood. Several important questions remain unan- swered. By what process do monofunctional adducts block pol II transcription? Does the mechanism differ from that of tran- scription inhibition by 1,2- and 1,3-intrastrand cross-links that comprise the major adducts of cisplatin? Why do pyriplatin and its homologues, which violate the classical structure-activity relationships (SARs) for active, bifunctional platinum drugs (8), show such promise by comparison to related monofunctional complexes like ½PtðNH3Þ3Clþ? Would knowledge of the struc- ture of pyriplatin-modified DNA at its site(s) of biological action inform the design of more potent analogues? In the present work we take a combined biochemical and x-ray structural approach to investigate the molecular mechanism of pol II transcription inhibition by a site-specific monofunctional platinum(II)-DNA adduct of pyriplatin. An unprecedented mo- lecular mechanism for pol II transcription inhibition is revealed, providing insight into structure-activity relationships that may ap- ply to the entire family of monofunctional DNA-damaging agents, whether or not they contain platinum. Results A Different Configuration of a Pyriplatin-DNA Adduct Accommodated in the Pol II Active Site. To understand how a monofunctional pyriplatin-DNA adduct is accommodated in the active site of the transcribing pol II elongation complex, we designed and pre- pared a DNA template containing a site-specific DNA lesion of this complex, as described previously (7). A transcribing pol II complex was then assembled in which the pyriplatin-DNA lesion occupies the active (þ1) site (Complex B, Table 1). The crystal structure of this complex reveals that the platinated nucleotide is captured as a pol II complex in the post-translocation state, in which the addition site is empty and ready for NTP loading (Dashed Ring, Fig. 2A and Fig. S2). Fig. 2A reveals that the Author contributions: D.W. and S.J.L. designed research; D.W., G.Z., and X.H. performed research; D.W., G.Z., X.H., and S.J.L. analyzed data; and D.W., X.H., and S.J.L. wrote the paper. The authors declare no conflict of interest. Data deposition: The atomic coordinates have been deposited in the Protein Data Bank, www.pdb.org (PDB ID codes 3M4O and 3M3Y). 1To whom correspondence may be addressed. E-mail: dongwang@ucsd.edu or lippard@ mit.edu. This article contains supporting information online at www.pnas.org/cgi/content/full/ 1002565107/DCSupplemental. 9584–9589 ∣PNAS ∣May 25, 2010 ∣vol. 107 ∣no. 21 www.pnas.org/cgi/doi/10.1073/pnas.1002565107 positioning of the pyriplatin-damaged guanosine residue is lo- cated above the bridge helix. This structure requires rotation of the cis-fPtðNH3Þ2ðpyÞg2þ moiety and its bound guanosine re- sidue into a different configuration compared to that adopted in the pyriplatin-duplex DNA structure, in order to avoid a steric clash with bridge helix (7). Fig. 2B depicts this comparison. The rotation is energetically facilitated by the formation of hydro- gen bonds between the ammine ligands on platinum with the phosphodiester moiety of the backbone between positions þ1 and þ2, with concomitant loss of a hydrogen bond between O6 of the platinated guanosine residue and an ammine ligand. An additional feature is that the pyridine group of the cis- fPtðNH3Þ2ðpyÞg2þ fragment, which points downstream toward the 50-direction of the template DNA, forms van der Waals inter- actions with bridge helix residues Val 829 and Ala 832. The purine base of the guanosine residue at position þ1 is displaced toward the major groove of the RNA–DNA duplex by comparison with structures having an undamaged base at this site in the post-trans- location state (9–11). Transcription Elongation Inhibited by a Pyriplatin–DNA Adduct. Be- cause transcription inhibition is an important component in the mechanism of action of platinum anticancer drugs (12–20), we investigated the effect of a site-specific pyriplatin–DNA ad- duct on the kinetics of pol II transcription elongation. We per- formed an extension assay using platinated (Complex A, Table 1) and unplatinated (Complex A0, control, Table 1) pol II transcribing complexes having a 9mer RNA as primer. These complexes were then incubated with a mixture of ATP, CTP, and GTP. The RNA transcripts in A could be elongated from the 9mer to the 11mer, stopping at a position corresponding to the Pt– DNA lesion site observed in the pol II complex of the damaged template DNA, whereas RNA transcripts in A0 were extended much farther downstream on the undamaged template control DNA (Fig. 3A). In order to avoid the possibility of misincorpora- tion-induced transcription inhibition in this assay, we carried out a similar extension assay using an RNA containing a 30-end CMP matched against the damaged base (pol II complex C, 11mer) (Table 1). A single matching GTP was incubated with this pol II complex to test whether the enzyme could bypass the Pt– DNA lesion. Consistent with the results of the previous assay, RNA transcripts could not be extended beyond an 11mer in the pol II complex with the damaged DNA template, whereas RNA transcripts were efficiently extended farther downstream along the undamaged DNA template (Fig. 3B). Similar extension assay results were obtained using a chain-terminated GTP analo- gue 30-dGTP or an RNA primer of different length (complex B, 10mer) (Table 1) (Fig. 3 C and D). Finally, to investigate whether the presence of the damaged base affects the rate of NTP incor- poration in a single round, we used complex B (10mer) and com- plex C (11mer), incubating with CTP and 30-dGTP, respectively. For CTP incorporation, RNA transcripts could be efficiently ex- tended from the 10mer to the 11mer using both damaged and nondamaged templates at a comparable rate (Fig. 3E), whereas no obvious extension of RNA transcripts from the 11mer to a 12mer was observed on the damaged DNA template (Fig. 3C). UTP failed to incorporate at the damaged template under the same conditions (Fig. S3A). No obvious intrinsic cleavage was observed for a pol II complex containing the 11mer RNA and Pt-damaged DNA template in the presence of 20 mM Mg2þ ion, suggesting that most of complex C (11mer) is not in the back- tracked state (Fig. S3B) (21–23). X-ray Structure of Pol II stalled at a Pyriplatin–DNA Adduct. To under- stand the nature of the pol II complex stalled at the pyriplatin- induced Pt–DNA adduct, we solved the x-ray crystal structure Fig. 1. Scheme depicting the formation of a monofunctional platinum-DNA adduct by pyriplatin on double-stranded duplex DNA. The structure of the pyriplatin-damaged DNA duplex used coordinates from the PDB (code 3CO3). The damaged and nondamaged DNA strands are shown in cyan and green, respectively. The pyridine ligand and two ammine groups of the cis-fPtðNH3Þ2ðpyÞg2þ moiety are depicted in magenta and blue, respec- tively. The platinum atom and nitrogen atoms of the cis-fPtðNH3Þ2ðpyÞg2þ moiety are highlighted in yellow and as a blue ball, respectively. The termini of the DNA strands are labeled. A B +1 -1 +1 -1 5’ 3’ 3’ 5’ 5’ 3’ Non-template DNA Bridge Helix Bridge Helix Addition Site Addition Site 5’ V829 A832 RNA RNA Template DNA Template DNA 3’ Fig. 2. Structure of a pol II transcribing complex encountering a site-specific pyriplatin-dG adduct in DNA. (A) A site-specific pyriplatin-DNA adduct is ac- commodated in the pol II active site. The view is a standard one, from the “Rpb2 side,” as described elsewhere (9–11, 39). The RNA transcript, template DNA strand, and nontemplate DNA strand are depicted in red, cyan, and green, respectively. Parts of the bridge helix (Rpb1 825–848) are shown in gray. The pyriplatin-damaged guanosine is colored magenta. The platinum atom of the cis-fPtðNH3Þ2ðpyÞg2þ moiety is denoted as a yellow ball and the two ammine groups are in blue. The dashed oval represents the empty nucleotide addition site in the post-translocation state. The positions of the RNA strand are labeled. (B) cis-fPtðNH3Þ2ðpyÞg2þ-dG in the pol II active site adopts a different configuration in comparison with its conformation in the structure of pyriplatin-modified duplex DNA. The superimposed geometry of the cis-fPtðNH3Þ2ðpyÞg2þ-guanosine unit from the DNA duplex structure (3CO3) is shown in light blue. Side chains of Val 829 and Ala 832 are depicted in orange. The remainder of the figure is the same as in A. Wang et al. PNAS ∣ May 25, 2010 ∣ vol. 107 ∣ no. 21 ∣ 9585 BIOCHEMISTRY of the enzyme in complex with a platinated DNA using an RNA- containing CTP matched against the damaged guanosine residue. In this structure, pol II is in pre-translocation state, with the newly added CMP still occupying the addition site without transloca- tion. The platinated guanosine residue forms Watson–Crick base pairs with the newly added CMP (Fig. 4 A and B and Fig. S4). The cis-fPtðNH3Þ2ðpyÞg2þ moiety is surrounded by the bridge helix at the bottom, part of the Rpb2 fork region (528–534) on the left side, and the sugar-phosphate backbone connecting template DNA positions þ1 and þ2 on the right side (Fig. 4B). Interest- ingly, upon CMP incorporation, the cis-fPtðNH3Þ2ðpyÞg2þ moiety adopts a different conformation. The pyridine group of this unit now faces toward 30-direction of template DNA (Fig. 4 A and B). The ammine group trans to pyridine is directed toward the bridge helix and forms hydrogen bonds with main chain atoms of Ala 828 and the side chain of Thr 831 (Fig. 4B). The residues in the bridge helix are highly conserved from yeast to humans. Because Thr 831 and Ala 828 are absolutely conserved between S. cerevisiae and humans, the interactions we observe in the S. cerevisiae pol II structure will also occur in human pol II. These structural results provide important insights into the transcription stalling process at a monofunctional pyriplatin– DNA adduct. The adduct adopts a significantly different confor- mation within the pol II active site compared to that in duplex DNA (7). The present structural and biochemical evidence reveals that pol II stalls after efficient incorporation of CTP against the damaged guanosine residue. The conformation of the pyriplatin–DNA adduct changes significantly upon incorpora- tion of CTP. The modified guanosine rotates into the pol II active site and serves as a template for base pairing with the matched substrate, and the cis-fPtðNH3Þ2ðpyÞg2þ moiety is now directed toward 30-end of the platinated DNA. Table 1. RNA and DNA scaffold of pol II transcribing complexes Complex A: (Damaged template 29mer with 9mer RNA) RNA: 5′ AUGGAGAGG 3′ DNA: 3′ CTACCTCTCCTG*CCCCACCAATACCCATC 5′ DNA: 5′ GTGGTTATGGGTAG 3′ Complex A′: (Nondamaged template 29mer with 9mer RNA) RNA: 5′ AUGGAGAGG 3′ DNA: 3′ CTACCTCTCCTGCCCCACCAATACCCATC 5′ DNA: 5′ GTGGTTATGGGTAG 3′ Complex B: (Damaged template 29mer with 10mer RNA) RNA: 5′ AUGGAGAGGA 3′ DNA: 3′ CTACCTCTCCTG*CCCCACCAATACCCATC 5′ DNA: 5′ GTGGTTATGGGTAG 3′ Complex B′: (Nondamaged template 29mer with 10mer RNA) RNA: 5′ AUGGAGAGGA 3′ DNA: 3′ CTACCTCTCCTGCCCCACCAATACCCATC 5′ DNA: 5′ GTGGTTATGGGTAG 3′ Complex C: (Damaged template 29mer with 11mer RNA) RNA: 5′ AUGGAGAGGAC3′ DNA: 3′ CTACCTCTCCTG*CCCCACCAATACCCATC 5′ DNA: 5′ GTGGTTATGGGTAG 3′ Complex C′: (Non-damaged Template 29mer with 11mer RNA) RNA: 5′ AUGGAGAGGAC3′ DNA: 3′ CTACCTCTCCTGCCCCACCAATACCCATC 5′ DNA: 5′ GTGGTTATGGGTAG 3′ G*: cDPCP-dG. Fig. 3. Pol II transcription elongation blocked by a site-specific pyriplatin- DNA adduct. (A) In vitro transcription with preformed pol II elongation com- plexes A and A0 incubated with a mixture of ATP, CTP, and GTP (25 μM each). Time points were taken after 0, 0.5, 1, 2, 3, 4, 8, 16, 32, or 64 min incubation. The RNA transcripts in lanes 1–10 were taken from reactions of the pol II com- plex with a nondamaged DNA template, whereas the RNA transcripts in lanes 11–20 were taken from reactions of the pol II complex with a site-specifically damaged DNA template. The stalled RNA transcript is indicated by a black arrow (Right), and the extended RNA transcript is visible to the left. The length and sequences of RNA transcripts are given at the left margin of the gel. (B) In vitro transcription with preformed pol II elongation complexes C and C’ incubated with 25 μM GTP. The remainder of gel is the same as in A. (C) In vitro transcription with preformed pol II elongation complexes C and C’ incubated with 25 μM of 30-dGTP. The rest of gel is same as in A. (D) In vitro transcription with preformed pol II elongation complexes B and B’ incubated with a mixture of 25 μM CTP and 30-dGTP. Time points were taken after 0, 0.5, 1, 2, 4, 8, 16, 32, 64 min of incubation. The rest of gel is the same as in A. (E) In vitro transcription with preformed pol II elongation complexes B and B’ incubated with 25 μM CTP. The remainder of the gel is the same as in A. 9586 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1002565107 Wang et al. The result is that the RNA transcript fails to extend beyond the site of damage, subsequent translocation and nucleotide addition being strongly inhibited. Several factors contribute to such translocation inhibition, including (i) stabilization of the initial pre-translocation state by interaction between the platinated guanosine and pol II residues (Fig. 4B); (ii) a high translocation energy barrier; and (iii) an unfavorable subsequent post-translo- cation state induced by the DNA lesion. Hydrogen bonding inter- actions between an ammine group of the cis-fPtðNH3Þ2ðpyÞg2þ moiety with bridge helix partially help to stabilize the initial pre- translocation state (Fig. 4B). To address the factors ii and iii, we modeled the pyriplatin-damaged guanosine residue at the −1 po- sition to mimic the state following translocation of the pyriplatin- modified guanosine from the þ1 to −1 position. The structure clearly reveals that the cis-fPtðNH3Þ2ðpyÞg2þ moiety serves as a strong steric block, narrowing the space between the DNA nucleotide base (−1) and the bridge helix and preventing the downstream undamaged nucleoside base on the DNA template strand from rotating into the canonical þ1 position (Fig. 5A). Moreover, the fact that the cis-fPtðNH3Þ2ðpyÞg2þ moiety at the −1 position sterically clashes with the downstream nucleotide base at the þ1 position suggests that this final state is unfavorable (Fig. 5 A and B). In summary, our results indicate that pyriplatin– DNA adducts inhibit pol II transcription elongation by prevent- ing subsequent translocation and nucleotide addition beyond the site of damage. Discussion Insights into Structure-Activity Relationships (SARs) for the Monofunc- tional Platinum Drug Family. The original SARs pertaining to bifunctional platinum compounds such as cisplatin (8) were for- mulated to explain why anticancer activity appeared to require neutral, cis-[PtA2X2] compositions, in which A is an amine ligand and X is a monoanionic leaving group. These rules are clearly violated by cationic, monofunctional platinum compounds such as pyriplatin (4, 5). Other monofunctional platinum complexes, including ½PtðdienÞClþ, ½PtðNH3Þ3Clþ, and trans-½PtðNH3Þ2 ðpyÞClþ, are inactive and do not arrest pol II transcription, whereas the cis-fPtðNH3Þ2ðpyÞg2þ unit bound to guanosine blocks pol II transcription and has significant anticancer proper- ties in mice when administered as pyriplatin (4, 5, 8, 24–32). The present structure of pol II in complex with DNA site- specifically modified by pyriplatin provides unique insight into SARs to be expected for monofunctional platinum drug candi- dates. We constructed models of potential stalled transcription complexes containing DNA modified by the following three representative units, fPtðNH3Þ3g2þ, trans-fPtðNH3Þ2ðpyÞg2þ, and cis-fPtðNH3Þ2ðpyÞg2þ bound to guanosine in DNA and posi- tioned in either the −1 or þ1 site of pol II, in order to mimic the A B +1 -1 +1 -1 Bridge Helix Bridge Helix Rpb2 528-534 +1 5’ -1 3’ 5’ 3’ T831 A828 5’ 3’ 5’ 3’ Non-template DNA 5’ 3’ +2 RNA Template DNA Template DNA RNA 3.9 Å 3.9 Å Fig. 4. Structure of pol II transcribing complex stalled at a site-specific pyriplatin-DNA adduct after CMP incorporation. (A) The newly incorporated matched CMP is highlighted in yellow. Other colors are as in Fig. 2. Interac- tions of the damaged nucleotide and pol II residues are highlighted in (B). The view is taken roughly from an ∼90 degree clockwise rotation along the RNA/DNA helix axis from A. Nitrogen and oxygen atoms are depicted in blue and red, respectively. Hydrogen bonds between ammine group of the cis-fPtðNH3Þ2ðpyÞg2þ moiety and bridge helix residues are shown as black dashed lines. The loop of Rpb2 828–834 is shown in green. X A +1 -1 -2 X +1 -1 -2 RNA Template DNA Bridge Helix Non-template DNA Addition Site 3’ 5’ 5’ 5’ 3’ 3’ +1 -1 -1 +1 +2 -2 Bridge Helix Addition Site 3’ 3’ 5’ Template DNA RNA 5’ 3’ B Fig. 5. Pol II translocation following CMP incorporation is inhibited by a site- specific pyriplatin-DNA adduct. (A) The cis-fPtðNH3Þ2ðpyÞg2þ-guanosine unit is superimposed with a nucleoside in −1 position shown in magenta and as a surface view. In the latter, the nitrogen and oxygen atoms are highlighted in blue and red, respectively. CMP at the 30-end of RNA chain is highlighted in yellow. The bridge helix is shown in gray as a surface view. The nucleosides at the þ1 and þ2 position of the template DNA are drawn in wheat and orange, respectively. The rotation of the downstream nucleoside base during trans- location, from the þ2 position to the þ1 position, is blocked by the cis- fPtðNH3Þ2ðpyÞg2þ moiety, as indicated. Other colors are as in Fig. 2. (B) The cis-fPtðNH3Þ2ðpyÞg2þ moiety of pyriplatin-dG adduct modeled at −1 posi- tion clashes with the base at the þ1 position. Colors are as in A, and the view as in Fig. 4B. Wang et al. PNAS ∣ May 25, 2010 ∣ vol. 107 ∣ no. 21 ∣ 9587 BIOCHEMISTRY post- and pre-translocation states, respectively (Fig. S1). For each modeled structure, we rotated the platinum unit about the Pt-N7 bond by 360° and computed van der Waals energies arising from contacts between platinum ligands and the rest of the pol II complex (Figs. S5–S9). The fPtðNH3Þ3g2þ and trans-fPtðNH3Þ2ðpyÞg2þ moieties could be readily accommo- dated within the pol II active site over wide energy minima. The lack of a significant steric clash for these two groups, in either the −1 or þ1 position of the pol II transcribing complex, indicates the absence of a barrier to transcriptional bypass (Figs. S6–S9). This finding agrees with experiment. In contrast, the energy bar- rier is prohibitively high for cis-fPtðNH3Þ2ðpyÞg2þ platinated DNA modeled at −1 position, which is consistent with its ability to block pol II bypass and the failure of pol II to reach the sub- sequent post-translocation state (Figs. S5 and S8). The presence of a pyridine or other bulky group in the cis configuration is important for restricting the rotation range of the cis- fPtðNH3Þ2ðpyÞg2þ moiety and thus rendering it a strong steric block to translocation. For a fPtðNH3Þ3g2þ or trans-fPtðNH3Þ2 ðpyÞg2þ adduct at the −1 position, such a steric clash can be avoided by rotation about the Pt-N7 bond, facilitating subsequent pol II translocation. These results are fully consistent with previous biochemical studies revealing that the latter two DNA adducts are inactive and fail to block transcription (5, 7, 12, 26–33). A Unique Molecular Mechanism of Pol II Transcription Inhibition. The stalling mechanism of monofunctional platinum drugs of the pyriplatin family is dramatically different from transcription inhi- bition by cisplatin and UV-induced 1,2-intrastrand cross-links. For the latter two DNA-modifications, a translocation barrier prevents delivery of damaged bases to the active site and/or leads to misincorporation of NTPs against the damage site, respectively (19, 34). Monofunctional platinum-damaged residues, on the other hand, can cross over the bridge helix and be accommodated in the pol II active site. For Pt–dG adducts, the correct CMP nucleotide can be efficiently incorporated against the damaged guanosine. It is blockage of the subsequent translocation from this position after incorporation of the cytosine nucleotide that leads to inhibition of the RNA polymerase, but only when a bulky pyridine ligand is present in the cis coordination site. In conclusion, we report here the structure of a pol II transcri- bing complex stalled at a site-specific monofunctional DNA adduct, revealing a unique mechanism of transcription inhibition by this kind of genome damage. The results establish a basis for SARs that govern the anticancer drug potential of monofunc- tional platinum-based DNA-damaging agents. Specific inter- actions between pol II active site residues and the platinum ligands are revealed, providing a structural framework for rational design of more potent monofunctional pyriplatin analo- gues. Because the spectrum of activity of pyriplatin is dramatically different from that of cisplatin against an extensive panel of can- cer cell lines but with reduced potency (7), this information will be valuable for increasing the anticancer drug potential of this family of compounds based on pol II stalling with concomitant induction of apoptosis. Methods Preparation of Pol II Transcribing Complexes. Ten-subunit S. cerevisiae pol II was purified as described (35). RNA oligonucleotides were purchased from Dharmacon and DNA oligonucleotides were obtained from IDT. cis- ½PtðNH3Þ2ðpyÞClCl was prepared by Ryan Todd at MIT. The site-specifically platinated template DNA was obtained as described (7). Pol II transcribing complexes were assembled with the use of synthetic oli- gonucleotides (10). Briefly, DNA and RNA oligonucleotides were annealed and mixed with pol II in 20 mM Tris (pH 7.5), 40 mM KCl, and 5 mM DTT. The final mixture contained 2 μM pol II, 10 μM site-specific pyriplatin- damaged template DNA strand, and 20 μM nontemplate DNA and RNA oli- gonucleotides. The mixture was kept for 1 h at room temperature, and excess oligonucleotides were removed by ultrafiltration. Crystals were obtained from solutions containing 390 mM ðNH4Þ2HPO4∕NaH2PO4, pH 5.9–6.3, 50 mM dioxane, 10 mM DTT, and 9–11% PEG6000. Crystals of transcribing complexes were transferred in a stepwise manner to cryobuffer as described (10, 11). For the structure of the pol II complex with CTP incorporation, 10 mM CTP was added to the cryobuffer (10, 11). Crystal Structure Determination and Analysis. Diffraction data were collected on beam line 11-1 at the Stanford Synchrotron Radiation Laboratory. Data were processed in DENZO and SCALEPACK (HKL2000) (36). Model building was performed with the program Coot (37), and refinement was done with REFMAC with TLS (CCP4i) (Table S1). In the structure of pol II complex with a CTP incorporation against damaged guanosine residue, we also observed additional weaker density within the second channel in comparison to the nucleoside residue at the þ1 position, which may correspond to nonspecific binding of a second CTP molecule through the soaking process. All structure models in the figures were superimposed with nucleoside residues near the active site using PYMOL (38). Transcription Assay. Transcription assays were performed essentially as de- scribed (11). In a typical reaction, 32P-labeled RNA oligonucleotide (10 pmol) was annealed with template DNA 29mer (20 pmol, damaged or nondamaged template) and nontemplate DNA 14mer (20 pmol) in elongation buffer (20 mM Tris-HCl, pH 7.5, 40 mM KCl, 0.5 mM MgCl2) in a final volume of 20 μL. An aliquot of the annealed RNA/DNA hybrid was incubated with a five- fold excess of pol II (final concentration of pol II 1.1 μM, of RNA oligonucleo- tide 0.22 μM, and of DNA oligonucleotides 0.44 μM) for 10 min at room tem- perature. Equal volumes of the NTP mixture solution were added (final concentrations 25 μM) and the mixture was then incubated for 0, 0.5, 1, 2, 3, 4, 8, 16, 32, or 64 min at room temperature before addition of stop solu- tion (final concentrations 5 M urea, 44.5 mM Tris-HCl, 44.5 mM boric acid, 26 mM EDTA, pH 8.0, Xylene Cyanol and Bromophenol Blue dyes). RNA pro- ducts were analyzed by PAGE in the presence of urea. Visualization and quantification of products were performed with the use of a PhosphorIma- ger (Molecular Dynamics). Computer Modeling Analysis. Three representative platinum units, fPtðNH3Þ3g2þ, trans-fPtðNH3Þ2ðpyÞg2þ, and cis-fPtðNH3Þ2ðpyÞg2þ bound to guanosine in DNA and positioned in either the −1 or þ1 site of pol II were modeled to mimic the post- and pre-translocation states, respectively. The vdW interaction energies between the three ligands at different orientations and the rest of the pol II complex were systematically computed and taken as direct indicators of steric effects. The structure of the cis-fPtðNH3Þ2ðpyÞg2þ fragment on DNA in pol II is available from the current study. Initial configurations for the other two units, fPtðNH3Þ3g2þ and trans-fPtðNH3Þ2ðpyÞg2þ, were obtained by modeling. Briefly, the same configuration of pol II, DNA, and RNA as found in the struc- ture containing cis-fPtðNH3Þ2ðpyÞg2þ was used for these two complexes. The geometry of the fPtðNH3Þ3g2þ moiety was taken from a previous structure where it binds to a B-DNA dodecamer (PDB ID: 5BNA) (39). Docking was achieved by aligning the damaged guanosine base of the two structures. Finally, the trans-ammine group in fPtðNH3Þ3g2þ was replaced with a pyridine ligand, and the Pt-N bond length was appropriately adjusted to obtain the structure for trans-fPtðNH3Þ2ðpyÞg2þ. The same procedure was used to generate structures at both þ1 and −1 positions. The vdW energies were computed for different configurations generated by rotating about the Pt-N7 bond from −180° to 180° for each platinum modi- fication (see Figs. S5–S7). The rotation angle (φ) was defined to be positive when rotating in the anticlockwise direction. In the configuration with φ ¼ 0°, the plane composed of two Pt-N bonds of the ligand which are per- pendicular to the Pt-N7 bond was set to be parallel to the damaged guano- sine base. We noticed that, for fPtðNH3Þ3g2þ and trans-fPtðNH3Þ2ðpyÞg2þ, two trans ammine groups were accommodated at slightly different configura- tions, with φ ¼ 0° due to the different local environment, which leads to slightly different energies between conformations with φ and φ  180°. Because the purpose of our modeling study is to identify major steric clashes instead of accurately computing free energy changes associated with rota- tion of the ligand, which requires extensive conformational sampling, we performed a simple average of the two energies (E1ðφÞ and E2ðφ  180°Þ) based on their Boltzmann weights (T 298 K), eq 1, ¯E ¼ ðe−βE1E1 þ e−βE2E2Þ∕ðe−βE1 þ e−βE2Þ [1] to get a better estimate of vdW energy profiles. 9588 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1002565107 Wang et al. The GROMACS simulation package was used to compute vdW energies between the ligands and the pol II complex (40). A 20-Å cutoff was adopted for computing the vdW interactions. The AMBER03 force field was used for the pol II complex including protein, RNA, and DNA (41). The vdW force field (Leonard–Jones potential) parameters for ligands were generated from the AMTECHAMBER module of the AMBER 9 package (42) using the general AMBER force field (GAFF) (43) developed for rational drug design. Since the Pt atom is not in direct contact with the pol II complex and does not con- tribute significantly to any steric effects, we excluded it from our vdW energy calculations. ACKNOWLEDGMENTS. This research was supported by the National Institute of General Medical Sciences (NIH Pathway to Independence Award GM085136 to D.W. and GM49985 to R.D. Kornberg) and by the National Cancer Institute (Grant CA034992 to S.J.L.). Portions of the research were carried out at the Stanford Synchrotron Radiation Laboratory, a national user facility operated by Stanford University on behalf of the U.S. Department of Energy, Office of Basic Energy Sciences. The SSRL Structural Molecular Biology Program is sup- ported by the Department of Energy, Office of Biological and Environmental Research, and by the National Institutes of Health, National Center for Research Resources, Biomedical Technology Program, and the National Insti- tute of General Medical Sciences. 1. Jamieson ER, Lippard SJ (1999) Structure, recognition, and processing of cisplatin–DNA adducts. Chem Rev 99:2467–2498. 2. Wang D, Lippard SJ (2005) Cellular processing of platinum anticancer drugs. Nat Rev Drug Discov 4:307–320. 3. Roy S, et al. (2008) Phenanthroline derivatives with improved selectivity as DNA- targeting anticancer or antimicrobial drugs. ChemMedChem 3:1427–1434. 4. Hollis LS, Amundsen AR, Stern EW (1989) Chemical and biological properties of a new series of cis-diammineplatinum(II) antitumor agents containing three nitrogen donors: cis-½PtðNH3Þ2ðN-donorÞClþ. J Med Chem 32:128–136. 5. Hollis LS, et al. (1991) Mechanistic studies of a novel class of trisubstituted platinum(II) antitumor agents. Cancer Res 51:1866–1875. 6. Bierbach U, Sabat M, Farrell N (2000) Inversion of the cis geometry requirement for cytotoxicity in structurally novel platinum(II) complexes containing the bidentate N,O-donor pyridin-2-yl-acetate. Inorg Chem 39:1882–1890. 7. Lovejoy KS, et al. (2008) cis-Diammine(pyridine)chloroplatinum(II), a monofunctional platinum(II) antitumor agent: Uptake, structure, function, and prospects. Proc Natl Acad Sci USA 105:8902–8907. 8. Cleare MJ, Hoeschele JD (1973) Studies on the antitumor activity of group VIII transi- tion metal complexes. Part I. Platinum (II) complexes. Bioinorg Chem 2:187–210. 9. Westover KD, Bushnell DA, Kornberg RD (2004) Structural basis of transcription: separation of RNA from DNA by RNA polymerase II. Science 303:1014–1016. 10. Westover KD, Bushnell DA, Kornberg RD (2004) Structural basis of transcription: nu- cleotide selection by rotation in the RNA polymerase II active center. Cell 119:481–489. 11. Wang D, et al. (2006) Structural basis of transcription: Role of the trigger loop in sub- strate specificity and catalysis. Cell 127:941–954. 12. Corda Y, et al. (1993) Spectrum of DNA–platinum adduct recognition by prokaryotic and eukaryotic DNA-dependent RNA polymerases. Biochemistry 32:8582–8588. 13. Mello JA, Lippard SJ, Essigmann JM (1995) DNA adducts of cis-diamminedichloropla- tinum(II) and its trans isomer inhibit RNA polymerase II differentially in vivo. Biochem- istry 34:14783–14791. 14. Sandman KE, Marla SS, Zlokarnik G, Lippard SJ (1999) Rapid fluorescence-based repor- ter-gene assays to evaluate the cytotoxicity and antitumor drug potential of platinum complexes. Chem Biol 6:541–551. 15. Lee KB, Wang D, Lippard SJ, Sharp PA (2002) Transcription-coupled and DNA damage- dependent ubiquitination of RNA polymerase II in vitro. Proc Natl Acad Sci USA 99:4239–4244. 16. Tornaletti S, Patrick SM, Turchi JJ, Hanawalt PC (2003) Behavior of T7 RNA polymerase and mammalian RNA polymerase II at site-specific cisplatin adducts in the template DNA. J Biol Chem 278:35791–35797. 17. Tremeau-Bravard A, Riedl T, Egly JM, Dahmus ME (2004) Fate of RNA polymerase II stalled at a cisplatin lesion. J Biol Chem 279:7751–7759. 18. Jung Y, Lippard SJ (2006) RNA polymerase II blockage by cisplatin-damaged DNA. Stability and polyubiquitylation of stalled polymerase. J Biol Chem 281:1361–1370. 19. Damsma GE, et al. (2007) Mechanism of transcriptional stalling at cisplatin-damaged DNA. Nat Struct Mol Biol 14:1127–1133. 20. Jung Y, Lippard SJ (2007) Direct cellular responses to platinum-induced DNA damage. Chem Rev 107:1387–1407. 21. Surratt CK, Milan SC, Chamberlin MJ (1991) Spontaneous cleavage of RNA in ternary complexes of Escherichia coli RNA polymerase and its significance for the mechanism of transcription. Proc Natl Acad Sci USA 88:7983–7987. 22. Orlova M, et al. (1995) Intrinsic transcript cleavage activity of RNA polymerase. Proc Natl Acad Sci USA 92:4596–4600. 23. Wang D, et al. (2009) Structural basis of transcription: backtracked RNA polymerase II at 3.4 angstrom resolution. Science 324:1203–1206. 24. Cohen GL, Bauer WR, Barton JK, Lippard SJ (1979) Binding of cis- and trans- dichlorodiammineplatinum(II) to DNA: evidence for unwinding and shortening of the double helix. Science 203:1014–1016. 25. Lecointe P, Macquet JP, Butour JL (1979) Correlation between the toxicity of platinum drugs to L1210 leukemia cells and their mutagenic properties. Biochem Biophys Res Commun 90:209–213. 26. Macquet JP, Butour JL (1983) Platinum-amine compounds: importance of the labile and inert ligands for their pharmacological activities toward L1210 leukemia cells. J Natl Cancer Inst 70:899–905. 27. Calvert AH (1986) Clinical applications of platinum metal complexes. Biochemical Mechanisms of Platinum Antitumor Drugs (IRI Press, Washington). 28. Balcarová Z, et al. (1998) DNA interactions of a novel platinum drug, cis-½PtClðNH3Þ2 ðN7-acyclovirÞþ. Mol Pharmacol 53:846–855. 29. Brabec V, Leng M (1993) DNA interstrand cross-links of trans-diamminedichloroplati- num(II) are preferentially formed between guanine and complementary cytosine residues. Proc Natl Acad Sci USA 90:5345–5349. 30. Lemaire MA, Schwartz A, Rahmouni AR, Leng M (1991) Interstrand cross-links are preferentially formed at the d(GC) sites in the reaction between cis-diamminedichlor- oplatinum (II) and DNA. Proc Natl Acad Sci USA 88:1982–1985. 31. Brabec V, Boudny V (1994) Monofunctional and interstrand DNA adducts of platinum (II) complexes. Met Based Drugs 1:195–200. 32. Brabec V, Boudný V, Balcarová Z (1994) Monofunctional adducts of platinum(II) pro- duce in DNA a sequence-dependent local denaturation. Biochemistry 33:1316–1322. 33. Novakova O, et al. (2009) Energetics, conformation, and recognition of DNA duplexes modified by methylated analogues of ½PtClðdienÞþ. Chemistry 15:6211–6221. 34. Brueckner F, Hennecke U, Carell T, Cramer P (2007) CPD damage recognition by tran- scribing RNA polymerase II. Science 315:859–862. 35. Cramer P, Bushnell DA, Kornberg RD (2001) Structural basis of transcription: RNA polymerase II at 2.8 angstrom resolution. Science 292:1863–1876. 36. Otwinowski Z, Minor W (1997) Processing of x-ray diffraction data collected in oscilla- tion mode. Method Enzymol 276:307–326. 37. Emsley P, Cowtan K (2004) Coot: Model-building tools for molecular graphics. Acta Cryst D60:2126–2132. 38. DeLano WL (2002) The PyMOL Molecular Graphics System (DeLano Scientific, Palo Alto, CA). 39. Wing RM, Pjura P, Drew HR, Dickerson RE (1984) The primary mode of binding of cisplatin to a B-DNA dodecamer: C-G-C-G-A-A-T-T-C-G-C-G. EMBO J 3:1201–1206. 40. Lindahl E, Hess B, van der Spoel D (2001) GROMACS 3.0: A package for molecular simulation and trajectory analysis. J Mol Model 7:306–317. 41. Duan Y, et al. (2003) A point-charge force field for molecular mechanics simulations of proteins based on condensed-phase quantum mechanical calculations. J Comput Chem 24:1999–2012. 42. Wang JM, Wang W, Kollman PA, Case DA (2006) Automatic atom type and bond type perception in molecular mechanical calculations. J Mol Graphics Modell 25:247–260. 43. Wang J, et al. (2004) Development and testing of a general Amber force field. J Comput Chem 25:1157–1174. Wang et al. PNAS ∣ May 25, 2010 ∣ vol. 107 ∣ no. 21 ∣ 9589 BIOCHEMISTRY
3M41
Crystal structure of the mutant V182A of orotidine 5'-monophosphate decarboxylase from Methanobacterium thermoautotrophicum
Conformational Changes in Orotidine 5′-Monophosphate Decarboxylase: “Remote” Residues that Stabilize the Active Conformation† B. McKay Wood‡, Tina L. Amyes§, Alexander A. Fedorov∥, Elena V. Fedorov∥, Andrew Shabila‡, Steven C. Almo∥, John P. Richard§, and John A. Gerlt*,‡ ‡Departments of Biochemistry and Chemistry, University of Illinois at Urbana-Champaign §Department of Chemistry, University at Buffalo, Buffalo, NY 14260 ∥Department of Biochemistry, Albert Einstein College of Medicine, Bronx, NY 10461 Abstract The structural factors responsible for the extraordinary rate enhancement (~1017) of the reaction catalyzed by orotidine 5′-monophosphate decarboxylase (OMPDC) have not been defined. Catalysis requires a conformational change that closes an active site loop and “clamps” the orotate base proximal to hydrogen-bonded networks that destabilize the substrate and stabilize the intermediate. In the OMPDC from Methanobacter thermoautotrophicus, a “remote” structurally conserved cluster of hydrophobic residues that includes Val 182 in the active site loop is assembled in the closed, catalytically active conformation. Substitution of these residues with Ala decreases kcat/Km with a minimal effect on kcat, providing evidence that the cluster stabilizes the closed conformation. The intrinsic binding energies of the 5′-phosphate group of OMP for the mutant enzymes are similar to that for the wild type, supporting this conclusion. Orotidine 5′-monophosphate decarboxylase (OMPDC1) is one of Nature’s best catalysts: the reaction occurs with a rate enhancement of ~1017 and a proficiency of ~1023 M (2). The reaction coordinate includes a vinyl carbanion intermediate (Scheme 1) (3, 4). Structures of OMPDCs complexed with UMP or 6-hydroxyUMP, a n intermediate analog (5-9), reveal hydrogen-bonded networks proximal to 1) C6 of the pyrimidine (Asp 70-Lys 72-Asp 75 in the OMPDC from Methanobacter thermoautotrophicus; MtOMPDC); and 2) O2, N3, and O4 of the base (Ser 127-Gln 185). These destabilize the substrate (9) and stabilize the intermediate, although the structural strategy for the latter is uncertain. The 5′-phosphate group binds in a conserved motif at the ends of the seventh and eighth β- strands of the (β/α)8-barrel structure, with interactions to backbone NH groups as well as the guanidinium group of a conserved Arg (Arg 203 in MtOMPDC). These interactions are important for catalysis: 1) kcat/Km for OMP exceeds that for orotidine by a factor of ~1011 for the OMPDC from Saccharomyces cerevisiae (ScOMPDC) (10); 2) kcat/Km for OMP †This work was supported by NIH Grants GM039754 to J.P.R. and GM065155 to J.A.G. *To whom correspondence should be addressed: J.A.G.: telephone, (217) 244-7414; fax: (217) 244-6538; j-gerlt@uiuc.edu.. SUPPORTING INFORMATION AVAILABLE Descriptions of the experimental procedures. This material is available free of charge via the Internet at http://pubs.acs.org. 1Abbreviations: OMP, orotidine 5′-monophosphate; OMPDC, OMP decarboxylase; MtOMPDC, OMPDC from Methanobacter thermoautotrophicus; ScOMPDC, OMPDC from Saccharomyces cerevisiae; EO, 1-(β-D-erythrofuranosyl)-orotic acid; HPi, phosphate dianion; IBE, intrinsic binding energy. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 December 26. Published in final edited form as: Biochemistry. 2010 May 4; 49(17): 3514–3516. doi:10.1021/bi100443a. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript exceeds that for 1-(β-D-erythrofuranosyl)orotic acid (EO; 5′-truncated OMP analog) by factors of 5.2 × 108 and 3.6 × 108 for ScOMPDC and MtOMPDC, respectively (11, 12); and 3 ) phosphite dianion ( H Pi) activates decarboxylation of EO by factors of 5.6 × 105 M−1 and 2.9 × 105 M−1 for ScOMPDC and MtOMPDC, respectively (11, 12). The “intrinsic binding energy” [IBE; (13)] of the 5′-phosphate/HPi 1) increases the affinity for the substrate (e.g., OMP vs. orotidine/EO); and 2) enables decarboxylation by juxtaposition of the substrate with the active site hydrogen-bonded networks (substrate destabilization and intermediate stabilization). How the IBE promotes catalysis is unknown but required to understand the structural basis for the rate enhancement. A loop located at the end of the seventh β-strand closes over the active site when OMP binds (Figure 1). Although the active site loops differ in both length and sequence in divergent OMPDCs (12), each includes a spatially conserved Gln (Gln 185 in MtOMPDC) hydrogen- bonded to both the 5′-phosphate and O2 of the pyrimidine as well as a conserved Ser at the end of the fifth β-strand that also is hydrogen-bonded to N3 of the pyrimidine (Ser 127). We characterized the importance of these “clamp” residues in ScOMPDC (Gln 215-Ser 154) using EO/HPi and confirmed that the 5′-phosphate/HPi “switch” is required to activate the enzyme (14). The structures of divergent OMPDCs reveal that OMP binding is always accompanied by a conformational change (Figure 1). The most obvious component is closure of the active site loop. However, the (β/α)8-barrel structure can be divided into two domains, one formed from the second, third, fourth, and fifth β-strands (where the hydrogen-bonded Asp 70-Lys 72-Asp 75 motif and Ser 127 are located) and the second from the sixth, seventh, eighth, and first β-strands (where the phosphate binding motif and the active site loop, including Gln 185, are located) (15). OMP binding reorients the domains, with the latter domain moving toward the former, enforcing the orotate carboxylate group to be juxtaposed vis a vis Asp 70- Lys 72-Asp 75 and, also, allowing formation of the Ser 127-Gln 185 “clamp”. Thus, the transition between the open and closed conformations is more complicated than “simple” hinge motion of the loop on the rigid framework of the (β/α)8-barrel structure. In this Report we identify “remote” residues involved in this conformational change and quantitate their importance in promoting and stabilizing the catalytically competent form of the enzyme. The active site loop of MtOMPDC, P180-G181-V182-G183-A184-Q185-G186-G187-D188, is disordered in the absence of substrate but ordered and closed in its presence (Figure 1). In the liganded structure two residues in the loop make contacts with the (β/α)8-barrel scaffold: 1) Gln 185 is hydrogen-bonded to Ser 127 (vide infra); and 2) Val 182 is embedded in a hydrophobic cluster also formed by Ile 199, Val 201, and Ile 218. The conservation of this hydrophobic cluster in all OMPDCs suggests its importance in a common catalytic strategy. We probed this strategy by mutagenesis of these hydrophobic residues. Ala substitutions for residues in the hydrophobic cluster cause substantial decreases in kcat/ Km for decarboxylation of OMP but have little effect on kcat (Table 1). We determined high resolution X-ray structures (≤ 1.4 Å) for each mutant in the presence of 6-hydroxyUMP (Figure 2). The liganded structures superimpose well with that of wild type, with only small differences observed at the sites of the substitutions (panel A). The active sites are identical to that of wild type (panel B), explaining the minimal impact on kcat. The mutated residues are remote from the active site (Figures 1 and 2). Thus, the effects of the substitutions on kcat/Km cannot be explained by altered direct interactions with the substrate. Instead, the effects can be explained by decreased stabilities of the closed conformation in which the substrate is destabilized (9) and the anionic intermediate is stabilized. A consistent model (Scheme 2) is that an equilibrium of open (Eo) and closed Wood et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (Ec) conformations exists in the absence or presence of substrate (KC’ or KC; KC’ << 1). The substitutions destabilize Ec (decrease KC and KC ’), so kcat/Km is decreased (the energy difference between Eo + S and [Ec•S]‡ is increased). However, the invariance of kcat establishes that the substitutions do not alter the reactivitiy of Ec•S (the energy difference between Ec•S and [Ec•S]‡). The energy to form Ec•S from Eo + S is achieved from 1) interactions of the 5′-phosphate group with its binding motif (IBE; vide infra), and 2) an increased concentration of OMP to form Ec•S (KC/KS or, equivalently, KC’/KS’, although the former is expected to be the relevant pathway). We also used the two part EO•HPi substrate. The values of kcat/Km for EO are decreased relative to that for wild type (Table 1); these can be explained by decreased populations of Ec (Scheme 2), assuming that EO, without a 5′-phosphate group, is unable to promote the transition from Eo to Ec. The values of kcat/Km for OMP and kcat/Km for EO allow calculation of the IBE for the 5′-phosphate group of OMP (Table 1). HPi activates the mutants as judged by the values of the third-order rate constant, (kcat/ Km)EO•HPi/Kd. The equivalent changes (ΔΔG‡) in kcat/Km for both OMP and EO and the third-order rate constant indicate that all three measure the effects of the substitutions on the values of KC. The values of the IBEs for the 5′-phosphate group for the mutants are the same as that for the wild type, establishing that decarboxylation in the Ec•S complex occurs with equivalent amounts of ground state destabilization (9) and transition state stabilization (as also reflected by the invariant values of kcat). The IBEs provide further support for the role of the “remote” hydrophobic cluster in stabilizing Ec relative to Eo but do not directly participating in catalysis. Our experiments implicate a structurally conserved hydrophobic cluster, Val 182, Ile 199, Val 201, and Ile 218 in MtOMPDC, in stabilizing the closed conformation required for catalysis. Its identification provides evidence that structural elements distal from the active site, in addition to the proximal active site loop that closes to “clamp” the substrate, are required for OMPDC’s extraordinary catalytic efficiency and proficiency. Supplementary Material Refer to Web version on PubMed Central for supplementary material. REFERENCES (1). Wood BM, Chan KK, Amyes TL, Richard JP, Gerlt JA. Biochemistry. 2009; 48:5510–5517. [PubMed: 19435313] (2). Radzicka A, Wolfenden R. Science. 1995; 267:90–93. [PubMed: 7809611] (3). Toth K, Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2007; 129:12946–12947. [PubMed: 17918849] (4). Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2008; 130:1574–1575. [PubMed: 18186641] (5). Miller BG, Hassell AM, Wolfenden R, Milburn MV, Short SA. Proc. Natl. Acad. Sci. U A. 2000; 97:2011–2016. (6). Appleby TC, Kinsland C, Begley TP, Ealick SE. Proc atl Acad Sci U S A. 2000; 97:2005–2010. (7). Harris P, Navarro Poulsen JC, Jensen KF, Larsen S. Biochemistry. 2000; 39:4217–4224. [PubMed: 10757968] (8). Wu N, Mo Y, Gao J, Pai EF. Proc. Natl. Acad. Sci. U S A. 2000; 97:2017–2022. [PubMed: 10681441] Wood et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (9). Chan KK, Wood BM, Fedorov AA, Fedorov EV, Imker HJ, Amyes TL, Richard JP, Almo SC, Gerlt JA. Biochemistry. 2009; 48:5518–5531. [PubMed: 19435314] (10). Sievers A, Wolfenden R. Bioorg. Chem. 2005; 33:45–52. [PubMed: 15668182] (11). Amyes TL, Richard JP, Tait JJ. J. Am. Chem. Soc. 2005; 127:15708–15709. [PubMed: 16277505] (12). Toth K, Amyes TL, Wood BM, Chan KK, Gerlt JA, Richard JP. Biochemistry. 2009; 48:8006– 8013. [PubMed: 19618917] (13). Jencks WP. Adv. Enzymol. Rel. Areas. Mol. Biol. 1975; 43:219–410. (14). Barnett SA, Amyes TL, Wood BM, Gerlt JA, Richard JP. Biochemistry. 2008; 47:7785–4487. [PubMed: 18598058] (15). Harris P, Poulsen JC, Jensen KF, Larsen S. J. Mol. Biol. 2002; 318:1019–1029. [PubMed: 12054799] (16). Go MK, Amyes TL, Richard JP. Biochemistry. 2009; 48:5769–5778. [PubMed: 19425580] Wood et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Superposition of the active site of wild type MtOMPDC in the absence (cyan; disordered loop depicted with the dotted line ) and presence ( grey ) of 6-hydroxyUMP. The carbons of 6-hydroxyUMP are highlighted in yellow; the carbons of Val 182 and Arg 203 in the liganded structure are highlighted in orange. Wood et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Superpositions of the 6-hydroxyUMP-liganded structures of wild type MtOMPDC and the single mutants in the hydrophobic cluster. Panel A, hydrophobic cluster; panel B, active sites. Wood et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. Wood et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 2. Wood et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Wood et al. Page 9 Table 1 Kinetic Constants for OMP, EO and EO/HPi and Intrinsic Binding Energies of the 5′-Phosphate Group of OMP at pH 7.1 and 25 °C. MtOMPDC kcat OMP s−1 kcat/Km OMP M−1 s−1 ΔΔG‡ kcal/mola kcat/Km EO M−1 s−1 ΔΔG‡ kcal/mola (kcat/Km)/K D EO•HPib M−2 s−1 ΔΔG‡ kcal/mola 5′-Phosphate IBEc kcal/mol Wild type 4.6 2.9 × 106 8.7 × 10−3 2500 11.6 d V182A 3.4 1.4 × 105 1.8 1.3 × 10−3 1.1 190 1.5 10.9 I199A 3.9 9.1 × 105 0.7 1.9 × 10−3 0.9 980 0.6 11.8 V201A 4.0 9.5 × 105 0.7 3.1 × 10−3 0.6 690 0.8 11.5 I218A 3.3 2.8 × 105 1.4 2.3 × 10−3 0.8 340 1.2 11.0 V182A/I199A 3.1 4.9 × 104 2.4 3.9 × 10−4 1.8 81 2.0 11.0 V182A/V201A 2.5 4.9 × 104 2.4 5.0 × 10−4 1.7 30 2.6 10.9 aCalculated from the ratio of the second-order or third-order rate constants for the wild type and mutant enzyme. bThird-order rate constant for reaction of EO/HPi. cTransition state stabilization by the 5′-phosphate group of OMP, calculated from the ratio of the values of kcat/Km for OMP and EO. IBEs for phosphite dianion can be calculated from the ratio of (kcat/ Km)/KD for EOHPi and kcat/Km for EO. Biochemistry. Author manuscript; available in PMC 2011 December 26.
3M43
Crystal structure of the mutant I199A of orotidine 5'-monophosphate decarboxylase from Methanobacterium thermoautotrophicum
Conformational Changes in Orotidine 5′-Monophosphate Decarboxylase: “Remote” Residues that Stabilize the Active Conformation† B. McKay Wood‡, Tina L. Amyes§, Alexander A. Fedorov∥, Elena V. Fedorov∥, Andrew Shabila‡, Steven C. Almo∥, John P. Richard§, and John A. Gerlt*,‡ ‡Departments of Biochemistry and Chemistry, University of Illinois at Urbana-Champaign §Department of Chemistry, University at Buffalo, Buffalo, NY 14260 ∥Department of Biochemistry, Albert Einstein College of Medicine, Bronx, NY 10461 Abstract The structural factors responsible for the extraordinary rate enhancement (~1017) of the reaction catalyzed by orotidine 5′-monophosphate decarboxylase (OMPDC) have not been defined. Catalysis requires a conformational change that closes an active site loop and “clamps” the orotate base proximal to hydrogen-bonded networks that destabilize the substrate and stabilize the intermediate. In the OMPDC from Methanobacter thermoautotrophicus, a “remote” structurally conserved cluster of hydrophobic residues that includes Val 182 in the active site loop is assembled in the closed, catalytically active conformation. Substitution of these residues with Ala decreases kcat/Km with a minimal effect on kcat, providing evidence that the cluster stabilizes the closed conformation. The intrinsic binding energies of the 5′-phosphate group of OMP for the mutant enzymes are similar to that for the wild type, supporting this conclusion. Orotidine 5′-monophosphate decarboxylase (OMPDC1) is one of Nature’s best catalysts: the reaction occurs with a rate enhancement of ~1017 and a proficiency of ~1023 M (2). The reaction coordinate includes a vinyl carbanion intermediate (Scheme 1) (3, 4). Structures of OMPDCs complexed with UMP or 6-hydroxyUMP, a n intermediate analog (5-9), reveal hydrogen-bonded networks proximal to 1) C6 of the pyrimidine (Asp 70-Lys 72-Asp 75 in the OMPDC from Methanobacter thermoautotrophicus; MtOMPDC); and 2) O2, N3, and O4 of the base (Ser 127-Gln 185). These destabilize the substrate (9) and stabilize the intermediate, although the structural strategy for the latter is uncertain. The 5′-phosphate group binds in a conserved motif at the ends of the seventh and eighth β- strands of the (β/α)8-barrel structure, with interactions to backbone NH groups as well as the guanidinium group of a conserved Arg (Arg 203 in MtOMPDC). These interactions are important for catalysis: 1) kcat/Km for OMP exceeds that for orotidine by a factor of ~1011 for the OMPDC from Saccharomyces cerevisiae (ScOMPDC) (10); 2) kcat/Km for OMP †This work was supported by NIH Grants GM039754 to J.P.R. and GM065155 to J.A.G. *To whom correspondence should be addressed: J.A.G.: telephone, (217) 244-7414; fax: (217) 244-6538; j-gerlt@uiuc.edu.. SUPPORTING INFORMATION AVAILABLE Descriptions of the experimental procedures. This material is available free of charge via the Internet at http://pubs.acs.org. 1Abbreviations: OMP, orotidine 5′-monophosphate; OMPDC, OMP decarboxylase; MtOMPDC, OMPDC from Methanobacter thermoautotrophicus; ScOMPDC, OMPDC from Saccharomyces cerevisiae; EO, 1-(β-D-erythrofuranosyl)-orotic acid; HPi, phosphate dianion; IBE, intrinsic binding energy. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 December 26. Published in final edited form as: Biochemistry. 2010 May 4; 49(17): 3514–3516. doi:10.1021/bi100443a. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript exceeds that for 1-(β-D-erythrofuranosyl)orotic acid (EO; 5′-truncated OMP analog) by factors of 5.2 × 108 and 3.6 × 108 for ScOMPDC and MtOMPDC, respectively (11, 12); and 3 ) phosphite dianion ( H Pi) activates decarboxylation of EO by factors of 5.6 × 105 M−1 and 2.9 × 105 M−1 for ScOMPDC and MtOMPDC, respectively (11, 12). The “intrinsic binding energy” [IBE; (13)] of the 5′-phosphate/HPi 1) increases the affinity for the substrate (e.g., OMP vs. orotidine/EO); and 2) enables decarboxylation by juxtaposition of the substrate with the active site hydrogen-bonded networks (substrate destabilization and intermediate stabilization). How the IBE promotes catalysis is unknown but required to understand the structural basis for the rate enhancement. A loop located at the end of the seventh β-strand closes over the active site when OMP binds (Figure 1). Although the active site loops differ in both length and sequence in divergent OMPDCs (12), each includes a spatially conserved Gln (Gln 185 in MtOMPDC) hydrogen- bonded to both the 5′-phosphate and O2 of the pyrimidine as well as a conserved Ser at the end of the fifth β-strand that also is hydrogen-bonded to N3 of the pyrimidine (Ser 127). We characterized the importance of these “clamp” residues in ScOMPDC (Gln 215-Ser 154) using EO/HPi and confirmed that the 5′-phosphate/HPi “switch” is required to activate the enzyme (14). The structures of divergent OMPDCs reveal that OMP binding is always accompanied by a conformational change (Figure 1). The most obvious component is closure of the active site loop. However, the (β/α)8-barrel structure can be divided into two domains, one formed from the second, third, fourth, and fifth β-strands (where the hydrogen-bonded Asp 70-Lys 72-Asp 75 motif and Ser 127 are located) and the second from the sixth, seventh, eighth, and first β-strands (where the phosphate binding motif and the active site loop, including Gln 185, are located) (15). OMP binding reorients the domains, with the latter domain moving toward the former, enforcing the orotate carboxylate group to be juxtaposed vis a vis Asp 70- Lys 72-Asp 75 and, also, allowing formation of the Ser 127-Gln 185 “clamp”. Thus, the transition between the open and closed conformations is more complicated than “simple” hinge motion of the loop on the rigid framework of the (β/α)8-barrel structure. In this Report we identify “remote” residues involved in this conformational change and quantitate their importance in promoting and stabilizing the catalytically competent form of the enzyme. The active site loop of MtOMPDC, P180-G181-V182-G183-A184-Q185-G186-G187-D188, is disordered in the absence of substrate but ordered and closed in its presence (Figure 1). In the liganded structure two residues in the loop make contacts with the (β/α)8-barrel scaffold: 1) Gln 185 is hydrogen-bonded to Ser 127 (vide infra); and 2) Val 182 is embedded in a hydrophobic cluster also formed by Ile 199, Val 201, and Ile 218. The conservation of this hydrophobic cluster in all OMPDCs suggests its importance in a common catalytic strategy. We probed this strategy by mutagenesis of these hydrophobic residues. Ala substitutions for residues in the hydrophobic cluster cause substantial decreases in kcat/ Km for decarboxylation of OMP but have little effect on kcat (Table 1). We determined high resolution X-ray structures (≤ 1.4 Å) for each mutant in the presence of 6-hydroxyUMP (Figure 2). The liganded structures superimpose well with that of wild type, with only small differences observed at the sites of the substitutions (panel A). The active sites are identical to that of wild type (panel B), explaining the minimal impact on kcat. The mutated residues are remote from the active site (Figures 1 and 2). Thus, the effects of the substitutions on kcat/Km cannot be explained by altered direct interactions with the substrate. Instead, the effects can be explained by decreased stabilities of the closed conformation in which the substrate is destabilized (9) and the anionic intermediate is stabilized. A consistent model (Scheme 2) is that an equilibrium of open (Eo) and closed Wood et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (Ec) conformations exists in the absence or presence of substrate (KC’ or KC; KC’ << 1). The substitutions destabilize Ec (decrease KC and KC ’), so kcat/Km is decreased (the energy difference between Eo + S and [Ec•S]‡ is increased). However, the invariance of kcat establishes that the substitutions do not alter the reactivitiy of Ec•S (the energy difference between Ec•S and [Ec•S]‡). The energy to form Ec•S from Eo + S is achieved from 1) interactions of the 5′-phosphate group with its binding motif (IBE; vide infra), and 2) an increased concentration of OMP to form Ec•S (KC/KS or, equivalently, KC’/KS’, although the former is expected to be the relevant pathway). We also used the two part EO•HPi substrate. The values of kcat/Km for EO are decreased relative to that for wild type (Table 1); these can be explained by decreased populations of Ec (Scheme 2), assuming that EO, without a 5′-phosphate group, is unable to promote the transition from Eo to Ec. The values of kcat/Km for OMP and kcat/Km for EO allow calculation of the IBE for the 5′-phosphate group of OMP (Table 1). HPi activates the mutants as judged by the values of the third-order rate constant, (kcat/ Km)EO•HPi/Kd. The equivalent changes (ΔΔG‡) in kcat/Km for both OMP and EO and the third-order rate constant indicate that all three measure the effects of the substitutions on the values of KC. The values of the IBEs for the 5′-phosphate group for the mutants are the same as that for the wild type, establishing that decarboxylation in the Ec•S complex occurs with equivalent amounts of ground state destabilization (9) and transition state stabilization (as also reflected by the invariant values of kcat). The IBEs provide further support for the role of the “remote” hydrophobic cluster in stabilizing Ec relative to Eo but do not directly participating in catalysis. Our experiments implicate a structurally conserved hydrophobic cluster, Val 182, Ile 199, Val 201, and Ile 218 in MtOMPDC, in stabilizing the closed conformation required for catalysis. Its identification provides evidence that structural elements distal from the active site, in addition to the proximal active site loop that closes to “clamp” the substrate, are required for OMPDC’s extraordinary catalytic efficiency and proficiency. Supplementary Material Refer to Web version on PubMed Central for supplementary material. REFERENCES (1). Wood BM, Chan KK, Amyes TL, Richard JP, Gerlt JA. Biochemistry. 2009; 48:5510–5517. [PubMed: 19435313] (2). Radzicka A, Wolfenden R. Science. 1995; 267:90–93. [PubMed: 7809611] (3). Toth K, Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2007; 129:12946–12947. [PubMed: 17918849] (4). Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2008; 130:1574–1575. [PubMed: 18186641] (5). Miller BG, Hassell AM, Wolfenden R, Milburn MV, Short SA. Proc. Natl. Acad. Sci. U A. 2000; 97:2011–2016. (6). Appleby TC, Kinsland C, Begley TP, Ealick SE. Proc atl Acad Sci U S A. 2000; 97:2005–2010. (7). Harris P, Navarro Poulsen JC, Jensen KF, Larsen S. Biochemistry. 2000; 39:4217–4224. [PubMed: 10757968] (8). Wu N, Mo Y, Gao J, Pai EF. Proc. Natl. Acad. Sci. U S A. 2000; 97:2017–2022. [PubMed: 10681441] Wood et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (9). Chan KK, Wood BM, Fedorov AA, Fedorov EV, Imker HJ, Amyes TL, Richard JP, Almo SC, Gerlt JA. Biochemistry. 2009; 48:5518–5531. [PubMed: 19435314] (10). Sievers A, Wolfenden R. Bioorg. Chem. 2005; 33:45–52. [PubMed: 15668182] (11). Amyes TL, Richard JP, Tait JJ. J. Am. Chem. Soc. 2005; 127:15708–15709. [PubMed: 16277505] (12). Toth K, Amyes TL, Wood BM, Chan KK, Gerlt JA, Richard JP. Biochemistry. 2009; 48:8006– 8013. [PubMed: 19618917] (13). Jencks WP. Adv. Enzymol. Rel. Areas. Mol. Biol. 1975; 43:219–410. (14). Barnett SA, Amyes TL, Wood BM, Gerlt JA, Richard JP. Biochemistry. 2008; 47:7785–4487. [PubMed: 18598058] (15). Harris P, Poulsen JC, Jensen KF, Larsen S. J. Mol. Biol. 2002; 318:1019–1029. [PubMed: 12054799] (16). Go MK, Amyes TL, Richard JP. Biochemistry. 2009; 48:5769–5778. [PubMed: 19425580] Wood et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Superposition of the active site of wild type MtOMPDC in the absence (cyan; disordered loop depicted with the dotted line ) and presence ( grey ) of 6-hydroxyUMP. The carbons of 6-hydroxyUMP are highlighted in yellow; the carbons of Val 182 and Arg 203 in the liganded structure are highlighted in orange. Wood et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Superpositions of the 6-hydroxyUMP-liganded structures of wild type MtOMPDC and the single mutants in the hydrophobic cluster. Panel A, hydrophobic cluster; panel B, active sites. Wood et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. Wood et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 2. Wood et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Wood et al. Page 9 Table 1 Kinetic Constants for OMP, EO and EO/HPi and Intrinsic Binding Energies of the 5′-Phosphate Group of OMP at pH 7.1 and 25 °C. MtOMPDC kcat OMP s−1 kcat/Km OMP M−1 s−1 ΔΔG‡ kcal/mola kcat/Km EO M−1 s−1 ΔΔG‡ kcal/mola (kcat/Km)/K D EO•HPib M−2 s−1 ΔΔG‡ kcal/mola 5′-Phosphate IBEc kcal/mol Wild type 4.6 2.9 × 106 8.7 × 10−3 2500 11.6 d V182A 3.4 1.4 × 105 1.8 1.3 × 10−3 1.1 190 1.5 10.9 I199A 3.9 9.1 × 105 0.7 1.9 × 10−3 0.9 980 0.6 11.8 V201A 4.0 9.5 × 105 0.7 3.1 × 10−3 0.6 690 0.8 11.5 I218A 3.3 2.8 × 105 1.4 2.3 × 10−3 0.8 340 1.2 11.0 V182A/I199A 3.1 4.9 × 104 2.4 3.9 × 10−4 1.8 81 2.0 11.0 V182A/V201A 2.5 4.9 × 104 2.4 5.0 × 10−4 1.7 30 2.6 10.9 aCalculated from the ratio of the second-order or third-order rate constants for the wild type and mutant enzyme. bThird-order rate constant for reaction of EO/HPi. cTransition state stabilization by the 5′-phosphate group of OMP, calculated from the ratio of the values of kcat/Km for OMP and EO. IBEs for phosphite dianion can be calculated from the ratio of (kcat/ Km)/KD for EOHPi and kcat/Km for EO. Biochemistry. Author manuscript; available in PMC 2011 December 26.
3M44
Crystal structure of the mutant V201A of orotidine 5'-monophosphate decarboxylase from Methanobacterium thermoautotrophicum
Conformational Changes in Orotidine 5′-Monophosphate Decarboxylase: “Remote” Residues that Stabilize the Active Conformation† B. McKay Wood‡, Tina L. Amyes§, Alexander A. Fedorov∥, Elena V. Fedorov∥, Andrew Shabila‡, Steven C. Almo∥, John P. Richard§, and John A. Gerlt*,‡ ‡Departments of Biochemistry and Chemistry, University of Illinois at Urbana-Champaign §Department of Chemistry, University at Buffalo, Buffalo, NY 14260 ∥Department of Biochemistry, Albert Einstein College of Medicine, Bronx, NY 10461 Abstract The structural factors responsible for the extraordinary rate enhancement (~1017) of the reaction catalyzed by orotidine 5′-monophosphate decarboxylase (OMPDC) have not been defined. Catalysis requires a conformational change that closes an active site loop and “clamps” the orotate base proximal to hydrogen-bonded networks that destabilize the substrate and stabilize the intermediate. In the OMPDC from Methanobacter thermoautotrophicus, a “remote” structurally conserved cluster of hydrophobic residues that includes Val 182 in the active site loop is assembled in the closed, catalytically active conformation. Substitution of these residues with Ala decreases kcat/Km with a minimal effect on kcat, providing evidence that the cluster stabilizes the closed conformation. The intrinsic binding energies of the 5′-phosphate group of OMP for the mutant enzymes are similar to that for the wild type, supporting this conclusion. Orotidine 5′-monophosphate decarboxylase (OMPDC1) is one of Nature’s best catalysts: the reaction occurs with a rate enhancement of ~1017 and a proficiency of ~1023 M (2). The reaction coordinate includes a vinyl carbanion intermediate (Scheme 1) (3, 4). Structures of OMPDCs complexed with UMP or 6-hydroxyUMP, a n intermediate analog (5-9), reveal hydrogen-bonded networks proximal to 1) C6 of the pyrimidine (Asp 70-Lys 72-Asp 75 in the OMPDC from Methanobacter thermoautotrophicus; MtOMPDC); and 2) O2, N3, and O4 of the base (Ser 127-Gln 185). These destabilize the substrate (9) and stabilize the intermediate, although the structural strategy for the latter is uncertain. The 5′-phosphate group binds in a conserved motif at the ends of the seventh and eighth β- strands of the (β/α)8-barrel structure, with interactions to backbone NH groups as well as the guanidinium group of a conserved Arg (Arg 203 in MtOMPDC). These interactions are important for catalysis: 1) kcat/Km for OMP exceeds that for orotidine by a factor of ~1011 for the OMPDC from Saccharomyces cerevisiae (ScOMPDC) (10); 2) kcat/Km for OMP †This work was supported by NIH Grants GM039754 to J.P.R. and GM065155 to J.A.G. *To whom correspondence should be addressed: J.A.G.: telephone, (217) 244-7414; fax: (217) 244-6538; j-gerlt@uiuc.edu.. SUPPORTING INFORMATION AVAILABLE Descriptions of the experimental procedures. This material is available free of charge via the Internet at http://pubs.acs.org. 1Abbreviations: OMP, orotidine 5′-monophosphate; OMPDC, OMP decarboxylase; MtOMPDC, OMPDC from Methanobacter thermoautotrophicus; ScOMPDC, OMPDC from Saccharomyces cerevisiae; EO, 1-(β-D-erythrofuranosyl)-orotic acid; HPi, phosphate dianion; IBE, intrinsic binding energy. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 December 26. Published in final edited form as: Biochemistry. 2010 May 4; 49(17): 3514–3516. doi:10.1021/bi100443a. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript exceeds that for 1-(β-D-erythrofuranosyl)orotic acid (EO; 5′-truncated OMP analog) by factors of 5.2 × 108 and 3.6 × 108 for ScOMPDC and MtOMPDC, respectively (11, 12); and 3 ) phosphite dianion ( H Pi) activates decarboxylation of EO by factors of 5.6 × 105 M−1 and 2.9 × 105 M−1 for ScOMPDC and MtOMPDC, respectively (11, 12). The “intrinsic binding energy” [IBE; (13)] of the 5′-phosphate/HPi 1) increases the affinity for the substrate (e.g., OMP vs. orotidine/EO); and 2) enables decarboxylation by juxtaposition of the substrate with the active site hydrogen-bonded networks (substrate destabilization and intermediate stabilization). How the IBE promotes catalysis is unknown but required to understand the structural basis for the rate enhancement. A loop located at the end of the seventh β-strand closes over the active site when OMP binds (Figure 1). Although the active site loops differ in both length and sequence in divergent OMPDCs (12), each includes a spatially conserved Gln (Gln 185 in MtOMPDC) hydrogen- bonded to both the 5′-phosphate and O2 of the pyrimidine as well as a conserved Ser at the end of the fifth β-strand that also is hydrogen-bonded to N3 of the pyrimidine (Ser 127). We characterized the importance of these “clamp” residues in ScOMPDC (Gln 215-Ser 154) using EO/HPi and confirmed that the 5′-phosphate/HPi “switch” is required to activate the enzyme (14). The structures of divergent OMPDCs reveal that OMP binding is always accompanied by a conformational change (Figure 1). The most obvious component is closure of the active site loop. However, the (β/α)8-barrel structure can be divided into two domains, one formed from the second, third, fourth, and fifth β-strands (where the hydrogen-bonded Asp 70-Lys 72-Asp 75 motif and Ser 127 are located) and the second from the sixth, seventh, eighth, and first β-strands (where the phosphate binding motif and the active site loop, including Gln 185, are located) (15). OMP binding reorients the domains, with the latter domain moving toward the former, enforcing the orotate carboxylate group to be juxtaposed vis a vis Asp 70- Lys 72-Asp 75 and, also, allowing formation of the Ser 127-Gln 185 “clamp”. Thus, the transition between the open and closed conformations is more complicated than “simple” hinge motion of the loop on the rigid framework of the (β/α)8-barrel structure. In this Report we identify “remote” residues involved in this conformational change and quantitate their importance in promoting and stabilizing the catalytically competent form of the enzyme. The active site loop of MtOMPDC, P180-G181-V182-G183-A184-Q185-G186-G187-D188, is disordered in the absence of substrate but ordered and closed in its presence (Figure 1). In the liganded structure two residues in the loop make contacts with the (β/α)8-barrel scaffold: 1) Gln 185 is hydrogen-bonded to Ser 127 (vide infra); and 2) Val 182 is embedded in a hydrophobic cluster also formed by Ile 199, Val 201, and Ile 218. The conservation of this hydrophobic cluster in all OMPDCs suggests its importance in a common catalytic strategy. We probed this strategy by mutagenesis of these hydrophobic residues. Ala substitutions for residues in the hydrophobic cluster cause substantial decreases in kcat/ Km for decarboxylation of OMP but have little effect on kcat (Table 1). We determined high resolution X-ray structures (≤ 1.4 Å) for each mutant in the presence of 6-hydroxyUMP (Figure 2). The liganded structures superimpose well with that of wild type, with only small differences observed at the sites of the substitutions (panel A). The active sites are identical to that of wild type (panel B), explaining the minimal impact on kcat. The mutated residues are remote from the active site (Figures 1 and 2). Thus, the effects of the substitutions on kcat/Km cannot be explained by altered direct interactions with the substrate. Instead, the effects can be explained by decreased stabilities of the closed conformation in which the substrate is destabilized (9) and the anionic intermediate is stabilized. A consistent model (Scheme 2) is that an equilibrium of open (Eo) and closed Wood et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (Ec) conformations exists in the absence or presence of substrate (KC’ or KC; KC’ << 1). The substitutions destabilize Ec (decrease KC and KC ’), so kcat/Km is decreased (the energy difference between Eo + S and [Ec•S]‡ is increased). However, the invariance of kcat establishes that the substitutions do not alter the reactivitiy of Ec•S (the energy difference between Ec•S and [Ec•S]‡). The energy to form Ec•S from Eo + S is achieved from 1) interactions of the 5′-phosphate group with its binding motif (IBE; vide infra), and 2) an increased concentration of OMP to form Ec•S (KC/KS or, equivalently, KC’/KS’, although the former is expected to be the relevant pathway). We also used the two part EO•HPi substrate. The values of kcat/Km for EO are decreased relative to that for wild type (Table 1); these can be explained by decreased populations of Ec (Scheme 2), assuming that EO, without a 5′-phosphate group, is unable to promote the transition from Eo to Ec. The values of kcat/Km for OMP and kcat/Km for EO allow calculation of the IBE for the 5′-phosphate group of OMP (Table 1). HPi activates the mutants as judged by the values of the third-order rate constant, (kcat/ Km)EO•HPi/Kd. The equivalent changes (ΔΔG‡) in kcat/Km for both OMP and EO and the third-order rate constant indicate that all three measure the effects of the substitutions on the values of KC. The values of the IBEs for the 5′-phosphate group for the mutants are the same as that for the wild type, establishing that decarboxylation in the Ec•S complex occurs with equivalent amounts of ground state destabilization (9) and transition state stabilization (as also reflected by the invariant values of kcat). The IBEs provide further support for the role of the “remote” hydrophobic cluster in stabilizing Ec relative to Eo but do not directly participating in catalysis. Our experiments implicate a structurally conserved hydrophobic cluster, Val 182, Ile 199, Val 201, and Ile 218 in MtOMPDC, in stabilizing the closed conformation required for catalysis. Its identification provides evidence that structural elements distal from the active site, in addition to the proximal active site loop that closes to “clamp” the substrate, are required for OMPDC’s extraordinary catalytic efficiency and proficiency. Supplementary Material Refer to Web version on PubMed Central for supplementary material. REFERENCES (1). Wood BM, Chan KK, Amyes TL, Richard JP, Gerlt JA. Biochemistry. 2009; 48:5510–5517. [PubMed: 19435313] (2). Radzicka A, Wolfenden R. Science. 1995; 267:90–93. [PubMed: 7809611] (3). Toth K, Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2007; 129:12946–12947. [PubMed: 17918849] (4). Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2008; 130:1574–1575. [PubMed: 18186641] (5). Miller BG, Hassell AM, Wolfenden R, Milburn MV, Short SA. Proc. Natl. Acad. Sci. U A. 2000; 97:2011–2016. (6). Appleby TC, Kinsland C, Begley TP, Ealick SE. Proc atl Acad Sci U S A. 2000; 97:2005–2010. (7). Harris P, Navarro Poulsen JC, Jensen KF, Larsen S. Biochemistry. 2000; 39:4217–4224. [PubMed: 10757968] (8). Wu N, Mo Y, Gao J, Pai EF. Proc. Natl. Acad. Sci. U S A. 2000; 97:2017–2022. [PubMed: 10681441] Wood et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (9). Chan KK, Wood BM, Fedorov AA, Fedorov EV, Imker HJ, Amyes TL, Richard JP, Almo SC, Gerlt JA. Biochemistry. 2009; 48:5518–5531. [PubMed: 19435314] (10). Sievers A, Wolfenden R. Bioorg. Chem. 2005; 33:45–52. [PubMed: 15668182] (11). Amyes TL, Richard JP, Tait JJ. J. Am. Chem. Soc. 2005; 127:15708–15709. [PubMed: 16277505] (12). Toth K, Amyes TL, Wood BM, Chan KK, Gerlt JA, Richard JP. Biochemistry. 2009; 48:8006– 8013. [PubMed: 19618917] (13). Jencks WP. Adv. Enzymol. Rel. Areas. Mol. Biol. 1975; 43:219–410. (14). Barnett SA, Amyes TL, Wood BM, Gerlt JA, Richard JP. Biochemistry. 2008; 47:7785–4487. [PubMed: 18598058] (15). Harris P, Poulsen JC, Jensen KF, Larsen S. J. Mol. Biol. 2002; 318:1019–1029. [PubMed: 12054799] (16). Go MK, Amyes TL, Richard JP. Biochemistry. 2009; 48:5769–5778. [PubMed: 19425580] Wood et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Superposition of the active site of wild type MtOMPDC in the absence (cyan; disordered loop depicted with the dotted line ) and presence ( grey ) of 6-hydroxyUMP. The carbons of 6-hydroxyUMP are highlighted in yellow; the carbons of Val 182 and Arg 203 in the liganded structure are highlighted in orange. Wood et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Superpositions of the 6-hydroxyUMP-liganded structures of wild type MtOMPDC and the single mutants in the hydrophobic cluster. Panel A, hydrophobic cluster; panel B, active sites. Wood et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. Wood et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 2. Wood et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Wood et al. Page 9 Table 1 Kinetic Constants for OMP, EO and EO/HPi and Intrinsic Binding Energies of the 5′-Phosphate Group of OMP at pH 7.1 and 25 °C. MtOMPDC kcat OMP s−1 kcat/Km OMP M−1 s−1 ΔΔG‡ kcal/mola kcat/Km EO M−1 s−1 ΔΔG‡ kcal/mola (kcat/Km)/K D EO•HPib M−2 s−1 ΔΔG‡ kcal/mola 5′-Phosphate IBEc kcal/mol Wild type 4.6 2.9 × 106 8.7 × 10−3 2500 11.6 d V182A 3.4 1.4 × 105 1.8 1.3 × 10−3 1.1 190 1.5 10.9 I199A 3.9 9.1 × 105 0.7 1.9 × 10−3 0.9 980 0.6 11.8 V201A 4.0 9.5 × 105 0.7 3.1 × 10−3 0.6 690 0.8 11.5 I218A 3.3 2.8 × 105 1.4 2.3 × 10−3 0.8 340 1.2 11.0 V182A/I199A 3.1 4.9 × 104 2.4 3.9 × 10−4 1.8 81 2.0 11.0 V182A/V201A 2.5 4.9 × 104 2.4 5.0 × 10−4 1.7 30 2.6 10.9 aCalculated from the ratio of the second-order or third-order rate constants for the wild type and mutant enzyme. bThird-order rate constant for reaction of EO/HPi. cTransition state stabilization by the 5′-phosphate group of OMP, calculated from the ratio of the values of kcat/Km for OMP and EO. IBEs for phosphite dianion can be calculated from the ratio of (kcat/ Km)/KD for EOHPi and kcat/Km for EO. Biochemistry. Author manuscript; available in PMC 2011 December 26.
3M45
Crystal structure of Ig1 domain of mouse SynCAM 2
N-Glycosylation at the SynCAM (Synaptic Cell Adhesion Molecule) Immunoglobulin Interface Modulates Synaptic Adhesion*□ S Received for publication,March 8, 2010, and in revised form, August 3, 2010 Published, JBC Papers in Press,August 25, 2010, DOI 10.1074/jbc.M110.120865 Adam I. Fogel‡1, Yue Li‡, Joanna Giza‡, Qing Wang‡2, TuKiet T. Lam§, Yorgo Modis‡, and Thomas Biederer‡3 From the ‡Department of Molecular Biophysics and Biochemistry and the §W. M. Keck Foundation Biotechnology Resource Laboratory, Yale University, New Haven, Connecticut 06520 Select adhesion molecules connect pre- and postsynaptic membranes and organize developing synapses. The regulation of these trans-synaptic interactions is an important neurobio- logical question. We have previously shown that the synaptic cell adhesion molecules (SynCAMs) 1 and 2 engage in homo- and heterophilic interactions and bridge the synaptic cleft to induce presynaptic terminals. Here, we demonstrate that site- specific N-glycosylation impacts the structure and function of adhesive SynCAM interactions. Through crystallographic anal- ysis of SynCAM 2, we identified within the adhesive interface of its Ig1 domain an N-glycan on residue Asn60. Structural model- ing of the corresponding SynCAM 1 Ig1 domain indicates that its glycosylation sites Asn70/Asn104 flank the binding interface of this domain. Mass spectrometric and mutational studies con- firm and characterize the modification of these three sites. These site-specific N-glycans affect SynCAM adhesion yet act in a differential manner. Although glycosylation of SynCAM 2 at Asn60 reduces adhesion, N-glycans at Asn70/Asn104 of SynCAM 1 increase its interactions. The modification of SynCAM 1 with sialic acids contributes to the glycan-dependent strengthening of its binding. Functionally, N-glycosylation promotes the trans- synaptic interactions of SynCAM 1 and is required for synapse induction. These results demonstrate that N-glycosylation of SynCAM proteins differentially affects their binding interface and implicate post-translational modification as a mechanism to regulate trans-synaptic adhesion. Synapses in the central nervous system are highly specialized sites of neuronal adhesion. They are morphologically defined by a presynaptic terminal filled with synaptic vesicles, an apposed postsynaptic specialization that contains neurotrans- mitter receptors, and a synaptic cleft of 20-nm width that sep- arates pre- and postsynaptic sites (1, 2). This cleft is filled with proteinaceous material (3). The proteins spanning the synaptic cleft not only tie pre- and postsynaptic membranes together. Select synaptic surface mol- ecules can also instruct the organization of nascent synapses (4–6). This was first demonstrated for neuroligins, postsynap- tic membrane proteins that bind the presynaptic neurexins (7–9). Adhesion molecules of the Ig superfamily and proteins containing leucine-rich repeats additionally mediate synaptic differentiation (10–14). Similarly, receptor tyrosine kinases, including EphB receptors, instruct synaptogenesis through trans-synaptic signaling (15, 16). These synapse-inducing pro- teins act in conjunction with N-cadherins that set the pace of synaptic maturation (17, 18). Among these synapse-organizing proteins, SynCAMs4 form a family of four Ig superfamily members that are single-span- ning membrane proteins with three extracellular Ig-like domains (19). SynCAMs, also known as nectin-like molecules, are prominently expressed throughout the brain and are enriched in synaptic plasma membranes (10, 20). They are N-glycosylated proteins, consistent with the presence of multiple predicted N-glycosylation sites in their extracellu- lar Ig domains (19, 20). SynCAMs 1, 2, and 3 can bind them- selves through homophilic binding, but SynCAMs 1/2 and 3/4 preferentially engage each other in specific heterophilic interactions (20, 21). SynCAM 1 and 2 form a trans-synaptic adhesion complex and promote the number of functional excitatory synapses (20). Consistent with the critical importance of synapse organiza- tion for brain functions, trans-synaptic adhesion molecules need to be regulated. Mechanisms include the control of their sorting by intracellular interactions as shown for neuroligins (22) and alternative splicing within sequences encoding extra- cellular domains, which specifies neurexin-neuroligin inter- actions and has been analyzed at atomic resolution (23–26). Post-translational modifications regulating synaptic adhesion molecules are less understood, but a negative effect of glycosy- lation on neuroligin 1 binding has been reported (27). * Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth Grant R01 DA018928 (to T. B.). This work was also supported by a Bur- roughs Wellcome Investigator in the Pathogenesis of Infectious Disease grant (to Y. M.), National Institutes of Health Predoctoral Program in Cellu- lar and Molecular Biology Grant T32 GM007223, and by National Institutes of Health Grant P30 DA018343 from the National Institute on Drug Abuse. Use of the National Synchrotron Light Source is supported by the Offices of Biological and of Basic Energy Sciences of the U.S. Department of Energy. □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental text, Table S1, and Figs. S1–S3. 1 Present address: NINDS, National Institutes of Health, 35 Convent Dr., Bethesda, MD 20892. 2 Present address: Program in Neurobiology and Behavior, Columbia Univer- sity, New York, NY 10032. 3 To whom correspondence should be addressed: 333 Cedar St., New Haven CT06520.Tel.:203-785-5465;Fax:203-785-6404;E-mail:thomas.biederer@ yale.edu. 4 The abbreviations used are: SynCAM, synaptic cell adhesion molecule; GPI, glycosylphosphatidylinositol; PNGase F, peptide:N-glycosidase F; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid; Pn, postnatal day n; FT-ICR, Fourier transform ion cyclotron resonance. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 45, pp. 34864–34874, November 5, 2010 © 2010 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. 34864 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 45•NOVEMBER 5, 2010 Here, we address the molecular properties that underlie and regulate SynCAM adhesion and function. Our crystallographic, mass spectrometric, and biochemical analyses of SynCAM 1 and 2 demonstrate that they carry N-glycans adjacent to and within the first Ig domain that provides their extracellular bind- ing interface. Unexpectedly, the glycosylation of these two SynCAM family members serves different roles. Although N-glycans within the Ig1 binding interface of SynCAM 2 reduce its binding, glycosylation at the SynCAM 1 Ig1 domain pro- motes its adhesion. Consequently, the ability to N-glycosylate SynCAM 1 increases its trans-synaptic interactions and syn- apse inducing activity. Together, these results identify glycosy- lation as a novel mechanism for positively and negatively regu- lating the trans-synaptic SynCAM adhesion complex in the brain. EXPERIMENTAL PROCEDURES Antibodies—Specific antibodies against SynCAM 1 (YUC8) and SynCAM 2 (YU524) were described previously (20). For immunostaining of SynCAMs 1–3, we utilized a pleio-SynCAM antibody (T2412) raised against the SynCAM 1 C terminus that equally recognizes this conserved sequence in full-length SynCAM 2 and 3 (20) but not the GPI-anchored SynCAM con- structs used in this study. Monoclonal antibodies to synapto- physin (7.2) and GDI (81.2) were obtained from Synaptic Sys- tems (Go¨ttingen, Germany), monoclonal antibodies to CASK were from Millipore (Billerica, MA), and monoclonal antibod- ies to FLAG (M2) were from Sigma. Monoclonal antibodies to SV2 (developed by Kathleen Buckley) were obtained from the Developmental Studies Hybridoma Bank maintained by the University of Iowa. Expression Vectors—A pCMV5 GPI vector backbone was generated by amplifying the GPI targeting sequence from GPI- VAMP2 (a gift from Dr. James Rothman, Department of Cell Biology, Yale University) with 5 SalI and 3 BamHI sites and subcloning into pCMV5. pCMV5 FLAG-GPI was generated analogously, adding the FLAG epitope DYKDDDDK N-termi- nal of the GPI anchoring sequence. Full-length SynCAM extra- cellular sequences or sequences lacking select Ig domains were amplified from pCMV IG9 vectors described previously (10) and subcloned into pCMV5 GPI or pCMV5 FLAG-GPI vectors. Point mutants were generated using site-directed PCR mu- tagenesis (QuikChange; Stratagene, La Jolla, CA). Expression vectors for full-length SynCAM 1 carrying an extracellular FLAG epitope N-terminal of the transmembrane region and for extracellular SynCAM sequences fused to a thrombin cleavage and IgG1-Fc sequence were described previously (20). GFP was expressed from pCAG GFP, a gift from Dr. Nenad Sestan (Department of Neurobiology, Yale University). Cell Culture—COS7 cells were maintained using standard procedures and transfected with FuGENE 6 (Roche Applied Science) for transient expression. HEK293 cell lines stably expressing the SynCAM 1 or 2 extracellular domains were selected in the presence of geneticin (American Bioanalytical, Natick, MA) after transfection of HEK293 cells with the vector pcDNA3.1 SynCAM 1 extracellular domain IgG1 linearized with BglII. Dissociated cultures of hippocampal neurons were prepared as described (28). Expression and Purification of SynCAM 1 and SynCAM 2 Extracellular Sequences—The extracellular sequences of mouse SynCAM 1 or SynCAM 2 were purified as described previously (20). Briefly, HEK293 cell lines stably expressing the full-length SynCAM 1 or SynCAM 2 extracellular sequences fused to IgG1-Fc were grown in DMEM low glucose medium supple- mented with 5% FBS and 50 mg/ml geneticin. The culture supernatant was collected and replaced with fresh medium every 72 h. 2 liters of culture supernatant were filtered, concen- trated, and then applied to 2 ml of protein A-agarose resin (Invitrogen) equilibrated in buffer A (50 mM Tris, pH 8.0, 150 mM NaCl, 2 mM -mercaptoethanol). SynCAM extracellular domains were eluted from the resin by the addition of bovine -thrombin protease (Hematologic Technologies, Essex Junc- tion, VT) in a 1:300 molar ratio at 16 °C overnight to cleave the resin-bound N-terminal human IgG1-Fc tag. The proteins were further purified by size exclusion chromatography on a Super- dex 200 column (GE Healthcare) in buffer A. Isothermal Titration Calorimetry—The binding of SynCAM 1 to SynCAM 2 was studied by isothermal titration calorimetry in 20 mM Tris, pH 8.0, 50 mM NaCl, at 25 °C, using an iTC200 system (MicroCal, Piscataway, NJ). The sample cell contained the purified SynCAM 2 extracellular domain protein at 5 M, and the syringe contained the SynCAM 1 extracellular domain at 50 M, with the IgG1-Fc tags cleaved off. Typically, one initial injection of 1.5 l and 19 serial injections of 2.0 l of SynCAM 1 were performed at 180-s intervals. The stirring speed was maintained at 1000 rpm, and the reference power was kept constant at 5 cal/s. The heat associated with each injection of SynCAM 1 was integrated and plotted against the molar ratio of SynCAM 1 to SynCAM 2. Thermodynamic parameters were extracted from a curve fit to the data using the Origin 7.0 soft- ware provided by MicroCal. The experiments were performed in triplicate with excellent reproducibility (10% variation in thermodynamic parameters). Preparation of SynCAM Extracellular Domain Complexes— SynCAM 2 was first expressed and applied to protein A-agarose resin as described above. The protein was then eluted from the resin with 0.2 M glycine, pH 3.0, and dialyzed into 20 mM Tris, pH 8.0, 50 mM NaCl, 2 mM -mercaptoethanol. A 2-fold molar excess of the purified SynCAM 1 extracellular domain with the IgG1-Fc tag cleaved off was then added to obtain heteromeric SynCAM 1-SynCAM 2 complexes in addition to the homo- meric SynCAM 2 complexes present in this preparation. The resulting mixture was incubated at 16 °C for 3 h and applied to 2 ml of protein A-agarose resin in buffer A. The SynCAM com- plexes were eluted from the resin with thrombin protease (1:300 molar ratio, 16 °C overnight) and were purified on a Superdex 200 column in buffer A. To aid in the subsequent crystallization step, the samples were partially deglycosy- lated under native conditions with PNGase F and neuramin- idase (New England Biolabs) at 37 °C for 48 h and separated from the endoglycosidases and cleaved glycans on a Super- dex 200 column in 50 mM Tris, pH 8.0, 50 mM NaCl, 2 mM -mercaptoethanol. Crystallization of the Ig1 Domain of SynCAM 2—Crystals were grown at 20 °C using the hanging drop vapor diffusion technique. The preparation of the SynCAM extracellular Glycans Modulate SynCAM Adhesion NOVEMBER 5, 2010•VOLUME 285•NUMBER 45 JOURNAL OF BIOLOGICAL CHEMISTRY 34865 domain complex was concentrated to 9.0 mg/ml in 20 mM Tris, pH 8.0, 50 mM NaCl, 2 mM -mercaptoethanol. The protein solution was mixed with an equal volume of well solution (0.1 M HEPES, pH 6.5–7.0, 21% PEG5000 monomethyl ether). Irregu- lar bulky crystals grew after 3 months. The crystals were tri- clinic (space group P1) with unit cell dimensions a  42.8 Å, b  50.5 Å, c  79.9 Å,   75.9°,   77.0°, and   65.2°. With four molecules/asymmetric unit, the Matthews co-efficient VM is 3.36 Å3 Da1, which corresponds to a solvent content of 63.4% (29). For data collection, the crystals were transferred to a cryoprotectant containing 0.1 M HEPES, pH 6.8, 21% PEG5000 monomethyl ether and 18% glycerol (v/v) and immediately fro- zen in liquid nitrogen. Data Collection and Processing—Crystallographic data were collected at 100 K on Beamline X29A of the National Synchro- tron Light Source at Brookhaven National Laboratory. The data were indexed, integrated, and scaled using the HKL2000 pro- gram suite (30). The data collection statistics are summarized in supplemental Table S1. Structure Determination and Refinement—The crystal struc- ture of SynCAM 2 Ig1 was determined by molecular replace- ment using a monomer of human SynCAM 3 (nectin-like mol- ecule 1) Ig1, Protein Data Bank Code 1Z9M (31), as the search model in the program PHASER 2.1 (32). The details of structure determination and refinement are described in the supplemen- tal materials. Glycopeptide Mapping—The purified SynCAM 1 extracellu- lar domain with the IgG1-Fc cleaved off was digested with the combined endoproteinases Lys C and trypsin. The samples were C18 RP ZipTip-cleaned and desalted prior to collecting MS data on a 9.4T Apex Qe FT-ICR MS instrument. Eluted peptides were directly infused into the mass spectrometer via nanoelectrospray at 250 nL/min into an Apollo II dual ion fun- nel ESI source. The spray shield voltage was set at 3500, and a 4000-V potential was applied on the glass capillary end cap. The instrument (running Compass Software with APEX control acquisition component (v.1.2) is set up to acquire single free induction decay signal (512,000) data with a mass range (m/z) from 450 to 2000. Enrichment of glycopeptides was confirmed using albumin, ovalbumin, -casein, and RNase B as standard proteins. All of the data were processed utilizing DA analysis software v. 3.4, online GlycoMod (Expasy), and MASCOT search engine. Glycan Profiling—Glycosidase treatment of the SynCAM 1 extracellular sequence with the IgG1-Fc cleaved off was per- formed using the glycosidases PNGase F or endoglycosidase H. Cleaved glycans were enriched with a Carbograph column, fol- lowed by C18 RP ZipTip prior to direct infusion into a 9.4T FT-ICR MS instrument. The electrospray source was config- ured with a capillary (low flow) sprayer optimized for positive mode. Ions were detected in the 450–2500 m/z with 512K data points/MS scan. All of the data were processed utilizing DA analysis software v.3.4, online GlycoMod (Expasy). Tissue Preparation—Samples from rat brain regions were prepared by rapid homogenization in 8 M urea. Protein concen- trations were determined using the Pierce BCA assay. ProteinDeglycosylation—Enzymaticdeglycosylationwasper- formed using neuraminidase (sialidase; Roche Applied Science) and PNGase F (New England Biolabs) according to the manu- facturers’ instructions. Affinity Chromatography—The SynCAM 1 extracellular do- main was immobilized on protein A beads to serve as affinity matrix. Rat forebrain proteins were solubilized with 1% CHAPS (Roche Applied Science), and affinity chromatography and quantitative immunoblotting were performed as described (10, 20). Surface Expression Control—COS7 cells expressing SynCAM constructs tagged with an extracellular FLAG epitope were fixed, labeled with anti-FLAG antibodies to detect surface-ex- pressed epitopes (antibody M2; 1:1000), washed, and then per- meabilized using 0.1% Triton X-100 to perform immuno- staining for total SynCAM protein (antibody T2412; 1:1000). The images were acquired on a Zeiss LSM 510 META laser scanning confocal microscope. Cell Overlay Experiments—Cell overlay assays were per- formed as described (21). Briefly, COS7 cells were co-trans- fected with expression vectors encoding extracellularly FLAG- tagged SynCAM constructs and soluble GFP or GFP alone as negative control. After 2 days, live cells were overlaid for 20 min at 25 °C with the purified SynCAM 1 extracellular domain at 2 g/ml or the SynCAM 2 extracellular domain at 10 g/ml. The IgG1-Fc fusion tag of these overlaid fusion proteins was directly detected by including Alexa 546-conjugated protein A (6 g/ml; Invitrogen) in this step. Surface-expressed SynCAM proteins were detected in these live cells by simultaneously add- ing anti-FLAG (antibody M2; 1:1000) and secondary anti- mouse antibodies conjugated to Alexa 488 (Invitrogen) (1:1000). The medium was then replaced with DMEM without phenol red, and the cells were immediately imaged with a Hamamatsu Orca camera attached to a Nikon Eclipse TE2000-U microscope. The signal of the secondary Alexa 488 antibody detecting anti-FLAG antibodies was used to define regions of interest, within which the fluorescence from the Alexa 546-conjugated protein A was measured and normalized to the anti-FLAG signal. Signals were quantified using a custom Matlab (MathWorks) script that is available upon request. Mixed Co-culture Assay for Synapse Induction—Co-culture assays were performed as described (28). Briefly, COS7 cells co-expressing GPI-anchored SynCAM 1 constructs and soluble GFP or GFP alone as negative control were seeded atop neurons at 6–7 days in vitro. At 8–9 days in vitro, these mixed co-cul- tures were fixed and immunostained for the presynaptic marker SV2 and for neuronal SynCAM proteins with the anti- body T2412. The images were acquired on a Zeiss LSM 510 META laser scanning confocal microscope. The surface area of COS7 cells immunopositive for neuronal SynCAMs and SV2 was quantified using a Matlab script that is available upon request. The images were collected blind to the synaptic marker channel. Miscellaneous Procedures—Sequence similarities were ana- lyzed using the T-Coffee method (33). Amino acid numbers refer to the position in the protein including the signal peptide. Statistical analyses were performed using the two-tailed t test, with statistical errors corresponding to the standard errors of mean. Glycans Modulate SynCAM Adhesion 34866 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 45•NOVEMBER 5, 2010 RESULTS High Affinity Binding of SynCAM 1 to SynCAM 2 Requires the Ig1 Domain—To define the molecular properties of SynCAM interactions, we measured the affinity between the SynCAM 1 and SynCAM 2 extracellular sequences by isothermal titration calorimetry. The resulting isotherm was consistent with a sin- gle binding interface between the two proteins in a 1:1 complex, with a tight apparent dissociation constant (Kd) of 78 nM (Fig. 1A). This Kd is very similar to the neuroligin 1/neurexin 1 interaction (34). We next mapped this single bind- ing interface within the three extra- cellular Ig-like domains. Utilizing constructs comprised of subsets of SynCAM 1 Ig domains, we mea- sured their adhesive interaction with the SynCAM 2 extracellular domain using a cell overlay approach (Fig. 1B). These experiments extended previous affinity chromatography studies (20) and allowed us to ana- lyze SynCAM interactions as they occur on the cell surface. We expressed an array of SynCAM 1 Ig constructs carrying an extracellular FLAG epitope and labeled the expressed proteins in live COS7 cells with anti-FLAG antibodies. To quantify adhesive binding, we over- laid these cells with the soluble SynCAM 2 extracellular sequence fused to IgG1-Fc and detected retained protein using fluorophore- labeled protein A. This signal was divided by the fluorescence mea- sured with anti-FLAG antibodies, which normalized for each cell the extent of SynCAM 2 retention to the amount of its surface-expressed SynCAM 1. All of the SynCAM 1 Ig constructs were properly N-glyco- sylated and sorted to the plasma membrane, with the Ig1 domain carrying N-glycans to the highest apparent extent (supplemental Fig. S1). These experiments showed that the tandem Ig1  2 domains of SynCAM 1 were sufficient for strong binding (Fig. 1, B and C). Moreover, the SynCAM 1 Ig1 domain was required for binding because the SynCAM 1 Ig2  3 construct did not retain SynCAM 2. The SynCAM 1 Ig1 domain alone was sufficient for SynCAM 2 binding, albeit at a lower strength. This reduced interaction of the SynCAM 1 Ig1 domain in the absence of the Ig2 and Ig3 domains is possibly due to a role of these domains in conferring a steric orientation to SynCAM 1 Ig1 that is favor- able for its interaction with SynCAM 2. Together, these results show that the first Ig domain of SynCAM 1 provides its binding interface. Crystal Structure of the SynCAM 2 Ig1 Domain—Aiming to characterize the extracellular SynCAM interactions at atomic resolution, we performed crystallization trials of the SynCAM 1/2 extracellular domain complex. The crystal structure, which was refined at 2.21 Å resolution (r  0.197, Rfree  0.245), FIGURE1.ThefirstIgdomainmediatestightheterophilicbindingofSynCAM1toSynCAM2.A,isothermal titration calorimetry analysis of the binding of the SynCAM 1 extracellular sequence to SynCAM 2. Left panel, enthalpicheatreleasedat25 °CduringthetitrationoftheSynCAM1extracellularsequenceintotheisothermal titration calorimetry cell containing the SynCAM 2 extracellular sequence. Right panel, integrated binding isotherms of the titration and best fit to a single-site model. The best fit yielded a dissociation constant Kd  78.0 nM, enthalphy H  9.1 kcal/mol, and binding stoichiometry n  1. B, analysis of adhesive SynCAM binding by cell overlay. COS7 cells expressed full-length SynCAM 1 or variants containing the indicated Ig domains, with an extracellular FLAG epitope inserted proximal to the transmembrane region. These surface- expressed proteins were detected in live cells by the addition of anti-FLAG antibodies and secondary antibod- ies conjugated to Alexa 488 (top row, green in the merge). The cells were simultaneously overlaid with the SynCAM 2 extracellular domain fused to IgG1-Fc together with protein A conjugated to Alexa 546 (second row, red in the merge) to label the retained protein. The first Ig domain of SynCAM 1 is required for adhesive binding to SynCAM 2 as depicted in the model below. C, quantification of the results in B. The results are expressed as a protein A signal detecting retained SynCAM 2 normalized to the signal from the indicated anti-FLAG labeled SynCAM 1 constructs expressed on COS7 cells. *, p  0.05; **, p  0.01; ***, p  0.001. Glycans Modulate SynCAM Adhesion NOVEMBER 5, 2010•VOLUME 285•NUMBER 45 JOURNAL OF BIOLOGICAL CHEMISTRY 34867 showed that the crystals contained only the Ig1 domain of SynCAM 2. This was consistent with the presence of a major protein band at 17 kDa in these crystals (data not shown), cor- responding to the size of one Ig domain. Upon closer examina- tion of the drop that produced the crystals, fungal growth was observed. This suggested that secreted fungal proteases may have cleaved the SynCAM 1/2 extracellular domain complex, allowing SynCAM 2 Ig1 to crystallize by itself. Several other proteins have been crystallized as a result of either intentional proteolytic cleavage or serendipitous cleavage by secreted fun- gal proteases (35, 36). The crystal structure of the SynCAM 2 Ig1 monomer (residues 35–131) showed that it adopts an Ig- like fold of the variable type (37) as predicted by sequence analysis (19), comprising two -sheets with nine antiparallel -strands (denoted A to G; Fig. 2A). Hydrophobic interac- tions between the two sheets form the core of the domain. A disulfide bridge between Cys53 and Cys113 links -strands B and G, further sta- bilizing the domain. Two N-linked N-acetylglucosamine residues are visible in the structure on Asn40 and Asn60, respectively. The remainder of these N-glycans had been re- moved during sample preparation to aid crystallization (see “Experi- mental Procedures”). Interestingly, SynCAM 2 Ig1 forms homodimers in the crystals, with each asymmetric unit contain- ing two dimers. The SynCAM 2 Ig1 homodimer has approximate dimensions of 60  42  33 Å (Fig. 2A). Dimer formation buries a total of 698 Å2 (11.5%) of solvent-accessi- ble surface/monomer. The N and C termini of each subunit in the dimer are antiparallel, indicating that the dimer corresponds to the trans-ad- hesion complex. The dimer inter- face is mostly hydrophobic (35% of the residues are nonpolar) and closely resembles the dimer inter- face of the Ig1 domain of SynCAM 3, also known as nectin-like mole- cule 1 (31), which is its closest struc- tural homolog. SynCAM 3 also par- ticipates in cell adhesion (20, 38). The Ig1 domains of SynCAM 2 and SynCAM 3 have high sequence identity (63%) and structural simi- larity (root mean square deviation, 0.7 Å over 96 equivalent C atoms). However, although the SynCAM 3 structure lacks glycans, in SynCAM 2 the N-acetylglucosamine on Asn60 forms a weak intersubunit contact in the trans-dimer interface of SynCAM 2 Ig1 (Fig. 2, A and B). Specifically, the carbonyl oxygen atom in the acetyl moiety of the first residue of the glycan is within 3.5 Å of the N atom of the Arg82 side chain in the other monomer and of a structured water molecule located in the dimer interface. The location of Asn60 at the dimer inter- face leaves little room for a bulky glycan, however, suggesting that full glycosylation at Asn60 may interfere with adhesive dimer formation. Homology Model of the SynCAM 1/2 Ig1 trans-Heterodimer— Like SynCAM 2, SynCAM 1 engages in both homo- and het- erophilic adhesion complexes (20, 21). The high sequence C N N A B C C’ D E F G a1 B C A C’ D E F G a1 A A B B C C C’ C’ D D E E F F G G a1 a1 N60 R82 N60 R82 3.5 Å 3.5 Å N60 N60 F92 F92 K96 K96 N60 N60 N60 N60 N40 N40 N40 N40 C 90o N N C C FIGURE 2. Structure of the SynCAM Ig1 domain interface. A, crystallographic results show that the SynCAM 2 Ig1 domain forms a dimer, characterized by mostly hydrophobic interactions across a noncrystallographic 2-fold axis, as shown in this ribbon representation. The N and C termini are marked. The N-acetylglucosamine residues on Asn40 and Asn60 are marked as spheres. B, stereodiagram of the SynCAM 2 Ig1 trans-homodimer interface, with the hydrogen bonding interactions of the N-acetylglucosamine on Asn60 shown as dashed lines. Side chains in the interfaces are shown in stick representation. Water molecules are highlighted as red spheres. C, theoretical model of the trans-heterodimer interface of SynCAM 1 Ig1/SynCAM 2 Ig1 shown as stereodia- gram. Residues in the interface are depicted in stick representation, with SynCAM 1 and SynCAM 2 in green and magenta, respectively. Interfaces are displayed in the same orientation as in B. The N-linked glycan on Asn60 of SynCAM 2 participates in the SynCAM 1 Ig1/SynCAM 2 Ig1 trans-heterodimer interface. Glycans Modulate SynCAM Adhesion 34868 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 45•NOVEMBER 5, 2010 identity of 44% between the Ig1 domains of SynCAM 1 and SynCAM 2 enabled us to build a homology model for SynCAM 1 Ig1 based on our crystal structure of SynCAM 2 Ig1. Our model predicts that the trans-dimeric interface of the SynCAM 1 Ig1 homodimer is more hydrophobic than that of SynCAM 2 Ig1 (supplemental Fig. S2A). The residues that form additional hydrophobic contacts in the SynCAM 1 model are Val76, Phe92, and Pro94. Although capable of homophilic binding, SynCAMs pref- erentially assemble into specific heterophilic complexes, and SynCAM 1 strongly binds SynCAM 2 (20, 21, 38–40). To better understand dimerization specificities, we modeled SynCAM 1 Ig1/SynCAM 2 Ig1 trans-heterodimers using the homodimeric crystal structure of SynCAM 2 Ig1 as template. Interestingly, a glycan-mediated contact occurs in the het- erodimer model between the N-linked glycan on Asn60 of SynCAM 2 and the side chain of Lys96 of SynCAM 1 (Fig. 2C). The Asn60 site of SynCAM 2 is not conserved in SynCAM 1, and this glycan may contribute to regulating the heterophilic binding of SynCAM 2 to SynCAM 1. Conversely, two N-glycosylation sites of the Ig1 domain of SynCAM 1, Asn70 and Asn104, are located on one face of the Ig1 domain, in the loop between strands B and C and in the middle of strand E, respec- tively (supplemental Fig. S2B). Residues 70 and 104 are both 20 Å from the trans-dimer interface and face away from the interface. N-Glycosylation of the SynCAM 1 Ig1 Domain—These crystallo- graphic results map glycans to dif- ferent surfaces of the Ig1 domain in SynCAM 1 and SynCAM 2. To examine SynCAM 1 N-glycosyla- tion, we performed a mass spec- trometry analysis of the SynCAM 1 extracellular sequence purified from HEK293 cells, which glycosy- late SynCAM 1 to the same appar- ent extent as found in brain (20). Using several different enzymes or combinations thereof, we observed a very high number of extracellular SynCAM 1 peptides/glycopeptides, which provide 70% sequence cov- erage of the protein. Glycopeptide mapping identified the asparagines Asn70 and Asn104 in the first Ig1 domain as potential N-glycosylation sites based on FT-ICR high mass accuracy and GlycoMod prediction (41) (Fig. 3). The deconvoluted mass list of each spectrum was entered into GlycoMod to predict possible glycosylations sites along with their potential glycan composition based on mass accuracy and the consensus sequence for N-glycosylation (42). The GlycoMod output identifies the Asn104 site as glycosylated (in the CNBr  trypsin and LysC  trypsin digest conditions; Fig. 3, B and C), with several glycopeptides showing sialylated glycan (NeuAc) modifications at that site. N-Glycosylation at Asn70 was pre- dicted from the mass list of SynCAM 1 digested with CNBr  trypsin (Fig. 3B). To obtain a profile of these N-glycan struc- tures, we subjected SynCAM 1 to PNGase F or endoglycosidase H to cleave the glycans. Glycan masses observed by FT-ICR suggested the presence of Hex, HexNac, and NeuAc carbohy- drates as predicted by GlycoMod (data not shown). Complex Modification of SynCAM 1 and SynCAM 2 in the Brain—SynCAM 1 and 2 are heavily glycosylated in the adult brain (20). To obtain insight into the extent of post-transla- tional SynCAM modification during early postnatal develop- ment, when most synapses form, we analyzed SynCAM 1 and 2 in several rat brain regions (Fig. 4). SynCAM modification was examined by immunoblotting prior to, during, and subsequent FIGURE 3. Glycopeptide mapping and glycan profiling of the SynCAM 1 extracellular sequence. Broadband FT-ICR MS mass spectrum of four different enzymatic digestion of the purified, glycosylated SynCAM 1 extracellular sequence. 10 g of SynCAM 1 was utilized for each digestion with CNBr (A), CNBr  trypsin (B), Lys C  trypsin (C), and protease type XIII (D). The asterisks indicate potential glycosylation sites based on exact mass measurements and GlycoMod prediction (41). The inset in B shows an enlarged region illustrating a predicted glycopeptide at m/z 1606.413 (3) that corresponds to a modification at the Asn70 position. Note that internal calibrations were utilized to obtain mass accuracy of 5 ppm. The inset in C shows an enlarged region with a glycopeptide that corresponds to the modified Asn104 residue of SynCAM 1 at m/z 1341.616 (3). Glycans Modulate SynCAM Adhesion NOVEMBER 5, 2010•VOLUME 285•NUMBER 45 JOURNAL OF BIOLOGICAL CHEMISTRY 34869 to the peak of synapse formation in the rodent brain at postna- tal day 4 (P4), P16, and P30, respectively (43). At all stages and in all regions, the apparent molecular masses of SynCAM 1 and 2 proteins were notably higher than the 41–45 kDa predicted from their open reading frames (19). Interestingly, SynCAM 1 modifications changed during development. At P4, it was expressed as diverse species that ranged from 90 to 115 kDa. As development progressed, SynCAM 1 was detected both as an apparently uniform high molecular mass species of 100 kDa and as multiple low molecular mass species of 70–85 kDa. These changes in the modification of SynCAM 1 were accompanied by a shift of its predominant expression from hindbrain to fore- brain. This is consistent with roles of SynCAM 1 in synapse formation, which progresses during brain development from the hindbrain to the forebrain. Its binding partner SynCAM 2 also followed a developmental expression increase toward fore- brain, yet SynCAM 2 was expressed as the same diverse species at 62–76 kDa throughout. Other N-glycosylated proteins such as synaptophysin also did not exhibit changes in their modifi- cation (Fig. 4), consistent with the developmentally indepen- dent glycosylation of other neuronal membrane proteins. Modification of SynCAM 2 Ig1 at Asn60 Reduces Adhesion— To perform a biochemical analysis of SynCAM 2 glycosylation, we changed the asparagine at position 60 to glutamine, choos- ing this substitution because it prevents N-glycosylation with- out altering immunoglobulin folds (44). Consistent with the conservative nature of this mutation, all Asn 3 Gln glycosyla- tion mutants used in this study were sorted to the cell surface, indicating proper folding (see below, Fig. 5, B and D, and sup- plemental Fig. S3B). Furthermore, the slightly increased bulk of the glutamine residue in the N60Q SynCAM 2 mutant can be expected to be easily accommodated in the structures of its homodimer as well as the heterodimer with SynCAM 1. To focus our analysis on extracel- lular interactions, we developed a GPI-anchored SynCAM 2 construct that tethered its extracellular sequence to the outer leaflet of the plasma membrane. This construct maintained a complex glycosylation pattern comparable with that seen for SynCAM 2 expressed in brain (Fig. 5A, lanes 1 and 2). The GPI construct of the SynCAM 2 N60Q mutant, however, lacked the N-gly- cosylated wild-type fractions above 55 kDa, consistent with selectively reduced glycosylation (Fig. 5A, lanes 3 and 4). We next analyzed the role of modifications at Asn60 for the ad- hesive interactions of SynCAM 2 (Fig. 5, B–E). Using a cell overlay approach with soluble proteins, we expressed GPI-anchored SynCAM 2 carrying a FLAG epitope in COS7 cells while overlaying the cells with the soluble extracellular sequence of SynCAM 2. As described above, the COS7 cell expressed protein was labeled with anti-FLAG antibodies and the overlaid soluble protein with protein A. Notably, the absence of a glycan at amino acid 60 of SynCAM 2 strongly increased its interaction with the overlaid extracellular sequence of SynCAM 2, more than doubling its homophilic retention by 125 31% (Fig. 5, B and C). Similarly, the N60Q mutation increased the hetero- philic binding of SynCAM 2 to overlaid SynCAM 1 by 61 10% (Fig. 5, D and E). N-Glycosylation at Asn60 of SynCAM 2 Ig1 therefore restricts its adhesive binding to both SynCAM 2 and SynCAM 1, possibly because of steric hindrance of N-glycans or charge repulsion within the binding interface. Positive Modulation of SynCAM 1 Adhesion by N-Linked Modification of Its Ig1 Domain—To address whether N-glyco- sylation within the first Ig domain of SynCAM 1 similarly reg- ulates its adhesion, we generated a SynCAM 1 N70Q,N104Q double mutant. These two sites were selected because our structural models predicted them to flank the SynCAM 1 dimer interface (supplemental Fig. S2B) and because our mass spec- trometry data suggested that they were N-glycosylated (Fig. 3). The SynCAM 1 N70Q,N104Q mutant migrated in immunob- lots at a lower apparent molecular mass, consistent with reduced N-glycosylation, and was correctly sorted to the plasma membrane (supplemental Fig. S3). A SynCAM 1 N70Q,N104Q,N116Q triple mutant lacking all predicted N-glycosylation sites in the Ig1 domain could not be analyzed because it was not properly sorted to the cell surface (data not shown). Live cell overlay assays with these GPI-anchored constructs showed that the lack of glycans at the Asn70/Asn104 sites of the Ig1 domain reduced its homophilic binding to wild-type SynCAM 1 by 51 16% (Fig. 6, A and B). Similarly, the N70Q,N104Q mutations decreased the heterophilic binding of FIGURE 4. SynCAM post-translational modifications are regionally and developmentally regulated in the brain. The indicated brain regions were dissected from rats at P4, P16, or P30. Equal protein amounts of 30 g were analyzed by immunoblotting using the antibodies shown. SynCAM 1 and 2 exhibited distinct expres- sion and modification patterns as described in the text. The N-glycosylated synaptic protein synaptophysin and the scaffolding molecule CASK served as loading controls. The asterisks mark nonspecific bands. Glycans Modulate SynCAM Adhesion 34870 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 45•NOVEMBER 5, 2010 SynCAM 1 to wild-type SynCAM 2 by 30 7%. This inhibitory effect of the SynCAM 1 Ig1 N70Q,N104Q mutation on its adhesive interactions contrasted with the increased binding of the SynCAM 2 Ig1 N60Q mutant. We additionally per- formed affinity chromatographies to analyze the effect of the SynCAM 1 N70Q,N104Q mutation on the retention of Syn- CAM 2 from brain (Fig. 6, C and D). Our results show that the loss of these two N-glycosylation sites reduces hetero- philic binding to SynCAM 2 by 30 3%, in agreement with our cell overlay data. FIGURE 5. SynCAM 2 glycosylation within the Ig1 interface at Asn60 reduces adhesive binding. A, immunoblot analysis of the GPI-anchored SynCAM 2 extracellular sequence and its N60Q mutant expressed in COS7 cells. Lack of the Asn60 N-glycosylation site resulted in the absence of the higher molecular mass glycoforms marked by asterisks. Deglycosylation with PNGase F reduced both wild-type and mutant protein to the same apparent molecular mass predicted for the unmodified protein. Constructs carried a FLAG epitope for detection. B, loss of Asn60 glycosylation promotes homophilic SynCAM 2 binding. COS7 cells expressing GPI-anchored SynCAM 2 or its N60Q mutant carrying an extracellular FLAG epitope (green) were overlaid with the soluble extracellular domain of SynCAM 2 (red). Cells expressing FLAG-tagged, GPI-anchored SynCAM 1 Ig2  3 served as a nega- tive control. Construct expression and SynCAM 2 retention were detected as described in Fig. 1B. C, quantification of the results in B. The results are expressed as protein A signal detecting retained SynCAM 2 normalized to the signalofCOS7surface-expressedSynCAMFLAGconstructs.COS7cellsexpress- ing GPI-anchored SynCAM 1 Ig2  3 or GFP alone served as negative controls. Signals are expressed relative to GFP negative control cells. Retention of SynCAM 2 on cells expressing SynCAM 1 Ig2  3 was lower than on GFP- expressing cells for unknown reasons (SynCAM 2, n  24 cells; N60Q, n  46; SynCAM 1 Ig2  3, n  39; GFP  27). ***, p  0.001. D, loss of Asn60 glycosy- lation in SynCAM 2 promotes its heterophilic binding to SynCAM 1. COS7 cells expressing FLAG-tagged, GPI-anchored SynCAM 2, or its N60Q mutant (green) were overlaid with the soluble extracellular domain of SynCAM 1 (red). Construct expression and SynCAM 1 retention were detected as described in Fig. 1B. E, quantification of the results in D was performed as described for C (SynCAM 2, n  25 cells; N60Q, n  40; SynCAM 1 Ig2  3, n  28; GFP  28). IB, immunoblot. FIGURE 6. N-Glycosylation of SynCAM 1 at Ig1 sites Asn70/Asn104 pro- motes adhesive binding. A, absence of Asn70/Asn104 glycosylation weakens the homo- and heterophilic interactions of SynCAM 1. COS7 cells expressing GPI-anchored SynCAM 1 or SynCAM 2 carrying an extracellular FLAG epitope (green) were overlaid with the soluble extracellular sequence of wild-type SynCAM 1 or its N70Q,N104Q glycosylation mutant (red). Cells expressing solubleGFPservedasnegativecontrol.Constructexpressionandretentionof soluble SynCAM 1 were detected as described in Fig. 1B. B, quantification of the results in A. The results are expressed as fluorescence intensity of retained SynCAM 1 normalized to the fluorescence intensity of COS7 surface-ex- pressed SynCAMFLAG constructs. COS7 cells expressing GFP alone served as negative controls. **, p  0.01; ***, p  0.001. C, lack of SynCAM 1 glycosyla- tion at Asn70/Asn104 reduces binding to brain SynCAM 2. The extracellular SynCAM 1 sequence or the N70Q,N104Q mutant were expressed in COS7 cellsasfusionswithIgG1-Fc,andequalamountswereimmobilizedonprotein A beads. Retention of solubilized rat brain membrane proteins on the immo- bilized proteins was analyzed by affinity chromatography. SDS eluates obtained from two parallel affinity bindings are shown. SynCAM 2 signal was detected by quantified immunoblotting. D, quantification of results obtained as in C (n  3). E, sialic acid modification of SynCAM 1 promotes its hetero- philic binding. The SynCAM 1 extracellular sequence was expressed and immobilized as in C and treated without or with sialidase under native condi- tions. Retention of membrane proteins from rat brain was analyzed by affinity chromatography. Affinity matrices were first eluted with 800 mM potassium acetate and then with SDS. SynCAM 1Ig lacking all three Ig domains served as negative control, and the GDP dissociation inhibitor GDI and synaptophy- sin served as controls for nonspecific binding. SynCAM 2 signal was detected by quantified immunoblotting. FT, flow-through. Glycans Modulate SynCAM Adhesion NOVEMBER 5, 2010•VOLUME 285•NUMBER 45 JOURNAL OF BIOLOGICAL CHEMISTRY 34871 These observations raised the possibility that extracellular SynCAM 1 interactions involve the participation of specific carbohydrate types. Because SynCAM 1 is modified with sialic and polysialic acid in brain (20, 45), which we confirmed in our mass spectrometry analysis of the purified protein, we tested whether sialic acids on SynCAM 1 contribute to its heterophilic SynCAM 2 binding. Affinity chromatography of SynCAM 2 extracted from brain was performed on the extracellular domain of SynCAM 1 that was either fully glycosylated (Fig. 6E, lanes 1–4) or from which sialic acids had been removed enzy- matically under native conditions (lanes 5–7). A construct lack- ing all Ig domains served as a negative control (lanes 8–10). The removal of sialic acids from SynCAM 1 reduced its retention of SynCAM 2 by 34 9% (n  3). Although we did not determine the sites of SynCAM 1 sialylation, this result indicates that SynCAM 2 adhesion may involve specific interactions with sialic acids on SynCAM 1. Alternatively, the negative charge of sialic acids may mediate favorable electrostatic interactions across the Ig1/Ig1 trans-interface, but the observation that the SynCAM 1/2 interaction is resistant to high salt conditions does not support this (Fig. 6E). N-Glycosylation at the Asn70/Asn104 Sites of SynCAM 1 Ig1 Promotes Synapse Induction—Extending our structural and biochemical analysis of SynCAM 1 N-glycosylation, we asked whether modification at the Asn70/Asn104 sites of SynCAM 1 Ig1 alters its trans-synaptic interactions and synaptogenic func- tion. We expressed the GPI-anchored SynCAM 1 extracellular domain or the N70Q,N104Q mutant in COS7 cells and co-cultured them with hippocampal neurons (Fig. 7A). We then measured two activities, the ability of COS7-expressed SynCAM 1 to recruit neuronal SynCAM proteins upon contact and its induction of presynaptic specializations in contacted neurons (28). The recruitment of neuronal SynCAM proteins was deter- mined by quantified immunostaining, taking advantage of the fact that they can be selectively detected using an antibody that does not recognize the GPI-anchored SynCAM constructs. As expected, GPI-SynCAM 1 expressed in COS7 cells efficiently recruited neuronal SynCAMs to contact sites (Fig. 7, A and B). The N70Q,N104Q mutant, however, recruited neuronal SynCAMs 24 11% less, consistent with a weakening of its trans-synaptic adhesion (Fig. 7B). This weakened recruitment of neuronal SynCAMs by SynCAM 1 N70Q,N104Q correlated with its inability to induce presynaptic specializations in neu- ronal co-cultures (Fig. 7C). The stronger effect of this mutation on synapse induction than on SynCAM recruitment indicates that a select threshold of trans-synaptic SynCAM clustering may have to be met to induce synapses. The N60Q mutant of SynCAM 2 did not further promote the synaptogenic activity of SynCAM 2 in this mixed co-culture assay (data not shown), presumably because the activity of the wild-type protein already saturated the synapse-forming potential of neurons under the overexpression conditions of this approach. Such sat- uration may compromise the detection of positive modulatory effects. Together, our functional studies demonstrate that modification of the N-glycosylation sites Asn70/Asn104 of SynCAM 1 Ig1 increases both trans-synaptic adhesion and its synaptogenic activity. DISCUSSION Our biochemical, crystallographic, mass spectrometry, and cell biological analyses characterize N-glycosylation as a modi- fication that modulates SynCAM adhesion. Our results further indicate roles of SynCAM 1 glycosylation in the regulation of synapse induction. Four lines of evidence support these conclu- sions. First, crystallographic results and structural modeling show that N-glycosylation can occur within and adjacent to the adhesive Ig1 interface of SynCAM 2 and SynCAM 1, respec- tively. Second, the glycosylation of these sites within the Ig1 domain differentially affects SynCAM properties, reducing the adhesion of SynCAM 2 while increasing the binding of SynCAM 1. Third, the ability to glycosylate these sites increases not only SynCAM 1 adhesion but also its synaptogenic activity. Fourth, the post-translational modification of SynCAM 1 is developmentally regulated in the brain, suggesting functional roles in vivo. The post-translational modification of synaptic adhesion molecules could be an attractive mechanism to regulate them. Indeed, an inhibitory effect of N-glycosylation on neuroligin 1 binding to neurexin 1 has been previously reported (27). How can N-glycosylation differentially reduce SynCAM 2 Ig1 bind- ing and promote SynCAM 1 adhesion? Our crystal structure of SynCAM 2 Ig 1 and homology model of SynCAM 1/SynCAM 2 Ig1 show that these adhesive trans-dimer interfaces consist mainly of hydrophobic interactions. The location of Asn60 at FIGURE 7. Modification of SynCAM 1 at its N-glycosylation sites Asn70/ Asn104 increases trans-synaptic adhesion and synapse induction. A, wild- type SynCAM 1 recruits neuronal SynCAMs and the presynaptic marker SV2 in a mixed co-culture assay. COS7 cells co-expressing GFP with the GPI-an- chored SynCAM 1 extracellular sequence or its N70Q,N104Q mutant were seeded atop dissociated hippocampal cultures at 7 days in vitro. COS7 cells expressing GFP alone served as negative control. Co-cultures were analyzed at 11 days in vitro by immunostaining for neuronally expressed SynCAM pro- teins (red) and the presynaptic vesicle marker SV2 (blue). GFP marked trans- fected COS7 cells (green). Wild-type SynCAM 1 recruits and retains neuronal SynCAMs, and SV2 puncta were detected atop COS7 cells expressing GPI- anchored SynCAM 1. Cells expressing SynCAM 1 N70Q,N104Q exhibited less SynCAM and no SV2 recruitment. B, quantification of the SynCAM recruit- ment shown in A (SynCAM 1, n  30 cells; N70Q,N104Q, n  34; GFP, n  23; apply also to C). *, p  0.05; **, p  0.01; ***, p  0.001. C, quantification of the SV2 recruitment shown in A. n.s., not significant. Glycans Modulate SynCAM Adhesion 34872 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 45•NOVEMBER 5, 2010 this SynCAM 2 dimer interface leaves little room for a bulky glycan in the crystal structure at this site. Glycans at Asn60 of SynCAM 2 may therefore weaken the dominant hydrophobic interactions at the dimer interface and reduce SynCAM 2 adhe- sion (Fig. 8A). In contrast, the glycans at Asn70 and Asn104 in SynCAM 1 do not participate in Ig1 dimer contacts and face away from the adhesive dimer interface. Why then is the N70Q,N104Q mutant deficient in adhesive dimer formation? We consider it possible that N-glycans of SynCAM 1 favor adhesive binding through limiting the conformational space available to the pro- tein or by inhibiting nonspecific protein clustering. Both mech- anisms have been previously proposed for other Ig superfamily adhesion proteins (46). Specifically, the glycans on Asn70 and Asn104 may bias or restrict the relative orientations of the SynCAM 1 Ig1 domain to favor adhesive dimer formation, for example by limiting the conformational space available to the Ig1 domain (Fig. 8B). Our results complement a body of studies characterizing the role of glycosylation for Ig superfamily members. These studies have established that carbohydrates can modulate homophilic adhesion and function, such as shown for L1 and NCAM, and that specific carbohydrate structures on Ig proteins can regu- late extracellular interactions as demonstrated for polysialy- lated NCAM (47–49). Interestingly, a fraction of SynCAM 1 also carries polysialic acids, making it only the second protein next to NCAM that exhibits this modification in the brain (45). This polysialylation of SynCAM 1 occurs at the third N-glyco- sylation site, which was not analyzed in our study, and may serve as an additional mechanism regulating adhesive strength. Sialic acids can also specify protein interactions as shown for the Siglec family of Ig-like lectins (50, 51). However, SynCAMs do not conform to conserved sequence motif in Siglecs (52) and appear unlikely to belong to this protein family. The potential roles of carbohydrates in binding specificity and carbohydrate- carbohydrate interactions (53) can now be addressed in future studies of adhesive SynCAM recognition. The significant developmental changes in the post-transla- tional modification of SynCAM 1 indicate that specific, pres- ently unknown glycosyltransferases modify it in the brain. In contrast, only a minor fraction of SynCAM 2 may undergo regulated carbohydrate modification. Functionally, this differ- ential glycosylation could modulate SynCAM interactions between neuronal populations, refining the potential for adhe- sive coding provided by the distinct SynCAM gene expression patterns (20, 21). The modification of SynCAMs with glycans may not only adjust their synaptic adhesive strength during brain development. Glycosylation could also change the struc- tural organization of SynCAM complexes in the synaptic cleft, analogous to the role of N-glycans in patterning the trans-ad- hesion arrays formed by L1 (54). Future studies will determine whether glycans on residues other than those analyzed here further modulate SynCAM structure and function, including the O-glycans at the stalk of the SynCAM 1 extracellular domain (19). With respect to the roles of modulated adhesion, it is inter- esting to note that the glycosylation sites Asn60 of SynCAM 2 and Asn70 of SynCAM 1 are evolutionarily conserved between human and murine orthologs and that the Asn104 site of mam- malian SynCAM 1 is even present in the avian and fish orthologs (19). This indicates that the ability to modify these sites in SynCAM Ig1 domains is functionally relevant. Together, N-glycosylation alters the adhesive interactions and synapse-inducing functions of SynCAMs, demonstrating that this modification modulates trans-synaptic SynCAM interac- tions. Our findings support the notion that glycosylation plays important roles in synaptic surface interactions (55, 56). Acknowledgments—We thank the members of the Biederer and Modis laboratories for helpful discussions. We also thank Edward Voss (W. M. Keck Foundation Biotechnology Resource Laboratory) and Michael Easterling (Bruker Daltonics, Inc.) for running some of the samples on the FT-ICR MS. We thank Howard Robinson, Annie He´r- oux, and other staff at the X25 and X29A beamlines of the National Synchrotron Light Source at Brookhaven National Laboratory. FIGURE 8. Model of differential SynCAM modulation by N-glycosylation of the first Ig domain. A, N-glycans may reduce SynCAM 2 adhesion through steric hindrance within the Ig1 binding interface. B, in contrast, N-glycans facing away from the SynCAM 1 Ig1 domain may restrict its conformational freedom and position it toward binding. Note that we do not exclude addi- tional interactions between SynCAM Ig domains. Dark gray, SynCAM 1; light gray, SynCAM 2. Glycans Modulate SynCAM Adhesion NOVEMBER 5, 2010•VOLUME 285•NUMBER 45 JOURNAL OF BIOLOGICAL CHEMISTRY 34873 REFERENCES 1. Palay, S. L. (1956) J. Biophys. Biochem. Cytol. 2, 193–202 2. Gray, E. G. (1959) J. Anat. 93, 420–433 3. Lucic´, V., Yang, T., Schweikert, G., Fo¨rster, F., and Baumeister, W. (2005) Structure 13, 423–434 4. Jin, Y., and Garner, C. C. (2008) Annu. Rev. Cell Dev. Biol. 24, 237–262 5. Biederer, T., and Stagi, M. (2008) Curr. Opin. Neurobiol. 18, 261–269 6. Giagtzoglou, N., Ly, C. V., and Bellen, H. J. (2009) Cold Spring Harbor Perspect. Biol. 1, a003079 7. Su¨dhof, T. C. (2008) Nature 455, 903–911 8. Huang, Z. J., and Scheiffele, P. (2008) Curr Opin Neurobiol 18, 77–83 9. Craig, A. M., and Kang, Y. (2007) Curr. Opin. Neurobiol. 17, 43–52 10. Biederer, T., Sara, Y., Mozhayeva, M., Atasoy, D., Liu, X., Kavalali, E. T., and Su¨dhof, T. C. (2002) Science 297, 1525–1531 11. Woo, J., Kwon, S. K., Choi, S., Kim, S., Lee, J. R., Dunah, A. W., Sheng, M., and Kim, E. (2009) Nat. Neurosci. 12, 428–437 12. Linhoff, M. W., Laure´n, J., Cassidy, R. M., Dobie, F. A., Takahashi, H., Nygaard, H. B., Airaksinen, M. S., Strittmatter, S. M., and Craig, A. M. (2009) Neuron 61, 734–749 13. de Wit, J., Sylwestrak, E., O’Sullivan, M. L., Otto, S., Tiglio, K., Savas, J. N., Yates, J. R., 3rd, Comoletti, D., Taylor, P., and Ghosh, A. (2009) Neuron 64, 799–806 14. Ko, J., Fuccillo, M. V., Malenka, R. C., and Su¨dhof, T. C. (2009) Neuron 64, 791–798 15. Kayser, M. S., McClelland, A. C., Hughes, E. G., and Dalva, M. B. (2006) J. Neurosci. 26, 12152–12164 16. Lim, B. K., Matsuda, N., and Poo, M. M. (2008) Nat. Neurosci. 11, 160–169 17. Takeichi, M. (2007) Nat. Rev. Neurosci. 8, 11–20 18. Kwiatkowski, A. V., Weis, W. I., and Nelson, W. J. (2007) Curr. Opin. Cell Biol. 19, 551–556 19. Biederer, T. (2006) Genomics 87, 139–150 20. Fogel, A. I., Akins, M. R., Krupp, A. J., Stagi, M., Stein, V., and Biederer, T. (2007) J. Neurosci. 27, 12516–12530 21. Thomas, L. A., Akins, M. R., and Biederer, T. (2008) J. Comp. Neurol. 510, 47–67 22. Dresbach, T., Neeb, A., Meyer, G., Gundelfinger, E. D., and Brose, N. (2004) Mol. Cell. Neurosci. 27, 227–235 23. Graf, E. R., Kang, Y., Hauner, A. M., and Craig, A. M. (2006) J. Neurosci. 26, 4256–4265 24. Boucard, A. A., Chubykin, A. A., Comoletti, D., Taylor, P., and Su¨dhof, T. C. (2005) Neuron 48, 229–236 25. Chih, B., Gollan, L., and Scheiffele, P. (2006) Neuron 51, 171–178 26. Koehnke, J., Jin, X., Trbovic, N., Katsamba, P. S., Brasch, J., Ahlsen, G., Scheiffele, P., Honig, B., Palmer, A. G., 3rd, and Shapiro, L. (2008) Struc- ture 16, 410–421 27. Comoletti, D., Flynn, R., Jennings, L. L., Chubykin, A., Matsumura, T., Hasegawa, H., Su¨dhof, T. C., and Taylor, P. (2003) J. Biol. Chem. 278, 50497–50505 28. Biederer, T., and Scheiffele, P. (2007) Nat. Protoc. 2, 670–676 29. Matthews, B. W. (1968) J. Mol. Biol. 33, 491–497 30. Otwinowski, Z., Minor, W., and Carter, C. W., Jr. (1997) Methods Enzy- mol. 276, 307–326 31. Dong, X., Xu, F., Gong, Y., Gao, J., Lin, P., Chen, T., Peng, Y., Qiang, B., Yuan, J., Peng, X., and Rao, Z. (2006) J. Biol. Chem. 281, 10610–10617 32. McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Sto- roni, L. C., and Read, R. J. (2007) J. Appl. Crystallogr. 40, 658–674 33. Notredame, C., Higgins, D. G., and Heringa, J. (2000) J. Mol. Biol. 302, 205–217 34. Arac¸, D., Boucard, A. A., Ozkan, E., Strop, P., Newell, E., Su¨dhof, T. C., and Brunger, A. T. (2007) Neuron 56, 992–1003 35. Bai, Y., Auperin, T. C., and Tong, L. (2007) Acta Crystallogr. Sect. F Struct. Biol. Cryst. Commun. 63, 135–138 36. Mandel, C. R., Kaneko, S., Zhang, H., Gebauer, D., Vethantham, V., Man- ley, J. L., and Tong, L. (2006) Nature 444, 953–956 37. Bork, P., Holm, L., and Sander, C. (1994) J. Mol. Biol. 242, 309–320 38. Kakunaga, S., Ikeda, W., Itoh, S., Deguchi-Tawarada, M., Ohtsuka, T., Mizoguchi, A., and Takai, Y. (2005) J. Cell Sci. 118, 1267–1277 39. Maurel, P., Einheber, S., Galinska, J., Thaker, P., Lam, I., Rubin, M. B., Scherer, S. S., Murakami, Y., Gutmann, D. H., and Salzer, J. L. (2007) J. Cell Biol. 178, 861–874 40. Spiegel, I., Adamsky, K., Eshed, Y., Milo, R., Sabanay, H., Sarig-Nadir, O., Horresh, I., Scherer, S. S., Rasband, M. N., and Peles, E. (2007) Nat. Neu- rosci. 10, 861–869 41. Cooper, C. A., Joshi, H. J., Harrison, M. J., Wilkins, M. R., and Packer, N. H. (2003) Nucleic Acids Res. 31, 511–513 42. Morelle, W., Canis, K., Chirat, F., Faid, V., and Michalski, J. C. (2006) Proteomics 6, 3993–4015 43. Fiala, J. C., Feinberg, M., Popov, V., and Harris, K. M. (1998) J. Neurosci. 18, 8900–8911 44. Drescher, B., Witte, T., and Schmidt, R. E. (2003) Immunology 110, 335–340 45. Galuska, S. P., Rollenhagen, M., Kaup, M., Eggers, K., Oltmann-Norden, I., Schiff, M., Hartmann, M., Weinhold, B., Hildebrandt, H., Geyer, R., Mu¨- hlenhoff, M., and Geyer, H. (2010) Proc. Natl. Acad. Sci. U.S.A. 107, 10250–10255 46. Rudd, P. M., and Dwek, R. A. (1997) Crit. Rev. Biochem. Mol. Biol. 32, 1–100 47. Rutishauser, U., and Landmesser, L. (1996) Trends Neurosci. 19, 422–427 48. Acheson, A., Sunshine, J. L., and Rutishauser, U. (1991) J. Cell Biol. 114, 143–153 49. Kadmon, G., Kowitz, A., Altevogt, P., and Schachner, M. (1990) J. Cell Biol. 110, 209–218 50. Varki, A. (2007) Nature 446, 1023–1029 51. Crocker, P. R. (2002) Curr. Opin. Struct. Biol. 12, 609–615 52. Zaccai, N. R., May, A. P., Robinson, R. C., Burtnick, L. D., Crocker, P. R., Brossmer, R., Kelm, S., and Jones, E. Y. (2007) J. Mol. Biol. 365, 1469–1479 53. Bucior, I., and Burger, M. M. (2004) Curr. Opin. Struct. Biol. 14, 631–637 54. He, Y., Jensen, G. J., and Bjorkman, P. J. (2009) Structure 17, 460–471 55. Kleene, R., and Schachner, M. (2004) Nat. Rev. Neurosci. 5, 195–208 56. Martin, P. T. (2002) Glycobiology 12, 1R–7R Glycans Modulate SynCAM Adhesion 34874 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 45•NOVEMBER 5, 2010
3M47
Crystal structure of the mutant I218A of orotidine 5'-monophosphate decarboxylase from Methanobacterium thermoautotrophicum
Conformational Changes in Orotidine 5′-Monophosphate Decarboxylase: “Remote” Residues that Stabilize the Active Conformation† B. McKay Wood‡, Tina L. Amyes§, Alexander A. Fedorov∥, Elena V. Fedorov∥, Andrew Shabila‡, Steven C. Almo∥, John P. Richard§, and John A. Gerlt*,‡ ‡Departments of Biochemistry and Chemistry, University of Illinois at Urbana-Champaign §Department of Chemistry, University at Buffalo, Buffalo, NY 14260 ∥Department of Biochemistry, Albert Einstein College of Medicine, Bronx, NY 10461 Abstract The structural factors responsible for the extraordinary rate enhancement (~1017) of the reaction catalyzed by orotidine 5′-monophosphate decarboxylase (OMPDC) have not been defined. Catalysis requires a conformational change that closes an active site loop and “clamps” the orotate base proximal to hydrogen-bonded networks that destabilize the substrate and stabilize the intermediate. In the OMPDC from Methanobacter thermoautotrophicus, a “remote” structurally conserved cluster of hydrophobic residues that includes Val 182 in the active site loop is assembled in the closed, catalytically active conformation. Substitution of these residues with Ala decreases kcat/Km with a minimal effect on kcat, providing evidence that the cluster stabilizes the closed conformation. The intrinsic binding energies of the 5′-phosphate group of OMP for the mutant enzymes are similar to that for the wild type, supporting this conclusion. Orotidine 5′-monophosphate decarboxylase (OMPDC1) is one of Nature’s best catalysts: the reaction occurs with a rate enhancement of ~1017 and a proficiency of ~1023 M (2). The reaction coordinate includes a vinyl carbanion intermediate (Scheme 1) (3, 4). Structures of OMPDCs complexed with UMP or 6-hydroxyUMP, a n intermediate analog (5-9), reveal hydrogen-bonded networks proximal to 1) C6 of the pyrimidine (Asp 70-Lys 72-Asp 75 in the OMPDC from Methanobacter thermoautotrophicus; MtOMPDC); and 2) O2, N3, and O4 of the base (Ser 127-Gln 185). These destabilize the substrate (9) and stabilize the intermediate, although the structural strategy for the latter is uncertain. The 5′-phosphate group binds in a conserved motif at the ends of the seventh and eighth β- strands of the (β/α)8-barrel structure, with interactions to backbone NH groups as well as the guanidinium group of a conserved Arg (Arg 203 in MtOMPDC). These interactions are important for catalysis: 1) kcat/Km for OMP exceeds that for orotidine by a factor of ~1011 for the OMPDC from Saccharomyces cerevisiae (ScOMPDC) (10); 2) kcat/Km for OMP †This work was supported by NIH Grants GM039754 to J.P.R. and GM065155 to J.A.G. *To whom correspondence should be addressed: J.A.G.: telephone, (217) 244-7414; fax: (217) 244-6538; j-gerlt@uiuc.edu.. SUPPORTING INFORMATION AVAILABLE Descriptions of the experimental procedures. This material is available free of charge via the Internet at http://pubs.acs.org. 1Abbreviations: OMP, orotidine 5′-monophosphate; OMPDC, OMP decarboxylase; MtOMPDC, OMPDC from Methanobacter thermoautotrophicus; ScOMPDC, OMPDC from Saccharomyces cerevisiae; EO, 1-(β-D-erythrofuranosyl)-orotic acid; HPi, phosphate dianion; IBE, intrinsic binding energy. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 December 26. Published in final edited form as: Biochemistry. 2010 May 4; 49(17): 3514–3516. doi:10.1021/bi100443a. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript exceeds that for 1-(β-D-erythrofuranosyl)orotic acid (EO; 5′-truncated OMP analog) by factors of 5.2 × 108 and 3.6 × 108 for ScOMPDC and MtOMPDC, respectively (11, 12); and 3 ) phosphite dianion ( H Pi) activates decarboxylation of EO by factors of 5.6 × 105 M−1 and 2.9 × 105 M−1 for ScOMPDC and MtOMPDC, respectively (11, 12). The “intrinsic binding energy” [IBE; (13)] of the 5′-phosphate/HPi 1) increases the affinity for the substrate (e.g., OMP vs. orotidine/EO); and 2) enables decarboxylation by juxtaposition of the substrate with the active site hydrogen-bonded networks (substrate destabilization and intermediate stabilization). How the IBE promotes catalysis is unknown but required to understand the structural basis for the rate enhancement. A loop located at the end of the seventh β-strand closes over the active site when OMP binds (Figure 1). Although the active site loops differ in both length and sequence in divergent OMPDCs (12), each includes a spatially conserved Gln (Gln 185 in MtOMPDC) hydrogen- bonded to both the 5′-phosphate and O2 of the pyrimidine as well as a conserved Ser at the end of the fifth β-strand that also is hydrogen-bonded to N3 of the pyrimidine (Ser 127). We characterized the importance of these “clamp” residues in ScOMPDC (Gln 215-Ser 154) using EO/HPi and confirmed that the 5′-phosphate/HPi “switch” is required to activate the enzyme (14). The structures of divergent OMPDCs reveal that OMP binding is always accompanied by a conformational change (Figure 1). The most obvious component is closure of the active site loop. However, the (β/α)8-barrel structure can be divided into two domains, one formed from the second, third, fourth, and fifth β-strands (where the hydrogen-bonded Asp 70-Lys 72-Asp 75 motif and Ser 127 are located) and the second from the sixth, seventh, eighth, and first β-strands (where the phosphate binding motif and the active site loop, including Gln 185, are located) (15). OMP binding reorients the domains, with the latter domain moving toward the former, enforcing the orotate carboxylate group to be juxtaposed vis a vis Asp 70- Lys 72-Asp 75 and, also, allowing formation of the Ser 127-Gln 185 “clamp”. Thus, the transition between the open and closed conformations is more complicated than “simple” hinge motion of the loop on the rigid framework of the (β/α)8-barrel structure. In this Report we identify “remote” residues involved in this conformational change and quantitate their importance in promoting and stabilizing the catalytically competent form of the enzyme. The active site loop of MtOMPDC, P180-G181-V182-G183-A184-Q185-G186-G187-D188, is disordered in the absence of substrate but ordered and closed in its presence (Figure 1). In the liganded structure two residues in the loop make contacts with the (β/α)8-barrel scaffold: 1) Gln 185 is hydrogen-bonded to Ser 127 (vide infra); and 2) Val 182 is embedded in a hydrophobic cluster also formed by Ile 199, Val 201, and Ile 218. The conservation of this hydrophobic cluster in all OMPDCs suggests its importance in a common catalytic strategy. We probed this strategy by mutagenesis of these hydrophobic residues. Ala substitutions for residues in the hydrophobic cluster cause substantial decreases in kcat/ Km for decarboxylation of OMP but have little effect on kcat (Table 1). We determined high resolution X-ray structures (≤ 1.4 Å) for each mutant in the presence of 6-hydroxyUMP (Figure 2). The liganded structures superimpose well with that of wild type, with only small differences observed at the sites of the substitutions (panel A). The active sites are identical to that of wild type (panel B), explaining the minimal impact on kcat. The mutated residues are remote from the active site (Figures 1 and 2). Thus, the effects of the substitutions on kcat/Km cannot be explained by altered direct interactions with the substrate. Instead, the effects can be explained by decreased stabilities of the closed conformation in which the substrate is destabilized (9) and the anionic intermediate is stabilized. A consistent model (Scheme 2) is that an equilibrium of open (Eo) and closed Wood et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (Ec) conformations exists in the absence or presence of substrate (KC’ or KC; KC’ << 1). The substitutions destabilize Ec (decrease KC and KC ’), so kcat/Km is decreased (the energy difference between Eo + S and [Ec•S]‡ is increased). However, the invariance of kcat establishes that the substitutions do not alter the reactivitiy of Ec•S (the energy difference between Ec•S and [Ec•S]‡). The energy to form Ec•S from Eo + S is achieved from 1) interactions of the 5′-phosphate group with its binding motif (IBE; vide infra), and 2) an increased concentration of OMP to form Ec•S (KC/KS or, equivalently, KC’/KS’, although the former is expected to be the relevant pathway). We also used the two part EO•HPi substrate. The values of kcat/Km for EO are decreased relative to that for wild type (Table 1); these can be explained by decreased populations of Ec (Scheme 2), assuming that EO, without a 5′-phosphate group, is unable to promote the transition from Eo to Ec. The values of kcat/Km for OMP and kcat/Km for EO allow calculation of the IBE for the 5′-phosphate group of OMP (Table 1). HPi activates the mutants as judged by the values of the third-order rate constant, (kcat/ Km)EO•HPi/Kd. The equivalent changes (ΔΔG‡) in kcat/Km for both OMP and EO and the third-order rate constant indicate that all three measure the effects of the substitutions on the values of KC. The values of the IBEs for the 5′-phosphate group for the mutants are the same as that for the wild type, establishing that decarboxylation in the Ec•S complex occurs with equivalent amounts of ground state destabilization (9) and transition state stabilization (as also reflected by the invariant values of kcat). The IBEs provide further support for the role of the “remote” hydrophobic cluster in stabilizing Ec relative to Eo but do not directly participating in catalysis. Our experiments implicate a structurally conserved hydrophobic cluster, Val 182, Ile 199, Val 201, and Ile 218 in MtOMPDC, in stabilizing the closed conformation required for catalysis. Its identification provides evidence that structural elements distal from the active site, in addition to the proximal active site loop that closes to “clamp” the substrate, are required for OMPDC’s extraordinary catalytic efficiency and proficiency. Supplementary Material Refer to Web version on PubMed Central for supplementary material. REFERENCES (1). Wood BM, Chan KK, Amyes TL, Richard JP, Gerlt JA. Biochemistry. 2009; 48:5510–5517. [PubMed: 19435313] (2). Radzicka A, Wolfenden R. Science. 1995; 267:90–93. [PubMed: 7809611] (3). Toth K, Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2007; 129:12946–12947. [PubMed: 17918849] (4). Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2008; 130:1574–1575. [PubMed: 18186641] (5). Miller BG, Hassell AM, Wolfenden R, Milburn MV, Short SA. Proc. Natl. Acad. Sci. U A. 2000; 97:2011–2016. (6). Appleby TC, Kinsland C, Begley TP, Ealick SE. Proc atl Acad Sci U S A. 2000; 97:2005–2010. (7). Harris P, Navarro Poulsen JC, Jensen KF, Larsen S. Biochemistry. 2000; 39:4217–4224. [PubMed: 10757968] (8). Wu N, Mo Y, Gao J, Pai EF. Proc. Natl. Acad. Sci. U S A. 2000; 97:2017–2022. [PubMed: 10681441] Wood et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (9). Chan KK, Wood BM, Fedorov AA, Fedorov EV, Imker HJ, Amyes TL, Richard JP, Almo SC, Gerlt JA. Biochemistry. 2009; 48:5518–5531. [PubMed: 19435314] (10). Sievers A, Wolfenden R. Bioorg. Chem. 2005; 33:45–52. [PubMed: 15668182] (11). Amyes TL, Richard JP, Tait JJ. J. Am. Chem. Soc. 2005; 127:15708–15709. [PubMed: 16277505] (12). Toth K, Amyes TL, Wood BM, Chan KK, Gerlt JA, Richard JP. Biochemistry. 2009; 48:8006– 8013. [PubMed: 19618917] (13). Jencks WP. Adv. Enzymol. Rel. Areas. Mol. Biol. 1975; 43:219–410. (14). Barnett SA, Amyes TL, Wood BM, Gerlt JA, Richard JP. Biochemistry. 2008; 47:7785–4487. [PubMed: 18598058] (15). Harris P, Poulsen JC, Jensen KF, Larsen S. J. Mol. Biol. 2002; 318:1019–1029. [PubMed: 12054799] (16). Go MK, Amyes TL, Richard JP. Biochemistry. 2009; 48:5769–5778. [PubMed: 19425580] Wood et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Superposition of the active site of wild type MtOMPDC in the absence (cyan; disordered loop depicted with the dotted line ) and presence ( grey ) of 6-hydroxyUMP. The carbons of 6-hydroxyUMP are highlighted in yellow; the carbons of Val 182 and Arg 203 in the liganded structure are highlighted in orange. Wood et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Superpositions of the 6-hydroxyUMP-liganded structures of wild type MtOMPDC and the single mutants in the hydrophobic cluster. Panel A, hydrophobic cluster; panel B, active sites. Wood et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. Wood et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 2. Wood et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Wood et al. Page 9 Table 1 Kinetic Constants for OMP, EO and EO/HPi and Intrinsic Binding Energies of the 5′-Phosphate Group of OMP at pH 7.1 and 25 °C. MtOMPDC kcat OMP s−1 kcat/Km OMP M−1 s−1 ΔΔG‡ kcal/mola kcat/Km EO M−1 s−1 ΔΔG‡ kcal/mola (kcat/Km)/K D EO•HPib M−2 s−1 ΔΔG‡ kcal/mola 5′-Phosphate IBEc kcal/mol Wild type 4.6 2.9 × 106 8.7 × 10−3 2500 11.6 d V182A 3.4 1.4 × 105 1.8 1.3 × 10−3 1.1 190 1.5 10.9 I199A 3.9 9.1 × 105 0.7 1.9 × 10−3 0.9 980 0.6 11.8 V201A 4.0 9.5 × 105 0.7 3.1 × 10−3 0.6 690 0.8 11.5 I218A 3.3 2.8 × 105 1.4 2.3 × 10−3 0.8 340 1.2 11.0 V182A/I199A 3.1 4.9 × 104 2.4 3.9 × 10−4 1.8 81 2.0 11.0 V182A/V201A 2.5 4.9 × 104 2.4 5.0 × 10−4 1.7 30 2.6 10.9 aCalculated from the ratio of the second-order or third-order rate constants for the wild type and mutant enzyme. bThird-order rate constant for reaction of EO/HPi. cTransition state stabilization by the 5′-phosphate group of OMP, calculated from the ratio of the values of kcat/Km for OMP and EO. IBEs for phosphite dianion can be calculated from the ratio of (kcat/ Km)/KD for EOHPi and kcat/Km for EO. Biochemistry. Author manuscript; available in PMC 2011 December 26.
3M4A
Crystal structure of a bacterial topoisomerase IB in complex with DNA reveals a secondary DNA binding site
Crystal structure of a bacterial topoisomerase IB in complex with DNA reveals a secondary DNA binding site Asmita Patel1, Lyudmila Yakovleva2, Stewart Shuman2,3, and Alfonso Mondragón1,3 1 Department of Biochemistry, Molecular Biology and Cell Biology, Northwestern University, 2205 Tech Drive, Evanston, Illinois 60208 2 Molecular Biology Program, Sloan-Kettering Institute, 1275 York Avenue, New York, New York 10065 Summary Type IB DNA topoisomerases (TopIB) are enzymes that relax supercoils by cleaving and resealing one strand of duplex DNA within a protein clamp that embraces a DNA segment. A longstanding conundrum concerns the capacity of TopIB enzymes to stabilize intramolecular duplex DNA crossovers and, in the case of poxvirus TopIB, form protein-DNA synaptic filaments. Here we report a structure of D. radiodurans TopIB in complex with a 12-bp duplex DNA that demonstrates a secondary DNA binding site located on the C-terminal domain. It comprises a distinctive interface with one strand of the DNA duplex and is conserved in all TopIB enzymes. Modeling of a TopIB with both DNA sites suggests that the secondary site could account for DNA crossover binding, nucleation of DNA synapsis, and generation of a filamentous plectoneme. In support of this, mutations of the secondary site eliminate synaptic plectoneme formation without affecting DNA cleavage or supercoil relaxation. Introduction Type IB DNA topoisomerases (TopIB) are encoded in the genomes of all eukarya, several eukaryal viruses (poxviruses and mimivirus; Benarroch et al., 2006), many bacteria (Krogh and Shuman, 2002), and several archaea (Forterre et al., 2007). They play important roles in relaxing supercoils generated during DNA replication and transcription. TopIB enzymes accomplish this task by repeatedly breaking and rejoining one strand of the DNA duplex through a covalent DNA-(3′-phosphotyrosyl)-enzyme intermediate (Corbett and Berger, 2004). Within the covalent TopIB–DNA complex, the noncovalently held 5′-OH DNA segment swivels about the protein-DNA nick before being religated to the 3′-phosphate of the covalently held strand. The number of supercoils removed by TopIB per cleavage-religation cycle follows an exponential distribution that depends on the torque stored in the supercoiled DNA and friction at the protein-DNA interface during the swivel (Koster et al., 2005). Crystal structures of DNA-bound cellular and poxvirus TopIB enzymes captured at sequential steps along the reaction pathway (pre-cleavage, transition-state, and post-cleavage covalent complex) have fully illuminated the DNA-protein interactions and reaction chemistry (Davies 3Corresponding authors: A.M. Phone: 847-491-7726, Fax: 847-467-6489, a-mondragon@northwestern.edu. S.S. Phone: 212-639-7145, Fax: 212-772-8410, s-shuman@ski.mskcc.org. Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. NIH Public Access Author Manuscript Structure. Author manuscript; available in PMC 2011 June 9. Published in final edited form as: Structure. 2010 June 9; 18(6): 725–733. doi:10.1016/j.str.2010.03.007. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript et al., 2006; Perry et al., 2006, 2010; Redinbo et al., 1998). These structures, together with biochemical studies (Krogh and Shuman, 2000; Tian et al., 2005), revealed how nucleophilic attack of the active-site tyrosine hydroxyl on the DNA phosphodiester bond is catalyzed by two arginines, a lysine and a histidine that stabilize the pentacoordinate transition-state and expel the 5′-OH leaving strand. The crystal structures also underscored how all TopIB enzymes envelop the duplex DNA cleavage site by forming a C-shaped protein clamp, wherein a C- terminal catalytic domain engages the DNA minor groove at and surrounding the scissile phosphodiester while an N-terminal domain module engages the DNA major groove on the face of the duplex opposite the cleavage site. Comparison of the structures of DNA-bound poxvirus TopIB (Perry et al., 2006) and the free apoenzyme (Cheng et al., 1998) highlighted that the active site is not preassembled prior to DNA binding. Indeed, in the poxvirus apoenzyme, three of the five catalytic residues (the lysine and arginine general acids and the tyrosine nucleophile) are either disordered or out of position to perform transesterification chemistry. The formation of a catalytically competent active site entails multiple conformational switches and disordered-to-ordered transitions within the catalytic domain that are triggered by recognition of specific nucleobases and backbone phosphates of the consensus 5′-CCCTT↓/3′-GGGAA cleavage site for poxvirus TopIB (Perry et al., 2006; Tian et al., 2004a; Tian et al., 2004b; Yakovleva et al., 2006). Many bacterial species encode a TopIB that resembles the poxvirus and mimivirus enzymes with respect to their small size, primary structures, and bipartite domain organization (Krogh and Shuman, 2002). It is speculated that horizontal transfer of TopIB genes among bacteria and eukaryal viruses occurred during their cohabitation in a unicellular eukaryal host, e.g., amoebae (Benarroch et al., 2006). The bacterial TopIB clade is exemplified by Deinococcus radiodurans TopIB (DraTopIB), the only member that has been characterized biochemically and structurally (Krogh and Shuman, 2002; Patel et al., 2006). Although the crystal structures and active sites of poxvirus TopIB and DraTopIB are quite similar (more so to each other than to the much larger eukaryal cellular TopIB enzymes), three features of DraTopIB stand out in comparison to the poxvirus TopIB: (i) DraTopIB does not transesterify at the poxvirus 5′- CCCTT↓ cleavage site, (ii) the five catalytic amino acids of DraTopIB are pre-assembled in the apoenzyme crystal structure, and (iii) a segment of the DraTopIB catalytic domain flanking a catalytic arginine that is disordered in the apoenzyme crystal structure corresponds to the “specificity helix” of poxvirus TopIB that is critical for DNA site recognition and cleavage (Patel et al., 2006; Perry et al., 2006; Yakovleva et al., 2008). It has long been suspected that the catalytic DNA binding mode seen in the available TopIB- DNA crystal structures might not be the only means by which TopIB interacts with DNA. In a pioneering study, Zechiedrich and Osheroff observed by electron microscopy that mammalian TopIB prefers to bind to relaxed circular and linear plasmid DNA molecules at the nodes created by the crossing of two duplex helices (Zechiedrich and Osheroff, 1990). They suggested that TopIB might initially bind to one DNA segment and then capture a second DNA segment at a distant site in the same plasmid molecule. Later, Madden et al. reported that the Y723F active site mutant of human TopIB binds preferentially to positively or negatively supercoiled plasmid DNA compared to relaxed DNA molecules (Madden et al., 1995). Crossover recognition provides a “topology sensor” and a potential means to direct TopIB action to plectonemic DNAs. Electron microscopy has also been used to visualize complexes formed by poxvirus TopIB on plasmid DNAs (Shuman et al., 1997). The poxvirus TopIB formed intramolecular loop structures in which non-contiguous DNA segments were synapsed at protein-containing nodes or within filamentous protein stems. The formation of filaments along the DNA suggested that poxvirus TopIB binds DNA cooperatively. At high TopIB concentrations, the DNA appeared to be “zipped up” within the protein filaments such that the duplex was folded back on itself. Formation of loops and filaments was also observed with an Patel et al. Page 2 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript active site mutant, TopIB-Phe274. The zipped-up poxvirus TopIB-DNA complexes formed on relaxed DNA were shown to be plectonemic supercoils, in which the two duplexes encompassed by the protein filaments are interwound in a right-handed helix (Shuman et al., 1997). Thus, TopIB binding to DNA directly imposes a higher order DNA structure. Further insights to the DNA binding properties of poxvirus TopIB were gained by applying atomic force microscopy (AFM) to the problem (Moreno-Herrero et al., 2005). AFM verified that poxvirus TopIB formed nodes and filaments on linear or nicked-circular DNAs by intramolecular synapsis of two distant DNA segments. Measuring the filament length as a function of TopIB concentration showed that synapsis is a highly cooperative process. The congruence of the EM and AFM studies suggested that TopIB-mediated DNA synapsis might contribute to organization of the 200-kbp vaccinia genome into a higher order structure conducive to transcription within virus cores. A key question is how TopIB bridges distant DNA sites. Is it via protein-protein interactions between two DNA-bound TopIB molecules? Or can a single molecule of TopIB bind simultaneously to two DNA segments? These two simple physical mechanisms to account for synapsis by the poxvirus TopIB are depicted in Fig. 1 as models A and B, respectively. In model A, protein-protein interactions between TopIB molecules provide the “glue” for synapsis of two TopIB-DNA filaments. (The model arbitrarily depicts the interaction between the C-terminal catalytic domains of DNA-bound protomers; it could just as well entail N- domain/N-domain interactions or N-domain/C-domain contacts.) In model B, the synapsed DNA duplex is captured at a putative secondary DNA binding site on the TopIB protomer. Similarly, the binding of TopIB to DNA crossovers can be explained by either TopIB-TopIB interactions or simultaneous occupancy of two DNA binding sites on TopIB. Here we report the crystal structure of the bacterial DraTopIB enzyme in a complex with DNA. The structure demonstrates a secondary DNA binding site located on the surface of the C- terminal domain. The secondary DNA site is ~30 Å from the catalytic DNA site and comprises an extensive network of direct and water-mediated hydrogen bonds from the enzyme to one strand of the DNA duplex. The secondary site appears to be conserved in the poxvirus and eukaryal cellular TopIB enzymes. A model of the poxvirus TopIB enzyme with both DNA sites filled suggests how second site capture might account for DNA crossover binding, nucleation of DNA synapsis, and plectonemic supercoiling within the synaptic filament. We provide biochemical evidence in support of this model by showing that mutations in the putative secondary DNA binding site of poxvirus TopIB affect the generation of plectonemic supercoils, but not supercoil relaxation. Results and Discussion Structure of a DraTopIB-DNA complex Crystals of DraTopIB were grown in the presence of a 12-bp DNA duplex (see Materials and Methods) that is not a substrate for cleavage by the enzyme. The crystals belong to space group C2, unlike the DraTopIB apoenzyme (Patel et al., 2006), which crystallized in space group P21. The structure of the DraTopIB-DNA complex was solved by Molecular Replacement using the structure of the DraTopIB apoenzyme catalytic domain as a search molecule and refined at 1.65 Å resolution to Rwork/Rfree of 0.193/0.222 (Table I). In the crystal lattice, the short DNA duplexes are stacked end-to-end to form a pseudo-continuous helix to which DraTopIB is bound. In prior TopIB-DNA structures, the DNA fits into an interdomain cleft, the domains form a circumferential clamp around the duplex, and the active site residues in the C-domain directly coordinate the scissile phosphodiester (Fig. 2A). By contrast, in the DraTopIB-DNA complex, the N- and C-domains are splayed apart (not shown) and the catalytic DNA binding site is unoccupied (Fig. 2B). The N-domain (aa 1–90) is partially disordered and makes no contacts to the 12-bp DNA ligand. The C-domain (aa 91–346) is fully Patel et al. Page 3 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript ordered and the DNA is docked on the basal surface far from the active site (Fig. 2B). The C- domains of the bacterial and viral TopIB enzymes comprise two globular lobes (lobes 1 and 2) (Fig. 2). Alignment of the C-domains of the poxvirus TopIB and DraTopIB DNA complexes reveals that the DraTopIB-DNA structure heralds the existence and position of a bona fide secondary DNA-binding site on TopIB. The secondary DNA binding site is located within lobe 1 (Fig. 2B). DNA binding triggers folding of the “specificity helix” Comparison of the C-domains of the DraTopIB apoenzyme and the DNA-bound DraTopIB reveals both subtle and profound changes correlated with DNA binding. The C-domains of the two structures superimpose with a rmsd of 1.07 Å for all main chain atoms. However, when the isolated lobes 1 and 2 are superimposed separately, the rmsd values are only 0.87 Å and 0.76 Å, respectively. Thus, whereas the individual lobes are nearly identical in the apoenzyme and DNA-bound DraTopIB, the DNA elicited a subtle reorientation of the lobes with respect to each other. The five catalytic amino acids (Arg137, Lys174, Arg239, Asn280 and Tyr289) are preassembled into an active site in the DNA complex (Fig. 2B) and are located in the same positions as in the apoenzyme. Yet, none of the five catalytic residues make contact with DNA in the crystal lattice. An important insight from the poxvirus TopIB-DNA cocrystal (Perry et al., 2006) was the identification of a “specificity helix” (131FGKMKYLKENETVG144) that binds the DNA target site in the major groove (Fig. 2B) and makes atomic contacts to nucleobases and phosphate oxygens that are important for cleavage site recognition and DNA transesterification (Yakovleva et al., 2008). This specificity helix is conserved among poxvirus and mimivirus TopIB enzymes and is a distinctive secondary structure feature of the viral/bacterial TopIB clade that is absent in human TopIB (Redinbo et al., 1998). The specificity helix of poxvirus TopIB is protease-sensitive and disordered in the apoenzyme, but adopts a defined secondary structure and becomes protease-resistant when the poxvirus TopIB is in the DNA-bound state (Cheng et al., 1998; Perry et al., 2006; Sekiguchi and Shuman, 1995). Tight docking of the specificity helix in the major groove 5′ of the scissile phosphate aids in placing the catalytic Arg130 residue in the active site. The DraTopIB equivalent of the poxvirus specificity helix is 138VGSDIYARQHKTYG151. Structure probing of free DraTopIB by limited proteolysis delineated a single trypsin-sensitive site within this segment between Arg145 and Gln146 (Krogh and Shuman, 2002). Moreover, the 139GSDIYARQHK148 peptide is disordered in the crystal structure of apo-DraTopIB (Patel et al., 2006). By contrast, we find presently that in the DraTopIB-DNA cocrystal, the previously disordered peptide segment forms a well-ordered α-helix that mimics the specificity helix of poxvirus TopIB (Fig. 2B). In the poxvirus TopIB, the specificity helix penetrates deeply into the DNA major groove, where it makes multiple side chain and main-chain contacts to the DNA phosphates and bases (Fig. S1A). The equivalent α-helix in DraTopIB makes a single (nonspecific) contact to a backbone phosphate in a symmetry-related 12-mer DNA and one contact to a nucleobase. We attribute these limited contacts to crystal packing (Fig. S1B). The symmetry related 12-mer DNA is not situated in the active site. The structures of DraTopIB and poxvirus TopIB suggest that several mechanisms exist to trigger the folding of the specificity helix. In poxvirus TopIB, this occurs as a direct response to binding of the cleavage recognition sequence in the catalytic DNA site. In the case of DraTopIB, the equivalent conformation switch occurs either as: (i) a nonspecific response to duplex DNA that does not trigger catalysis, or (ii) an indirect, perhaps allosteric, response to occupancy of the secondary DNA binding site. Because the catalytic Arg137 is already poised in the active site in the DraTopIB apoenzyme, we presume that the induced folding of the Patel et al. Page 4 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript specificity helix is critical for recognition of the target site(s) for DNA transesterification, which (though clearly different from that of the poxvirus TopIB) are presently uncharted. Architecture of the secondary DNA binding site The secondary DNA binding site is located entirely within lobe 1 of the C-domain of DraTopIB. The DNA interface comprises a helix-loop-helix (aa 106–129) that contacts one strand of the 12mer duplex from the minor groove side and another α-helix (aa 186–194) that contacts the same strand from the major groove side (Fig. 3A). The component peptide elements form a groove on the basal surface of the C-domain into which only one strand of the DNA duplex fits. Most of the atomic contacts entail H-bond donation from protein side chain (Thr110, Lys122, Arg129, Thr187, Asn191, Lys194) or main chain (Gly117) atoms to three consecutive DNA phosphates of one DNA strand (Fig. 3B). In addition, Arg116 and Arg186 contact different guanine bases in the same DNA strand (Fig. 3A,B). The “footprint” of the secondary DNA site covers 4 nucleotides (Fig. 3B) and consists of 11 direct hydrogen bonds from protein to DNA (Fig. 3A). In addition, there are at least 8 water-mediated hydrogen bonds from protein to DNA (not shown). The interface area between the protein and DNA is 584 Å2, significantly larger than the 380 Å2 for the interface area between DNA and the specificity helix. The interface area is small in comparison with that of the poxvirus Top IB with DNA (1517 Å2), but this is to be expected for a non-sequence-specific secondary site with a relatively small footprint. Thus, we regard the DNA interactions at the secondary site of DraTopIB as too extensive to ascribe to incidental lattice contacts, unlike the case of the DraTopIB specificity helix discussed above that makes few contacts with the DNA. The significance of the secondary DNA binding site defined by the new DraTopIB structure is underscored by its conservation in the poxvirus and eukaryal TopIB. Superposition of the viral and human enzymes on DraTopIB shows that the component α-helices and loops of the secondary DNA site are preserved in all three TopIBs, affording a similar groove to accommodate one strand of duplex DNA (Fig. 4). Also, many of the basic and hydrophilic side chains that contact the DNA in DraTopIB (Fig. 4A) are either conserved in the poxvirus and human enzymes or substituted by a related side chain in a similar spatial position (Fig. 4B and C). The available structural and phylogenetic information suggest that the secondary DNA site is not a unique feature of DraTopIB. Whether other TopIB enzymes can bind DNA using this site remains to be determined (see below). A recent study implicates a cluster of four lysines on the surface of human TopIB as contributory to the preferential binding of TopIB to supercoiled DNA (Yang et al., 2009); these lysines (underlined in Fig. 4C) are located within the putative human TopIB equivalent of the DraTopIB secondary DNA binding site identified presently and three of them are conserved in poxviral and/or bacterial TopIB. These observations lend further credence to a secondary DNA site common to type IB enzymes. To gain a sense of whether the secondary site might be important for DNA relaxation and/or DNA site recognition by poxvirus TopIB, we engineered three alanine-cluster mutants that eliminated the putative equivalents of the secondary DNA-binding side chains. One cluster (N103A-K104A-K107A-K108A-Y115A; or 5xAla) targeted five amino acids in the helix- loop-helix that engages from the minor groove; the second cluster (R181A-K184A-K188A; or 3xAla) targeted three residues in the α-helix that binds from the major groove. A third cluster (N103A-K104A-K107A-K108A-Y115A-R181A-K184A-K188A; or 8xAla) simultaneously changed all eight amino acids to alanine. The mutated poxvirus TopIBs were produced in E. coli with a C-terminal His6-tag and purified from soluble bacterial lysates by Ni-agarose and phosphocellulose chromatography (Yakovleva et al., 2008) in parallel with wild-type TopIB. The rate of relaxation of supercoiled plasmid DNA by the 3xAla cluster mutant was indistinguishable from the wild-type TopIB, whereas the 5xAla and 8xAla cluster mutants relaxed at about half and one-third of the wild-type rate, respectively (Fig. 5A). The rates of Patel et al. Page 5 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript single-turnover cleavage of a short duplex DNA “suicide substrate” containing a single 5′- CCCTT↓ cleavage site for poxvirus TopIB were identical for the wild-type, 3xAla, 5xAla, and 8xAla cluster mutant proteins (not shown). We surmise that individual constituents or grouped subsets of the imputed secondary DNA site of poxvirus are not crucial for topoisomerase activity under the in vitro conditions we employ routinely to assess mutational effects. Implications for TopIB-mediated DNA synapsis The present discovery of a secondary DNA binding site on DraTopIB lends support for Model B in Figure 1, in which DNA synapsis by poxvirus TopIB reflects capture of a second DNA duplex by DNA-bound proteins. Instructive clues to the process of synapsis were gained by docking the 12-mer DNA duplex bound to DraTopIB into the structure of the poxvirus TopIB bound covalently to its CCCTT target site in duplex DNA as a vanadate transition state mimetic (Perry et al., 2010). This was performed by superimposing the DNA-bound TopIB structures and then subtracting the DraTopIB protein. The resulting model of poxvirus TopIB with the primary and secondary DNA sites occupied in shown in Fig. 6, in three different orientations. The image in Fig. 6C highlights the C-shaped clamp formed by the N- and C-domains around the DNA duplex engaged in the catalytic site. Fig. 6B highlights the groove between the α- helices of the secondary DNA site at the base of the C-domain into which one of the DNA strands fits. The primary and secondary DNA duplexes are oriented similarly, but separated by ~30 Å. The view in Fig. 6A illustrates that the DNAs in the two sites do not have a parallel trajectory; rather their paths cross in this and other planes. Indeed, the angular difference between the paths of the two helices could explain the plectonemic winding of one duplex abound the other that occurs within the synaptic filaments formed by poxvirus TopIB on initially relaxed circular DNAs (Shuman et al., 1997). A similar model built using human TopIB (not shown) also shows the possible formation of a complex with two DNA binding sites, despite the much larger size of the eukaryotic enzyme and the presence of additional protein domains. This binding mode could explain the way eukaryotic TopIB binds to nodes in positively or negatively supercoiled DNA. The secondary binding site is required for synaptic plectoneme formation by poxvirus TopIB The structure of DraTopIB in complex with DNA at a secondary binding site provides a blueprint for functional probing of the basis for DNA condensation and synapsis by poxvirus TopIB, especially the plausibility of the model depicted in Fig. 6. Thus, we queried whether the cluster mutations in the secondary site of vaccinia TopIB affect the formation of plectonemic DNA braids, a key biochemical manifestation of TopIB-mediated synapsis (Shuman et al., 1997). To perform this analysis, we modified the wild-type TopIB by changing the Tyr274 nucleophile to phenylalanine, a maneuver that abolishes transesterification without affecting noncovalent DNA binding or intramolecular synapsis (Shuman et al., 1997). As shown in Fig. 5B, the incubation of relaxed plasmid DNA circles with stoichiometric amounts of the Phe274 protein introduced torsional strain, which, after relaxation by catalytic amounts of wild-type TopIB, resulted in the acquisition of up to 8 negative supercoils. As discussed previously (Shuman et al., 1997), this reaction indicates that the TopIB-DNA synaptic complex is a plectonemic supercoil in which the two duplexes encompassed by the protein filament are interwound in a right-handed helix (Shuman et al., 1997). When the same F274 change was introduced into the 5xAla, 3xAla and 8xAla cluster mutants, we found that they were uniformly unable to promote such plectonemic braiding (Fig. 5B). We surmise that the poxvirus equivalent of the secondary DNA site is essential for intramolecular synapsis and its topological sequelae. Our results are consistent with model B in Fig. 1 as the basis for TopIB-mediated synapsis. In summary, we have demonstrated the existence, structure, and functional relevance of a secondary DNA binding site on bacterial and viral TopIB. This work gives impetus and affords Patel et al. Page 6 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript a structural guide to analogous studies of DNA crossover binding by eukaryal cellular TopIB enzymes. Materials and Methods DraTopIB was produced and purified as described previously (Patel et al., 2006). Complementary 12-mer DNA oligonucleotides (5′-GAATAAGGGCGC-3′ and 3′- CTTATTCCCGCG-5′) were purified by reverse phase HPLC (Aggarwal, 1990) and then annealed. Initial crystallization trials with mixtures of DraTopIB and 12-bp duplex DNA were performed by the sitting drop vapor diffusion method in 96 well plates set up with a Hydra-II crystallization robot. Small crystals grew in polyethylene glycol at 10°C with a 1:1 molar ratio of protein and DNA. Refined conditions using the hanging drop vapor diffusion method yielded larger crystals in 30% PEG 400 (w/v), 0.1 M sodium acetate (pH 4.5), 0.2 M calcium chloride. Prior to data collection, the crystals were cryoprotected with 25% glycerol by stepwise transfer through cooled (4° C) solutions of the well buffer containing increasing concentrations of glycerol (in 5% glycerol increments per step and soaking for 2 min/step). Crystals were harvested with a rayon crystal-mounting loop, and flash cooled in liquid nitrogen. A complete diffraction data set was collected initially using a laboratory x-ray source. Later, a higher resolution data set (to 1.65 Å) was collected using synchrotron radiation at DND CAT at the Advanced Photon Source (Argonne National Laboratory). All data were processed using XDS (Kabsch, 1993) and scaled using SCALA (Collaborative Computational Project 4, 1994). The crystals belong to space group C2 with unit cell dimensions a=119.7 Å, b=53.4 Å, and c=77.4 Å, β = 96.3° and had one DraTopIB protomer in the asymmetric unit. Data collection statistics are listed in Table I. The structure was solved by molecular replacement with the program PHASER (McCoy et al., 2007) using the C-terminal domain of the DraTopIB apoenzyme (PDB 2F4Q) as a search model. Initial electron density maps clearly showed Fo−Fc difference density for DNA, but only weak density for the N-terminal domain. After refinement, the entire DNA and most of the C-terminal domain were built, but some regions of the N-terminal domain were disordered. Refinement was performed with REFMAC5 (Murshudov et al., 1997) and model rebuilding was executed in COOT (Emsley and Cowtan, 2004). The structure has a final Rwork and Rfree of 19.3% and 22.2% respectively with 99.7% of the residues in the favored regions of the Ramachandran plot and no outliers, and good rotamer distributions (Davis et al., 2004). Refinement statistics for the complex are compiled in Table I. The coordinates of the final model and the structure factors have been deposited in the PDB with accession code 3M4A. Figures were made with PyMOL (DeLano 2002). Wild-type vaccinia virus TopIB and three mutants with clustered alanine substitutions at the secondary DNA binding site (5xAla, 3xAla, and 8xAla) were produced in E. coli with C- terminal His6 tags and purified from soluble bacterial lysates by Ni-agarose and phosphocellulose chromatography (Yakovleva et al., 2008). Topoisomerase reaction mixtures containing (per 20 μl) 50 mM Tris-HCl (pH 7.5), 100 mM NaCl, 2.5 mM EDTA, 250 ng supercoiled pUC19 plasmid DNA, and 2.5 ng of TopIB were incubated at 37° C. Aliquots (20 μl) were withdrawn at 5, 10, 20, 30, 60, 120 and 240 s and then quenched immediately with SDS. “Time 0” samples were taken prior to adding TopIB. The mixtures were analyzed by electrophoresis through a 1% horizontal agarose gel in TBE buffer (90 mM Tris-borate, 2.5 mM EDTA). DNA was visualized by staining with ethidium bromide and UV transillumination. Biochemical analysis of the topological changes in relaxed circular DNA accompanying intramolecular synapsis was performed as described previously (Shuman et al., 1997). To Patel et al. Page 7 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript prepare the relaxed plasmid DNA, a reaction mixture (100 μl) containing 50 mM Tris-HCl (pH 7.5), 100 mM NaCl, 5 mM MgCl2, 25 μg supercoiled pUC19 DNA, and 0.5 μg wild-type vaccinia TopIB was incubated for 30 min at 37° C. The reaction was quenched by adding SDS to 0.2% and EDTA to 20 mM. The mixture was digested with 1 μg proteinase K for 2 h at 37° C, then extracted twice with phenol: chloroform and once with chloroform. The relaxed DNA was recovered by ethanol precipitation and resuspended in 10 mM Tris-HCl (pH 7.5), 1 mM EDTA. Vaccinia TopIB-F274 and F274 variants of the 5xAla, 3xAla, and 8xAla cluster mutants were produced in E. coli with C-terminal His6 tags and purified by Ni-agarose and phosphocellulose chromatography. To assay synaptic plectoneme formation, reaction mixtures (20 μl) containing 50 mM Tris-HCl (pH 7.5), 100 mM NaCl, 2.5 mM EDTA, 250 ng relaxed pUC19 DNA, and 100 or 200 ng of vaccinia TopIB-F274 or its Ala-cluster variants were incubated for 15 min at 37° C. The mixtures were then supplemented where with 5 ng wild- type vaccinia TopIB where specified and incubated for another 15 min at 37° C. The reactions were halted by adding SDS to 0.5%. The mixtures were digested with 10 μg proteinase K for 1 h at 37° C and then analyzed by electrophoresis through a 1% horizontal agarose gel in TBE buffer, in parallel with samples of supercoiled and relaxed pUC19. The DNA was visualized by staining with ethidium bromide. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments Support from the R.H. Lurie Comprehensive Cancer Center of Northwestern University to the Structural Biology Facility is acknowledged. This research was supported by NIH grants GM51350 (to A.M.) and GM46330 (to S.S.). S.S. is an American Cancer Society Research Professor. Portions of this work were performed at the DuPont- Northwestern-Dow Collaborative Access Team (DND-CAT) Synchrotron Research Center at the Advanced Photon Source (APS). DND-CAT is supported by Dupont, DOW and the NSF. Use of the APS is supported by the Department of Energy (DOE). We thank Greg Van Duyne for providing the coordinates of the poxvirus TopIB–DNA vanadate transition state mimetic in advance of publication. References Aggarwal AK. Crystallization of DNA binding proteins with oligodeoxynucleotides. METHODS: A Companion to Meth Enzym 1990;1:83–90. Benarroch D, Claverie JM, Raoult D, Shuman S. Characterization of mimivirus DNA topoisomerase IB suggests horizontal gene transfer between eukaryal viruses and bacteria. J Virol 2006;80:314–321. [PubMed: 16352556] Cheng C, Kussie P, Pavletich N, Shuman S. Conservation of structure and mechanism between eukaryotic topoisomerase I and site-specific recombinases. Cell 1998;92:841–850. [PubMed: 9529259] Collaborative Computational Project 4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D 1994;50:760–763. [PubMed: 15299374] Corbett KD, Berger JM. Structure, molecular mechanisms, and evolutionary relationships in DNA topoisomerases. Annu Rev Biophys Biomol Struct 2004;33:95–118. [PubMed: 15139806] Davies DR, Mushtaq A, Interthal H, Champoux JJ, Hol WG. The structure of the transition state of the heterodimeric topoisomerase I of Leishmania donovani as a vanadate complex with nicked DNA. J Mol Biol 2006;357:1202–1210. [PubMed: 16487540] Davis IW, Murray LW, Richardson JS, Richardson DC. MOLPROBITY: structure validation and all- atom contact analysis for nucleic acids and their complexes. Nucleic Acids Res 2004;32:W615–619. [PubMed: 15215462] Diederichs K, Karplus PA. Improved R-factors for diffraction data analysis in macromolecular crystallography. Nat Struct Biol 1997;4:269–275. [PubMed: 9095194] Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 2004;60:2126–2132. [PubMed: 15572765] Patel et al. Page 8 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Forterre P, Gribaldo S, Gadelle D, Serre MC. Origin and evolution of DNA topoisomerases. Biochimie 2007;89:427–446. [PubMed: 17293019] Kabsch W. Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Crystallogr 1993;26:795–800. Koster DA, Croquette V, Dekker C, Shuman S, Dekker NH. Friction and torque govern the relaxation of DNA supercoils by eukaryotic topoisomerase IB. Nature 2005;434:671–674. [PubMed: 15800630] Krogh BO, Shuman S. Catalytic mechanism of DNA topoisomerase IB. Mol Cell 2000;5:1035–1041. [PubMed: 10911997] Krogh BO, Shuman S. A poxvirus-like type IB topoisomerase family in bacteria. Proc Natl Acad Sci U S A 2002;99:1853–1858. [PubMed: 11830640] Madden KR, Stewart L, Champoux JJ. Preferential binding of human topoisomerase I to superhelical DNA. Embo J 1995;14:5399–5409. [PubMed: 7489729] McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software. J Appl Crystallogr 2007;40:658–674. [PubMed: 19461840] Moreno-Herrero F, Holtzer L, Koster DA, Shuman S, Dekker C, Dekker NH. Atomic force microscopy shows that vaccinia topoisomerase IB generates filaments on DNA in a cooperative fashion. Nucleic Acids Res 2005;33:5945–5953. [PubMed: 16237128] Murshudov GN, Vagin AA, Dodson EJ. Refinement of macromolecular structures by the maximum- likelihood method. Acta Crystallogr D 1997;53:240–255. [PubMed: 15299926] Patel A, Shuman S, Mondragon A. Crystal structure of a bacterial type IB DNA topoisomerase reveals a preassembled active site in the absence of DNA. J Biol Chem 2006;281:6030–6037. [PubMed: 16368685] Perry K, Hwang Y, Bushman FD, Van Duyne GD. Structural basis for specificity in the poxvirus topoisomerase. Mol Cell 2006;23:343–354. [PubMed: 16885024] Perry K, Hwang Y, Bushman FD, Van Duyne GD. Insights from the structure of a smallpox virus topoisomerase-DNA transition state mimic. Structure 2010;18:127–137. [PubMed: 20152159] Redinbo MR, Stewart L, Kuhn P, Champoux JJ, Hol WG. Crystal structures of human topoisomerase I in covalent and noncovalent complexes with DNA. Science 1998;279:1504–1513. [PubMed: 9488644] Sekiguchi J, Shuman S. Proteolytic footprinting of vaccinia topoisomerase bound to DNA. J Biol Chem 1995;270:11636–11645. [PubMed: 7744804] Shuman S, Bear DG, Sekiguchi J. Intramolecular synapsis of duplex DNA by vaccinia topoisomerase. Embo J 1997;16:6584–6589. [PubMed: 9351838] Tian L, Claeboe CD, Hecht SM, Shuman S. Remote phosphate contacts trigger assembly of the active site of DNA topoisomerase IB. Structure 2004a;12:31–40. [PubMed: 14725763] Tian L, Claeboe CD, Hecht SM, Shuman S. Mechanistic plasticity of DNA topoisomerase IB: phosphate electrostatics dictate the need for a catalytic arginine. Structure 2005;13:513–520. [PubMed: 15837190] Tian L, Sayer JM, Jerina DM, Shuman S. Individual nucleotide bases, not base pairs, are critical for triggering site-specific DNA cleavage by vaccinia topoisomerase. J Biol Chem 2004b;279:39718– 39726. [PubMed: 15252055] Yakovleva L, Chen S, Hecht SM, Shuman S. Chemical and traditional mutagenesis of vaccinia DNA topoisomerase provides insights to cleavage site recognition and transesterification chemistry. J Biol Chem 2008;283:16093–16103. [PubMed: 18367446] Yakovleva L, Lai J, Kool ET, Shuman S. Nonpolar nucleobase analogs illuminate requirements for site- specific DNA cleavage by vaccinia topoisomerase. J Biol Chem 2006;281:35914–35921. [PubMed: 17005552] Yang Z, Carey JF, Champoux JJ. Mutational analysis of the preferential binding of human topoisomerase I to supercoiled DNA. FEBS J 2009;276:5906–5919. [PubMed: 19740104] Zechiedrich EL, Osheroff N. Eukaryotic topoisomerases recognize nucleic acid topology by preferentially interacting with DNA crossovers. EMBO J 1990;9:4555–4562. [PubMed: 2176156] Patel et al. Page 9 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Models for TopIB-mediated DNA synapsis Single molecule imaging by EM and AFM has shown that poxvirus TopIB binds cooperatively to linear plasmid DNA and forms protein-DNA filaments in which distant segments in the same DNA molecule are synapsed. The TopIB protomer consists of a small N-terminal domain (depicted as a cyan sphere) and a larger C-terminal domain (magenta sphere) that contains the active site. The primary DNA binding site resides within a protein clamp formed by the N and C domains. Two possible mechanisms to account for synapsis are depicted as models A and B. In model A, TopIB-TopIB interactions promote synapsis of two TopIB-DNA filaments. In model B, the synapsed DNA duplex is captured at a distinct secondary DNA binding site on the TopIB protomer. Patel et al. Page 10 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. A secondary DNA binding site on DraTopIB The C-terminal catalytic domains of poxvirus TopIB (from the structure of the covalent TopIB- DNA intermediate; PDB 2H7F) and DraTopIB (from the present DraTopIB-DNA cocrystal structure) were superimposed and offset horizontally. The tertiary structures, depicted as beige ribbons traces, are homologous throughout. The C-domains consist of two globular lobes (1 and 2, denoted for poxvirus TopIB in panel A). The active sites are located on the superior surface of the C-domain (denoted by the arrow for DraTopIB in panel B); the catalytic amino acid side chains are shown as stick models. The specificity helix is shown in magenta. The duplex DNA segments of the respective DNA ligands are shown as gray spacing-filling models. The DNA duplex is covalently linked to poxvirus TopIB is the primary (catalytic) DNA binding site. By contrast, DraTopIB binds its DNA ligand on the opposite face of the C-domain ~30 Å away from the primary DNA site. See also Figure S1 for a comparison of the DNA contacts of the specificity helix in poxvirus TopIB versus DraTopIB. Patel et al. Page 11 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Close-up view of the interactions between DNA and DraTopIB (A) A stereo image of the secondary DNA binding site is shown. Side chains and a main chain amide that contact the DNA ligand are depicted as sticks. Hydrogen bonds are denoted by dashed lines. (B) The 12mer duplex DNA ligand is depicted as a two-dimensional base-paired ladder. Atomic contacts between the indicated amino acid side chains (or the Gly117 amide) and the DNA phosphates or nucleobases are indicated by arrows. Patel et al. Page 12 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. Conservation of the secondary DNA binding site (A) A stereo image of the secondary DNA binding site in DraTopIB in shown, viewed from the opposite side of the C-domain as the image in Fig. 3. The DNA is depicted as a cartoon with the phosphate backbone chain shown in yellow and the nucleobases as blue sticks. (B) The corresponding region of the C-domain of poxvirus TopIB is shown in the same orientation as in panel A. The stereo image highlights the basic and polar side chains that are candidates to comprise the secondary DNA site in poxvirus TopIB. (C) This panel shows the aligned amino acids sequences of the two protein segments comprising the actual or imputed secondary DNA binding sites in exemplary bacterial, eukaryal cellular, and poxvirus TopIB enzymes – D. radiodurans (Dra) TopIB, Homo sapiens (Hsa) TopIB, and vaccinia (vac) TopIB, respectively – for which crystal structures are available. The aligment is based on superposition of the tertiary structures. The nine DraTopIB amino acids that contact the DNA in the secondary site are denoted by ●. The amino acids clusters in vaccinia TopIB that, when simultaneously mutated to alanine abolished synaptic plectoneme formation, are denoted by |. The four lysines in HsaTopIB implicated in crossover recognition on the basis of the effects of clustered lysine- to-glutamate changes (Yang et al., 2009) are underlined. Positions of side chain identity/ similarity at the actual or imputed DNA binding residues are highlighted in yellow shading. Patel et al. Page 13 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Mutations at the secondary binding site of poxvirus TopIB abolish formation of plectonemic synaptic complexes but have little or no effect on supercoil relaxation (A) Kinetics of relaxation of 250 ng pUC19 DNA by 2.5 ng of vaccinia TopIB, either wild- type (WT) or the indicated mutants with clustered alanine substitutions at the putative secondary DNA binding site. The DNA products were resolved by native agarose gel electrophoresesis. The supercoiled (S) and relaxed (R) circular DNAs are indicated on the left. (B) Plectonemic supercoiling. Relaxed circular pUC19 DNA was incubated with 100 or 200 ng of vaccinia TopIB-Phe274 or the indicated Ala-cluster mutants thereof, then treated where indicated with 5 ng of wild-type TopIB to relax any supercoils introduced by binding of the F274 protein(s). The products were resolved by native agarose gel electrophoresis. Patel et al. Page 14 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. Model of TopIB with primary and secondary DNA sites occupied The diagram shows a model of poxvirus TopIB with both DNA binding sites occupied. The model was built by aligning the DraTopIB-DNA complex to the crystal structure of poxvirus TopIB bound covalently to its DNA target site as a vanadate transition state mimetic (Perry et al., 2010; PDB 3IGC) and then deleting the DraTopIB protein to leave just the 12-mer DNA in the secondary site. Three different views of the model are shown in A, B and C. The primary DNA ligand is enveloped within a circumferential protein clamp formed by the N- and C- domains. The secondary DNA ligand is docked at the base of the C-domain with one strand of the duplex fitting into the secondary binding site (B). Patel et al. Page 15 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Patel et al. Page 16 Table I Diffraction data and refinement statistics Data collection Space group C2 Cell 119.7 Å, 53.4 Å, 77.4 Å, β=96.33 Detector type/source MAR-CCD/APS Wavelength (Å) 1.00 Resolution (Å) 1.65 Detector type/source MAR345/home Wavelength (Å) 1.5418 Resolution (Å) 1.92 Merged data sets Measured reflections 265,853 Unique reflections 56,385 Completeness (%)a 96.2 (89.5) Rsym (%)a,b 4.8 (27.6) Rmeas (%)a,c 5.3 (32.5) Redundancy 4.7 (3.4) Mean(I/σ(I)) 24.1 (4.1) Refinement Resolution (Å) 76 – 1.65 (1.65 – 1.693) Number of reflections: working set/test set 53,565/2,820 (3,633/209) Rworkf 19.3 (22.0) Rfreed 22.2 (23.2) Protein atoms 2,495 DNA atoms 487 Water molecules 278 Other 12 r.m.s.d. from target values Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Patel et al. Page 17  Bond lengths (Å) 0.008  Bond angles (°) 1.29 Average B-factor (Å2):  Main chain 16.0  Side chain 17.7  DNA 15.2  Solvent 24.6 Ramachandran plot  Favored (%)f 99.7  Outliers (%)f 0.0 Rotamer outliers (%)f 0.4 aNumbers in parenthesis represent values in the highest resolution shell. bRsym=Σ|I−<I>|/ΣI, where I=observed intensity, and <I>=average intensity obtained from multiple measurements. cRmeas as defined by (Diederichs and Karplus, 1997). dR-factor=Σ||Fo| − |Fc||/Σ|Fo|, where |Fo|=observed structure factor amplitude and |Fc|=calculated structure factor amplitude. eRfree: R-factor based on 5% of the data excluded from refinement. fCalculated with MolProbity (Davis et al., 2004). Structure. Author manuscript; available in PMC 2011 June 9.
3M4D
Crystal structure of the M113N mutant of alpha-hemolysin
Molecular bases of cyclodextrin adapter interactions with engineered protein nanopores Arijit Banerjeea, Ellina Mikhailovaa, Stephen Cheleya, Li-Qun Gub,2, Michelle Montoyac,3, Yasuo Nagaokaa,4, Eric Gouauxd, and Hagan Bayleya,1 aDepartment of Chemistry, University of Oxford, Oxford, OX1 3TA, United Kingdom; bDepartment of Medical Biochemistry & Genetics, Texas A&M University System Health Science Center, College Station, TX 77843-1114; cDepartment of Biochemistry and Molecular Biophysics and Howard Hughes Medical Institute, Columbia University, New York, NY 10032; and dVollum Institute and Howard Hughes Medical Institute, Oregon Health and Science University, Portland, OR 97239 Edited by Gregory A. Petsko, Brandeis University, Waltham, MA, and approved March 3, 2010 (received for review December 15, 2009) Engineered protein pores have several potential applications in biotechnology: as sensor elements in stochastic detection and ultrarapid DNA sequencing, as nanoreactors to observe single- molecule chemistry, and in the construction of nano- and micro- devices. One important class of pores contains molecular adapters, which provide internal binding sites for small molecules. Mutants of the α-hemolysin (αHL) pore that bind the adapter β-cyclodextrin (βCD) ∼104 times more tightly than the wild type have been ob- tained. We now use single-channel electrical recording, protein en- gineering including unnatural amino acid mutagenesis, and high- resolution x-ray crystallography to provide definitive structural in- formation on these engineered protein nanopores in unparalleled detail. alpha-hemolysin ∣single molecule ∣stochastic sensing ∣structure ∣ unnatural amino acid M any research groups have used protein engineering to obtain enzymes and antibodies with new properties suited for specific tasks (1–6). Fewer groups have taken on the difficult problem of engineering membrane proteins (7). We have engi- neered the α-hemolysin protein pore, mindful of several potential applications in biotechnology, including its ability to act as a de- tector in stochastic sensing (8) and ultrarapid DNA sequencing (9), to serve as a nanoreactor for the observation of single- molecule chemistry (10) and to act as a component for the con- struction of nano- and microdevices (11). An important breakthrough in this area, which enabled the sto- chastic sensing of organic molecules including the detection of DNA bases in the form of nucleoside monophosphates (12, 13), was the discovery of internal molecular adapters, a form of non- covalent protein modification (14). Most useful have been cyclo- dextrin (CD) adapters, which have until now been used in the absence of detailed structural information about how they work. The present paper is a definitive investigation, which provides such information through the application of a wide variety of technical approaches: single-channel electrical recording, protein engineering including unnatural amino acid mutagenesis, and x-ray crystallography. The studies employing mutagenesis show that the striking interactions seen in the crystal structures also occur in individual pores in lipid bilayers. We reveal that the tight-binding αHL mutants (15) M113N7 and M113F7 bind βCD in different orientations within the hep- tameric pore. In the case of M113N7, the top (primary hydroxyls) of the CD ring faces the trans entrance of the pore. In the case of M113F7, the bottom (secondary hydroxyls) of the CD ring faces the trans entrance, while the top of the ring is bonded to the pore through remarkable CH-π interactions. Another tight-binding mutant, M113V7, can bind the CD in both orientations. These results illustrate the exquisite level of engineering that can be achieved with protein nanopores, which is, for example, far be- yond what is possible with solid-state pores. The work also pro- vides information valuable for the design of new binding sites within the lumen of the αHL pore or within other β-barrel pro- teins. Our results will be of interest to others exploring the inter- actions of CDs with the αHL pore (16, 17), including groups involved in computational studies (18, 19). In addition CDs bind to a variety of other pores, including porins (20, 21) and connex- ins (22), and are being tested in vivo as blockers of the anthrax protective antigen pore (23, 24). The CD adapter concept has also been incorporated into other formats, e.g., with glass nano- pores (25), and artificial pores based on CDs have been made by several groups (26–28). Our work is pertinent to these studies. Results Kinetics and Thermodynamics of the Interactions of βCD with αHL Pores Containing Met, Phe and Asn at Position 113. We showed earlier that position 113 in the αHL pore (Fig. 1A) is critical for the bind- ing of βCD (14). Subsequently, residue 113, which is Met in the WT protein, was changed to each of the remaining 19 naturally occurring amino acids by site-directed mutagenesis (15). We found that 11 of these mutants, expressed as homoheptamers, bound βCD with a similar affinity and with similar kinetics to the WT homoheptamer. Two mutants (P, W) bound βCD about 10 times more strongly than the WT homoheptamer, while six of them (V, H, Y, D, N, F) bound with high affinity, i.e., with a Kd value 103 to 104 times lower than the WT. Remarkably, the side chains of the latter six amino acids bear little resemblance to one another, and this issue is addressed in the present paper. We first examined the two amino acids with the most disparate side chains (Fand N) by making heteromeric pores containing WT (Met-113), M113F, and M113N subunits. Three series of heteroheptamers were produced: WT7−nM113Nn, WT7−nM113Fn, and M113F7−nM113Nn. The heteroheptamers were separated by SDS-polyacrylamide gel electrophoresis aided by an oligoaspartate (D8) tail on the first of the two types of sub- unit (Fig. 1B) (29). All 21 combinations of WT, M113F, and M113N subunits formed αHL pores that interacted with βCD as shown by single-channel current recordings, which revealed the extent of block by βCD (Fig. S1), the association and dissociation Author contributions: A.B., S.C., E.G., and H.B. designed research; A.B., E.M., S.C., L.-Q.G., M.M., and Y.N. performed research; A.B., E.G., and H.B. analyzed data; and A.B. and H.B. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. 1To whom correspondence should be addressed. E-mail: hagan.bayley@chem.ox.ac.uk. 2Present address: Department of Biological Engineering and Dalton Cardiovascular Research Center, University of Missouri, Columbia, MO 65211. 3Present address: Nature Structural & Molecular Biology, 75 Varick Street, 9th Floor, New York NY 10013-1917. 4Present address: Department of Biotechnology, Faculty of Engineering, Kansai University, 3-3-35 Yamate-cho, Suita, Osaka 564-8680, Japan. This article contains supporting information online at www.pnas.org/cgi/content/full/ 0914229107/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.0914229107 PNAS ∣May 4, 2010 ∣vol. 107 ∣no. 18 ∣8165–8170 BIOCHEMISTRY rate constants for βCD (kon and koff), and (from the latter) the equilibrium dissociation constant for βCD (Kd ¼ koff∕kon) (15). The kon values for βCD for the 21 combinations of subunits were all similar at ∼5 × 105 M−1 s−1 (Fig. 1C, Upper). By contrast, the koff values differed widely, ranging from ∼5 × 10−2 s−1 to ∼103 s−1. For WT7−nM113Nn and WT7−nM113Fn, the koff values decreased as M113N or M113F subunits were added. In the case of M113N, there was a steep drop in the value of koff after the fifth subunit had been incorporated. In the case of M113F, the decrease in the value of koff occurred less precipitously as the M113F subunits were added (Fig. 1C, Lower). Intriguingly, with M113F7−nM113Nn, koff first increased as M113N subunits were added to M113F7 until n ¼ 4 (M113F3M113N4) and then de- creased for larger values of n (Fig. 1C, Lower). We recognize that there is more than one permutation of heteromers containing two to five mutant subunits (Fig. 1B), but we have ignored this fact here because no significant differences in the properties of indi- vidual heteromers were observed. For example, 42 recordings were made of WT5M113N2, which has three permutations. Because, kon showed little variation with subunit composition, the variation in Kd was similar to the variation in koff (Fig. 1C). While these studies were in progress, the crystal structures of βCD complexed to M113N7 (Fig. 2B) and M113F7 (Fig. 2C) were solved (Table S1) (30). High-resolution structures could be obtained because the CD and the αHL pore have the same C7 symmetry. In the case of M113N7, βCD is bound with the second- ary hydroxyl face “upward” (Fig. 2B). In each glucose unit of the βCD, the 2-hydroxyl is hydrogen bonded to the side-chain amide of an Asn-113 (the residue introduced by mutagenesis) and the 3-hydroxyl is hydrogen bonded to the ϵ-amino group of Lys-147. In the case of M113F7, two βCDs are bound to the αHL pore (Fig. 2C). It is the top βCD in the structure that concerns us, be- cause it is in contact with the Phe-113 residues introduced by mu- tagenesis. It is immediately apparent that the top βCD in M113F7 is in the opposite orientation to the βCD in M113N7 with each 6-hydroxyl group in a CH-π bonding interaction (31–35) with a Phe-113 side chain. The opposite orientations of the βCDs in M113N7 and M113F7 immediately explain why heteromers formed from similar numbers of M113N and M113F subunits (e.g., M113N4M113F3) bind βCD weakly (see also Discussion). Unnatural Amino Acid Mutagenesis. To further explore the range of noncovalent interactions that are available when βCD binds to the αHL pore, five unnatural amino acids (Fig. 3A and Fig. S2) were incorporated at position 113, by using the in vitro nonsense codon suppression method (36). In particular, we had noted that M113V7 containing the β-branched Val binds βCD tightly (15), and therefore we compared cyclopropylglycine (Cpg) and cyclo- propylalanine (Cpa). We also further examined the means by which M113F7 binds βCD tightly, by comparing the properties of 4-fluorophenylalanine (f1F), pentafluorophenylalanine (f5F), and cyclohexylalanine (Cha) at position 113. The five homomeric pores all produced single-channel cur- rents with unitary conductance values in the range expected for properly assembled heptamers (Fig. S3). All five bound βCD (Fig. 3B, Level 2), either tightly (f1F, Cpg) or weakly (f5F, Cha, Cpa) as described in detail below. During the long βCD binding events, additional current spikes were seen (Fig. 3B). Similar Fig. 1. Binding of βCD by heteromeric αHL pores formed by WT, M113F and M113N subunits. (A) Crystal structure of WT-αHL (61) showing residue 113 (Met, yellow). Left panel, side view and right panel, top view. (B) Separation of 35S-labeled αHL heteroheptamers by SDS-polyacrylamide electrophoresis. The separation of the M113F7-nM113Nn heteromers is shown as detected by autoradiography of a dried gel. The M113F subunits carried a D8 tail. Lane 1, M113N7; lane 2, M113F7−nM113Nn (the heteromers formed from several preparations made with differing ratios of M113F and M113N subunits were mixed to give roughly equal amounts of each subunit combination); lane 3, M113F7. A diagram of the eight different combinations of subunits and their permuta- tions is shown to the right of the autoradiogram. The various permutations are not separated by electrophoresis. (C) Kinetics of the interaction of βCD with single heteromeric αHL pores as determined by bilayer recording. Values of kon were calculated by using kon ¼ 1∕ðτon½βCDÞ, where τon is the mean interevent interval. Values of koff were determined by using koff ¼ 1∕τoff, where τoff is the mean dwell time of βCD in the pore. Values of Kd were calculated by using Kd ¼ koff∕kon. Each point represents the mean  s:d: for three or more determinations. Where they cannot be seen, the s.d. values lie within the symbol. Black squares, WT7−nM113Nn; gray squares, M113F7−nM113Nn; empty squares, M113F7−nWTn. 8166 ∣ www.pnas.org/cgi/doi/10.1073/pnas.0914229107 Banerjee et al. events had been observed previously with certain Met-113 repla- cement mutants and may represent movement of the βCD at its binding site (e.g., rotation about axes perpendicular to the C7 axis) (15). The additional current spikes were more prevalent for M113V7 and M113Cpg7, which may take part in more con- formationally labile interactions with βCD, compared with say M113F7 (Fig. S4). Interactions of βCD with Homoheptamers Bearing Aromatic Residues at Position 113. To further understand the nature of the binding of βCD to aromatic side chains, we examined the kinetics of βCD binding to the homoheptamers containing f1F or f5F at position 113, M113f1F7 and M113f5F7 (Fig. 3C). For both mutants, the value of kon was very similar to that of WT7, but the values of koff and therefore Kd for M113f1F7 differed dramatically from WT7 and were close to the values for the tight-binding mutant M113F7 (Table S2A). By contrast, koff and Kd for M113f5F7 were similar to the values for WT7 (Table S2A). To determine whether M113f1F7 binds βCD in the same orien- tation as M113F7 (Fig. 2C), we made heteromers of the M113f1F subunit with M113N or M113F and examined M113F4M113f1F3 and M113N4M113f1F3. M113F4M113f1F3 binds βCD as strongly as either M113F7 or M113f1F7, but M113N4M113f1F3 binds βCD weakly with a similar affinity to WT7 (Fig. 3D and Table S3). Therefore, it is reasonable to infer that M113F7 and M113f1F7 bind βCD in the same orientation with the 6- hydroxyl groups of the CD in proximity to the aromatic rings on the protein. Cyclohexylalanine (Cha) was used to replace the aromatic side chains with a roughly isosteric hydrophobic group. Again the va- lue of kon for βCD was little changed, but koff for M113Cha7 had an intermediate value of 42  6 s−1. Therefore, M113Cha7 binds βCD more weakly than M113F7 but distinctly more strongly than the WT7 pore (Table S2A and Fig. 3C). Interactions of βCD with Homoheptamers Bearing Hydrophobic Resi- dues at Position 113. M113V7 binds βCD very strongly, and there- fore we compared αHL pores with Cpg or Cpa at position 113. Cpg is roughly isosteric with Val, and like Val has a β-branched side chain. Gratifyingly, M113Cpg7 has a kon value similar to the other αHL pores, and koff and Kd values close to those of M113V7 (Table S2B and Fig. 3E). Cyclopropylalanine (Cpa), with an additional methylene group compared to Cpg, is roughly isosteric with Leu, a weak binder, and M113Cpa7 also binds βCD weakly with kon, koff and Kd values similar to those of WT7 (Table S2B and Fig. 3E). M113I7 and M113T7, which are β-branched, are also weak binders, but Ile and Thr are less closely related to Val than Cpg. To determine whether M113V7 binds βCD in the same orien- tation as M113F7 or M113N7 (Fig. 2), we made heteromers of M113V and the M113N or M113F subunits. M113V3M113F4, M113V4M113F3, M113V3M113N4, and M113V4M113N3 were examined in detail. All four heteroheptamers bound βCD more weakly than M113V7, M113F7 or M113N7 (Fig. 3F and Table S4), suggesting that Val at position 113 interacts with βCD strongly but in a different manner to either Phe or Asn. Each heteromer exhibited a range of Kd values, perhaps reflecting the various pos- sible permutations of the two different subunits around the cen- tral axis of the heptamer, although this heterogeneity was not seen for heteromers made from WT, M113F and M113N (Fig. 1). Discussion Soon after we discovered that βCD binds to the WT-αHL pore for around a millisecond, we found a mutant pore, M113N7, that re- leases βCD ∼104 times more slowly (14). This prompted us to examine all 19 mutants in which residue 113 is replaced by a nat- ural amino acid, with the surprising result that a collection of ami- no acids with structurally unrelated side chains (V, H, Y, D, N, F) are tight binders (15). Here, we have examined the nature of the binding interactions more closely by single-channel electrical re- cording, protein engineering including unnatural amino acid mu- tagenesis, and high-resolution x-ray crystallography, and we provide the first definitive structural information on an engi- neered protein nanopore. We find that βCD can bind tightly to the αHL pore in three different ways depending on the residue at 113, as exemplified by Asn, Phe, and Val. Because Asn and Phe have quite different side chains, we first compared the ability of M113N and M113F subunits to take part in binding the CD. The examination of het- eromeric proteins containing WT (Met-113), M113N and M113F subunits showed that the replacement of WT subunits in WT7 with M113N or M113F subunits led to increased affinity for βCD. The more M113N or M113F subunits that were added, the tighter binding became. By contrast, when subunits in M113N7 were replaced with M113F subunits, binding became weaker, reaching a minimum at three to four M113F subunits, and then increasing in strength with five M113F subunits or more (Fig. 1C). Parallel structural studies (30) revealed the basis of the “oppos- ing” effects of the M113N and M113F subunits. βCD binds to M113N7 in the opposite orientation to that in which it binds to M113F7. In M113N7, the secondary hydroxyls in the βCD ring are hydrogen bonded to Lys-147 and Asn-113 (Fig. 2A). By con- trast, βCD interacts with M113F7 through its primary hydroxyl face (Fig. 2B). It seemed likely that M113V7, bound βCD in yet another way, and this was examined by forming heteromers between M113V and M113N or M113F. The presence of three or four subunits of either M113N or M113F greatly decreases the affinity of the pore for βCD (Fig. 3F), with an average koff of 7.3 × 101 s−1, indicating that a third binding mode is indeed operating Fig. 2. X-ray structures of M113N and M113F homoheptamers with βCD bound. (A) Side view of heptameric αHL. βCD binds in the blue highlighted region. (B) βCD bound to M113N7 (dotted lines indicate hy- drogen bonding). The side chains of Lys-147 are in pale brown and the side chains of Asn- 113 in yellow. (C) βCD bound to M113F7 (dotted lines indicate CH-π bonding). The side chains of Phe-113 are in yellow. The sec- ond βCD in the M113F7 · ðβCDÞ2 structure is hydrogen bonded to the top βCD in a head- to-head arrangement and has no apparent interactions with the protein. For both (B) and (C), four β strands were omitted from the barrel to give a better view. Banerjee et al. PNAS ∣ May 4, 2010 ∣ vol. 107 ∣ no. 18 ∣ 8167 BIOCHEMISTRY (Table S4). In summary, the three groups of tight-binding mutants comprise αHL pores incorporating, at position 113: (i) the hydro- gen-bonding amino acids N, D (the latter would have to be largely in the protonated form), and possibly H; (ii) the aromatics F, Y, f1F, and possibly H, and more weakly W; (iii) the β-branched ami- no acids V, Cpg. There may be yet other means by which CDs can bind to the αHL pore. For example, we earlier found that hepta- 6-sulfato-βCD can bind tightly to αHL pores containing the N139Q mutation (37). Presumably, this CD is bound at a site low- er down in the β barrel in a fashion that includes hydrogen bond- ing to the Gln at position 139. While the various mutants exhibited widely different koff values, the value of kon was almost invariant and averaged ∼2.3 × 105 M−1 s−1 (Table S2) (15). Ap- parently, transport to the binding site is rate limiting, through a route unaffected by mutagenesis. koff increased precipitously with the addition of WTsubunits to M113N7 (Fig. 1C). Crystal structures of M113N7 show that resi- dues 111, 113, and 147 are reorganized by compari- son with WT7 and then undergo a more limited rearrangement when βCD binds (Fig. S5). For example, the side chain of Lys-147 shifts position to form a bifurcated hydrogen bond with a 3-hydroxyl group of βCD and the side chain carbonyl of an Asn- 113 (Fig. S6). Therefore, the side chains of residues 111, 113, and 147 might be in a variety of conformations in WT7−nM113Nn het- eromers and offer less well preorganized binding sites for βCD than they do in M113N7. Further, the intramolecular hydrogen bonds of the secondary hydroxyls in βCD (38) must be disrupted upon binding as both hydroxyls on each glucose ring form hydro- gen bonds to the mutant subunits (Fig. 2B). Because the hydrogen bonds that are broken in βCD are arranged in a circle, the break- age of bonds involving a single glucose (three bonds in all) will be relatively more disruptive than those involving adjoining glucose residues or the entire circle. The overall binding cooperativity in M113N7 could be attributed to enthalpic cooperativity outweigh- ing entropic penalties to binding (39). Positive cooperativity has been observed previously in fairly rigid model systems (40). By contrast with M113N7, there is little movement of side chains in ðM113FÞ7 by comparison with WT7 and little move- ment, including Phe-113, upon binding βCD (Fig. S7A). Further, the crystal structure of M113F7 · βCD suggests that each Phe re- sidue interacts independently with the βCD through what appear to be CH-π interactions (Fig. S7B). These interactions are ex- pected to be weak and not strongly directional and hence offer less enthalpic cooperativity, as supported by the B-factors (crys- tallographic temperatures factors) at the primary βCD binding site, which are between ∼40 and 50. Positive cooperativity is ob- served, but it is less pronounced than in the case of M113N7 (Table S5). In the case of M113N7, the B-factors of the residues that bind βCD are in the 20s implying that the βCD is more rigidly held than it is in M113F7. The binding of sugars to aromatic residues in proteins can in- clude CH-π bonding (41) or OH-π bonding or a finely balanced complement of both (42, 43). However, we have dismissed the possibility of an OH-π interaction between Phe-113 and the 6-hydroxyl groups of βCD as the distance between the center of the phenyl rings to the nearest hydroxyl oxygen is higher (5.2  0.65 Å, n ¼ 7) than that expected for a favorable OH-π interaction (33). While we propose that βCD binds to Phe-113 through a C-6 CH-π interaction (Fig. S7B), the distances between the center of the Phe-113 ring and the nearest C-6 of βCD ob- served in the M113F7 · βCD structure (4.66  0.24 Å, n ¼ 7) are in the upper range of the expected distance for a strong inter- action, which is ∼4.5 Å (33). The angle between the normal to the aromatic rings and the line connecting the C-6 atoms to the aro- matic midpoint is 8.0  5.6°, which is well within the expected range (44). The measurements with M113f5F7 argue against a hydrophobic interaction between Phe residues at position 113 and the βCD ring. In f5F, the hydrophobicity of the phenyl ring is significantly increased (45) yet M113f5F7 binds βCD weakly, like WT7 (Fig. 3C and Table S2A). By contrast with F, f1F, Y and N, homomeric αHL pores with f5F and W at position 113 bound βCD relatively weakly (Fig. 3C and Table S2A). In the case of f5F, the powerful electron with- drawing action of the five fluorine atoms leaves a highly increased positive charge at the center of the ring (46, 47), mitigating against a hydrogen-bonding interaction. The electron-rich Trp Fig. 3. Properties of pores containing natural and unnatural amino acid sub- stitutions at position 113. The data were recorded at þ40 mV in 1.0 M NaCl, 10 mM sodium phosphate, pH 7.5. (A) Unnatural amino acids used in this study: 4-fluorophenylalanine, f1F; pentafluorophenylalanine, f5F; cyclohex- ylalanine, Cha; cyclopropylglycine, Cpg; cyclopropylalanine, Cpa. (B) Repre- sentative current traces from single homoheptameric αHL pores, containing unnatural amino acids at position 113, in the presence of βCD. βCD (40 μM final) was added to the trans chamber. Level 1, open pore current; level 2, pore occupied by βCD. The broken line indicates zero current. (C) In- teraction of βCD with homomeric αHL pores containing aromatic amino acids at position 113. Kd values for the interaction between βCD and the αHL pore were calculated by using Kd ¼ koff∕kon. Each column represents the mean  s:d: for 10 or more determinations: dark gray, natural amino acids; light gray, unnatural amino acids. Data adapted from Gu and colleagues (15) are marked (*). (D) Representative current traces from single-channel recordings of βCD binding to M113F4M113f1F3 and M113N4M113f1F3. βCD (40 μM final) was added to the trans chamber. The broken line indicates zero current. (E) Interaction of βCD with homomeric αHL pores containing hydrophobic amino acids at position 113. Kd values for the interaction between βCD and the αHL pore were calculated by using Kd ¼ koff∕kon. Each column represents the mean  s:d: for ten or more determinations: dark gray, natural amino acids; light gray, unnatural amino acids. Data adapted from Gu and colleagues (15) are marked (*). (F) koff values for βCD from heteroheptamers formed with M113F and M113V subunits and with M113N and M113V subunits. βCD (40 μM final) was added to the trans chamber. The kon values for βCD for all these mutants are similar, at ∼3 × 105 M−1 s−1. Empty square: average koff values for the mutant (bar is s:d). Filled square: M113V3M113F4; filled circle: M113V4M113F3; filled upright triangle: M113V3M113N4; filled in- verted triangle: M113V4M113N3. 8168 ∣ www.pnas.org/cgi/doi/10.1073/pnas.0914229107 Banerjee et al. ring (44, 46, 47) should favor hydrogen bonding, but here we can- not make a direct comparison with the crystal structure of M113F7 as the indole ring is far larger than benzene. It is possible that it cannot become oriented in the same manner and that it is misaligned for hydrogen bonding. Our experiments suggest that M113V7 and M113Cpg7 bind βCD in a third way. In heteromers with M113V, both M113F and M113N reduce the affinity of the pore for βCD suggesting that neither the CH-π interaction with Phe-113 nor the hydrogen- bonding interactions with Asn-113 and Lys-147 are compatible with binding to Val. Close interactions of Val with glucose rings have been noted previously (48). Therefore, we propose that the Val side-chain interacts with the side of the glucose ring. This in- teraction might occur in one or both orientations of the CD ring (Fig. 4). Conclusion We provide structural information on engineered protein nano- pores and describe three distinct ways in which βCD can bind within the lumen of mutant αHL pores in atomic detail. Our re- sults will be useful in several areas of basic science and biotech- nology. By using host molecules lodged within the αHL pore, host-guest interactions can be investigated in fine detail at the single-molecule level (17, 49). The present work will now permit us to examine binding events at a known face of a CD. The work also provides information for designing new binding sites within the lumen of the αHL pore (37) or within other β barrel proteins (21, 50) and for using molecular design to devise ways in which to covalently attach CDs within pores (13, 51). These areas impact practical applications of nanopore technology including stochas- tic sensing (8), single-molecule DNA sequencing (9, 12, 13, 52), the use of nanoreactors for the observation of single-molecule chemistry (10), and the construction of nano- and microdevices (11, 53), as well as the design of CDs as therapeutic agents (23, 24). Methods Full details of the experimental procedures can be found in SI Appendix. Materials L-Amino acids were obtained as follows: 4-fluorophenylalanine (f1F) (Fluka); pentafluorophenylalanine (f5F) (PepTech Corp.); cyclopropylglycine (Cpg) (Ty- ger); cyclopropylalanine (Cpa) (Tyger). 4-N-benzoyl-5′-O-(4,4′-dimethoxytri- tyl)-2′-deoxycytidine-3′-[(2-cyanoethyl)-(N,N-diisopropyl)]-phosphoramidite and bis(2-cyanoethyl)-N,N-diisopropylphosphoramidite for the synthesis of pdCpA were purchased from Glen Research and Toronto Research Chemicals, respectively. Preparation of NVOC-Protected Aminoacyl-pdCpA. NVOC-protected aminoacyl-pdCpAs were prepared as reported previously by reacting the dinucleotide pdCpA with N-protected, carboxylic acid-activated, amino acids (54–56). Preparation of NVOC-Protected Aminoacyl-tRNA. NVOC-protected aminoacyl- pdCpAs were ligated enzymatically with a truncated tRNA, prepared by using methods described elsewhere (57, 58). Genetic Constructs and Mutagenesis. All new αHL constructs were verified by DNA sequencing. Details of each construct can be found in SI Appendix. Synthesis, Assembly, and Purification of Mutant αHL pores. αHL monomers (WT and mutants) were prepared in vitro by coupled transcription and translation (IVTT) and assembled into homoheptamers on rabbit red blood cell membranes followed by purification by SDS–PAGE as described earlier (59). Heteroheptamers were prepared by mixing the two required DNAs (one encoding an αHL with a D8 tail) before IVTT and then oligomerizing the mixed translation products on rabbit red blood cell membranes. Pores with the desired combinations of subunits were purified by SDS–PAGE (59). Synthesis, Assembly, and Purification of αHL Mutants Containing Unnatural Ami- no Acids. αHL polypeptides containing unnatural amino acids were synthe- sized by IVTT in the presence of rabbit red blood cell membranes. The plasmid with a stop codon (TAG) at position 113 was used. Deprotected ami- noacyl-tRNAs (SI Appendix) were added to the IVTT mixtures. For heterohep- tamers with subunits containing unnatural amino acids in combination with M113N or M113F, monomers were first made, which were then coassembled on rabbit red blood cell membranes and subsequently purified by SDS–PAGE. Single-Channel Current Recordings in Planar Lipid Bilayers. (15, 60) Recordings were made with 1.0 M NaCl, 10 mM sodium phosphate, pH 7.5, in both cham- bers, at an applied potential of þ40 mV. Data were recorded at 22  2°C. The bilayer was formed from 1,2-diphytanoyl-sn-glycero-phosphocholine (Avanti Polar Lipids). Proteins were added to the cis chamber, and βCD to the trans chamber. Single-channel currents were recorded with an Axopatch 200B patch-clamp amplifier (Axon Instruments) and filtered at 2 kHz with a built-in 4-pole low-pass Bessel Filter. The data were acquired at a sampling rate of 10 kHz. For mutants that bind βCD strongly, the data were acquired for at least 30 min and for weak-binding mutants for at least 10 min. Kinetic Data Analysis. Current amplitude and dwell-time histograms were made by using ClampFit 9.0. The mean dwell times, τoff, were determined by fitting the dwell-time histograms to single exponentials. Values of kon and koff were obtained by using the mean dwell times and mean interevent intervals, as described previously (15, 60). This analysis assumes a binary in- teraction, which was supported in all cases examined by the finding of only one major blockade level and a single exponential distribution of dwell times (τoff). Fig. 4. Molecular model showing the three classes of interaction between the αHL pore and βCD identified in this work. The model identifies the region of βCD responsible for each interaction (H atoms interacting with Phe-113 or Asn-113 and Lys-147: gray). The first class of interaction is with aromatic residues and involves the seven -CH2OH groups of the βCD. The second class is typified by the interactions with Asn at position 113, which involve hydro- gen-bonds to the secondary 2-hydroxyls of the βCD. Structural studies show that this interaction is supported by hydrogen bonding between Lys-147 and the secondary 3-hydroxyls of the βCD. Structural studies and experiments with heteromers suggest that the βCD in M113F7 is in the opposite orienta- tion to the βCD in M113N7, in support of the model shown here. As the inter- action with Val is hydrophobic, it is not directional and βCD may not bind at the same position inside the β barrel as it does in M113F7 or M113N7. Banerjee et al. PNAS ∣ May 4, 2010 ∣ vol. 107 ∣ no. 18 ∣ 8169 BIOCHEMISTRY Protein Crystallography. Details can be found in SI Appendix. Protein Data Bank: The coordinates and structure factors of the described structures have been deposited with accession codes 3M2L ðM113F7Þ, 3M3R ðM113F7 · βCDÞ, 3M4D ðM113N7Þ, 3M4E ðM113N7 · βCDÞ. ACKNOWLEDGMENTS. We thank Dennis Dougherty for the plasmid pTHG73. This work was funded by a Royal Society Wolfson Research Merit Award (to H.B.), the Medical Research Council (H.B.), the National Institutes of Health (H.B.), and the Howard Hughes Medical Institute (E.G.). 1. Lu Y, Yeung N, Sieracki N, Marshall NM (2009) Design of functional metalloproteins. Nature 460:855–862. 2. Gebauer M, Skerra A (2009) Engineered protein scaffolds as next-generation antibody therapeutics. Curr Opin Chem Biol 13:245–255. 3. Gronwall C, Stahl S (2009) Engineered affinity proteins—Generation and applications. J Biotechnol 140:254–269. 4. Arnold U (2009) Incorporation of non-natural modules into proteins: Structural features beyond the genetic code. Biotechnol Lett 31:1129–1139. 5. Tracewell CA, Arnold FH (2009) Directed enzyme evolution: Climbing fitness peaks one amino acid at a time. Curr Opin Chem Biol 13:3–9. 6. Fruk L, Kuo CH, Torres E, Niemeyer CM (2009) Apoenzyme reconstitution as a chemical tool for structural enzymology and biotechnology. Angew Chem Int Ed Engl 48:1550–1574. 7. Bayley H, Jayasinghe L (2004) Functional engineered channels and pores. Mol Membr Biol 21:209–220. 8. Bayley H, Cremer PS (2001) Stochastic sensors inspired by biology. Nature 413:226–230. 9. Branton D, et al. (2008) The potential and challenges of nanopore sequencing. Nature Biotechnol 26:1146–1153. 10. Bayley H, Luchian T, Shin S-H, Steffensen MB (2008) Single Molecules and Nanotech- nology, eds R Rigler and H Vogel (Springer, Heidelberg), pp 251–277. 11. Maglia G, et al. (2009) Droplet networks with incorporated protein diodes show collective properties. Nat Nanotechnol 4:437–440. 12. Astier Y, Braha O, Bayley H (2006) Toward single molecule DNA sequencing: Direct identification of ribonucleoside and deoxyribonucleoside 5'-monophosphates by using an engineered protein nanopore equipped with a molecular adapter. J Am Chem Soc 128:1705–1710. 13. Clarke J, et al. (2009) Continuous base identification for single-molecule nanopore DNA sequencing. Nature Nanotechnol 4:265–270. 14. Gu L-Q, Braha O, Conlan S, Cheley S, Bayley H (1999) Stochastic sensing of organic analytes by a pore-forming protein containing a molecular adapter. Nature 398:686–690. 15. Gu L-Q, Cheley S, Bayley H (2001) Prolonged residence time of a noncovalent molecular adapter, β-cyclodextrin, within the lumen of mutant α-hemolysin pores. J Gen Physiol 118:481–494. 16. Ervin EN, Kawano R, White RJ, White HS (2008) Simultaneous alternating and direct current readout of protein ion channel blocking events using glass nanopore membranes. Anal Chem 80:2069–2076. 17. Gurnev PA, Harries D, Parsegian VA, Bezrukov SM (2009) The dynamic side of the Hof- meister effect: A single-molecule nanopore study of specific complex formation. ChemPhysChem 10:1445–1449. 18. Mamonova T, Kurnikova M (2006) Structure and energetics of channel-forming protein-polysaccharide complexes inferred via computational statistical thermody- namics. J Phys Chem B 110:25091–25100. 19. Egwolf B, Luo Y, Walters DE, Roux B (2010) Ion selectivity of alpha-hemolysin with beta-cyclodextrin adapter. II. Multi-ion effects studied with grand canonical monte carlo/brownian dynamics simulations. J Phys Chem B 114:2901–2909. 20. Orlik F, et al. (2003) CymA of Klebsiella oxytoca outer membrane: Binding of cyclodex- trins and study of the current noise of the open membrane. Biophys J 85:876–885. 21. Chen M, Khalid S, Sansom MSP, Bayley H (2008) OmpG: Engineering a quiet pore for biosensing. Proc Natl Acad Sci USA 105:6272–6277. 22. Locke D, Koreen IV, Liu JY, Harris AL (2004) Reversible pore block of connexin channels by cyclodextrins. J Biol Chem 279:22883–22892. 23. Karginov VA, Nestorovich EM, Moayeri M, Leppla SH, Bezrukov SM (2005) Blocking anthrax lethal toxin at the protective antigen channel by using structure-inspired drug design. Proc Natl Acad Sci USA 102:15075–15080. 24. Moayeri M, Robinson TM, Leppla SH, Karginov VA (2008) In vivo efficacy of beta- cyclodextrin derivatives against anthrax lethal toxin. Antimicrob Agents Chemother 52:2239–2241. 25. Gao C, Ding S, Tan Q, Gu LQ (2009) Method of creating a nanopore-terminated probe for single-molecule enantiomer discrimination. Anal Chem 81:80–86. 26. Pregel MJ, Jullien L, Lehn J-M (1992) Towards artificial ion channels: Transport of alkali metal ions across liposomal membranes by “bouquet” molecules. Angew Chem Int Edit 31:1637–1639. 27. Bacri L, Benkhaled A, Guegan P, Auvray L (2005) Ionic channel behavior of modified cyclodextrins inserted in lipid membranes. Langmuir 21:5842–5846. 28. Jog PV, Gin MS (2008) A light-gated synthetic ion channel. Org Lett 10:3693–3696. 29. Howorka S, Cheley S, Bayley H (2001) Sequence-specific detection of individual DNA strands using engineered nanopores. Nat Biotechnol 19:636–639. 30. Montoya M (2004) Insights into Membrane Association and Bioengineering of a Pore- Forming Toxin: Structural Studies of Staphylococcal α-Hemolysin (Columbia University, New York). 31. Steiner T (2002) The hydrogen bond in the solid state. Angew Chem Int Ed 41:49–76. 32. Steiner T (2002) Hydrogen bonds from water molecules to aromatic acceptors in very high-resolution protein crystal structures. Biophys Chem 95:195–201. 33. Steiner T, Koellner G (2001) Hydrogen bonds with pi-acceptors in proteins: Frequencies and role in stabilizing local 3D structures. J Mol Biol 305:535–557. 34. Brandl M, Weiss MS, Jabs A, Sühnel J, Hilgenfeld R (2001) CH...π-interactions in proteins. J Mol Biol 307:357–377. 35. Weiss MS, Brandl M, Suhnel J, Pal D, Hilgenfeld R (2001) More hydrogen bonds for the (structural) biologist. Trends Biochem Sci 26:521–523. 36. Wang L, Xie J, Schultz PG (2006) Expanding the genetic code. Annu Rev Biophys Biomol Struct 35:225–249. 37. Gu L-Q, Cheley S, Bayley H (2001) Capture of a single molecule in a nanocavity. Science 291:636–640. 38. Saenger W, et al. (1998) Structures of the common cyclodextrins and their larger analogues—Beyond the doughnut. Chem Rev 98:1787–1802. 39. Hunter CA, Tomas S (2003) Cooperativity, partially bound states, and enthalpy-entropy compensation. Chem Biol 10:1023–1032. 40. Bisson AP, Hunter CA (1996) Cooperativity in the assembly of zipper complexes. Chem Commun 1723–1724. 41. del Carmen Fernandez-Alonso M, Canada FJ, Jimenez-Barbero J, Cuevas G (2005) Molecular recognition of saccharides by proteins. Insights on the origin of the carbohydrate-aromatic interactions. J Am Chem Soc 127:7379–7386. 42. Jimenez-Barbero J, Asensio JL, Canada FJ, Poveda A (1999) Free and protein-bound carbohydrate structures. Curr Opin Struct Biol 9:549–555. 43. Stanca-Kaposta EC, et al. (2007) Carbohydrate molecular recognition: A spectroscopic investigation of carbohydrate-aromatic interactions. Phys Chem Chem Phys 9:4444–4451. 44. Brandl M, Weiss MS, Jabs A, Suhnel J, Hilgenfeld R (2001) CH...π-interactions in proteins. J Mol Biol 307:357–377. 45. Woll MG, Hadley EB, Mecozzi S, Gellman SH (2006) Stabilizing and destabilizing effects of phenylalanine –>F5-phenylalanine mutations on the folding of a small protein. J Am Chem Soc 128:15932–15933. 46. Mecozzi S, West AP, Jr, Dougherty DA (1996) Cation-pi interactions in aromatics of biological and medicinal interest: Electrostatic potential surfaces as a useful qualitative guide. Proc Natl Acad Sci USA 93:10566–10571. 47. Dougherty DA (2008) Cys-loop neuroreceptors: Structure to the rescue?. Chem Rev 108:1642–1653. 48. Hondoh H, et al. (2008) Substrate recognition mechanism of alpha-1,6-glucosidic linkage hydrolyzing enzyme, dextran glucosidase from Streptococcus mutans. J Mol Biol 378:913–922. 49. Kang XF, Cheley S, Guan X, Bayley H (2006) Stochastic detection of enantiomers. J Am Chem Soc 128:10684–10685. 50. Chen M, Li QH, Bayley H (2008) Orientation of the monomeric porin OmpG in planar lipid bilayers. ChemBioChem 9:3029–3036. 51. Wu H-C, Astier Y, Maglia G, Mikhailova E, Bayley H (2007) Protein nanopores with covalently attached molecular adapters. J Am Chem Soc 129:16142–16148. 52. Bayley H (2006) Sequencing single molecules of DNA. Curr Opin Chem Biol 10:628–637. 53. Astier Y, Bayley H, Howorka S (2005) Protein components for nanodevices. Curr Opin Chem Biol 9:576–584. 54. Robertson SA, Noren CJ, Anthony-Cahill SJ, Griffith MC, Schultz PG (1989) The use of 5'-phospho-2 deoxyribocytidylylriboadenosine as a facile route to chemical aminoacy- lation of tRNA. Nucleic Acids Res 17:9649–9660. 55. Ellman J, Mendel D, Anthony-Cahill S, Noren CJ, Schultz PG (1991) Biosynthetic method for introducing unnatural amino acids site-specifically into proteins. Method Enzymol 202:301–336. 56. Kearney PC, et al. (1996) Dose-response relations for unnatural amino acids at the agonist binding site of the nicotinic acetylcholine receptor: Tests with novel side chains and with several agonists. Mol Pharmacol 50:1401–1412. 57. England TE, Bruce AG, Uhlenbeck OC (1980) Specific labeling of 3′ termini of RNA with T4 RNA ligase. Method Enzymol 65:65–74. 58. Nowak MW, et al. (1998) In vivo incorporation of unnatural amino acids into ion channels in xenopus oocyte expression system. Method Enzymol 293:504–529. 59. Cheley S, Braha O, Lu X, Conlan S, Bayley H (1999) A functional protein pore with a “retro” transmembrane domain. Protein Sci 8:1257–1267. 60. Gu L-Q, et al. (2000) Reversal of charge selectivity in transmembrane protein pores by using non-covalent molecular adapters. Proc Natl Acad Sci USA 97:3959–3964. 61. Song L, et al. (1996) Structure of staphylococcal α-hemolysin, a heptameric transmem- brane pore. Science 274:1859–1865. 8170 ∣ www.pnas.org/cgi/doi/10.1073/pnas.0914229107 Banerjee et al.
3M4E
Crystal structure of the M113N mutant of alpha-hemolysin bound to beta-cyclodextrin
Molecular bases of cyclodextrin adapter interactions with engineered protein nanopores Arijit Banerjeea, Ellina Mikhailovaa, Stephen Cheleya, Li-Qun Gub,2, Michelle Montoyac,3, Yasuo Nagaokaa,4, Eric Gouauxd, and Hagan Bayleya,1 aDepartment of Chemistry, University of Oxford, Oxford, OX1 3TA, United Kingdom; bDepartment of Medical Biochemistry & Genetics, Texas A&M University System Health Science Center, College Station, TX 77843-1114; cDepartment of Biochemistry and Molecular Biophysics and Howard Hughes Medical Institute, Columbia University, New York, NY 10032; and dVollum Institute and Howard Hughes Medical Institute, Oregon Health and Science University, Portland, OR 97239 Edited by Gregory A. Petsko, Brandeis University, Waltham, MA, and approved March 3, 2010 (received for review December 15, 2009) Engineered protein pores have several potential applications in biotechnology: as sensor elements in stochastic detection and ultrarapid DNA sequencing, as nanoreactors to observe single- molecule chemistry, and in the construction of nano- and micro- devices. One important class of pores contains molecular adapters, which provide internal binding sites for small molecules. Mutants of the α-hemolysin (αHL) pore that bind the adapter β-cyclodextrin (βCD) ∼104 times more tightly than the wild type have been ob- tained. We now use single-channel electrical recording, protein en- gineering including unnatural amino acid mutagenesis, and high- resolution x-ray crystallography to provide definitive structural in- formation on these engineered protein nanopores in unparalleled detail. alpha-hemolysin ∣single molecule ∣stochastic sensing ∣structure ∣ unnatural amino acid M any research groups have used protein engineering to obtain enzymes and antibodies with new properties suited for specific tasks (1–6). Fewer groups have taken on the difficult problem of engineering membrane proteins (7). We have engi- neered the α-hemolysin protein pore, mindful of several potential applications in biotechnology, including its ability to act as a de- tector in stochastic sensing (8) and ultrarapid DNA sequencing (9), to serve as a nanoreactor for the observation of single- molecule chemistry (10) and to act as a component for the con- struction of nano- and microdevices (11). An important breakthrough in this area, which enabled the sto- chastic sensing of organic molecules including the detection of DNA bases in the form of nucleoside monophosphates (12, 13), was the discovery of internal molecular adapters, a form of non- covalent protein modification (14). Most useful have been cyclo- dextrin (CD) adapters, which have until now been used in the absence of detailed structural information about how they work. The present paper is a definitive investigation, which provides such information through the application of a wide variety of technical approaches: single-channel electrical recording, protein engineering including unnatural amino acid mutagenesis, and x-ray crystallography. The studies employing mutagenesis show that the striking interactions seen in the crystal structures also occur in individual pores in lipid bilayers. We reveal that the tight-binding αHL mutants (15) M113N7 and M113F7 bind βCD in different orientations within the hep- tameric pore. In the case of M113N7, the top (primary hydroxyls) of the CD ring faces the trans entrance of the pore. In the case of M113F7, the bottom (secondary hydroxyls) of the CD ring faces the trans entrance, while the top of the ring is bonded to the pore through remarkable CH-π interactions. Another tight-binding mutant, M113V7, can bind the CD in both orientations. These results illustrate the exquisite level of engineering that can be achieved with protein nanopores, which is, for example, far be- yond what is possible with solid-state pores. The work also pro- vides information valuable for the design of new binding sites within the lumen of the αHL pore or within other β-barrel pro- teins. Our results will be of interest to others exploring the inter- actions of CDs with the αHL pore (16, 17), including groups involved in computational studies (18, 19). In addition CDs bind to a variety of other pores, including porins (20, 21) and connex- ins (22), and are being tested in vivo as blockers of the anthrax protective antigen pore (23, 24). The CD adapter concept has also been incorporated into other formats, e.g., with glass nano- pores (25), and artificial pores based on CDs have been made by several groups (26–28). Our work is pertinent to these studies. Results Kinetics and Thermodynamics of the Interactions of βCD with αHL Pores Containing Met, Phe and Asn at Position 113. We showed earlier that position 113 in the αHL pore (Fig. 1A) is critical for the bind- ing of βCD (14). Subsequently, residue 113, which is Met in the WT protein, was changed to each of the remaining 19 naturally occurring amino acids by site-directed mutagenesis (15). We found that 11 of these mutants, expressed as homoheptamers, bound βCD with a similar affinity and with similar kinetics to the WT homoheptamer. Two mutants (P, W) bound βCD about 10 times more strongly than the WT homoheptamer, while six of them (V, H, Y, D, N, F) bound with high affinity, i.e., with a Kd value 103 to 104 times lower than the WT. Remarkably, the side chains of the latter six amino acids bear little resemblance to one another, and this issue is addressed in the present paper. We first examined the two amino acids with the most disparate side chains (Fand N) by making heteromeric pores containing WT (Met-113), M113F, and M113N subunits. Three series of heteroheptamers were produced: WT7−nM113Nn, WT7−nM113Fn, and M113F7−nM113Nn. The heteroheptamers were separated by SDS-polyacrylamide gel electrophoresis aided by an oligoaspartate (D8) tail on the first of the two types of sub- unit (Fig. 1B) (29). All 21 combinations of WT, M113F, and M113N subunits formed αHL pores that interacted with βCD as shown by single-channel current recordings, which revealed the extent of block by βCD (Fig. S1), the association and dissociation Author contributions: A.B., S.C., E.G., and H.B. designed research; A.B., E.M., S.C., L.-Q.G., M.M., and Y.N. performed research; A.B., E.G., and H.B. analyzed data; and A.B. and H.B. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. 1To whom correspondence should be addressed. E-mail: hagan.bayley@chem.ox.ac.uk. 2Present address: Department of Biological Engineering and Dalton Cardiovascular Research Center, University of Missouri, Columbia, MO 65211. 3Present address: Nature Structural & Molecular Biology, 75 Varick Street, 9th Floor, New York NY 10013-1917. 4Present address: Department of Biotechnology, Faculty of Engineering, Kansai University, 3-3-35 Yamate-cho, Suita, Osaka 564-8680, Japan. This article contains supporting information online at www.pnas.org/cgi/content/full/ 0914229107/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.0914229107 PNAS ∣May 4, 2010 ∣vol. 107 ∣no. 18 ∣8165–8170 BIOCHEMISTRY rate constants for βCD (kon and koff), and (from the latter) the equilibrium dissociation constant for βCD (Kd ¼ koff∕kon) (15). The kon values for βCD for the 21 combinations of subunits were all similar at ∼5 × 105 M−1 s−1 (Fig. 1C, Upper). By contrast, the koff values differed widely, ranging from ∼5 × 10−2 s−1 to ∼103 s−1. For WT7−nM113Nn and WT7−nM113Fn, the koff values decreased as M113N or M113F subunits were added. In the case of M113N, there was a steep drop in the value of koff after the fifth subunit had been incorporated. In the case of M113F, the decrease in the value of koff occurred less precipitously as the M113F subunits were added (Fig. 1C, Lower). Intriguingly, with M113F7−nM113Nn, koff first increased as M113N subunits were added to M113F7 until n ¼ 4 (M113F3M113N4) and then de- creased for larger values of n (Fig. 1C, Lower). We recognize that there is more than one permutation of heteromers containing two to five mutant subunits (Fig. 1B), but we have ignored this fact here because no significant differences in the properties of indi- vidual heteromers were observed. For example, 42 recordings were made of WT5M113N2, which has three permutations. Because, kon showed little variation with subunit composition, the variation in Kd was similar to the variation in koff (Fig. 1C). While these studies were in progress, the crystal structures of βCD complexed to M113N7 (Fig. 2B) and M113F7 (Fig. 2C) were solved (Table S1) (30). High-resolution structures could be obtained because the CD and the αHL pore have the same C7 symmetry. In the case of M113N7, βCD is bound with the second- ary hydroxyl face “upward” (Fig. 2B). In each glucose unit of the βCD, the 2-hydroxyl is hydrogen bonded to the side-chain amide of an Asn-113 (the residue introduced by mutagenesis) and the 3-hydroxyl is hydrogen bonded to the ϵ-amino group of Lys-147. In the case of M113F7, two βCDs are bound to the αHL pore (Fig. 2C). It is the top βCD in the structure that concerns us, be- cause it is in contact with the Phe-113 residues introduced by mu- tagenesis. It is immediately apparent that the top βCD in M113F7 is in the opposite orientation to the βCD in M113N7 with each 6-hydroxyl group in a CH-π bonding interaction (31–35) with a Phe-113 side chain. The opposite orientations of the βCDs in M113N7 and M113F7 immediately explain why heteromers formed from similar numbers of M113N and M113F subunits (e.g., M113N4M113F3) bind βCD weakly (see also Discussion). Unnatural Amino Acid Mutagenesis. To further explore the range of noncovalent interactions that are available when βCD binds to the αHL pore, five unnatural amino acids (Fig. 3A and Fig. S2) were incorporated at position 113, by using the in vitro nonsense codon suppression method (36). In particular, we had noted that M113V7 containing the β-branched Val binds βCD tightly (15), and therefore we compared cyclopropylglycine (Cpg) and cyclo- propylalanine (Cpa). We also further examined the means by which M113F7 binds βCD tightly, by comparing the properties of 4-fluorophenylalanine (f1F), pentafluorophenylalanine (f5F), and cyclohexylalanine (Cha) at position 113. The five homomeric pores all produced single-channel cur- rents with unitary conductance values in the range expected for properly assembled heptamers (Fig. S3). All five bound βCD (Fig. 3B, Level 2), either tightly (f1F, Cpg) or weakly (f5F, Cha, Cpa) as described in detail below. During the long βCD binding events, additional current spikes were seen (Fig. 3B). Similar Fig. 1. Binding of βCD by heteromeric αHL pores formed by WT, M113F and M113N subunits. (A) Crystal structure of WT-αHL (61) showing residue 113 (Met, yellow). Left panel, side view and right panel, top view. (B) Separation of 35S-labeled αHL heteroheptamers by SDS-polyacrylamide electrophoresis. The separation of the M113F7-nM113Nn heteromers is shown as detected by autoradiography of a dried gel. The M113F subunits carried a D8 tail. Lane 1, M113N7; lane 2, M113F7−nM113Nn (the heteromers formed from several preparations made with differing ratios of M113F and M113N subunits were mixed to give roughly equal amounts of each subunit combination); lane 3, M113F7. A diagram of the eight different combinations of subunits and their permuta- tions is shown to the right of the autoradiogram. The various permutations are not separated by electrophoresis. (C) Kinetics of the interaction of βCD with single heteromeric αHL pores as determined by bilayer recording. Values of kon were calculated by using kon ¼ 1∕ðτon½βCDÞ, where τon is the mean interevent interval. Values of koff were determined by using koff ¼ 1∕τoff, where τoff is the mean dwell time of βCD in the pore. Values of Kd were calculated by using Kd ¼ koff∕kon. Each point represents the mean  s:d: for three or more determinations. Where they cannot be seen, the s.d. values lie within the symbol. Black squares, WT7−nM113Nn; gray squares, M113F7−nM113Nn; empty squares, M113F7−nWTn. 8166 ∣ www.pnas.org/cgi/doi/10.1073/pnas.0914229107 Banerjee et al. events had been observed previously with certain Met-113 repla- cement mutants and may represent movement of the βCD at its binding site (e.g., rotation about axes perpendicular to the C7 axis) (15). The additional current spikes were more prevalent for M113V7 and M113Cpg7, which may take part in more con- formationally labile interactions with βCD, compared with say M113F7 (Fig. S4). Interactions of βCD with Homoheptamers Bearing Aromatic Residues at Position 113. To further understand the nature of the binding of βCD to aromatic side chains, we examined the kinetics of βCD binding to the homoheptamers containing f1F or f5F at position 113, M113f1F7 and M113f5F7 (Fig. 3C). For both mutants, the value of kon was very similar to that of WT7, but the values of koff and therefore Kd for M113f1F7 differed dramatically from WT7 and were close to the values for the tight-binding mutant M113F7 (Table S2A). By contrast, koff and Kd for M113f5F7 were similar to the values for WT7 (Table S2A). To determine whether M113f1F7 binds βCD in the same orien- tation as M113F7 (Fig. 2C), we made heteromers of the M113f1F subunit with M113N or M113F and examined M113F4M113f1F3 and M113N4M113f1F3. M113F4M113f1F3 binds βCD as strongly as either M113F7 or M113f1F7, but M113N4M113f1F3 binds βCD weakly with a similar affinity to WT7 (Fig. 3D and Table S3). Therefore, it is reasonable to infer that M113F7 and M113f1F7 bind βCD in the same orientation with the 6- hydroxyl groups of the CD in proximity to the aromatic rings on the protein. Cyclohexylalanine (Cha) was used to replace the aromatic side chains with a roughly isosteric hydrophobic group. Again the va- lue of kon for βCD was little changed, but koff for M113Cha7 had an intermediate value of 42  6 s−1. Therefore, M113Cha7 binds βCD more weakly than M113F7 but distinctly more strongly than the WT7 pore (Table S2A and Fig. 3C). Interactions of βCD with Homoheptamers Bearing Hydrophobic Resi- dues at Position 113. M113V7 binds βCD very strongly, and there- fore we compared αHL pores with Cpg or Cpa at position 113. Cpg is roughly isosteric with Val, and like Val has a β-branched side chain. Gratifyingly, M113Cpg7 has a kon value similar to the other αHL pores, and koff and Kd values close to those of M113V7 (Table S2B and Fig. 3E). Cyclopropylalanine (Cpa), with an additional methylene group compared to Cpg, is roughly isosteric with Leu, a weak binder, and M113Cpa7 also binds βCD weakly with kon, koff and Kd values similar to those of WT7 (Table S2B and Fig. 3E). M113I7 and M113T7, which are β-branched, are also weak binders, but Ile and Thr are less closely related to Val than Cpg. To determine whether M113V7 binds βCD in the same orien- tation as M113F7 or M113N7 (Fig. 2), we made heteromers of M113V and the M113N or M113F subunits. M113V3M113F4, M113V4M113F3, M113V3M113N4, and M113V4M113N3 were examined in detail. All four heteroheptamers bound βCD more weakly than M113V7, M113F7 or M113N7 (Fig. 3F and Table S4), suggesting that Val at position 113 interacts with βCD strongly but in a different manner to either Phe or Asn. Each heteromer exhibited a range of Kd values, perhaps reflecting the various pos- sible permutations of the two different subunits around the cen- tral axis of the heptamer, although this heterogeneity was not seen for heteromers made from WT, M113F and M113N (Fig. 1). Discussion Soon after we discovered that βCD binds to the WT-αHL pore for around a millisecond, we found a mutant pore, M113N7, that re- leases βCD ∼104 times more slowly (14). This prompted us to examine all 19 mutants in which residue 113 is replaced by a nat- ural amino acid, with the surprising result that a collection of ami- no acids with structurally unrelated side chains (V, H, Y, D, N, F) are tight binders (15). Here, we have examined the nature of the binding interactions more closely by single-channel electrical re- cording, protein engineering including unnatural amino acid mu- tagenesis, and high-resolution x-ray crystallography, and we provide the first definitive structural information on an engi- neered protein nanopore. We find that βCD can bind tightly to the αHL pore in three different ways depending on the residue at 113, as exemplified by Asn, Phe, and Val. Because Asn and Phe have quite different side chains, we first compared the ability of M113N and M113F subunits to take part in binding the CD. The examination of het- eromeric proteins containing WT (Met-113), M113N and M113F subunits showed that the replacement of WT subunits in WT7 with M113N or M113F subunits led to increased affinity for βCD. The more M113N or M113F subunits that were added, the tighter binding became. By contrast, when subunits in M113N7 were replaced with M113F subunits, binding became weaker, reaching a minimum at three to four M113F subunits, and then increasing in strength with five M113F subunits or more (Fig. 1C). Parallel structural studies (30) revealed the basis of the “oppos- ing” effects of the M113N and M113F subunits. βCD binds to M113N7 in the opposite orientation to that in which it binds to M113F7. In M113N7, the secondary hydroxyls in the βCD ring are hydrogen bonded to Lys-147 and Asn-113 (Fig. 2A). By con- trast, βCD interacts with M113F7 through its primary hydroxyl face (Fig. 2B). It seemed likely that M113V7, bound βCD in yet another way, and this was examined by forming heteromers between M113V and M113N or M113F. The presence of three or four subunits of either M113N or M113F greatly decreases the affinity of the pore for βCD (Fig. 3F), with an average koff of 7.3 × 101 s−1, indicating that a third binding mode is indeed operating Fig. 2. X-ray structures of M113N and M113F homoheptamers with βCD bound. (A) Side view of heptameric αHL. βCD binds in the blue highlighted region. (B) βCD bound to M113N7 (dotted lines indicate hy- drogen bonding). The side chains of Lys-147 are in pale brown and the side chains of Asn- 113 in yellow. (C) βCD bound to M113F7 (dotted lines indicate CH-π bonding). The side chains of Phe-113 are in yellow. The sec- ond βCD in the M113F7 · ðβCDÞ2 structure is hydrogen bonded to the top βCD in a head- to-head arrangement and has no apparent interactions with the protein. For both (B) and (C), four β strands were omitted from the barrel to give a better view. Banerjee et al. PNAS ∣ May 4, 2010 ∣ vol. 107 ∣ no. 18 ∣ 8167 BIOCHEMISTRY (Table S4). In summary, the three groups of tight-binding mutants comprise αHL pores incorporating, at position 113: (i) the hydro- gen-bonding amino acids N, D (the latter would have to be largely in the protonated form), and possibly H; (ii) the aromatics F, Y, f1F, and possibly H, and more weakly W; (iii) the β-branched ami- no acids V, Cpg. There may be yet other means by which CDs can bind to the αHL pore. For example, we earlier found that hepta- 6-sulfato-βCD can bind tightly to αHL pores containing the N139Q mutation (37). Presumably, this CD is bound at a site low- er down in the β barrel in a fashion that includes hydrogen bond- ing to the Gln at position 139. While the various mutants exhibited widely different koff values, the value of kon was almost invariant and averaged ∼2.3 × 105 M−1 s−1 (Table S2) (15). Ap- parently, transport to the binding site is rate limiting, through a route unaffected by mutagenesis. koff increased precipitously with the addition of WTsubunits to M113N7 (Fig. 1C). Crystal structures of M113N7 show that resi- dues 111, 113, and 147 are reorganized by compari- son with WT7 and then undergo a more limited rearrangement when βCD binds (Fig. S5). For example, the side chain of Lys-147 shifts position to form a bifurcated hydrogen bond with a 3-hydroxyl group of βCD and the side chain carbonyl of an Asn- 113 (Fig. S6). Therefore, the side chains of residues 111, 113, and 147 might be in a variety of conformations in WT7−nM113Nn het- eromers and offer less well preorganized binding sites for βCD than they do in M113N7. Further, the intramolecular hydrogen bonds of the secondary hydroxyls in βCD (38) must be disrupted upon binding as both hydroxyls on each glucose ring form hydro- gen bonds to the mutant subunits (Fig. 2B). Because the hydrogen bonds that are broken in βCD are arranged in a circle, the break- age of bonds involving a single glucose (three bonds in all) will be relatively more disruptive than those involving adjoining glucose residues or the entire circle. The overall binding cooperativity in M113N7 could be attributed to enthalpic cooperativity outweigh- ing entropic penalties to binding (39). Positive cooperativity has been observed previously in fairly rigid model systems (40). By contrast with M113N7, there is little movement of side chains in ðM113FÞ7 by comparison with WT7 and little move- ment, including Phe-113, upon binding βCD (Fig. S7A). Further, the crystal structure of M113F7 · βCD suggests that each Phe re- sidue interacts independently with the βCD through what appear to be CH-π interactions (Fig. S7B). These interactions are ex- pected to be weak and not strongly directional and hence offer less enthalpic cooperativity, as supported by the B-factors (crys- tallographic temperatures factors) at the primary βCD binding site, which are between ∼40 and 50. Positive cooperativity is ob- served, but it is less pronounced than in the case of M113N7 (Table S5). In the case of M113N7, the B-factors of the residues that bind βCD are in the 20s implying that the βCD is more rigidly held than it is in M113F7. The binding of sugars to aromatic residues in proteins can in- clude CH-π bonding (41) or OH-π bonding or a finely balanced complement of both (42, 43). However, we have dismissed the possibility of an OH-π interaction between Phe-113 and the 6-hydroxyl groups of βCD as the distance between the center of the phenyl rings to the nearest hydroxyl oxygen is higher (5.2  0.65 Å, n ¼ 7) than that expected for a favorable OH-π interaction (33). While we propose that βCD binds to Phe-113 through a C-6 CH-π interaction (Fig. S7B), the distances between the center of the Phe-113 ring and the nearest C-6 of βCD ob- served in the M113F7 · βCD structure (4.66  0.24 Å, n ¼ 7) are in the upper range of the expected distance for a strong inter- action, which is ∼4.5 Å (33). The angle between the normal to the aromatic rings and the line connecting the C-6 atoms to the aro- matic midpoint is 8.0  5.6°, which is well within the expected range (44). The measurements with M113f5F7 argue against a hydrophobic interaction between Phe residues at position 113 and the βCD ring. In f5F, the hydrophobicity of the phenyl ring is significantly increased (45) yet M113f5F7 binds βCD weakly, like WT7 (Fig. 3C and Table S2A). By contrast with F, f1F, Y and N, homomeric αHL pores with f5F and W at position 113 bound βCD relatively weakly (Fig. 3C and Table S2A). In the case of f5F, the powerful electron with- drawing action of the five fluorine atoms leaves a highly increased positive charge at the center of the ring (46, 47), mitigating against a hydrogen-bonding interaction. The electron-rich Trp Fig. 3. Properties of pores containing natural and unnatural amino acid sub- stitutions at position 113. The data were recorded at þ40 mV in 1.0 M NaCl, 10 mM sodium phosphate, pH 7.5. (A) Unnatural amino acids used in this study: 4-fluorophenylalanine, f1F; pentafluorophenylalanine, f5F; cyclohex- ylalanine, Cha; cyclopropylglycine, Cpg; cyclopropylalanine, Cpa. (B) Repre- sentative current traces from single homoheptameric αHL pores, containing unnatural amino acids at position 113, in the presence of βCD. βCD (40 μM final) was added to the trans chamber. Level 1, open pore current; level 2, pore occupied by βCD. The broken line indicates zero current. (C) In- teraction of βCD with homomeric αHL pores containing aromatic amino acids at position 113. Kd values for the interaction between βCD and the αHL pore were calculated by using Kd ¼ koff∕kon. Each column represents the mean  s:d: for 10 or more determinations: dark gray, natural amino acids; light gray, unnatural amino acids. Data adapted from Gu and colleagues (15) are marked (*). (D) Representative current traces from single-channel recordings of βCD binding to M113F4M113f1F3 and M113N4M113f1F3. βCD (40 μM final) was added to the trans chamber. The broken line indicates zero current. (E) Interaction of βCD with homomeric αHL pores containing hydrophobic amino acids at position 113. Kd values for the interaction between βCD and the αHL pore were calculated by using Kd ¼ koff∕kon. Each column represents the mean  s:d: for ten or more determinations: dark gray, natural amino acids; light gray, unnatural amino acids. Data adapted from Gu and colleagues (15) are marked (*). (F) koff values for βCD from heteroheptamers formed with M113F and M113V subunits and with M113N and M113V subunits. βCD (40 μM final) was added to the trans chamber. The kon values for βCD for all these mutants are similar, at ∼3 × 105 M−1 s−1. Empty square: average koff values for the mutant (bar is s:d). Filled square: M113V3M113F4; filled circle: M113V4M113F3; filled upright triangle: M113V3M113N4; filled in- verted triangle: M113V4M113N3. 8168 ∣ www.pnas.org/cgi/doi/10.1073/pnas.0914229107 Banerjee et al. ring (44, 46, 47) should favor hydrogen bonding, but here we can- not make a direct comparison with the crystal structure of M113F7 as the indole ring is far larger than benzene. It is possible that it cannot become oriented in the same manner and that it is misaligned for hydrogen bonding. Our experiments suggest that M113V7 and M113Cpg7 bind βCD in a third way. In heteromers with M113V, both M113F and M113N reduce the affinity of the pore for βCD suggesting that neither the CH-π interaction with Phe-113 nor the hydrogen- bonding interactions with Asn-113 and Lys-147 are compatible with binding to Val. Close interactions of Val with glucose rings have been noted previously (48). Therefore, we propose that the Val side-chain interacts with the side of the glucose ring. This in- teraction might occur in one or both orientations of the CD ring (Fig. 4). Conclusion We provide structural information on engineered protein nano- pores and describe three distinct ways in which βCD can bind within the lumen of mutant αHL pores in atomic detail. Our re- sults will be useful in several areas of basic science and biotech- nology. By using host molecules lodged within the αHL pore, host-guest interactions can be investigated in fine detail at the single-molecule level (17, 49). The present work will now permit us to examine binding events at a known face of a CD. The work also provides information for designing new binding sites within the lumen of the αHL pore (37) or within other β barrel proteins (21, 50) and for using molecular design to devise ways in which to covalently attach CDs within pores (13, 51). These areas impact practical applications of nanopore technology including stochas- tic sensing (8), single-molecule DNA sequencing (9, 12, 13, 52), the use of nanoreactors for the observation of single-molecule chemistry (10), and the construction of nano- and microdevices (11, 53), as well as the design of CDs as therapeutic agents (23, 24). Methods Full details of the experimental procedures can be found in SI Appendix. Materials L-Amino acids were obtained as follows: 4-fluorophenylalanine (f1F) (Fluka); pentafluorophenylalanine (f5F) (PepTech Corp.); cyclopropylglycine (Cpg) (Ty- ger); cyclopropylalanine (Cpa) (Tyger). 4-N-benzoyl-5′-O-(4,4′-dimethoxytri- tyl)-2′-deoxycytidine-3′-[(2-cyanoethyl)-(N,N-diisopropyl)]-phosphoramidite and bis(2-cyanoethyl)-N,N-diisopropylphosphoramidite for the synthesis of pdCpA were purchased from Glen Research and Toronto Research Chemicals, respectively. Preparation of NVOC-Protected Aminoacyl-pdCpA. NVOC-protected aminoacyl-pdCpAs were prepared as reported previously by reacting the dinucleotide pdCpA with N-protected, carboxylic acid-activated, amino acids (54–56). Preparation of NVOC-Protected Aminoacyl-tRNA. NVOC-protected aminoacyl- pdCpAs were ligated enzymatically with a truncated tRNA, prepared by using methods described elsewhere (57, 58). Genetic Constructs and Mutagenesis. All new αHL constructs were verified by DNA sequencing. Details of each construct can be found in SI Appendix. Synthesis, Assembly, and Purification of Mutant αHL pores. αHL monomers (WT and mutants) were prepared in vitro by coupled transcription and translation (IVTT) and assembled into homoheptamers on rabbit red blood cell membranes followed by purification by SDS–PAGE as described earlier (59). Heteroheptamers were prepared by mixing the two required DNAs (one encoding an αHL with a D8 tail) before IVTT and then oligomerizing the mixed translation products on rabbit red blood cell membranes. Pores with the desired combinations of subunits were purified by SDS–PAGE (59). Synthesis, Assembly, and Purification of αHL Mutants Containing Unnatural Ami- no Acids. αHL polypeptides containing unnatural amino acids were synthe- sized by IVTT in the presence of rabbit red blood cell membranes. The plasmid with a stop codon (TAG) at position 113 was used. Deprotected ami- noacyl-tRNAs (SI Appendix) were added to the IVTT mixtures. For heterohep- tamers with subunits containing unnatural amino acids in combination with M113N or M113F, monomers were first made, which were then coassembled on rabbit red blood cell membranes and subsequently purified by SDS–PAGE. Single-Channel Current Recordings in Planar Lipid Bilayers. (15, 60) Recordings were made with 1.0 M NaCl, 10 mM sodium phosphate, pH 7.5, in both cham- bers, at an applied potential of þ40 mV. Data were recorded at 22  2°C. The bilayer was formed from 1,2-diphytanoyl-sn-glycero-phosphocholine (Avanti Polar Lipids). Proteins were added to the cis chamber, and βCD to the trans chamber. Single-channel currents were recorded with an Axopatch 200B patch-clamp amplifier (Axon Instruments) and filtered at 2 kHz with a built-in 4-pole low-pass Bessel Filter. The data were acquired at a sampling rate of 10 kHz. For mutants that bind βCD strongly, the data were acquired for at least 30 min and for weak-binding mutants for at least 10 min. Kinetic Data Analysis. Current amplitude and dwell-time histograms were made by using ClampFit 9.0. The mean dwell times, τoff, were determined by fitting the dwell-time histograms to single exponentials. Values of kon and koff were obtained by using the mean dwell times and mean interevent intervals, as described previously (15, 60). This analysis assumes a binary in- teraction, which was supported in all cases examined by the finding of only one major blockade level and a single exponential distribution of dwell times (τoff). Fig. 4. Molecular model showing the three classes of interaction between the αHL pore and βCD identified in this work. The model identifies the region of βCD responsible for each interaction (H atoms interacting with Phe-113 or Asn-113 and Lys-147: gray). The first class of interaction is with aromatic residues and involves the seven -CH2OH groups of the βCD. The second class is typified by the interactions with Asn at position 113, which involve hydro- gen-bonds to the secondary 2-hydroxyls of the βCD. Structural studies show that this interaction is supported by hydrogen bonding between Lys-147 and the secondary 3-hydroxyls of the βCD. Structural studies and experiments with heteromers suggest that the βCD in M113F7 is in the opposite orienta- tion to the βCD in M113N7, in support of the model shown here. As the inter- action with Val is hydrophobic, it is not directional and βCD may not bind at the same position inside the β barrel as it does in M113F7 or M113N7. Banerjee et al. PNAS ∣ May 4, 2010 ∣ vol. 107 ∣ no. 18 ∣ 8169 BIOCHEMISTRY Protein Crystallography. Details can be found in SI Appendix. Protein Data Bank: The coordinates and structure factors of the described structures have been deposited with accession codes 3M2L ðM113F7Þ, 3M3R ðM113F7 · βCDÞ, 3M4D ðM113N7Þ, 3M4E ðM113N7 · βCDÞ. ACKNOWLEDGMENTS. We thank Dennis Dougherty for the plasmid pTHG73. This work was funded by a Royal Society Wolfson Research Merit Award (to H.B.), the Medical Research Council (H.B.), the National Institutes of Health (H.B.), and the Howard Hughes Medical Institute (E.G.). 1. Lu Y, Yeung N, Sieracki N, Marshall NM (2009) Design of functional metalloproteins. Nature 460:855–862. 2. Gebauer M, Skerra A (2009) Engineered protein scaffolds as next-generation antibody therapeutics. Curr Opin Chem Biol 13:245–255. 3. Gronwall C, Stahl S (2009) Engineered affinity proteins—Generation and applications. J Biotechnol 140:254–269. 4. Arnold U (2009) Incorporation of non-natural modules into proteins: Structural features beyond the genetic code. Biotechnol Lett 31:1129–1139. 5. Tracewell CA, Arnold FH (2009) Directed enzyme evolution: Climbing fitness peaks one amino acid at a time. Curr Opin Chem Biol 13:3–9. 6. Fruk L, Kuo CH, Torres E, Niemeyer CM (2009) Apoenzyme reconstitution as a chemical tool for structural enzymology and biotechnology. Angew Chem Int Ed Engl 48:1550–1574. 7. Bayley H, Jayasinghe L (2004) Functional engineered channels and pores. Mol Membr Biol 21:209–220. 8. Bayley H, Cremer PS (2001) Stochastic sensors inspired by biology. Nature 413:226–230. 9. Branton D, et al. (2008) The potential and challenges of nanopore sequencing. Nature Biotechnol 26:1146–1153. 10. Bayley H, Luchian T, Shin S-H, Steffensen MB (2008) Single Molecules and Nanotech- nology, eds R Rigler and H Vogel (Springer, Heidelberg), pp 251–277. 11. Maglia G, et al. (2009) Droplet networks with incorporated protein diodes show collective properties. Nat Nanotechnol 4:437–440. 12. Astier Y, Braha O, Bayley H (2006) Toward single molecule DNA sequencing: Direct identification of ribonucleoside and deoxyribonucleoside 5'-monophosphates by using an engineered protein nanopore equipped with a molecular adapter. J Am Chem Soc 128:1705–1710. 13. Clarke J, et al. (2009) Continuous base identification for single-molecule nanopore DNA sequencing. Nature Nanotechnol 4:265–270. 14. Gu L-Q, Braha O, Conlan S, Cheley S, Bayley H (1999) Stochastic sensing of organic analytes by a pore-forming protein containing a molecular adapter. Nature 398:686–690. 15. Gu L-Q, Cheley S, Bayley H (2001) Prolonged residence time of a noncovalent molecular adapter, β-cyclodextrin, within the lumen of mutant α-hemolysin pores. J Gen Physiol 118:481–494. 16. Ervin EN, Kawano R, White RJ, White HS (2008) Simultaneous alternating and direct current readout of protein ion channel blocking events using glass nanopore membranes. Anal Chem 80:2069–2076. 17. Gurnev PA, Harries D, Parsegian VA, Bezrukov SM (2009) The dynamic side of the Hof- meister effect: A single-molecule nanopore study of specific complex formation. ChemPhysChem 10:1445–1449. 18. Mamonova T, Kurnikova M (2006) Structure and energetics of channel-forming protein-polysaccharide complexes inferred via computational statistical thermody- namics. J Phys Chem B 110:25091–25100. 19. Egwolf B, Luo Y, Walters DE, Roux B (2010) Ion selectivity of alpha-hemolysin with beta-cyclodextrin adapter. II. Multi-ion effects studied with grand canonical monte carlo/brownian dynamics simulations. J Phys Chem B 114:2901–2909. 20. Orlik F, et al. (2003) CymA of Klebsiella oxytoca outer membrane: Binding of cyclodex- trins and study of the current noise of the open membrane. Biophys J 85:876–885. 21. Chen M, Khalid S, Sansom MSP, Bayley H (2008) OmpG: Engineering a quiet pore for biosensing. Proc Natl Acad Sci USA 105:6272–6277. 22. Locke D, Koreen IV, Liu JY, Harris AL (2004) Reversible pore block of connexin channels by cyclodextrins. J Biol Chem 279:22883–22892. 23. Karginov VA, Nestorovich EM, Moayeri M, Leppla SH, Bezrukov SM (2005) Blocking anthrax lethal toxin at the protective antigen channel by using structure-inspired drug design. Proc Natl Acad Sci USA 102:15075–15080. 24. Moayeri M, Robinson TM, Leppla SH, Karginov VA (2008) In vivo efficacy of beta- cyclodextrin derivatives against anthrax lethal toxin. Antimicrob Agents Chemother 52:2239–2241. 25. Gao C, Ding S, Tan Q, Gu LQ (2009) Method of creating a nanopore-terminated probe for single-molecule enantiomer discrimination. Anal Chem 81:80–86. 26. Pregel MJ, Jullien L, Lehn J-M (1992) Towards artificial ion channels: Transport of alkali metal ions across liposomal membranes by “bouquet” molecules. Angew Chem Int Edit 31:1637–1639. 27. Bacri L, Benkhaled A, Guegan P, Auvray L (2005) Ionic channel behavior of modified cyclodextrins inserted in lipid membranes. Langmuir 21:5842–5846. 28. Jog PV, Gin MS (2008) A light-gated synthetic ion channel. Org Lett 10:3693–3696. 29. Howorka S, Cheley S, Bayley H (2001) Sequence-specific detection of individual DNA strands using engineered nanopores. Nat Biotechnol 19:636–639. 30. Montoya M (2004) Insights into Membrane Association and Bioengineering of a Pore- Forming Toxin: Structural Studies of Staphylococcal α-Hemolysin (Columbia University, New York). 31. Steiner T (2002) The hydrogen bond in the solid state. Angew Chem Int Ed 41:49–76. 32. Steiner T (2002) Hydrogen bonds from water molecules to aromatic acceptors in very high-resolution protein crystal structures. Biophys Chem 95:195–201. 33. Steiner T, Koellner G (2001) Hydrogen bonds with pi-acceptors in proteins: Frequencies and role in stabilizing local 3D structures. J Mol Biol 305:535–557. 34. Brandl M, Weiss MS, Jabs A, Sühnel J, Hilgenfeld R (2001) CH...π-interactions in proteins. J Mol Biol 307:357–377. 35. Weiss MS, Brandl M, Suhnel J, Pal D, Hilgenfeld R (2001) More hydrogen bonds for the (structural) biologist. Trends Biochem Sci 26:521–523. 36. Wang L, Xie J, Schultz PG (2006) Expanding the genetic code. Annu Rev Biophys Biomol Struct 35:225–249. 37. Gu L-Q, Cheley S, Bayley H (2001) Capture of a single molecule in a nanocavity. Science 291:636–640. 38. Saenger W, et al. (1998) Structures of the common cyclodextrins and their larger analogues—Beyond the doughnut. Chem Rev 98:1787–1802. 39. Hunter CA, Tomas S (2003) Cooperativity, partially bound states, and enthalpy-entropy compensation. Chem Biol 10:1023–1032. 40. Bisson AP, Hunter CA (1996) Cooperativity in the assembly of zipper complexes. Chem Commun 1723–1724. 41. del Carmen Fernandez-Alonso M, Canada FJ, Jimenez-Barbero J, Cuevas G (2005) Molecular recognition of saccharides by proteins. Insights on the origin of the carbohydrate-aromatic interactions. J Am Chem Soc 127:7379–7386. 42. Jimenez-Barbero J, Asensio JL, Canada FJ, Poveda A (1999) Free and protein-bound carbohydrate structures. Curr Opin Struct Biol 9:549–555. 43. Stanca-Kaposta EC, et al. (2007) Carbohydrate molecular recognition: A spectroscopic investigation of carbohydrate-aromatic interactions. Phys Chem Chem Phys 9:4444–4451. 44. Brandl M, Weiss MS, Jabs A, Suhnel J, Hilgenfeld R (2001) CH...π-interactions in proteins. J Mol Biol 307:357–377. 45. Woll MG, Hadley EB, Mecozzi S, Gellman SH (2006) Stabilizing and destabilizing effects of phenylalanine –>F5-phenylalanine mutations on the folding of a small protein. J Am Chem Soc 128:15932–15933. 46. Mecozzi S, West AP, Jr, Dougherty DA (1996) Cation-pi interactions in aromatics of biological and medicinal interest: Electrostatic potential surfaces as a useful qualitative guide. Proc Natl Acad Sci USA 93:10566–10571. 47. Dougherty DA (2008) Cys-loop neuroreceptors: Structure to the rescue?. Chem Rev 108:1642–1653. 48. Hondoh H, et al. (2008) Substrate recognition mechanism of alpha-1,6-glucosidic linkage hydrolyzing enzyme, dextran glucosidase from Streptococcus mutans. J Mol Biol 378:913–922. 49. Kang XF, Cheley S, Guan X, Bayley H (2006) Stochastic detection of enantiomers. J Am Chem Soc 128:10684–10685. 50. Chen M, Li QH, Bayley H (2008) Orientation of the monomeric porin OmpG in planar lipid bilayers. ChemBioChem 9:3029–3036. 51. Wu H-C, Astier Y, Maglia G, Mikhailova E, Bayley H (2007) Protein nanopores with covalently attached molecular adapters. J Am Chem Soc 129:16142–16148. 52. Bayley H (2006) Sequencing single molecules of DNA. Curr Opin Chem Biol 10:628–637. 53. Astier Y, Bayley H, Howorka S (2005) Protein components for nanodevices. Curr Opin Chem Biol 9:576–584. 54. Robertson SA, Noren CJ, Anthony-Cahill SJ, Griffith MC, Schultz PG (1989) The use of 5'-phospho-2 deoxyribocytidylylriboadenosine as a facile route to chemical aminoacy- lation of tRNA. Nucleic Acids Res 17:9649–9660. 55. Ellman J, Mendel D, Anthony-Cahill S, Noren CJ, Schultz PG (1991) Biosynthetic method for introducing unnatural amino acids site-specifically into proteins. Method Enzymol 202:301–336. 56. Kearney PC, et al. (1996) Dose-response relations for unnatural amino acids at the agonist binding site of the nicotinic acetylcholine receptor: Tests with novel side chains and with several agonists. Mol Pharmacol 50:1401–1412. 57. England TE, Bruce AG, Uhlenbeck OC (1980) Specific labeling of 3′ termini of RNA with T4 RNA ligase. Method Enzymol 65:65–74. 58. Nowak MW, et al. (1998) In vivo incorporation of unnatural amino acids into ion channels in xenopus oocyte expression system. Method Enzymol 293:504–529. 59. Cheley S, Braha O, Lu X, Conlan S, Bayley H (1999) A functional protein pore with a “retro” transmembrane domain. Protein Sci 8:1257–1267. 60. Gu L-Q, et al. (2000) Reversal of charge selectivity in transmembrane protein pores by using non-covalent molecular adapters. Proc Natl Acad Sci USA 97:3959–3964. 61. Song L, et al. (1996) Structure of staphylococcal α-hemolysin, a heptameric transmem- brane pore. Science 274:1859–1865. 8170 ∣ www.pnas.org/cgi/doi/10.1073/pnas.0914229107 Banerjee et al.
3M4G
H57A HFQ from Pseudomonas Aeruginosa
structural communications 760 doi:10.1107/S1744309110017331 Acta Cryst. (2010). F66, 760–764 Acta Crystallographica Section F Structural Biology and Crystallization Communications ISSN 1744-3091 The structures of mutant forms of Hfq from Pseudomonas aeruginosa reveal the importance of the conserved His57 for the protein hexamer organization Olga Moskaleva, Bogdan Melnik, Azat Gabdulkhakov, Maria Garber, Stanislav Nikonov, Elena Stolboushkina and Alexei Nikulin* Institute of Protein Research, RAS, Russia Correspondence e-mail: nikulin@vega.protres.ru Received 12 March 2010 Accepted 11 May 2010 PDB References: P. aeruginosa Hfq, H57A mutant, 3inz; H57T mutant, 3m4g. The bacterial Sm-like protein Hfq forms homohexamers both in solution and in crystals. The monomers are organized as a continuous -sheet passing through the whole hexamer ring with a common hydrophobic core. Analysis of the Pseudomonas aeruginosa Hfq (PaeHfq) hexamer structure suggested that solvent-inaccessible intermonomer hydrogen bonds created by conserved amino-acid residues should also stabilize the quaternary structure of the protein. In this work, one such conserved residue, His57, in PaeHfq was replaced by alanine, threonine or asparagine. The crystal structures of His57Thr and His57Ala Hfq were determined and the stabilities of all of the mutant forms and of the wild-type protein were measured. The results obtained demonstrate the great importance of solvent-inaccessible conserved hydrogen bonds between the Hfq monomers in stabilization of the hexamer structure. 1. Introduction In bacteria, Hfq protein acts as a global post-transcriptional regulator which binds small regulatory RNAs and promotes their interaction with mRNAs (Valentin-Hansen et al., 2004; Brennan & Link, 2007). It controls the expression of many genes by its action on mRNA translation, stability or polyadenylation (Zhang et al., 1998; Vytvytska et al., 2000; Hajnsdorf & Re´gnier, 2000; Sledjeski et al., 2001). Hfq is a small (70–110 amino-acid residues) thermostable protein which exists in a homohexameric form in solution (Brennan & Link, 2007; Zhang et al., 2002; Møller et al., 2002). A hexameric organization has also been observed in the crystal structures of Hfq from Staphylococcus aureus (Schumacher et al., 2002), the core part of Escherichia coli Hfq (Sauter et al., 2003) and Hfq from Pseudomonas aeruginosa (Nikulin et al., 2005). All of these proteins formed doughnut-shaped rings with outer and inner diameters of about 65 and 10 A˚ and a thickness of 25–30 A˚ . Hfq belongs to the Sm/Sm-like protein family. This family includes eukaryotic Sm and Sm-like (Lsm) proteins and archaeal Lsm proteins (Wilusz & Wilusz, 2005). Eukaryotic Sm/Lsm proteins are involved in RNA processing in the cell (Kufel et al., 2004; Verdone et al., 2004). In crystals they form heptamers (Achsel et al., 1999; Mayes et al., 1999; Walke et al., 2001) or octamers (Naidoo et al., 2008), whereas in cytoplasm the Sm proteins are found as heterodimers or trimers and are only able to form heptamers in the presence of U-rich small nuclear RNAs (UsnRNAs; Achsel et al., 1999; Will & Lu¨hrmann, 2001). Archaeal genomes usually encode one or two distinct Lsm proteins called Lsm1 and Lsm2 (Salgado-Garrido et al., 1999). Lsm3 proteins have only been identified in a few archaeal species (Mura et al., 2003; Kilic et al., 2006). Lsm1 is the most abundant of the archaeal species. Archaeal Lsm1 and Lsm2 proteins form stable homoheptamers, with the exception of Archaeoglobus fulgidus AF-Sm2, which can exist in hexameric or heptameric forms depending on the pH or the presence of RNA (To¨ro¨ et al., 2001; Achsel et al., 2001; Kilic et al., 2006). # 2010 International Union of Crystallography All rights reserved The Sm-protein family is characterized by a conserved motif of about 70 amino acids, which is called the Sm-domain. It is a -barrel- type structure consisting of five -strands, which are capped by an N-terminal -helix (Fig. 1a). The Sm-domain contains two conserved sequence motifs (Sm1 and Sm2) linked by a loop that differs in length and sequence depending on the species (Valentin-Hansen et al., 2004; Se´raphin, 1995). Strands 1, 2 and 3 form the Sm1 motif and strands 4 and 5 constitute the Sm2 motif of the domain. The sequence of the Sm1 motif is conserved among all bacteria, archaea and eukarya (Kambach et al., 1999; Hajnsdorf & Re´gnier, 2000; Valentin-Hansen et al., 2004; Kilic et al., 2006). In contrast, the Sm2 motif has different consensus sequences in bacterial Hfq and eukaryal/archaeal Sm/Lsm proteins (Sauter et al., 2003). Analysis of the known crystal structures (To¨ro¨ et al., 2002; Nikulin et al., 2005; Brennan & Link, 2007) has shown that the -strands of the Sm2 motifs organize the protein ring by means of hydrogen bonds formed by the main-chain O and N atoms. The quaternary structure is additionally stabilized by contacts between strongly conserved amino acids. Previously, we have suggested (Nikulin et al., 2005) that the conserved YKHI consensus sequence of the Sm2 motif in Hfq should define its hexamer formation and that His57 could play a very important role in the stabilization of the hexamer structure. To prove this hypothesis, we mutated His57 in P. aeruginosa Hfq (PaeHfq) to alanine, threonine and asparagine, measured the stability of the wild- type hexamer and the obtained mutant forms and solved the crystal structures of PaeHfq with His57Thr and His57Ala mutations. 2. Materials and methods 2.1. Site-directed mutagenesis, gene expression and recombinant protein purification To prepare mutant forms of PaeHfq, site-directed mutagenesis was carried out by PCR using oligonucleotides which contained the desired mutations. All of the mutants (hfqH57A, hfqH57N and hfqH57T) were constructed in two steps. In step 1, fragments carrying a mutation were amplified from pET22b(+)/Hfq DNA by PCR with the reverse primer 50-CGGGATCCTCAAGCGTTGCCC-30 and corresponding oligonucleotides for each fragment (H57A, 50-GTT- TACAAGGCGGCGATCTCC-30; H57N, 50-GTTTACAAGAAC- GCGATCTCC-30; H57T, 50-GTTTACAAGACCGCGATCTCC-30). The fragments were then completed by PCR using the forward primer 50-GGGAATTCCATATGTCAAAAGGGCAT-30 and the PCR products obtained in step 1. The final PCR products were inserted into pET22b(+) plasmid DNA and verified by sequencing. All of the mutant proteins were purified as described previously (Nikulin et al., 2005). 2.2. Circular-dichroism (CD) measurements CD measurements were performed on a Jasco J600 spectro- polarimeter equipped with a Julabo F25 computer-controlled thermostat. All spectra and melting experiments were measured structural communications Acta Cryst. (2010). F66, 760–764 Moskaleva et al.  Hfq 761 Figure 1 (a) Overall structure of the Hfq hexamer from P. aeruginosa. One monomer is coloured according to the conserved sequence motifs: the Sm1 motif (1, 2 and 3) is shown in yellow, the Sm2 motif (4 and 5) in red and the N-terminal 1 helix in blue. The position of amino-acid residue 57 is shown by a black sphere. (b) Superposition of wild- type PaeHfq (cyan), H57T PaeHfq (magenta) and H57A PaeHfq (green). The C-atom r.m.s. deviations of H57T PaeHfq and H57A PaeHfq from the wild-type protein are 0.43 and 0.49 A˚ , respectively. (c) The amino-acid sequence of PaeHfq with corresponding secondary-structure elements. Amino-acid residues that are conserved in Lsm proteins from bacteria, archaea and eukarya are shown in green; those conserved in bacteria only are shown in cyan. structural communications 762 Moskaleva et al.  Hfq Acta Cryst. (2010). F66, 760–764 Figure 2 The interface of two adjacent monomers in the PaeHfq hexamer. The main chains of the monomers are shown in green and yellow. Side chains are shown for residue 57 only. Hydrogen bonds are shown as dotted lines. (a) The wild-type PaeHfq crystal structure. (b) The H57A PaeHfq crystal structure. (c) The H57T PaeHfq crystal structure. using a cell with a 0.1 mm path length. The melting experiments were performed by monitoring the change in ellipticity at 220 nm. 2.3. Crystallization and data collection Protein crystals were obtained using the hanging-drop vapour- diffusion technique at 295 K. All drops were set up by mixing 2.0 ml protein solution (8 mg ml1 protein, 100 mM NaCl, 50 mM Tris–HCl pH 8.0) with 2.0 ml reservoir solution (200 mM NH4Cl, 15% PEG MME 2000, 50 mM Tris–HCl pH 8.5, 20 mM CdCl2 or ZnCl2). Crystals appeared after 1 d and reached maximum dimensions of 300  100  50 mm within one week. Before freezing, the crystals were transferred to 15% PEG MME 2000, 15% PEG 400, 200 mM ammonium chloride, 50 mM Tris–HCl pH 8.5. X-ray diffraction data were collected from the crystals on EMBL beamline X12 (DESY, Hamburg) or the BL14.1 beamline at BESSY (Berlin) and were processed using XDS (Kabsch, 2010). Detailed data-collection statistics are given in Table 1. 2.4. Structure determination and refinement The protein structures were solved by the molecular-replacement method using the PHENIX package (Adams et al., 2002) with a hexamer of wild-type PaeHfq as the initial model (PDB code 1u1s; Nikulin et al., 2005). The simulated-annealing protocol following conventional residual refinement in combination with manual inspection in Coot (Emsley & Cowtan, 2004) was used to refine the model. Water molecules were introduced into the model using the ‘water pick’ function of Coot and the highest peaks in the Fo  Fc map were assigned to ions. At the final stage anisotropic ADP refinement of H57T PaeHfq was implemented, improving the R and Rfree factors from 0.199 and 0.244 to 0.149 and 0.218, respectively. The structure coordinates of H57A PaeHfq and H57T PaeHfq have been deposited in the Protein Data Bank (PDB codes 3inz and 3m4g, respectively). 3. Results and discussion 3.1. Crystal structures of H57A PaeHfq and H57T PaeHfq The crystal structures of H57A PaeHfq and H57T PaeHfq were solved and refined to 2.05 and 1.7 A˚ resolution, respectively (Table 1). The substitutions did not change the overall shape of the hexamer or the conformations of the monomers (Fig. 1b). In the wild-type protein the side chain of His57 formed two hydrogen bonds to the main-chain O atoms of the adjacent monomer (Fig. 2a). We supposed that the mutations would result in the disappearance of one or both of these hydrogen bonds. Indeed, the substitution of His57 by alanine led to a loss of the hydrogen bonds (Fig. 2b). In contrast, the replacement of His57 by threonine gave rise to the formation of new hydrogen bonds between adjacent monomers that replaced those in the wild-type protein. Two water molecules acted as bridges connecting the hydroxyl of the threonine of one monomer to the main-chain carbonyl O atoms of Thr57 and Ile59 of another molecule (Fig. 2c). Nevertheless, the compensation was not completely equivalent. In the wild-type protein one of the hydrogen bonds formed by His57 is inaccessible to solvent, whereas in H57T PaeHfq the water-bridge hydrogen bonds are accessible. In this case the protein atoms could easily form new hydrogen bonds to solvent. At higher temperature the water molecules could even escape from their sites. In this case, the hydroxyl group of Thr57 could be posi- tioned at a short distance from the two carbonyl O atoms of the neighbouring monomer, which is not desirable. To prove this hypothesis, we measured the stability of the Hfq mutant proteins. 3.2. Stability of the Hfq mutant forms To evaluate the influence of the His57 substitutions on PaeHfq hexamer stability, CD spectra of the wild-type protein and its mutant forms were measured. At room temperature all these proteins had similar spectra corresponding to an / structure (Fig. 3a). It was found that wild-type PaeHfq possesses extreme stability: its CD spectrum did not change during heating to 366 K or on the addition of urea up to 8 M. Difference scanning calorimetric experiments showed that the denaturation peak of wild-type PaeHfq appeared near 393 K (V. V. Filimonov, personal communication). Therefore, PaeHfq has one of the highest denaturation temperatures of known proteins (Tanaka et al., 2006). The secondary structure of wild-type PaeHfq, as well as those of its H57A, H57T and H57N mutants, was completely destroyed in the presence of 5 M GdnHCl (Fig. 3a). The GdnHCl- induced unfolding of the proteins under equilibrium conditions demonstrated that all of the substitutions changed the stability of the protein considerably but in a similar way (Fig. 3b). structural communications Acta Cryst. (2010). F66, 760–764 Moskaleva et al.  Hfq 763 Figure 3 (a) CD spectrum of wild-type and mutant (H57A, H57T, H57N) PaeHfq proteins under nondenaturing conditions (lower lines) and in the presence of 5 M GdnHCl (upper lines). (b) Relative change of ellipticity at 220 nm during equilibrium unfolding of the proteins by GdnHCl. (c) Relative change of ellipticity at 220 nm during temperature unfolding of the mutant proteins in the presence of 1 M GdnHCl. Table 1 Data-collection and refinement statistics. Values in parentheses are for the highest resolution shell. H57T PaeHfq H57A PaeHfq Macromolecule details PDB code 3inz 3m4g No. of residues per monomer 82 82 Molecular assembly Hexamer Hexamer Molecular weight of the hexamer (Da) 54411 54489 Data-collection statistics Wavelength (A˚ ) 1.00 0.91841 Resolution range (A˚ ) 30.0–1.7 (1.74–1.7) 30.0–2.05 (2.16–2.05) Space group P21212 P1 Unit-cell parameters (A˚ , ) a = 61.3, b = 71.2, c = 104.4,  =  =  = 90 a = 66.5, b = 66.6, c = 68.7,  = 91.8,  = 115.3,  = 119.9 Total reflections 337107 (10581) 234040 (34408) Unique reflections 50597 (2983) 53606 (7783) Redundancy 6.7 (3.5) 4.4 (4.4) Completeness (%) 94.1 (80.4) 97.4 (96.5) Rmerge (%) 4.0 (36.2) 5.5 (49.0) Average I/(I) 27.9 (3.9) 14.4 (3.0) Wilson B factor (A˚ 2) 31.2 32.9 Refinement statistics Resolution (A˚ ) 30.0–1.70 (1.74–1.70) 30.0–2.05 (2.09–2.05) Completeness (%) 94.1 (80.4) 97.4 (96.5) Reflections 50571 (2983) 53561 (2704) Test reflections 2528 (131) 2725 (127) Rwork (%) 0.149 (0.178) 0.194 (0.296) Rfree (%) 0.218 (0.247) 0.261 (0.393) No. of waters 379 331 No. of ions 11 18 R.m.s. deviation from ideal geometry Bonds (A˚ ) 0.010 0.005 Angles () 1.315 0.903 Chirality () 0.103 0.059 Planarity () 0.006 0.004 Average B value (A˚ 2) Main chain 31.39 45.24 Side chain and water 38.74 50.82 MolProbity results Ramachandran favoured (%) 95.20 95.76 Ramachandran allowed (%) 98.74 99.88 Ramachandran outliers (%) 1.26 0.12 To reveal the difference in stability of the PaeHfq mutants, the relative changes in ellipticity at 220 nm were measured during temperature unfolding (Fig. 3c). The presence of 1 M GdnHCl in the buffer was important in order to melt the proteins within the oper- ating range of the spectropolarimeter. The H57N, H57A and H57T mutant forms of PaeHfq had melting temperatures of 346, 343 and 341 K, respectively, whereas wild-type PaeHfq retained its structure up to 366 K. Compared with the other mutants, the H57T PaeHfq had the lowest melting temperature, which was accompanied by a deterioration of melting-process cooperativity. The reason for this behaviour of H57T PaeHfq appears to be a consequence of the incorporation of water molecules between the side chain of the threonine and the main chain of the adjacent protein monomer as discussed above. In the H57N PaeHfq protein stereochemical analysis showed that the asparagine residue is able to organize a direct but water-accessible hydrogen bond to the main-chain atoms of the neighbouring monomer. Therefore, this substitution resulted in a decreased melting temperature for the mutant protein forms but did not lead to deterioration of the melting cooperativity. The research was supported by the Russian Academy of Sciences, the Russian Federal Agency for Science and Innovation (02.740.11.0295), the Russian Foundation for Basic Research (10-04- 00818) and the Program of the RAS on Molecular and Cellular Biology. References Achsel, T., Brahms, H., Kastner, B., Bachi, A., Wilm, M. & Lu¨hrmann, R. (1999). EMBO J. 18, 5789–5802. Achsel, T., Stark, H. & Lu¨hrmann, R. (2001). Proc. Natl Acad. Sci. USA, 98, 3685–3689. Adams, P. D., Grosse-Kunstleve, R. W., Hung, L.-W., Ioerger, T. R., McCoy, A. J., Moriarty, N. W., Read, R. J., Sacchettini, J. C., Sauter, N. K. & Terwilliger, T. C. (2002). Acta Cryst. D58, 1948–1954. Brennan, R. G. & Link, T. M. (2007). Curr. Opin. Microbiol. 10, 125–133. Emsley, P. & Cowtan, K. (2004). Acta Cryst. D60, 2126–2132. Hajnsdorf, E. & Re´gnier, P. (2000). Proc. Natl Acad. Sci. USA, 97, 1501–1505. Kabsch, W. (2010). Acta Cryst. D66, 125–132. Kambach, C., Walke, S., Young, R., Avis, J., de La Fortelle, E., Raker, V., Lu¨hrmann, R., Li, J. & Nagai, K. (1999). Cell, 96, 375–387. Kilic, T., Sanglier, S., Van Dorsselaer, A. & Suck, D. (2006). Protein Sci. 15, 2310–2317. Kufel, J., Bousquet-Antonelli, C., Beggs, J. D. & Tollervey, D. (2004). Mol. Cell. Biol. 24, 9646–9657. Mayes, A. E., Verdone, L., Legrain, P. & Beggs, J. D. (1999). EMBO J. 18, 4321–4331. Møller, T., Franch, T., Hojrup, P., Keene, D. R., Bachinger, H. P., Brennan, R. G. & Valentin-Hansen, P. (2002). Mol. Cell, 9, 23–30. Mura, C., Phillips, M., Kozhukhovsky, A. & Eisenberg, D. (2003). Proc. Natl Acad. Sci. USA, 100, 4539–4544. Naidoo, N., Harrop, S. J., Sobti, M., Haynes, P. A., Szymczyna, B. R., Williamson, J. R., Curmi, P. M. G. & Mabbutt, B. C. (2008). J. Mol. Biol. 377, 1357–1371. Nikulin, A., Stolboushkina, E., Perederina, A., Vassilieva, I., Blaesi, U., Moll, I., Kachalova, G., Yokoyama, S., Vassylyev, D., Garber, M. & Nikonov, S. (2005). Acta Cryst. D61, 141–146. Salgado-Garrido, J., Bragado-Nilsson, E., Kandels-Lewis, S. & Se´raphin, B. (1999). EMBO J. 18, 3451–3462. Sauter, C., Basquin, J. & Suck, D. (2003). Nucleic Acids Res. 31, 4091–4098. Schumacher, M. A., Pearson, R. F., Møller, T., Valentin-Hansen, P. & Brennan, R. G. (2002). EMBO J. 21, 3546–3556. Se´raphin, B. (1995). EMBO J. 14, 2089–2098. Sledjeski, D. D., Whitman, C. & Zhang, A. (2001). J. Bacteriol. 183, 1997–2005. Tanaka, T., Sawano, M., Ogasahara, K., Sakaguchi, Y., Bagautdinov, B., Katoh, E., Kuroishi, C., Shinkai, A., Yokoyama, S. & Yutani, K. (2006). FEBS Lett. 580, 4224–4230. To¨ro¨, I., Basquin, J., Teo-Dreher, H. & Suck, D. (2002). J. Mol. Biol. 320, 129–142. To¨ro¨, I., Thore, S., Mayer, C., Basquin, J., Se´raphin, B. & Suck, D. (2001). EMBO J. 20, 2293–2303. Valentin-Hansen, P., Eriksen, M. & Udesen, C. (2004). Mol. Microbiol. 51, 1525–1533. Verdone, L., Galardi, S., Page, D. & Beggs, J. D. (2004). Curr. Biol. 14, 1487– 1491. Vytvytska, O., Moll, I., Kaberdin, V. R., von Gabain, A. & Bla¨si, U. (2000). Genes Dev. 14, 1109–1118. Walke, S., Bragado-Nilsson, E., Se´raphin, B. & Nagai, K. (2001). J. Mol. Biol. 308, 49–58. Will, C. L. & Lu¨hrmann, R. (2001). Curr. Opin. Cell Biol. 13, 290–301. Wilusz, C. J. & Wilusz, J. (2005). Nature Struct. Mol. Biol. 12, 1031–1036. Zhang, A., Altuvia, S., Tiwari, A., Argaman, L., Hengge-Aronis, R. & Storz, G. (1998). EMBO J. 17, 6061–6068. Zhang, A., Wassarman, K. M., Ortega, J., Steven, A. C. & Storz, G. (2002). Mol. Cell, 9, 11–22. structural communications 764 Moskaleva et al.  Hfq Acta Cryst. (2010). F66, 760–764
3M4I
Crystal structure of the second part of the Mycobacterium tuberculosis DNA gyrase reaction core: the TOPRIM domain at 1.95 A resolution
Structural Insights into the Quinolone Resistance Mechanism of Mycobacterium tuberculosis DNA Gyrase Je´re´mie Piton1,2,3, Ste´phanie Petrella4, Marc Delarue1,2, Gwe´nae¨lle Andre´-Leroux2,5, Vincent Jarlier4, Alexandra Aubry4, Claudine Mayer1,2,6* 1 Unite´ de Dynamique Structurale des Macromole´cules, De´partement de Biologie Structurale et Chimie, Institut Pasteur, Paris, France, 2 URA 2185, CNRS, Paris, France, 3 UPMC Univ Paris 06, Paris, France, 4 UPMC Univ Paris 06, EA1541, Bacte´riologie-Hygie`ne, Paris, France, 5 Unite´ de Biochimie Structurale, De´partement de Biologie Structurale et Chimie, Institut Pasteur, Paris, France, 6 Universite´ Paris Diderot Paris 7, Paris, France Abstract Mycobacterium tuberculosis DNA gyrase, an indispensable nanomachine involved in the regulation of DNA topology, is the only type II topoisomerase present in this organism and is hence the sole target for quinolone action, a crucial drug active against multidrug-resistant tuberculosis. To understand at an atomic level the quinolone resistance mechanism, which emerges in extensively drug resistant tuberculosis, we performed combined functional, biophysical and structural studies of the two individual domains constituting the catalytic DNA gyrase reaction core, namely the Toprim and the breakage- reunion domains. This allowed us to produce a model of the catalytic reaction core in complex with DNA and a quinolone molecule, identifying original mechanistic properties of quinolone binding and clarifying the relationships between amino acid mutations and resistance phenotype of M. tuberculosis DNA gyrase. These results are compatible with our previous studies on quinolone resistance. Interestingly, the structure of the entire breakage-reunion domain revealed a new interaction, in which the Quinolone-Binding Pocket (QBP) is blocked by the N-terminal helix of a symmetry-related molecule. This interaction provides useful starting points for designing peptide based inhibitors that target DNA gyrase to prevent its binding to DNA. Citation: Piton J, Petrella S, Delarue M, Andre´-Leroux G, Jarlier V, et al. (2010) Structural Insights into the Quinolone Resistance Mechanism of Mycobacterium tuberculosis DNA Gyrase. PLoS ONE 5(8): e12245. doi:10.1371/journal.pone.0012245 Editor: Hendrik W. van Veen, University of Cambridge, United Kingdom Received April 22, 2010; Accepted July 21, 2010; Published August 18, 2010 Copyright:  2010 Piton et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: J.P. is funded by the ‘‘Ministere de l’enseignement superieur et de la recherche.’’ The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: mayer@pasteur.fr Introduction Type II topoisomerases are essential and ubiquitous nucleic acid- dependent nanomachines involved in the regulation of DNA topology and especially in the regulation of DNA supercoiling [1]. Type II topoisomerases act by an ATP-dependant double-stranded DNA break [1]. Except archaeal topoisomerase VI [2,3], they all belong to a single protein superfamily, the type IIA topoisomerases, sharing homologous sequences and overall structures [4]. However, they have acquired distinct functions during evolution [1]. Bacterial genomes usually encode two type IIA enzymes, DNA gyrase and topoisomerase IV. DNA gyrase facilitates DNA unwinding at replication forks and topoisomerase IV has a specialized function in mediating the decatenation of interlocked daughter chromosomes [5]. Mycobacterium tuberculosis, the aetiologic agent of tuberculosis, is unusual in possessing only one type II topoisomerase, DNA gyrase [6]. Consequently, the M. tuberculosis DNA gyrase exhibits a different activity spectrum as compared to other DNA gyrases, namely it supercoils DNA with an efficiency comparable to that of other DNA gyrases but shows enhanced relaxation, DNA cleavage, and decatenation activities [7]. DNA gyrase and topoisomerase IV consist of two subunits (GyrA and GyrB in DNA gyrase, ParC and ParE in topoisomerase IV), which form the catalytically active heterotetrameric complex (i.e. A2B2 and C2E2, respectively). Subunit A consists of two domains, the N-terminal breakage-reunion domain and a carboxy- terminal domain, termed CTD. Subunit B consists of the ATPase domain followed by the Toprim domain. The GyrB Toprim and GyrA breakage-reunion domains come from separate subunits and cooperatively form the enzyme core (Figure 1A). The breakage- reunion domain contains the catalytic tyrosine responsible for the cleavage and religation of the DNA double helix. Although the structure of a fully intact, active type IIA topoisomerase has yet to be determined, structural and biochemical studies of the individual fragments have led several authors to propose a model of its global quaternary structure and a catalytic mechanism of the holoenzyme [8]. The breakage-reunion domain binds a DNA segment termed the ‘gate’ or G-segment at the DNA-gate. The N-terminal ATPase domains dimerize upon ATP binding, capturing the DNA duplex to be transported (T-segment). The T-segment is then passed through a transient break in the G-segment opened by the breakage-reunion domains, the DNA is resealed and the T- segment released through a protein gate, the C-gate, prior to resetting of the enzyme to the open clamp form. Quinolones, which target the two bacterial type II topoisom- erases, exert their powerful antibacterial activity by interfering with the enzymatic reaction cycle. Specifically, they bind to the enzyme-DNA binary complex, thereby stabilizing the covalent enzyme tyrosyl-DNA phosphate ester. The resulting ternary complexes block DNA replication and lead to cell death [9]. PLoS ONE | www.plosone.org 1 August 2010 | Volume 5 | Issue 8 | e12245 Quinolones are one of the most effective second-line drugs in the treatment of multidrug-resistant tuberculosis (MDR-TB; strains resistant to the two main antituberculous drugs, rifampicin and isoniazid) [10] and are currently under study for shortening treatment duration of drug-susceptible tuberculosis [11]. Tuber- culosis still remains the leading cause of death from a curable infectious disease causing millions of deaths annually (http://www. who.int). Unfortunately, due to the long and complex nature of TB treatment, inappropriate use of first line antituberculous drugs is common, leading to the emergence of drug-resistant bacilli, especially MDR strains. Widespread dissemination of these bacilli poses a serious threat to global TB control [12]. Compared with E. coli, the ‘‘intrinsic resistance’’ of M. tuberculosis to quinolones is relatively high, mainly due to the primary structure of DNA gyrase. Namely, amino acids at positions 81 and 90 in GyrA and 482 in GyrB have been demonstrated to be involved in ‘‘intrinsic quinolone resistance’’ [13]. Nonetheless, quinolones, and in particular fluoroquinolones, are essential antibiotics for MDR- Figure 1. Domain organization and structures of the individual domains from the M. tuberculosis DNA gyrase catalytic core. A. Domain organization of the M. tuberculosis DNA gyrase. The catalytic core is composed by the Toprim domain and the breakage-reunion domain. B. Three orthogonal views of the dimeric Toprim domain from M. tuberculosis colored by regions. The crystal structure of the complete Toprim domain (TopBK) encompasses residues T448 to E654. The schematically represented primary sequence is colored as in the structure. The N-terminal residue numbers of the regions (Toprim, tail and hinge) and the TopBK C-terminal residue number are indicated. The Toprim region, constituted by discontinuous N- and C-terminal sequence segments and containing the magnesium-binding site (E459, D532 and D534) and the QRDR-B (Quinolone Resistance Determining Region in GyrB) is colored in yellow, the Tail region in purple and the hinge between the two regions in blue. The second monomer generated by a crystallographic two-fold axis is represented in grey. C. Three views of the dimeric breakage-reunion domain from M. tuberculosis colored by regions. The crystal structure of the complete breakage-reunion domain (GA57BK) extends from D9 to A501. The N-terminal helix is colored in red, the DNA-gate containing the catalytic residues R128 and Y129 and the QRDR-A in blue, the ‘tower’ in green, the helix-bundle in orange and the C-gate in purple. doi:10.1371/journal.pone.0012245.g001 M. tuberculosis DNA Gyrase PLoS ONE | www.plosone.org 2 August 2010 | Volume 5 | Issue 8 | e12245 TB [13,14]. However, M. tuberculosis develops ‘‘acquired resis- tance’’ to quinolones following prolonged exposure, leading to the emergence of extensively drug-resistant (XDR) strains (MDR-TB strains resistant to any fluoroquinolone and to at least one of three injectable second-line anti-TB drugs) [15,16,17]. This ‘‘acquired resistance’’ is mainly a result of mutations in the DNA gyrase sequence [18,19]. Mutations conferring bacterial resistance to quinolones occur in two short discrete segments termed the quinolone resistance-determining regions (QRDR) [20] located in the breakage-reunion domain of GyrA subunit (QRDR-A) and less frequently in the Toprim domain of GyrB (QRDR-B) [20,21,22]. Among the described mutations, we have unequivocally demon- strated that the nature of the amino acids at positions 88, 90 and 94 in GyrA plays a crucial role in the ‘‘acquired resistance’’ to quinolones (Table 1) [21,23]. The challenge of better understanding the complex mechanism of quinolone resistance in M. tuberculosis requires high-resolution structures of the antibiotic targets. Following our previous results, the aim of this work was to obtain a 3-dimensional understanding of the relationships between a given amino acid mutation and quinolone resistance phenotype in M. tuberculosis. Simultaneously to our results, two structures of M. tuberculosis DNA gyrase domains were published last year, the low resolution GyrB’ structure (PDB code 2ZJT, [24]) and the truncated MtGyrA59 domain (PDB code 3ILW, [25]). The first picture of the enzyme-quinolone interac- tions was given by the low resolution structures of Streptococcus pneumoniae ParC breakage-reunion and ParE Toprim domain in complex with DNA and quinolones (PDB codes 3FOF and 3K9F, [26]). Moreover, other efforts to develop new potent catalytic inhibitors of bacterial DNA gyrase were illustrated by the crystal structure of E. coli DNA gyrase in complex with the bifunctional antibiotic simocyclinone D8 [27]. Its mode of action is unique in that it directly interacts with DNA gyrase to prevent its binding to DNA. In this work, we combined X-ray crystallographic studies, sedimentation velocity experiments and activity assays of the two domains that form the enzyme core of M. tuberculosis DNA gyrase, the GyrB Toprim and GyrA breakage-reunion domains. We solved two high resolution structures of the Toprim domain displaying two different conformations of the metal-binding site, to 2.1 and 1.95 A˚ resolution, respectively. The crystal structure of the breakage-reunion domain we solved to 2.7 A˚ resolution, revealed a promising interaction that will be further exploited for drug design. This interaction involves the N-terminal helix, which is anchored in the active site of a symmetry-related molecule. Additionally, using the crystal structures of both domains, we modeled the catalytic reaction core in complex with DNA and a quinolone. This study brings the first structural explanation on quinolone resistance mechanism of M. tuberculosis DNA gyrase. Results Crystal structures of the Toprim and breakage-reunion domains are biologically relevant The C-terminal GyrB domain (Toprim domain, residues 448– 654) and the entire N-terminal GyrA domain (breakage-reunion domain, known as GyrA59 in E. coli, residues 1–502), hereafter named TopBK and GA57BK, respectively, were overproduced and purified. DNA cleavage activity assays show that TopBK is able to catalyze DNA breaks when associated to the full-length A subunit. Similarly, GA57BK is able to catalyze DNA breaks when associated with the full-length B subunit (Figure 2A and B). Interestingly, the GA57BK-TopBK complex has DNA cleavage activity, showing that these domains possess all determinants for DNA cleavage and confirming that these two domains form the catalytic reaction core of the M. tuberculosis DNA gyrase (Figure 2A and B). In addition to DNA cleavage, some nicking is also observed when TopBK is associated either with the full length GyrA, or with GA57BK (Figure 2B). This could be the result of a decrease in the complex stability when TopBK is used in the activity assays. TopBK was crystallized in presence of magnesium (crystal I) and calcium (crystal II) and the structures were solved at 2.1 A˚ and 1.95 A˚ resolution, respectively, with one monomer in the asymmetric unit in both cases. Slight modifications of the previously described crystallization conditions [24,28], e.g. mod- ifying the pH value and adding divalent cations, led to a space group change and a substantial increase in the resolution (2.8 to 1.95 A˚ ). A crystallographic two-fold axis generates a dimeric structure, similar to the dimer observed in the asymmetric unit of the GyrB’ structure (2ZJT). GA57BK corresponds to the entire N- terminal domain with a molecular mass of 57 kDa. The crystals belong to space group C2, with a dimer in the asymmetric unit. Clear electron density was observed for the N-terminal fragment that could be built either entirely (chain A) or partially (chain B) because of different crystal contacts. Consequently, the final model spans residues 9 to 499 for chain A and 29 to 501 for chain B. Both TopBK and GA57BK display a dimeric structure in the crystal (Figure 1B and C). The biological relevance of these dimeric forms was investigated using analytical ultracentrifugation. Sedimentation experiments reveal that TopBK and GA57BK exhibit different behaviour in solution. In the case of TopBK, two species are observed with a 50/50 distribution when using a protein concentration corresponding to the crystallization condi- tions (Figure 2C). The two species display a sedimentation coefficient of 2.360.1 S and 3.660.2 S, corresponding to the monomer and the dimer, respectively, according to the theoretical sedimentation coefficient values calculated from the crystallo- graphic structure (2.2 and 3.5 S, respectively). In contrast, GA57BK is mainly dimeric in solution (Figure 2C). Sedimentation experiments showed that one species was observed with a sedimentation coefficient of 5.460.2 S, compatible with the value calculated from the crystallographic dimer structure (5.6 S). The good agreement between these experimental and theoretical values indicates that the dimeric conformation of GA57BK is stable in solution. These results suggest that the biological unit is a dimer. Table 1. Mutations described in M. tuberculosis strains implicated in ‘‘acquired’’ resistance to quinolones. Mutation Effect on quinolone susceptibility Reference GyrA GyrB G88A resistance 13 A90V resistance 21 D94A, G, N resistance 21 N499D resistance 21 T80A no effect 21 T80A+A90G hypersusceptibility 21 Summary of mutations described in M. tuberculosis strains (e.g. clinical strains or strains cultured in vitro in presence of quinolone in order to select a resistant strain), which have been unequivocally demonstrated as implicated in ‘‘acquired’’ resistance. doi:10.1371/journal.pone.0012245.t001 M. tuberculosis DNA Gyrase PLoS ONE | www.plosone.org 3 August 2010 | Volume 5 | Issue 8 | e12245 The crystal structure of the isolated Toprim domain is a dimer The overall fold of the TopBK structure is very similar in both crystal forms, and also similar to the previously published GyrB’ structure (2ZJT) [24] and to the Toprim domain of the known eukaryotic counterpart, the yeast topoisomerase II [29]. The structure displays a two-domain organization, a globular domain constituted by discontinuous segments (residues 448–564 and 633– 654) and the Tail domain (residues 565–608) connected by a loop- helix-loop hinge region (residues 609–632) (Figure 1B). The globular domain, organized in a Rossmann-like fold, contains the Toprim domain described by Aravind and collaborators [30] and the QRDR-B (residues 461–499) (Figure 3A and S1). The Tail domain comprises a three-stranded antiparallel b-sheet and an a- helix. In the globular domain, the conserved acidic triad (E459, D532, D534), which constitutes the signature of the Toprim domain, binds the magnesium ion essential for the catalysis of the cleavage-ligation reaction. In the structure of crystal I, the magnesium ion is not visible, despite being present in the crystallization condition. However, side chains of the catalytic triad are in conformations which would allow ion coordination, as observed in the yeast topoisomerase II in complex with DNA (Figure 3B). Presumably, the ion is not bound due to the absence of the DNA. When magnesium is substituted in the crystallization conditions by calcium (TopBK crystal II), side chains of the triad are observed in an inactive conformation similar to the one observed for the low resolution M. tuberculosis Toprim domain structure [24]. The Toprim domain forms a dimer with a symmetry related molecule in both crystal structures (crystal I and II), burying 1017 A˚ 2 at the protein-protein interface, indicative of a biologi- cally relevant interaction. The two species observed in sedimen- tation experiments with a 50/50 distribution are identified as the monomeric TopBK domain and the crystallographic dimer suggesting that this crystallographic dimer exists in solution outside the context of the full-length subunit. Surprisingly, the high resolution structures of TopBK, revealed two disordered regions, between b1 and b2 (residues 460–474) and between b2 and a2 (residues 484–492) (Figure 3A). These regions are structured in the context of the catalytic core or in presence of DNA. The first disordered region corresponds to the a1-helix [30], as observed in the three structures of the yeast topoisomerase II [29,31,32] and in the structure of the S. pneumoniae reaction core [26]. Interestingly, this region is located at the dimer interface and placing an a-helix would generate steric hindrance between the two helices of the crystallographic related monomers (Figure 3C). The second disordered region, the loop between b2 and a2, is exposed to the solvent explaining its high flexibility. In the structures of type II topoisomerases in complex with DNA, this loop (hereafter named DBL for DNA-Binding Loop) constitutes the interface between the Toprim domain and DNA and is stabilized through protein-DNA interactions. The breakage-reunion domain is in a closed conformation GA57BK forms a biological dimer in the asymmetric unit, generating a heart-like shaped structure with outer dimensions of 1006100690 A˚ (Figure 1C) and a central hole of 30 A˚ diameter allows the passage of the T-segment from the DNA-gate to the C- gate. GA57BK forms a biological dimer in a ‘closed’ conformation in the asymmetric unit, as the C-gate, which constitutes the so- called primary dimer interface, and the DNA-gate, the secondary protein-protein interface, are both closed (Figure S2). This closed conformation is observed in all isolated breakage-reunion domain structures, the MtGyrA59 from M. tuberculosis [25], GyrA59 from E. coli and of the two topoisomerase IV structures from S. pneumoniae [33] and from S. aureus [34]. This shows that the closed conformation is stable and energetically favorable. This stability is Figure 2. Activity assays and oligomerization of the TopBK and GA57BK domains. A. The quinolone-mediated DNA cleavage activity test measured on supercoiled pBR322 DNA (0.4 mg) as a substrate in the presence of moxifloxacin (50 mg/ml) and 2.5 mg of each subunit alone: full length subunit A (ABK), full length subunit B (BBK), GA57BK and TopBK. Lanes a and b are supercoiled pBR322 DNA and control of cleavage activity with WT M. tuberculosis DNA gyrase (ABK and BBK), respectively. B. The quinolone-mediated DNA cleavage activity test measured on supercoiled pBR322 DNA (0.4 mg) as a substrate in the presence of moxifloxacin (50 mg/ml) with various amounts (indicated by values in mg) of GA57BK associated with the full length subunit B (BBK, 1 mg), various amounts of TopBK with the full length subunit A (ABK, 1 mg), and various amounts (indicated by values in mg) of the binary complex constituted by GA57BK and TopBK. Lanes M, a and b are DNA size markers, supercoiled pBR322 DNA and control of cleavage activity with WT M. tuberculosis DNA gyrase (ABK and BBK), respectively. N, L and S denote nicked, linear and supercoiled DNA, respectively. C. Sedimen- tation experiments of GA57BK and TopBK. The single peak of GA57BK corresponds to the dimer, with a sedimentation coefficient of 5.460.2 S. The two peaks observed for TopBK correspond to the monomeric and dimeric form, with sedimentation coefficients of 2.360.1 S and 3.460.2 S, respectively. c(s) on the y-axis designates the distribution of the sedimentation coefficients observed for the experiment. doi:10.1371/journal.pone.0012245.g002 M. tuberculosis DNA Gyrase PLoS ONE | www.plosone.org 4 August 2010 | Volume 5 | Issue 8 | e12245 essential to generate the interface needed to trap the DNA G- fragment in order to start the topoisomerase cycle. This is in agreement with FRET experiments showing that the DNA-gate of the Bacillus subtilis DNA gyrase is predominantly in the closed conformation during the DNA relaxation and supercoiling reactions [35]. When comparing all five protein-protein interfaces (sum of DNA- and C-gate interfaces), the highest value is observed for both structures of M. tuberculosis DNA gyrase (Table S1). Whereas the C-gate displays similar values, ranging from 1029 to 1120 A˚ 2, differences in interface area are observed at the DNA- gate with a value of more than 800 A˚ 2 for M. tuberculosis, representing nearly one half of the total interface. Both structures confirm that the M. tuberculosis breakage-reunion domain has a compact closed conformation, especially at the level of the DNA- gate, whatever the crystal environment. Each monomer of GA57BK contains five distinct regions, the N-terminal fragment (residues 9–41), deleted in MtGyrA59 and disordered in the homologous structures, and the four typically observed regions in breakage-reunion domains of all type II topoisomerases (Figure 1C and S3). In this way, GA57BK resembles the type II topoisomerase structures in complex with Toprim, namely the yeast topoisomerase II or the structure of the complex between ParC, ParE, DNA and a fluoroquinolone (see below). The next four domains, the DNA-gate (residues 42–169), the ‘tower’ (residues 170–355 and 491–501), the C-gate (residues 401–444) and the three-helix bundle (residues 356–400 and 445– 490) (Figure 1C), exhibit an overall structural fold similar to that observed for other bacterial type II topoisomerases [33,34,36] and the yeast topoisomerase II [29,31,32]. The DNA-binding helix- turn-helix motif (a3 and a4 helices), the QRDR-A (residues 74– 113) and the catalytic residues involved in DNA cleavage, namely R128 and Y129, are localised in the DNA-gate. The active site is blocked through crystal contacts established by the N-terminal helix In contrast to other structures of the breakage-reunion domain alone (i.e. E. coli DNA gyrase, S. aureus and S. pneumoniae topoisomerase IV), the N-terminal segment of GA57BK (residues 9–41) is ordered and is organized in two distinct secondary structures (Figure 4A). Residues D9 to E16 form a loop whose B factors indicate high flexibility, followed by a 24-residue long a-helix (Figure 4B). Until now, this helix was only observed when the Toprim domain is also present, whether DNA is complexed (in the structure of the S. pneumoniae topoisomerase IV catalytic core in complex with DNA, 3FOF [26] and the yeast topoisomerase II catalytic core-DNA complex, 2RGR [29]) or not (in the two structures of the yeast topoisomerase II catalytic core, 1BJT [32] and 1BGW [31]). A previously unobserved feature of our crystal structure of GA57BK is the interaction between this N-terminal region with neighbouring molecules in the crystal packing. As shown in Figure 4, the N-terminal fragment residues of chain A in a given asymmetric unit clearly establish direct contacts with the active site residues of its nearest neighbour (chain A’) in the adjacent asymmetric unit. As these two molecules are related by the crystallographic two-fold axis, this interaction is reciprocal. The a- helix is deeply anchored in the active site of its neighbouring molecule. Several hydrogen-bonding interactions link E23 from the a-helix to the a3–a4 region, namely D89, A90 and S91 from the symmetry-related molecule (Figure 4C). R26 links the main chain carbonyl-group of H87 via a water molecule. In addition, D30 establishes hydrogen bonds with the hydroxyl group of the catalytic tyrosine (Y129) and a salt bridge with the catalytic arginine (R128). Finally, on the opposite face of the N-terminal helix, S27 and Y31 form an H-bonding network with R39 and R54 from the symmetry-related molecule (Figure 4C). This arrangement buries a surface area of 1227 A˚ 2, indicating a stable interaction. The resulting tetramer could explain the small peak observed in sedimentation experiments (Figure 2B). Further studies exploiting this interaction for drug design will be investigated. A peptide of 16 amino acids corresponding to residues 15 to 30 of the M. tuberculosis breakage-reunion domain will be used as an inhibitor for M. tuberculosis DNA gyrase in activity and binding assays in order to develop structure-activity relationships. Combined docking and molecular dynamics simu- Figure 3. The TopBK magnesium-binding site. A. Overall view of the dimeric structure of the Toprim domain from M. tuberculosis. One monomer constituting the asymmetric unit is represented in green, the second monomer generated by a crystallographic two-fold axis in grey. The secondary structures are indicated by black labels. The locations of the two disordered regions, the DNA Binding Loop (DBL) and the a1-helix, are indicated by the red labels ‘‘DBL’’ and ‘‘a1’’, respectively. B. The magnesium-binding site of both M. tuberculosis TopBK structures, TopBK crystal I (3IFZ, in green) and TopBK crystal II (3M4I, in purple) with the conserved residues, E459, D532 and D534. The active site of the S. cerevisiae Toprim domain (2RGR) is represented in blue and its bound magnesium ion in orange. C. Close view of the TopBK dimer interface. The two symmetry-related a1 helices (shown in red and grey) generate steric clashes. doi:10.1371/journal.pone.0012245.g003 M. tuberculosis DNA Gyrase PLoS ONE | www.plosone.org 5 August 2010 | Volume 5 | Issue 8 | e12245 lations will be used to design small molecules that mimic the peptide-active site interactions [37]. The M. tuberculosis breakage-reunion domain possesses two specific structural motifs Unexpectedly, structural comparison of M. tuberculosis DNA gyrase to other type II topoisomerases clearly reveals that there is no significant difference between a DNA gyrase from species containing only one type II topoisomerase and the other type II topoisomerases, DNA gyrase and topoisomerase IV, generally found in bacteria (Figure S4). However, we found that two regions could be correlated to the wider substrate spectrum of M. tuberculosis DNA gyrase function. First, a sequence motif (DPP) in the loop between the a3–a4 DNA-binding motif and the catalytic tyrosine residue resembles the sequence observed in topoisomerases IV and is rarely observed in DNA gyrase sequences. Localised at the side of Figure 4. The active site of M. tuberculosis DNA gyrase is blocked by the N-terminal helix of a symmetry-related molecule. A. Two dimers of GA57BK, related by the crystallographic two-fold axis, interact through the N-terminal helix. B. Omit maps for the N-terminal helix. The (2Fobs – Fcalc) map shown in blue is contoured at 1.5 s whilst the (Fobs – Fcalc) map shown in green is contoured at 3 s. C. Detailed interactions of the N-terminal helix (chain A’, in hot pink) in the active site of the symmetry-related molecule (chain A, in light green). Y31 of the N-terminal helix and R54 of the symmetry-related molecule are located on the back-side of the helix and are not represented for better clarity. D. Based on the model discussed in the text, the N-terminal helix (chain A’, in hot pink) occupies the quinolone-binding pocket (QBP) and clashes with the modeled DNA, represented in orange, and the fluoroquinolone, in yellow, bound to the QBP. doi:10.1371/journal.pone.0012245.g004 M. tuberculosis DNA Gyrase PLoS ONE | www.plosone.org 6 August 2010 | Volume 5 | Issue 8 | e12245 the DNA-gate and in direct interaction with DNA (Figure 5A and B), this loop could contribute to the topoisomerase IV-like activity (i.e. decatenation) of M. tuberculosis DNA gyrase. Second, a specific insertion in the M. tuberculosis sequence consists in a negatively charged motif DEEE (residues 211–214) (Figure S3). In the structure, this motif is localised at the solvent-exposed surface of the tower domain in the a10-loop-a10’ region (Figure 5C). SAXS studies showed that this region interacts with the GyrA CTD [38]. Superimposition of the different breakage-reunion domains shows that the structures display significant differences in this region and can be clustered in three distinct groups according to the conformation of the loop (Figure 5C). First, the eukaryotic topoisomerase II group, represented by the three different structures of the S. cerevisiae topoisomerase II, is characterized by the absence of the helices a10’. The second group, which contains the bacterial type IIA topoisomerases (topoisomerase IV or DNA gyrase) from organisms containing two topoisomerases, possess a short a10’. The interaction between this region and the CTD might therefore be different in these two groups suggesting that this region may be implicated in functional specificity of type II topoisomerases, as the CTD plays a crucial role in DNA interaction. Finally, the two structures of M. tuberculosis constitute the third group. The DEEE motif creates an extension of the a10’ helix modifying the CTD interface and could thus play an important role during the catalytic cycle of the M. tuberculosis DNA gyrase. To confirm the relationships between these two specific structural motifs and the function of M. tuberculosis DNA gyrase, the role of the DPP and the DEEE motifs will be studied through site-directed mutagenesis. Structural modeling of the catalytic reaction core in complex with DNA and quinolone During the catalytic cycle of DNA gyrase, a ternary complex is formed between the Toprim and the breakage-reunion domains and DNA. Quinolones target this complex and inhibit the enzyme through stabilization of the covalent DNA-protein complex formed during catalysis. To explore the mechanistic implications of the M. tuberculosis DNA gyrase and to understand how the N- terminal helix would interfere in the context of the complex structure, we performed structural modeling of the cleavage complex based on the structure of a topoisomerase IV complex [26]. This quaternary complex is composed of the catalytic reaction core consisting of the breakage-reunion domain (GA57BK), the Toprim domain (TopBK), a 34-bp DNA duplex and one of the most promising fourth-generation fluoroquinolone, moxifloxacin. In the structure of the complex, DNA is settled on the DNA gate, is linked covalently to the two catalytic tyrosines 129, and is maintained on each side by the ‘tower’ of the breakage- reunion domain and the Toprim domain (Figure 6). The two Figure 5. Comparison of the M. tuberculosis breakage-reunion domain to other type II topoisomerase structures. A. Global view of the breakage-reunion domain. The boxes indicate the three close-up views shown in B, C and D. B. The DPP loop of GA57BK represented in light green is near the DNA phosphate backbone, in orange (see text for details of the model). C. Close view of the a10–a10’ loop. Both M. tuberculosis structures, GA57BK (represented in light green) and MtGyrA59 (in yellow) possess a DEEX sequence insertion in this loop. The conformation of this loop is different in other bacterial type II topoisomerases, namely the three topoisomerase IV structures represented in red and E. coli GyrA59 in green, and in the three yeast topoisomerase II structures in blue. D. Close-up view of the a3–a4 loop. The conformations of GA57BK chain B (light green), and MtGyrA59 (yellow) are different from the conformation of GA57BK chain A (light green) and E. coli GyrA59 (dark green). doi:10.1371/journal.pone.0012245.g005 M. tuberculosis DNA Gyrase PLoS ONE | www.plosone.org 7 August 2010 | Volume 5 | Issue 8 | e12245 catalytic sites related by the heterotetramer two-fold axis are separated by four base-pairs and each catalytic site contains one quinolone molecule (Figure 6A and B). The quinolone carboxylate group points towards the major groove, and the R7 group is localised in the minor groove. The interaction energy between each quinolone molecule and its devoted binding pocket is 2105 and 2112 kcal/mol, respectively. The slight discrepancy could reflect some sequential binding. However, those values evidence a very good binding affinity that is illustrated in Figures 6B and C. Discussion In the present work, we have structurally characterized the two components of the catalytic reaction core. The structure of the breakage-reunion domain (known as GyrA59 in E. coli) reveals a new interaction promising for drug design, whilst the high resolution structures of the Toprim domain highlights two disordered regions that play a crucial role during the catalytic reaction of DNA gyrase. The strong point of this study is that we could identify original mechanistic properties of quinolone binding that clarify relationships between amino acid mutations and resistance phenotype. These structure-mechanism relationships have been established from the modeling of the catalytic reaction core based on the two crystal structures, DNA and quinolone, using the crystal structure of the cleavage complex formed by the S. pneumoniae breakage-reunion and Toprim domains of topoisom- erase IV stabilized by a fluoroquinolone [26]. The Quinolone-Binding Pocket (QBP), a drug-binding pocket composed of protein and DNA residues Whereas the structures of the S. pneumoniae topoisomerase IV and the M. tuberculosis DNA gyrase reaction core are very similar, our model allowed us to establish clear relationships between Figure 6. Model of the catalytic reaction core in complex with DNA and moxifloxacin. A. Overall structure of the complex. GA57BK is represented in blue, TopBK in red, the DNA in orange and the moxifloxacin in green. B. Close-up view of the two quinolone-binding pockets (QBP). The purple arrow highlights the rise of the intercalated base step that constitutes the DNA walls of the QBP. Protein residues that constitute the QBP protein walls are indicated in red for TopBk and blue for GA57BK. The residues shown in sticks belong to the QRDR and are implicated in quinolone resistance. C. Close-up view along the DNA axis of one of the two QBP. The same residues as in B are represented in sticks. D. Schematic representation of the interactions between QBP residues and chemical groups of the quinolone. doi:10.1371/journal.pone.0012245.g006 M. tuberculosis DNA Gyrase PLoS ONE | www.plosone.org 8 August 2010 | Volume 5 | Issue 8 | e12245 amino acid mutation and resistance phenotype in M. tuberculosis DNA gyrase. We propose that the atypical quinolone-binding mode in the Quinolone-Binding Pocket (QBP), whose walls are constituted not only by regions of the Toprim and the breakage- reunion domains but also by DNA (Figure 6B), explains the effect of the amino acid nature at a given position on the observed resistance. The drug is intercalated between the dinucleotide step for which the DNA backbone of one strand is broken (the phosphorus atom is covalently linked to oxygen atom of the catalytic tyrosine). The intercalated dinucleotide step is strongly perturbed, with a twist of nearly 10u and a rise of 7.3 A˚ (36u and 3.4 A˚ for a canonical B-helix, 33u and 2.7 A˚ for A-DNA), typical of an intercalation mechanism (as observed, for example, in the structure of a DNA-nogalamycin complex [39]). The two intercalated base pairs form a saddle, where quinolone is stabilised through p-p interactions (Figure 6B and 7). The quinolone molecule is blocked in this DNA saddle mainly by Van der Waals contacts with residues of both protein domains (Figure 6C, D and 7). On one side, the carboxylate and the R2 groups (R2 is a hydrogen atom in the moxifloxacin) of the drug are maintained by the a3–a4 loop and the beginning of the a4-helix of the breakage- reunion domain (residues 86–91). On the other side, quinolone is immobilized by three regions of the Toprim domain. The b1-a1 loop (residues 459–462) interacts with the R1 group, the b2-DBL loop (residues 480–486) with the R7–R8 group and the beginning of a2 (residues 498–502) with the R7 group (Figure 6C and D). Consequently, both deformation (rise) of the intercalated dinucle- otide step forming the DNA saddle, and the specific sequence of the QRDR-A and B, are required to build up the QBP and determine the geometrical characteristics of the binding pocket (volume and shape). In addition, the conformation of the loop connecting helices a3 and a4 (residues 84–88) also affects the depth of the QBP (Figure 6C and D). Whereas this loop displays two different conformations in the two monomers in the GA57BK crystal structure (Figure 5D), our model clearly shows that the presence of DNA tends to push this loop towards the conformation observed in the E. coli structure, suggesting that only this conformation is observed when the QBP is formed. Structural insights into the mechanism of ‘‘intrinsic resistance’’ to quinolone The three residues, M81 and A90 in GyrA and R482 in GyrB have been shown to be implicated in ‘‘intrinsic’’ quinolone resistance of M. tuberculosis [13,14]. We have previously demon- strated that A90S and R482K substitutions (S and K are the corresponding residues in E. coli) have direct effect on resistance level [13]. In our model, A90 and R482 are part of the QBP, residing on the a4-helix and in the b2-DBL, respectively. As shown in Figure 6C, A90 side chain is oriented toward the carboxylate group of the quinolone. The substitution of this alanine for serine could increase the stability of the drug through a hydrogen bond between the serine side chain and the hydroxyl group of the quinolone, as observed in the S. pneumoniae complex (3K9F). Furthermore, the R482 side chain is located in the minor groove and forms a gate which blocks the quinolone in the pocket (Figure 7). It has been shown that removing a lysine from the minor groove energetically costs more than removing an arginine [40]. The gate will open more easily when the residue at this position is an arginine, in contrast to lysine, contributing to the destabilisation of the quinolone in the QBP. This open-close mechanism of the gate could play a role in the ‘‘intrinsic resistance’’ mechanism. Finally, in our previous study, we showed that M81I substitution (I in E. coli) alone had not any effect, but could raise the quinolone susceptibility when associated with the A90S mutation [13]. This correlates well with the fact that M81 is not directly located in the QBP. This residue is spatially too far from the quinolone-binding site, but it could affect the QBP by altering the conformation of the a4-helix through direct Figure 7. Two views of the Quinolone-Binding Pocket (QBP). The DNA-protein complex is represented in molecular surface and moxifloxacine in sticks. GA57BK is colored in dark blue, TopBK in firebrick, DNA in orange and moxifloxacin in green. The residues of TopBK belonging to the QBP are colored in yellow for the b1-a1 loop residues, in purple for the b2-DBL residues (including R482), and in pink for the DBL-a2 residues. The residues of GA57BK belonging to the QBP are represented in light green and correspond to the a3–a4 region. doi:10.1371/journal.pone.0012245.g007 M. tuberculosis DNA Gyrase PLoS ONE | www.plosone.org 9 August 2010 | Volume 5 | Issue 8 | e12245 interactions of the residue at position 81 and two residues of the helix, namely D89 and I92. All these observations show that direct interaction has direct effect on the resistance level, and gives a synergetic effect role to the amino acid nature at position 81. The role of amino acids at position 81 and 90 in ‘‘intrinsic resistance’’ will be further investigated through structural studies of the M. tuberculosis DNA gyrase double mutant A90S-M81I in complex with DNA and a quinolone. Structural insights into the mechanism of ‘‘acquired resistance’’ to quinolone A number of mutations that lead to fluoroquinolone ‘‘acquired resistance’’ have been described in the literature [15,16]. They are all localised in the QRDR-A and -B (residues 74–113 of the GA57BK structure and 461–499 of the TopBK structure, respectively). Interestingly, our model shows that all the residues in the QRDR implicated in the ‘‘acquired resistance’’ are localised in the QBP (as defined above), highlighting the relationships between the QRDR of both subunits and the structurally identified QBP, as previously suggested [41]. The model showed that the overall geometry of the QBP, rather than the network of H-bonding, is crucial for the recognition and binding of quinolone in the pocket. Consequently, amino acid changes in the QBP will lead to modification of the pocket geometry, either (i) directly, for residues whose side chains point into the QBP or, most importantly, (ii) indirectly, through modification of the DNA structure, for residues interacting with the DNA moiety of the QBP. Mutations implicated in nearly ninety percent of the resistant strains are located in the QRDR-A at positions 90 and 94. Interestingly, only A90, which also contributes to the intrinsic resistance of M. tuberculosis, interacts through a CH-O bond between its methyl group with the quinolone carboxylate group. Substitution by a valine could generate steric hindrance and this could explain why this mutation is known to increase quinolone resistance [42]. Mutations at other positions on the a4-helix affect the DNA backbone structure by changing the major groove dimensions, as DNA stacks on the a4-helix (Figure 6B and C). Consequently, the size of the saddle formed by the intercalated base pairs will be modified (Figure 7). This size modification could affect the binding and the stability of the drug in the QBP. To illustrate this mechanism, the amino acid at position 94 has a paradoxical effect on the resistance level. Indeed, substitution by either smaller residues like glycine or alanine and bulky residues like tyrosine both increase the resistance level [21]. These residues will either expand or reduce the volume of the pocket, leading to instability of the quinolone in the QBP. Mutations in the QRDR- B, like N499, are much less frequent, but their effects on DNA gyrase activity can also be explained by this shape recognition mechanism. All these observations can be used to improve the efficacy of already existing quinolones. Conclusion Taken together with our previous work concerning the role of specific residues implicated in quinolone resistance [13,23], our structural results concerning the M. tuberculosis breakage-reunion and Toprim domains and the modeled complex of the catalytic reaction core provide key insights into the relationship between the amino acid sequence of the M. tuberculosis DNA gyrase and the resistance mechanism to quinolones, a major class of antibiotics against this pathogen. In addition, these results highlight two directions for future work. First, M. tuberculosis DNA gyrase, the single type II topoisomerase in this organism, possesses two specific structural motifs, the DEEE loop and the DPP loop, which could partially explain its different activity spectrum as compared to topoisomerase IV or DNA gyrase. Hence, this atypical activity spectrum could be explained by the unique nature of the amino acids present in the DNA gate. Second, the N-terminal helix of the GA57BK structure is structurally ordered and stabilised through crystal contacts. Interestingly, this helix blocks the active site of a symmetry-related molecule through interactions with residues of the a3–a4 loop. In the asymmetric unit, the dimeric structure displays two different conformations for this loop. In agreement with what was proposed by Tretter et al. [25], this suggests that this region is conformationally dynamic (Figure 5D). Furthermore, this helix contacts active site residues important for the catalysis of the breakage-ligation reaction. The presence of this N-terminal helix would prevent DNA binding (Figure 4D). These observations will be exploited for the design of a new inhibitor family using peptide- based approaches that target DNA gyrase by competitive inhibition of DNA binding. Thus, they open up new avenues for the development of novel peptide-based DNA gyrase inhibitors, providing valuable new strategies to combat this disease as strains resistant to the current repertoire of drugs are emerging. Materials and Methods Cloning, expression, purification and crystallization of GA57BK The breakage-reunion domain of DNA gyrase subunit A from M. tuberculosis (residues 1–502), hereafter named GA57BK because of its molecular weight of 57 kDa, was cloned, expressed and purified as reported previously [43]. Briefly, the PCR amplified construct was ligated into the pET-29a vector (Novagen) between the NdeI and XhoI sites. The C-terminal His-tagged protein was overproduced after transforming the plasmid into Rosetta 2(DE3) pLysS (novagen), and purified with a Ni-NTA column and a size exclusion chromatography using Superdex-75 10/300 (GE Healthcare). The protein was concentrated to 10–15 mg/ml in 100 mM Tris-HCl pH 8. Ga57BK crystals were prepared using the hanging drop vapor diffusion method, mixing 2 volumes of protein sample against 1 volume of reservoir solution [100 mM Sodium HEPES pH 7.5, 4% PEG 4000, 30% MPD]. Crystals grew after several days at 21uC to a maximum size of 2006200650 mm3. Data collection, structure determination and refinement of GA57BK Crystals were directly flash frozen in liquid nitrogen. Native diffraction data were collected at the SOLEIL PROXIMA-1 beamline to 2.7 A˚ resolution. The XDS package [44] was used for all data integration and scaling. The crystals belong to space group C2 with unit cell dimensions a = 163.9 A˚ , b = 109.6 A˚ , c = 102.0 A˚ , b = 120.4u and contain one biological dimer in the asymmetric unit corresponding to a Matthews coefficient value of 3.4 A˚ 3/Da [45]. Data collection statistics are shown in Table 2. The structure of GA57BK was determined by molecular replacement with AMoRe [46] implemented in CCP4 [47] using the breakage-reunion domain of the DNA gyrase from E. coli [36] (pdb accession code 1AB4) as a search model. Two distinct orientations and positions were found in the asymmetric unit. Structure refinement was carried out with BUSTER-TNT [48] using two-fold non-crystallographic symmetry restraints. Model building was performed manually with the program Coot [49]. Model refinement statistics are summarized in Table 2. The figures were prepared using PyMol [50], available at http://pymol. sourceforge.net/. Interface areas were calculated with the PISA server [51]. M. tuberculosis DNA Gyrase PLoS ONE | www.plosone.org 10 August 2010 | Volume 5 | Issue 8 | e12245 Cloning, expression, purification and crystallization of TopBK crystal I and II The Toprim domain of DNA gyrase subunit B from Mycobacterium tuberculosis (residues 448–675), hereafter named TopBK, was cloned into the expression vector pRSF-2 Ek/LIC (Novagen). The plasmid was transformed into Rosetta 2(DE3) pLysS (Novagen). The transformed cells were grown in LB medium in presence of chloramphenicol and kanamycin. Gene expression was induced by addition of IPTG (Sigma) to a final concentration of 1 mM at 22uC over night. Cells were harvested by centrifugation and stored at 220uC one night. Cells were resuspended in buffer B1 containing 20 mM Tris-HCl pH 8, 500 mM NaCl and 15 mM imidazole. The cells were lysed by sonication. Following the centrifugation, the protein was run over a Ni-NTA column (GE Healthcare) equilibrated with buffer B1 at 4uC. The TopBK protein was eluted using a linear gradient from 15 to 500 mM imidazole. Finally, the protein was loaded on a Superdex-75 10/300 (GE Healthcare) equilibrated with a buffer containing 20 mM Tris-HCl pH 8. The protein was then concentrated to 5 mg/ml in the same buffer. TopBK crystal I was obtained in 10% PEG 4K, 200 mM ammonium sulfate, 15 mM magnesium chloride, 100 mM Tris- HCl pH 8 by vapour diffusion with the hanging drop vapour diffusion method mixing 2 volumes of protein sample with 1 volume of reservoir solution [10% PEG 4K, 200 mM ammonium sulfate, 15 mM magnesium chloride, 100 mM Tris-HCl pH 8]. TopBK crystal II was obtained in similar conditions, except that magnesium chloride was substituted by calcium chloride. Data collection, structure determination and refinement of TopBK For TopBK crystal I, diffraction data were collected at ESRF on beamline id23eh1 to 2.1 A˚ resolution. The XDS package was used for data processing and scaling (Table 2). The crystals belong to Table 2. Data collection and refinement statistics. TopBK crystal I TopBK crystal II GA57BK Data Collection Beamline ESRF ID23eh1 SOLEIL PROXIMA 1 SOLEIL PROXIMA 1 Space group P43212 P43212 C2 Unit cell dimensions a, b, c (A˚) 52.9, 52.9, 190.2 52.8, 52.8, 190.5 163.9, 109.6, 102.0 a, b, c (u) 90, 90, 90 90, 90, 90 90, 120.4, 90 Wavelength (A˚) 0.9762 0.9800 0.9800 Resolution (A˚) 14–2.1 (2.3–2.1)a 29–1.95 (2.06–1.95) 35–2.7 (2.8–2.7) Rsym (%)b 13.0 (55.0) 7.5 (58.7) 8.6 (72.6) Redundancya 8.6 (4.0) 7.6 (7.8) 3.5 (3.5) Completeness (%)a 99.1 (86.5) 99.7 (99.2) 98.9 (99.0) I/sig(I)a 13.1 (3.3) 16.9 (3.4) 12.44 (2.24) Refinement Resolution (A˚) 14.0–2.1 17.0–1.95 19.9–2.7 No. Reflections 16487 20626 42396 No. Atoms Protein 1474 1482 7534 Water 107 147 238 Rwork/Rfree c 0.214, 0.249 0.210, 0.230 0.192, 0.233 B-factors Protein 38.3 37.8 52.5 Water 52.6 52.2 57.6 RMSD Bond length (A˚) 0.004 0.007 0.004 Bond angles (u) 0.835 0.97 0.719 Ramachandran analysis Most favored (%) 93.8 92.7 90.2 Additional allowed (%) 5.6 6.7 9.3 Generously allowed (%) 0.0 0.0 0.4 Disallowed (%) 0.6 0.6 0.1 aThe values in parentheses are statistics from the highest resolution shell. bRsym~P P DIhkl{Ihkl jð ÞD=P Ihkl where Ihkl(j) is the jth observed intensity of Ihkl and Ihkl is the final average value of intensity. cRwork~P DDFobsD{DFcalcDD=P DFobsD and Rfree~P DDFobsD{DFcalcDD=P DFobsD where the sum is restricted to reflections that belong to a test set of 5% randomly selected data. doi:10.1371/journal.pone.0012245.t002 M. tuberculosis DNA Gyrase PLoS ONE | www.plosone.org 11 August 2010 | Volume 5 | Issue 8 | e12245 space group P43212 with unit cell dimensions a = b = 52.87 A˚ , c = 190.22 A˚ . The structure of TopBK was determined by molecular replacement with Molrep [52] implemented in ccp4 using one monomer of the previously published structure (PDB accession code 2ZJT, [24]) as the starting model. The asymmetric unit contains one monomer corresponding to a Matthews coefficient value of 2.4 A˚ 3/Da. Structure refinement was carried out with BUSTER-TNT to 2.1 A˚ resolution. Model building was performed manually with the program coot. Model refinement statistics are summarized in Table 2. For TopBK crystal II, diffraction data were collected at SOLEIL PROXIMA 1 to 1.95 A˚ resolution. TopBK crystal II is isomorphous to crystal I and structure determination protocol was the same as for crystal I (Table 2). Analytical ultracentrifugation Sedimentation velocity experiments were performed in a Beckman XL-I analytical ultracentrifuge using a double sector charcoal-Epon cell at 20uC and 42000 rpm. Absorbance scans were taken at 276 nm every 6 min. The protein concentration was 1 mg/ml for GA57BK corresponding to 17.5 mM in 20 mM Tris pH 8. For TopBK, experiments were performed at three protein concentrations, 0.5, 1 and 4 mg/ml (corresponding to 18, 37 and 148 mM, respectively) in the same buffer. The program Sednterp 1.09 (available at http://www.rasmb.bbri.org) was used to calculate solvent density (0.9988 g/cm3), solvent viscosity (0.010069 Poise) and partial specific volume (0.7340 ml/g for GA57BK and 0.7390 for TopBK) using the amino-acid compo- sition. The sedimentation data were analyzed with the program Sedfit [53] using the continuous c(s) and c(M) distributions. Theoretical sedimentation coefficients were calculated from the crystal structure PDB file using Hydropro 7c [54] with a hydrated radius of 3.4 A˚ for the atomic elements. The same experiments were performed for GA57BK and TopBK in 20 mM Tris pH 8 and 100 mM NaCl. Sedimentation data were analyzed with appropriate values of solvent density and viscosity. Activity assays DNA supercoiling and cleavage assays were carried out as previously described [7,13,23,55]. Briefly, DNA cleavage assays were performed with various ratios of purified M. tuberculosis GyrA and GyrB subunits or GA57BK and TopBK domains. The reaction mixture (total volume 20 ml) contained DNA gyrase assay buffer (40 mM Tris-HCl pH 7.5, 25 mM KCl, 6 mM magnesium acetate, 2 mM spermidine, 4 mM DTT, 0.1 mg/ml E. coli tRNA, BSA (0.36 mg/ml), 100 mM potassium glutamate), supercoiled pBR322 DNA (0.4 mg) as the substrate and moxifloxacin (50 mg/ ml). Proteins were added and reaction mixtures were incubated at 25uC for 1 h. Three ml of 2% SDS and 3 ml of a 1 mg/ml solution of proteinase K were added, and incubation was continued for 30 min at 37uC. Reactions were terminated by the addition of 50% glycerol containing 0.25% bromophenol blue, and the total reaction mixture was subjected to electrophoresis in 1% agarose gel in TBE 0.56 buffer (Tris-Borate-EDTA, pH 8.3). After running for 3.5 hrs at 50 V, the gel was stained with ethidium bromide (0.7 mg/ml), photographed and quantified with an Alpha Innotech digital camera and associated software. All enzyme assays were done at least twice, with reproducible results. Molecular modeling The catalytic core model (GA57BK2+TopBK2+DNA) was generated by superposition onto the crystal structure of the Streptococcus pneumoniae topoisomerase IV catalytic core [26] (pdb accession code 3FOF). Chains A and B from 3IFZ (GA57BK) were superposed to the corresponding chains from 3FOF, respectively, using SSM implemented in coot. The two disordered regions of the TopBK structure were modeled using the Toprim domain of 3K9F as a template. The amino acid torsion angles in these regions were validated using the Ramachandran plot. The two monomers of TOPBK were superposed using the same method to the chains C and D of the S. pneumoniae topoisomerase IV catalytic core structure. The DNA coordinates (chain E, F, G, H) without moxifloxacin were inserted in the complex and defined as fixed atoms. The complex was then energy minimized. Energy minimization was performed with the NAMD2 program [56] using CHARMM27 force field. The system was minimized by 300 000 steps of conjugate gradient minimization. Non bonded interaction parameters were set such that electrostatic interaction is shifted to zero at 12 A˚ and the van der Waals interaction is switched off from 10 A˚ to 12 A˚ . For the docking, the two fluoroquinolone moieties were extracted from 3FOF coordinates. They were positioned in the minimized catalytic core with respect to their respective positions in the 3FOF structure. The system was further minimized using the Minimization module of Discovery Studio (Accelrys), the CHARMM forcefield and a cascade of Steepest Descent, Gradient Conjugate and Adopted Basis Newton Raphson minimizations, during which the backbone of the protein complex plus the DNA atoms were constrained while the side chains and ligand moieties were allowed to relax (6,000 iterations with final RMS gradient 0.01). We computed energetic criteria as the potential energy of the complex. The minimised model deviates from the crystal structure of the Streptococcus pneumoniae topoisomerase IV catalytic core with an rmsd of 2.4 A˚ over 1033 Ca atoms. Finally, we computed the interaction energy (which corresponds to the sum of VDW and electrostatics non-bonded interactions) between each moxifloxacin and its devoted quinolone-binding pocket with the Calculate Interaction Energy module of Discovery Studio (Accelrys). Accession numbers Co-ordinates and structure factors of TopBK crystal I have been deposited in the protein data bank with the code 3IG0, TopBK crystal II with the code 3M4I and GA57BK with the code 3IFZ. Supporting Information Table S1 Values of the interfaces calculated by PISA for the five structures of the breakage-reunion domain dimer in closed conformation. The PDB codes for the five structures are given: 3IFZ (this work) and 3ILW (25) correspond to M. tuberculosis DNA gyrase, 1AB4 (36) to E. coli DNA gyrase, 2INR (34) to S. aureus topoisomerase IV, 2NOV (33) to S. pneumoniae topoisomerase IV. Nat, Nres correspond to the number of atoms and residues, respectively, in interaction between the two monomers. Found at: doi:10.1371/journal.pone.0012245.s001 (1.36 MB DOC) Figure S1 Structure-based sequence alignment of the Toprim domain from type II topoisomerases. The sequence names are as follows: MtGyr (PDB code 3IFZ) (this work), M. tuberculosis DNA gyrase; SpTopIV (PDB code 3FOF) (26), S. pneumoniae topoisom- erase IV and ScTopII (PDB code 2RGR) (29), S. cerevisiae topoisomerase II. alpha-helices (cylinders) and beta-strands (arrows) of M. tuberculosis GA57BK are shown with the sequences and color-coded according to Figure 1 (Toprim region in yellow, the hinge in blue and the Tail region in purple). Residues emphasized by black shading are 100% conserved. The magne- sium binding site residues are underlined by red stars (E and DxD). M. tuberculosis DNA Gyrase PLoS ONE | www.plosone.org 12 August 2010 | Volume 5 | Issue 8 | e12245 The disordered regions are emphasized in pale grey and indicated as alpha1 and DBL for DNA Binding Loop. The QRDR-B is delimited by a blue frame. Found at: doi:10.1371/journal.pone.0012245.s002 (0.04 MB DOC) Figure S2 The three different conformations of the breakage- reunion domain. A. The breakage-reunion domain of M. tuberculosis (PDB id 3IFZ) (this work), representing the closed conformation with the DNA-gate and the C-gate closed. This closed conformation is also observed in the E. coli DNA gyrase (36), S. pneumoniae and S. aureus topoisomerase IV breakage-reunion domain structures (33,34). B. The breakage-reunion domain of S. cerevisiae in complex with DNA (PDB id 2RGR) (29), representing an open conformation with the DNA-gate open and the C-gate closed. C. The breakage-reunion domain of S. cerevisiae (PDB id 1BGW) (31), representing an open conformation with the DNA- gate closed and the C-gate open. Found at: doi:10.1371/journal.pone.0012245.s003 (0.90 MB DOC) Figure S3 Structure-based sequence alignment of the breakage- reunion domain from type II topoisomerases. The sequence names are as follows: MtGyr (PDB code 3IFZ) (this work), M. tuberculosis DNA gyrase; EcGyr (PDB code 1AB4) (36), E. coli DNA gyrase; SaTopIV (PDB code 2INR) (34), S. aureus topoisomerase IV; SpTopIV (PDB code 2NOV) (33), S. pneumoniae topoisomerase IV; EcTopIV (PDB code 1ZVU), E. coli topoisomerase IV and ScTopII (PDB code 2RGR) (29), S. cerevisiae topoisomerase II. alpha-helices (cylinders) and beta-strands (arrows) of M. tuberculosis GA57BK are shown with the sequences and color-coded according to Figure 1 (N-terminal helix in red, DNA-gate in blue, Tower in green, helix bundle in orange and C-gate in purple). Residues emphasized by black shading are 100% conserved. The catalytic residues are underlined by red stars (R128 and Y129) and GA57BK specific motifs by black stars (the DPP and DEEX motifs). The QRDR-A is delimited by a blue frame. Found at: doi:10.1371/journal.pone.0012245.s004 (0.06 MB DOC) Figure S4 Superimposition of the different monomer structures of the breakage-reunion domain. M. tuberculosis DNA gyrase GA57BK (3IFZ) (this work) in light green, M. tuberculosis DNA gyrase MtGyrA59 (3ILW, 25) in pale green, E. coli DNA gyrase (1AB4) (36) in dark green, S. pneumoniae topoisomerase IV (2NOV) (33) in red, S. aureus topoisomerase IV (2INR) (34) in pale red, S. pneumoniae complexed with DNA (3FOF) (26) in dark red and E. coli topoisomerase IV (1ZVU) in firebrick. The rmsd (in Ang.) after superimposition and the number of common Ca (in parenthesis) are indicated in the table. The color code is conserved. Found at: doi:10.1371/journal.pone.0012245.s005 (0.43 MB DOC) Acknowledgments We greatly acknowledge the help of Ahmed Haouz and the PF6 facility (Plate-Forme de cristalloge´ne`se et diffraction des Rayons X), Bertrand Raynal and the PFBMI facility (Plate-Forme de Biophysique des Macromole´cules et de leurs Interactions) from the Pasteur Institute. We thank Fre´de´ric Poitevin for help with modeling and Nathalie Barilone for generous donation of M. tuberculosis H37Rv genomic DNA. We are especially grateful to Olivier Poch for helpful discussions. We thank Joseph Cockburn for careful reading of the manuscript. We want to dedicate this manuscript to the memory of Warren DeLano, the developer of PyMol, who passed away in November 2009. Author Contributions Conceived and designed the experiments: JP AA CM. Performed the experiments: JP GAL AA. Analyzed the data: JP SP MD CM. Contributed reagents/materials/analysis tools: VJ. Wrote the paper: JP AA CM. Contributed to the writing of the paper: SP MD GAL. References 1. Champoux JJ (2001) DNA topoisomerases: structure, function, and mechanism. Annu Rev Biochem 70: 369–413. 2. Buhler C, Gadelle D, Forterre P, Wang JC, Bergerat A (1998) Reconstitution of DNA topoisomerase VI of the thermophilic archaeon Sulfolobus shibatae from subunits separately overexpressed in Escherichia coli. Nucleic Acids Res 26: 5157–5162. 3. Gadelle D, Filee J, Buhler C, Forterre P (2003) Phylogenomics of type II DNA topoisomerases. Bioessays 25: 232–242. 4. Schoeffler AJ, Berger JM (2005) Recent advances in understanding structure- function relationships in the type II topoisomerase mechanism. Biochem Soc Trans 33: 1465–1470. 5. Levine C, Hiasa H, Marians KJ (1998) DNA gyrase and topoisomerase IV: biochemical activities, physiological roles during chromosome replication, and drug sensitivities. Biochim Biophys Acta 1400: 29–43. 6. Cole ST, Brosch R, Parkhill J, Garnier T, Churcher C, et al. (1998) Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393: 537–544. 7. Aubry A, Fisher LM, Jarlier V, Cambau E (2006) First functional character- ization of a singly expressed bacterial type II topoisomerase: the enzyme from Mycobacterium tuberculosis. Biochem Biophys Res Commun 348: 158–165. 8. Schoeffler AJ, Berger JM (2008) DNA topoisomerases: harnessing and constraining energy to govern chromosome topology. Q Rev Biophys 41: 41–101. 9. Hooper CD, Rubinstein E (2003) Quinolone Antimicrobial Agents: ASM Press. 10. Blumberg HM, Burman WJ, Chaisson RE, Daley CL, Etkind SC, et al. (2003) American Thoracic Society/Centers for Disease Control and Prevention/ Infectious Diseases Society of America: treatment of tuberculosis. Am J Respir Crit Care Med 167: 603–662. 11. Conde MB, Efron A, Loredo C, De Souza GR, Graca NP, et al. (2009) Moxifloxacin versus ethambutol in the initial treatment of tuberculosis: a double- blind, randomised, controlled phase II trial. Lancet 373: 1183–1189. 12. World Health Organization (2008) Global tuberculosis control-epidemiology, strategy, financing. WHO report 2009 WHO/HTM/TB/2009. 411 p. 13. Matrat S, Aubry A, Mayer C, Jarlier V, Cambau E (2008) Mutagenesis in the alpha3alpha4 GyrA helix and in the Toprim domain of GyrB refines the contribution of Mycobacterium tuberculosis DNA gyrase to intrinsic resistance to quinolones. Antimicrob Agents Chemother 52: 2909–2914. 14. Guillemin I, Jarlier V, Cambau E (1998) Correlation between quinolone susceptibility patterns and sequences in the A and B subunits of DNA gyrase in mycobacteria. Antimicrob Agents Chemother 42: 2084–2088. 15. Duong DA, Nguyen TH, Nguyen TN, Dai VH, Dang TM, et al. (2009) Beijing genotype of Mycobacterium tuberculosis is significantly associated with high-level fluoroquinolone resistance in Vietnam. Antimicrob Agents Chemother 53: 4835–4839. 16. Sun Z, Zhang J, Zhang X, Wang S, Zhang Y, et al. (2008) Comparison of gyrA gene mutations between laboratory-selected ofloxacin-resistant Mycobacterium tuberculosis strains and clinical isolates. Int J Antimicrob Agents 31: 115–121. 17. van Doorn HR, An DD, de Jong MD, Lan NT, Hoa DV, et al. (2008) Fluoroquinolone resistance detection in Mycobacterium tuberculosis with locked nucleic acid probe real-time PCR. Int J Tuberc Lung Dis 12: 736–742. 18. Hooper DC (1999) Mechanisms of fluoroquinolone resistance. Drug Resist Updat 2: 38–55. 19. Mdluli K, Ma Z (2007) Mycobacterium tuberculosis DNA gyrase as a target for drug discovery. Infect Disord Drug Targets 7: 159–168. 20. Takiff HE, Salazar L, Guerrero C, Philipp W, Huang WM, et al. (1994) Cloning and nucleotide sequence of Mycobacterium tuberculosis gyrA and gyrB genes and detection of quinolone resistance mutations. Antimicrob Agents Chemother 38: 773–780. 21. Aubry A, Veziris N, Cambau E, Truffot-Pernot C, Jarlier V, et al. (2006) Novel gyrase mutations in quinolone-resistant and -hypersusceptible clinical isolates of Mycobacterium tuberculosis: functional analysis of mutant enzymes. Antimicrob Agents Chemother 50: 104–112. 22. Veziris N, Martin C, Brossier F, Bonnaud F, Denis F, et al. (2007) Treatment failure in a case of extensively drug-resistant tuberculosis associated with selection of a GyrB mutant causing fluoroquinolone resistance. Eur J Clin Microbiol Infect Dis 26: 423–425. 23. Matrat S, Veziris N, Mayer C, Jarlier V, Truffot-Pernot C, et al. (2006) Functional analysis of DNA gyrase mutant enzymes carrying mutations at position 88 in the A subunit found in clinical strains of Mycobacterium tuberculosis resistant to fluoroquinolones. Antimicrob Agents Chemother 50: 4170–4173. M. tuberculosis DNA Gyrase PLoS ONE | www.plosone.org 13 August 2010 | Volume 5 | Issue 8 | e12245 24. Fu G, Wu J, Liu W, Zhu D, Hu Y, et al. (2009) Crystal structure of DNA gyrase B’ domain sheds lights on the mechanism for T-segment navigation. Nucleic Acids Res 37: 5908–5916. 25. Tretter EM, Schoeffler AJ, Weisfield SR, Berger JM (2010) Crystal structure of the DNA gyrase GyrA N-terminal domain from Mycobacterium tuberculosis. Proteins 8(2): 492–495. 26. Laponogov I, Sohi MK, Veselkov DA, Pan XS, Sawhney R, et al. (2009) Structural insight into the quinolone-DNA cleavage complex of type IIA topoisomerases. Nat Struct Mol Biol 16: 667–669. 27. Edwards MJ, Flatman RH, Mitchenall LA, Stevenson CE, Le TB, et al. (2009) A crystal structure of the bifunctional antibiotic simocyclinone D8, bound to DNA gyrase. Science 326: 1415–1418. 28. Fu G, Wu J, Zhu D, Hu Y, Bi L, et al. (2009) Crystallization and preliminary crystallographic studies of Mycobacterium tuberculosis DNA gyrase B C-terminal domain, part of the enzyme reaction core. Acta Crystallogr Sect F Struct Biol Cryst Commun 65: 350–352. 29. Dong KC, Berger JM (2007) Structural basis for gate-DNA recognition and bending by type IIA topoisomerases. Nature 450: 1201–1205. 30. Aravind L, Leipe DD, Koonin EV (1998) Toprim–a conserved catalytic domain in type IA and II topoisomerases, DnaG-type primases, OLD family nucleases and RecR proteins. Nucleic Acids Res 26: 4205–4213. 31. Berger JM, Gamblin SJ, Harrison SC, Wang JC (1996) Structure and mechanism of DNA topoisomerase II. Nature 379: 225–232. 32. Fass D, Bogden CE, Berger JM (1999) Quaternary changes in topoisomerase II may direct orthogonal movement of two DNA strands. Nat Struct Biol 6: 322–326. 33. Laponogov I, Veselkov DA, Sohi MK, Pan XS, Achari A, et al. (2007) Breakage- reunion domain of Streptococcus pneumoniae topoisomerase IV: crystal structure of a gram-positive quinolone target. PLoS One 2: e301. 34. Carr SB, Makris G, Phillips SE, Thomas CD (2006) Crystallization and preliminary X-ray diffraction analysis of two N-terminal fragments of the DNA- cleavage domain of topoisomerase IV from Staphylococcus aureus. Acta Crystallogr Sect F Struct Biol Cryst Commun 62: 1164–1167. 35. Gubaev A, Hilbert M, Klostermeier D (2009) The DNA-gate of Bacillus subtilis gyrase is predominantly in the closed conformation during the DNA supercoiling reaction. Proc Natl Acad Sci U S A 106: 13278–13283. 36. Morais Cabral JH, Jackson AP, Smith CV, Shikotra N, Maxwell A, et al. (1997) Crystal structure of the breakage-reunion domain of DNA gyrase. Nature 388: 903–906. 37. Vagner J, Qu H, Hruby VJ (2008) Peptidomimetics, a synthetic tool of drug discovery. Curr Opin Chem Biol 12: 292–296. 38. Costenaro L, Grossmann JG, Ebel C, Maxwell A (2005) Small-angle X-ray scattering reveals the solution structure of the full-length DNA gyrase a subunit. Structure 13: 287–296. 39. Smith CK, Brannigan JA, Moore MH (1996) Factors affecting DNA sequence selectivity of nogalamycin intercalation: the crystal structure of d(TGTACA)2- nogalamycin2. J Mol Biol 263: 237–258. 40. Rohs R, West SM, Sosinsky A, Liu P, Mann RS, et al. (2009) The role of DNA shape in protein-DNA recognition. Nature 461: 1248–1253. 41. Heddle J, Maxwell A (2002) Quinolone-binding pocket of DNA gyrase: role of GyrB. Antimicrob Agents Chemother 46: 1805–1815. 42. Von Groll A, Martin A, Jureen P, Hoffner S, Vandamme P, et al. (2009) Fluoroquinolone resistance in Mycobacterium tuberculosis and mutations in gyrA and gyrB. Antimicrob Agents Chemother 53: 4498–4500. 43. Piton J, Matrat M, Petrella S, Jarlier V, Aubry A, Mayer C (2009) Purification, crystallization and preliminary X-ray diffraction experiments on the breakage- reunion domain of the DNA gyrase from Mycobacterium tuberculosis. Acta Crystallogr Sect F Struct Biol Cryst Commun 65: 1182–1186. 44. Kabsch W (1988) Automatic indexing of rotation diffraction patterns. Journal of Applied Crystallography 21: 67–72. 45. Matthews BW (1968) Solvent content of protein crystals. J Mol Biol 33: 491–497. 46. Navaza J (1994) AMoRe: an automated package for molecular replacement. Acta Crystallographica Section A 50: 157–163. 47. Collaborative Computational Project N (1994) The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 50: 760–763. 48. Blanc E, Roversi P, Vonrhein C, Flensburg C, Lea SM, et al. (2004) Refinement of severely incomplete structures with maximum likelihood in BUSTER-TNT. Acta Crystallogr D Biol Crystallogr 60: 2210–2221. 49. Emsley P, Cowtan K (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132. 50. DeLano WL (2002) The PyMOL Molecular Graphics System. DeLano Scientific, San Carlos, CA, USA http://www.pymol.org. 51. Krissinel E, Henrick K (2007) Inference of macromolecular assemblies from crystalline state. J Mol Biol 372: 774–797. 52. Vagin A, Teplyakov A (1997) MOLREP: an Automated Program for Molecular Replacement. Journal of Applied Crystallography 30: 1022–1025. 53. Brown PH, Schuck P (2006) Macromolecular size-and-shape distributions by sedimentation velocity analytical ultracentrifugation. Biophysical Journal 90: 4651–4661. 54. Garcı´a De La Torre J, Huertas ML, Carrasco B (2000) Calculation of hydrodynamic properties of globular proteins from their atomic-level structure. Biophysical Journal 78: 719–730. 55. Aubry A, Pan XS, Fisher LM, Jarlier V, Cambau E (2004) Mycobacterium tuberculosis DNA gyrase: interaction with quinolones and correlation with antimycobacterial drug activity. Antimicrob Agents Chemother 48: 1281–1288. 56. Phillips JC, Braun R, Wang W, Gumbart J, Tajkhorshid E, et al. (2005) Scalable molecular dynamics with NAMD. J Comput Chem 26: 1781–1802. M. tuberculosis DNA Gyrase PLoS ONE | www.plosone.org 14 August 2010 | Volume 5 | Issue 8 | e12245
3M4J
Crystal structure of N-acetyl-L-ornithine transcarbamylase complexed with PALAO
Reversible Post-Translational Carboxylation Modulates The Enzymatic Activity Of N-Acetyl-L-Ornithine Transcarbamylase† Yongdong Li1,2, Xiaolin Yu1, Jeremy Ho1, David Fushman3, Norma M. Allewell3, Mendel Tuchman1, and Dashuang Shi1,‡ 1Research Center for Genetic Medicine and Department of Integrative Systems Biology, Children’s National Medical Center, The George Washington University, Washington, DC 20010, USA. 2Key Laboratory of Organo-Pharmaceutical Chemistry, Jiangxi Province, Gannan Normal University, Ganzhou 341000, China. 3Department of Chemistry and Biochemistry, College of Chemical and Life Sciences, University of Maryland, College Park, MD 20742, USA. Abstract N-acetyl-L-ornithine transcarbamylase (AOTCase), rather than ornithine transcarbamylase (OTCase), is the essential carbamylase enzyme in the arginine biosynthesis of several plant and human pathogens. The specificity of this unique enzyme provides a potential target for controlling the spread of these pathogens. Recently, several crystal structures of AOTCase from Xanthomonas campestris (xc) have been determined. In these structures, an unexplained electron density at the tip of Lys302 side-chain was observed. Using 13C NMR spectroscopy, we show herein that Lys302 is post-translationally carboxylated. The structure of wild-type AOTCase complexed with the bisubstrate analogue, Nδ-(phosphonoacetyl)-Nα-acetyl-L-ornithine (PALAO), indicates that the carboxyl group on Lys302 forms a strong hydrogen bonding network with surrounding active site residues, Lys252, Ser253, His293, and Glu92 from the adjacent subunit either directly or via a water molecule. Furthermore, the carboxyl group is involved in binding N-acetyl-L-ornithine via a water molecule. Activity assays with the wild-type enzyme and several mutants demonstrate that the post translational modification of lysine 302 has an important role in catalysis. Post-translational modification of the ε-amino group of lysine residues in proteins is a common mechanism used by organisms to regulate protein functions including DNA-protein interactions, subcellular localization, transcriptional activity, and protein stability and activity (1). Lysine residues can be modified by the addition of functional groups to become acetylated, methylated, carbamylated or carboxylated. The role of histone lysine acetylation and methylation in affecting chromatin structure and gene expression has been well established for more than a decade (2). However, the biological roles for lysine carbamylation and carboxylation have rarely been investigated. †This work was supported by Public Health Service grants DK-47870 (MT) and DK-067935 (DS) from the National Institute of Diabetes, Digestive and Kidney Diseases. JH was supported by a Scholarship from the Doug and Lynn Parsons Family Foundation. The Cornell High Energy Synchrotron Source (CHESS) is supported by the National Science Foundation under award DMR 0225180 and the Macromolecular Diffraction Facility at CHESS (MacCHESS) is supported by award RR-01646 from the National Institutes of Health, through its National Center for Research Resources. ‡Corresponding author. dshi@cnmcresearch.org. Phone: 202-476-5817. Fax: 202-476-6014. SUPPORTING INFORMATION AVAILABLE Figure S1. Structure and hydrogen bonding network around residue 302 for previously determined AOTCase structures. This material is available free of charge via the Internet at http://pubs.acs.org. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 August 17. Published in final edited form as: Biochemistry. 2010 August 17; 49(32): 6887–6895. doi:10.1021/bi1007386. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript In vivo, lysine acetylation and methylation are usually carried out by acetyltransferase and methyltransferase enzymes, respectively (3). In addition, some proteins such as hemoglobin and human serum albumin can be acetylated non-enzymatically by chemicals such as aspirin, methyl acetyl phosphate, and other acetylating agents such as acetyl-CoA (4–8). Lysine can also be methylated by small chemicals in vitro, and this has routinely been used as a rescue method for protein crystallization (9). Lysine carbamylation and lysine carboxylation have only been achieved by using chemicals and no enzyme has yet been found to catalyze these modifications. Lysine carbamylation was one of the earliest post- translational modification of proteins to be elucidated when it was identified as a product of reversible denaturation-renaturation studies of proteins with urea (10,11). This carbamylation, which produces homocitrulline, has also been detected in uremic patients (12) and in patients with elevated plasma and/or urinary lysine levels (13). In contrast, lysine carboxylation is not as commonly reported, but has been identified in a number of proteins via crystal structure determinations. In most of these proteins, the carboxyl groups of modified lysines are involved in bridging two metal ions that play a structural role in the active site. In several other proteins, however, a direct role for a carboxylated lysine in the catalytic mechanism has been reported (14–16). N-acetyl-L-ornithine transcarbamylase (AOTCase, EC 2.1.3.9) was recently discovered to be part of a novel arginine biosynthesis pathway in plant pathogens of the Xanthomonadaceae family such as Xylella and Xanthomonas (17–19). These pathogens attack a variety of economically important crops including citrus fruits, cotton, tomatoes, and rice (20,21). Genome sequence analyses showed that an AOTCase-like gene is also present in some human pathogens such as Stenotrophomonas maltophilia and members of the genus Bacteroides (22). In the case of Bacteroides fragilis, this gene was later confirmed to encode another novel transcarbamylase, N-succinyl-L-ornithine transcarbamylase (SOTCase, EC 2.1.3.11) (23). Crystal structures of both AOTCase and SOTCase bound with substrate or substrate analogues have recently been determined (17,18,23). An extended density at the side-chain tip of Lys302 in AOTCase was observed suggesting a post- translational modification. Since Lys302 is located within the active site of AOTCase and is not present in SOTCase, it was proposed as one of three key signature residues to distinguish the two carbamylases (22). Here, we demonstrate that Lys302 is post- translationally modified by carboxylation and that this change affects the catalytic function of the enzyme. MATERIALS AND METHODS Materials All chemicals were purchased from Sigma Chemical Company unless otherwise specified. ANOR was purchased from Indofine Chemical Co., Inc. N-acetyl-L-citrulline was custom synthesized and purified by Chiral Quest Company. PALAO (>95% purity) was synthesized by IMI TAMI Institute of Research and Development Ltd. (19). xcAOTCase was prepared and purified as previously described (18). Mutants K302A (primer: 5’- CTGCGTCGCAACGTCGCGGCTACTGATGCGGTG-3’), K302E (primer:5’- CTGCGTCGCAACGTCGAGGCTACTGATGCGGTG-3’) and K302R (primer: 5’- CTGCGTCGCAACGTCAGGGCTACTGATGCGGTG-3’) were generated by site-directed mutagenesis using the “Quik Change” mutagenesis kit (Stratagene) according to the manufacturer’s protocol. The correct mutants were confirmed by DNA sequencing. Recombinant mutant proteins were expressed and purified in the same manner as the wild- type enzyme. Li et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Activity assay The modified colorimetric assay method, which detects the formation of the ureido group during the transcarbamylation reaction (24), was used to measure enzyme activity. CP and AORN concentration were kept constant as 4.0 mM. After an incubation of 5 minutes, the reaction was stopped by the addition of 1 ml of color reagent, as described previously (19). A set of tubes containing known amounts of N-acetyl-L-citrulline was included with each rack of enzyme assays to produce a standard curve for calculation of the enzyme specific activity. Mass spectrometric analysis In order to identify the post-translational modification, mass spectrometric analysis was carried out on a 4700 ABI TOF/TOF mass spectrometer (Applier Biosystems) based on the method described previously (25). In brief, 10 µg of native protein were digested overnight at 310 K using trypsin in 50 mM ammonium bicarbonate pH 7.4. After desalting using a C18 ZipTip micropipette tip, the resulting peptides were eluted in 10 µl of acetonitrile/0.1% TFA [70:30(v:v)]. The sample was mixed with matrix solution and spotted on a MALDI plate to be submitted to the mass spectrometric analysis. Chemical rescue experiments The assay in the presence of various selected chemical was conducted as described above. The stock solutions of small chemicals were titrated to the pH of the assay with KOH or HCl. 13C NMR experiments The wild-type and K302A mutant protein of AOTCase (~10 mg) was precipitated by degassed buffer (pH 4.5) containing 25 mM sodium acetate. After centrifugation, the precipitate was re-dissolved by adding a buffer containing 20 mM NaH13CO3, 100 mM Tris HCl (pH 8.0) and 50 mM NaCl. Before NMR experiments, 40 µl D2O was added to 500 µl protein sample. The 13C NMR spectra were collected on a Bruker Avance 600 spectrometer (operating at 14.1 T) equipped with a direct 13C-detection probe at 298 K. The experimental settings and processing parameters for the wild-type protein and K302A variant were identical. 512 transients were collected with 4K time domain points and a spectral width of 3019 Hz centered at 160 ppm. The spectra were processed using exponential multiplication with the line broadening factor set to 3Hz. The similarity of protein concentration in both samples was verified by 1H NMR (not shown). Crystallization, data collection and processing PALAO-bound wild-type and mutant AOTCase crystals were grown using the hanging-drop vapor diffusion method, with conditions similar to those used to produce native and ligand- complexed AOTCase crystals (18,23). 2.0 µl of ~10 mg/ml solution of AOTCase were mixed with 1.6 µl of reservoir solution and 0.4 µl PALAO solution (~0.01 M). The reservoir solution contained 20% (w/v) PEG 3350, 0.2 M lithium sulfate and 0.1 M bis-Tris, pH 6.0. Diffraction data for the PALAO-bound crystal were collected at 100 K at the F1 beam line of the Cornell High Energy Synchrotron Source. Data sets for the PALAO-bound mutant AOTCase crystals were obtained using a Rigaku anode x-ray generator in the Molecular Structure Section of the National Institute of Health. All data were processed using HKL2000 package (26) and reduced using the program TRUNCATE in the CCP4 suite (27). Data collection parameters are listed in Table 1. Li et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Structure solution and refinement Molecular replacement was used for phase determination of the PALAO-bound wild-type and mutant AOTCase structures. The coordinates of AOTCase (PDB code: 3KZO) after removal of ligands or water molecules were used for phase determination. Upon rigid-body refinement, electron density corresponding to the ligands could be clearly visualized. The ligands were built into the map using the graphics program O (28). Refinements were carried out using molecular annealing, energy minimization and restrained B factor refinement with the program CNS1.1 (29). During refinements, 5% of the reflections at various resolutions were randomly selected to set aside to calculate Rfree to monitor the progress of refinement (30). After every cycle of refinement, the model was manually adjusted using the program O (28). Water molecules were automatically assigned using the program WATERPICK of CNS. Model quality was checked using the program PROCHECK (31) to ensure good stereochemistry for all three models. The final refinement statistics are listed in Table 1. Figures 1 was drawn using the programs Pymol (http://www.pymol.org). Figure 3 was drawn using ChemDraw 8.0. The coordinates have been deposited with the RCSB PDB as entries 3M4J, 3M4N, 3M5C and 3M5D. RESULTS Lys302 in AOTCase is carboxylated To investigate the nature of the modification of Lys302 and how it affects catalytic activity, we revisited all AOTCase structures. In the PALAO-bound AOTCase structure, the electron density map clearly indicates that Lys302 is post-translationally modified (Figure 1A). The type of modification can include methylation, acetylation, carbamylation, and carboxylation. The shape of the electron density can been used to distinguish methyl groups from larger functional groups, but it is difficult to distinguish between acetyl, carbamyl, and carboxyl groups, all of which have three non-hydrogen atoms in a plane. Given the hydrogen bonding network with surrounding residues (Lys252, Ser253 and His293, Table 2), a carboxylated modification is the most likely choice for the modification of Lys302 in AOTCase. To exclude that the modification’s identity represents chemically stable moieties (methyl, acetyl, carbamyl), we analyzed trypsin digested fragments of purified AOTCase by TOF- TOF mass spectroscopy. As expected, only a peptide fragment with an unmodified Lys302 was observed, consistent with the lability of the carboxylic group in acidic solutions. At low pH, the carboxyl group is spontaneously released as carbon dioxide (14, 32), in contrast to other modified groups that are stably bound and can be observed by mass spectrometry analysis after proteolysis (33). The putative carboxyl group on the modified Lys302 forms direct hydrogen bonds with main-chain or side-chain nitrogen atoms of Lys252, Ser253 and His293 (Figure 1A and Table 2). Among these, Lys252 is involved in direct hydrogen bonding to the carboxyl group of the AORN moiety of PALAO, and His293 forms a strong hydrogen bond with the main-chain nitrogen atom of Leu295 in the conserved His293-Cys294-Leu295-Pro206 (HCLP) motif. The hydrogen bonding network between the carboxyl group of modified Lys302, His293 and the main-chain nitrogen atom of Leu295 is reminiscent of the similar hydrogen bonding network, Glu310-His302-Leu304 and Glu299-Leu272-Leu274, found in human and E. coli OTCase, respectively (34, 35). These three residues are conserved in all OTCase sequences, and the interactions between them are important for maintaining the HCLP motif in a specific conformation to orientate their main-chain oxygen atoms towards the active site. In all known transcarbamylase structures, a leucine residue corresponding to Leu295 is in an energetically unfavorable conformation and the peptide bond between this Li et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript leucine and Pro296 is in the cis conformation. In addition to the direct hydrogen bonding interaction above, the carboxyl group on the modified Lys302 interacts with the α-amino nitrogen atom of the AORN moiety of PALAO and Glu92 from the adjacent subunit via water molecules. When we revisited all previously determined AOTCase structures (see supplementary Figure S1) we found: (1) Lys302 was carboxylated in the absence of substrate binding, but substrate binding immobilizes the side-chain of Lys302 further by hydrogen bonding interaction via water molecules. (2) Water-mediated hydrogen bonding promotes interactions of carboxylated Lys302 with Glu92 from the adjacent subunit and the α-amino nitrogen atom of AORN. (3) Similarly to AOTCase, in the structure of SOTCase E92Z (Z = Ala, Ser, Pro, Val), mutant with N-succinyl-L-norvaline bound (22) the carboxylated Lys302 hydrogen bonds to the α-amino nitrogen atom and the succinyl carboxyl group of N- succinyl-L-norvaline via water molecules (Figure S1). To obtain direct, independent evidence for the carboxylation of Lys302, 13C NMR experiments were carried out with both wild-type protein and the K302A mutant. As observed for other proteins with carboxylated lysine (36,37), the strong 13C NMR signal at 164 ppm characteristic of a carboxyl group was clearly detectable in AOTCase wild-type protein labeled by 13C-bicarbonate, in contrast to the K302A mutant where the signal was weak (Figure 2). Since there are 17 other lysine residues in the protein, the weak signal seen for the K302A mutant might be caused by the adventitious carboxylation of another lysine with reduced pKa, as has been observed for the K392A mutant of the sensor domain of the BlaR protein (38). Functional and structural studies of Lys302 mutants To investigate the effect of lysine carboxylation on enzyme activity, Lys302 was mutated to alanine, glutamate or arginine. Each of these variants was expressed in E. coli and gave similar yields. Enzymatic assays demonstrated a significant decrease in enzymatic activity in all three mutants, reflecting the functional importance of Lys302 (Table 3). The level of enzymatic activity for the wild-type (WT) and three mutants was WT > K302A > K302E ≫ K302R. To determine the structural basis of these results, the WT and mutant enzymes bound with PALAO were crystallized and their structures were determined at 1.8–2.2 Å resolution. Only the K302R mutation had and appreciable effect on the structure of the protein. Since K302 is located near the AORN binding site, the mutations would weaken AORN binding to the active site. In the structure of the K302A mutant, three additional water molecules (labeled as w3, w4 and w5 in Figure 1B) replace the carboxylated lysine. The two water molecules (labeled w1 and w2 in Figure 1A–1D) that mediate the hydrogen bonding interaction of carboxylated Lys302 with PALAO and Glu92 from the adjacent subunit are also found in the K302A mutant structure. Furthermore, these water molecules maintain a similar hydrogen-bonding network to the wild-type enzyme. These results might explain why the K302A mutant retains significant catalytic activity (Table 3). To investigate whether adding short-chain carboxylic acids to the K302A mutant increases its activity as other enzymes (14, 15, 39, 40), the activity of the K302A mutant was measured in the presence of high formate and acetate concentration (0.5 M). Surprisingly, the activity of the K302A mutant was not significantly improved. The crystal structure of the K302A mutant soaking with the crystallization buffer in the presence of 0.5 M acetate was also determined (not shown) and it was observed that the same five water molecules were present in the cavity that replaced the side chain of the carboxylated lysine. This, the acetate’s inability to replace the water molecules in the crystal structure, is consistent with the unchanged activity assay results. Li et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The side-chain of Glu302 in the K302E mutant structure is well defined and anchored by hydrogen bonding interaction with the main-chain nitrogen atom of Arg298 and weakly hydrogen bonded to the main-chain nitrogen atom of Ser253 (Figure 1C). Two of three additional water molecules (w4 and w5) observed in the K302A mutant structure occupied the same position as the carboxyl oxygen atoms of Glu302 and form a similar hydrogen- bonding network. Relative to the PALAO-bound wild-type structure, there is only one more water molecule (w3) at the position of the carboxyl group of the carboxylated Lys302. This water molecule mediates a hydrogen bonding interaction between Glu302 and Lys252. Two common water molecules (w1 and w2) that interact with PALAO and Glu92 from the adjacent subunit, respectively, were also identified in the K302E structure. Our observation that the K302E mutant had lower enzymatic activity than that of the K302A mutant (Table 3) was surprising since the carboxyl group of the glutamate could conceivably function similarly to a carboxylated lysine. The explanation may be that, in the K302A mutant, the hydrogen bonding network is well maintained by water molecules in the cavity that replaces the carboxylated lysine. In particular, w3 is optimally located for strong hydrogen bonding to w1 (2.7 Å), which in turn binds AORN. The distances between w1 and the carboxyl oxygen of carboxylated Lys302 in all wild-type crystal structures are within 2.4–2.7 Å, but the distance between w1 and w2 in the K302E structure is significantly greater (3.2 Å). The weaker hydrogen bonding interaction may be a reason for lower enzymatic activity of the K302E mutant. In contrast to the K302A and K302E structures, the K302R structure shows a much larger reduction in enzyme activity relative to the wild-type enzyme. The electron density for the side-chain of Arg302 is weak and the temperature factor of its side-chain is 54.4 Å2, significantly higher than those of carboxylated Lys302 (44.7 Å2) and Glu302 (33.4 Å2), implying greater flexibility. Furthermore, the side-chain of Arg302 is oriented differently from the carboxyl group of carboxylated Lys302 and pushes the nearby residues His180, Pro181 and Lys182 outwards about 1.0 Å (Figure 1D). However, the water molecules involved in hydrogen bonding to the α-amino nitrogen atom of PALAO (w1) and the side- chains of Lys252 (w3) and Glu92 (w2) from the adjacent subunit are conserved. Consistent with the K302E structure, the distance between w1 and w2 is even greater (3.4 Å) than in the WT structure and the hydrogen bonding interaction between w2 and w3 is no longer observed. Thus, the almost undetectable enzymatic activity of the K302R mutant probably results from the changes at its active site, including the weakened hydrogen bonding network involved in substrate binding. DISCUSSION Several lines of evidence clearly indicate that Lys302 in AOTCase is carboxylated. First, the extra electron density indicates that the side-chain of Lys302 is modified. Second, the hydrogen bonding environment of Lys302 for hydrogen bonding interactions is compatible with a carboxyl group, but not for a positively charged lysine side-chain. Third, the modification is labile at low pH, since mass spectroscopy of samples prepared at low pH indicated that Lys302 was no longer modified. Fourth, the clear presence of the indicative 13C NMR signal at 164 ppm for wild-type protein and its absence in the K302A mutant confirms carboxylation of Lys302. It is well known that lysine carboxylation is non-enzymatic and reversible, while other post- translational modifications such as methylation, acetylation, and carbamylation are irreversible and detectable by mass spectroscopy. Furthermore, lysine methylation and acetylation usually require an enzyme-catalyzed reaction in vivo (41). Therefore, it is unlikely that such lysine modifications will be observed in recombinant proteins Li et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript overexpressed in a foreign host (e.g. E. coli). Lysine methylation can be achieved by using special chemicals in vitro, but these chemicals are not present in vivo. Lysine carbamylation and carboxylation use completely different mechanisms to form functionally different groups (Figure 3). Carbamylation can be achieved by cyanate produced from myeloperoxidase-catalyzed oxidation of thiocyanate, an anion abundant in plasma and increased in smokers, or from urea in the plasma. Lysine carboxylation, on the other hand, occurs readily in aqueous solution in the presence of carbon dioxide at a basic pH (32,42). Even though carbamylation and carboxylation use very different mechanisms, the two are confused in the literatures. Lysine carbamylation (or carbamoylation) is referred to in several publications (15,32,42–44), when the actual reaction is in fact carboxylation. The activity of the K302A mutant is almost half of that of the wild-type enzyme raising the question of why AOTCase retains a lysine in this position. Perhaps this lysine was maintained through evolution to distinguish AOTCase from SOTCase which uses N- succinyl-L-ornithine (SORN) rather than AORN (22), and OTCase which uses L-ornithine. An alternative explanation may be found in the very low activity of the K302R mutant. The side-chain of arginine has a positive charge while carboxylated lysine has a negative charge. The side chain of unmodified lysine is usually located in a similar position as that of arginine, as observed in the structure of UV damage endonuclease (14). It would be expected that the activity of AOTCase with an uncarboxylated lysine would be as low as the K302R mutant’s. It could further be surmised that, the respective organisms need to use carboxylation as a switch to turn “on” or “off” the arginine biosynthetic pathway. It has been well known that rubulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) in plant cells uses the carboxylation on Lys201 as a switch to turn the enzyme “on” during the day and “off” at night by removing the carboxyl group (45,46). Carbon dioxide and bicarbonate have been found to play an important biological role in modulating several biological processes including photosynthetic carbon fixation (47), pH homeostasis (48), carbon metabolism (49), activation of virulence in pathogenic organisms (50), sperm maturation (51), stimulation of mammalian G-protein-responsive adenylyl cyclase (52), and as an alarmone in Drosophila (53,54). Whether or not carboxylation of a key lysine in their related proteins is used as an underlying regulatory mechanism should be investigated further. There are 197 structures with carboxylated lysine residue (modified residue indicated as Kcx) in the Protein Data Bank (PDB). If structures with 90% identity are counted only once, there are still 52 unique structures remaining in this pool (Table 4). These proteins include hydantoinase (40,55), folylpolyglutamate synthase (43), UV damage endonuclease (14), OXA10, OXA-1 class D β-lactamase (38,56,57), urease (42), phosphotriesterase (58), dihydroorotase (59), dihydropyrimidinase (60), organophosphate hydrolase (61) and MurE and MurD ligases (44,62). In most of these proteins, the carboxylated lysine bridges two metal ions, similar to the role of glutamate or aspartate in proteins with two metal-binding sites (26 structures among 52). However, the urease apoenzyme can be activated in vitro only in the presence of carbon dioxide prior to nickel binding (63), suggesting that the carboxylated lysine may have other structural roles beyond binding metals. In some proteins such as β-lactamase, UV damage endonuclease, Rubisco, MurD and MurE ligase and BlaR signal transducer protein, a carboxylated lysine plays an essential catalytic role. More interestingly, in three structures (PDB ID: 1HL9, 1PU6 and 2UYN for fucosiadase, 3- methyladenine DNA glycosylase and TdcF protein of unknown function, respectively), the carboxylated lysines are located near the surface of proteins, presumably playing primarily a structure stabilizing role (64–66). Since the carboxyl group is labile at acidic pH, but easily formed in the presence of carbon dioxide at basic pH, the number of proteins with lysine carboxylation must be underestimated. Furthermore, the carboxylated lysine must be fixed in place by metal ions (either one or two) or hydrogen bonding with other protein residues (at least one). Therefore, any detection method involving denaturing the proteins will result Li et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript in release of the carboxyl group. With current technology, 13C NMR (38) and crystallography are the only methods that can detect this modification. However, these methods are not amenable to high-throughput investigations. The majority (49 out of 52 structures in the PDB) of known lysine carboxylation modifications were found to be located at or near the active site, probably because these sites receive the most attention. Revisiting the structures in PDB with more attention to surface lysines might reveal more structures with carboxylated lysines. In conclusion, we have shown that Lys302 in AOTCase is post-translationally modified by carboxylation and that this modification may be functionally important for enzymatic activity. Lysine carboxylation is likely to be a more common event than currently appreciated and may play a critical role in enzymatic activity and protein stability. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Abbreviations ACIT N-acteyl-L-citrulline ANOR N-acetyl-L-norvaline AORN N-acetyl-L-Ornithine AOTCase N-acetyl-L-ornithine transcarbamylase ATCase aspartate transcarbamlyase OTCase ornithine transcarbamylase CP carbamyl phosphate ORN L-ornithine PALAO Nδ-(phosphonacetyl)-Nα-acetyl-L-ornithine SORN N-succinyl-L-ornithine WT wild-type xc Xanthomonas campestris Acknowledgments We thank Dr. David Davies for facilitating our use of the diffraction equipment in the Molecular Structure Section of the National Institute of Health and Dr. Fred Dyda for help in data collection and processing, and Dr. Yui-Fai Lam in the University of Maryland for help in setting up NMR measurements. REFERENCES 1. Close P, Creppe C, Gillard M, Ladang A, Chapelle JP, Nguyen L, Chariot A. The emerging role of lysine acetylation of non-nuclear proteins. Cell Mol Life Sci. 2010; 67:1255–1264. [PubMed: 20082207] 2. Geiman TM, Robertson KD. Chromatin remodeling, histone modifications, and DNA methylation- how does it all fit together? J Cell Biochem. 2002; 87:117–125. [PubMed: 12244565] 3. An W. Histone acetylation and methylation: combinatorial players for transcriptional regulation. Subcell Biochem. 2007; 41:351–369. [PubMed: 17484136] Li et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 4. Yang F, Bian C, Zhu L, Zhao G, Huang Z, Huang M. Effect of human serum albumin on drug metabolism: structural evidence of esterase activity of human serum albumin. J Struct Biol. 2007; 157:348–355. [PubMed: 17067818] 5. Xu AS, Labotka RJ, London RE. Acetylation of human hemoglobin by methyl acetylphosphate. Evidence of broad regio-selectivity revealed by NMR studies. J Biol Chem. 1999; 274:26629– 26632. [PubMed: 10480863] 6. Ueno H, Pospischil MA, Manning JM. Methyl acetyl phosphate as a covalent probe for anion- binding sites in human and bovine hemoglobins. J Biol Chem. 1989; 264:12344–12351. [PubMed: 2745446] 7. Stavropoulos P, Nagy V, Blobel G, Hoelz A. Molecular basis for the autoregulation of the protein acetyl transferase Rtt109. Proc Natl Acad Sci U S A. 2008; 105:12236–12241. [PubMed: 18719104] 8. Bewley MC, Graziano V, Jiang J, Matz E, Studier FW, Pegg AE, Coleman CS, Flanagan JM. Structures of wild-type and mutant human spermidine/spermine N1-acetyltransferase, a potential therapeutic drug target. Proc Natl Acad Sci U S A. 2006; 103:2063–2068. [PubMed: 16455797] 9. Walter TS, Meier C, Assenberg R, Au KF, Ren J, Verma A, Nettleship JE, Owens RJ, Stuart DI, Grimes JM. Lysine methylation as a routine rescue strategy for protein crystallization. Structure. 2006; 14:1617–1622. [PubMed: 17098187] 10. Stark GRSW, Moore S. Reactions of the cyanante present in aqueous urea with amino acids and proteins. J Biol Chem. 1960; 235:3177–3181. 11. Bobb D, Hofstee BH. Gel isoelectric focusing for following the successive carbamylations of amino groups in chymotrypsinogen A. Anal Biochem. 1971; 40:209–217. [PubMed: 5550146] 12. Kraus LM, Kraus AP Jr. Carbamoylation of amino acids and proteins in uremia. Kidney Int Suppl. 2001; 78:S102–S107. [PubMed: 11168993] 13. Al-Dirbashi OY, Al-Hassnan ZN, Rashed MS. Determination of homocitrulline in urine of patients with HHH syndrome by liquid chromatography tandem mass spectrometry. Anal Bioanal Chem. 2006; 386:2013–2017. [PubMed: 17053917] 14. Meulenbroek EM, Paspaleva K, Thomassen EA, Abrahams JP, Goosen N, Pannu NS. Involvement of a carboxylated lysine in UV damage endonuclease. Protein Sci. 2009; 18:549–558. [PubMed: 19241382] 15. Dementin S, Bouhss A, Auger G, Parquet C, Mengin-Lecreulx D, Dideberg O, van Heijenoort J, Blanot D. Evidence of a functional requirement for a carbamoylated lysine residue in MurD, MurE and MurF synthetases as established by chemical rescue experiments. Eur J Biochem. 2001; 268:5800–5807. [PubMed: 11722566] 16. Cha J, Mobashery S. Lysine N(zeta)-decarboxylation in the BlaR1 protein from Staphylococcus aureus at the root of its function as an antibiotic sensor. J Am Chem Soc. 2007; 129:3834–3835. [PubMed: 17343387] 17. Shi D, Yu X, Roth L, Morizono H, Tuchman M, Allewell NM. Structures of N-acetylornithine transcarbamoylase from Xanthomonas campestris complexed with substrates and substrate analogs imply mechanisms for substrate binding and catalysis. Proteins. 2006; 64:532–542. [PubMed: 16741992] 18. Shi D, Morizono H, Yu X, Roth L, Caldovic L, Allewell NM, Malamy MH, Tuchman M. Crystal structure of N-acetylornithine transcarbamylase from Xanthomonas campestris: a novel enzyme in a new arginine biosynthetic pathway found in several eubacteria. J Biol Chem. 2005; 280:14366– 14369. [PubMed: 15731101] 19. Morizono H, Cabrera-Luque J, Shi D, Gallegos R, Yamaguchi S, Yu X, Allewell NM, Malamy MH, Tuchman M. Acetylornithine transcarbamylase: a novel enzyme in arginine biosynthesis. J Bacteriol. 2006; 188:2974–2982. [PubMed: 16585758] 20. da Silva FR, Vettore AL, Kemper EL, Leite A, Arruda P. Fastidian gum: the Xylella fastidiosa exopolysaccharide possibly involved in bacterial pathogenicity. FEMS Microbiol Lett. 2001; 203:165–171. [PubMed: 11583843] 21. da Silva AC, Ferro JA, Reinach FC, Farah CS, Furlan LR, Quaggio RB, Monteiro-Vitorello CB, Van Sluys MA, Almeida NF, Alves LM, do Amaral AM, Bertolini MC, Camargo LE, Camarotte G, Cannavan F, Cardozo J, Chambergo F, Ciapina LP, Cicarelli RM, Coutinho LL, Cursino-Santos Li et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript JR, El-Dorry H, Faria JB, Ferreira AJ, Ferreira RC, Ferro MI, Formighieri EF, Franco MC, Greggio CC, Gruber A, Katsuyama AM, Kishi LT, Leite RP, Lemos EG, Lemos MV, Locali EC, Machado MA, Madeira AM, Martinez-Rossi NM, Martins EC, Meidanis J, Menck CF, Miyaki CY, Moon DH, Moreira LM, Novo MT, Okura VK, Oliveira MC, Oliveira VR, Pereira HA, Rossi A, Sena JA, Silva C, de Souza RF, Spinola LA, Takita MA, Tamura RE, Teixeira EC, Tezza RI, Trindade dos Santos M, Truffi D, Tsai SM, White FF, Setubal JC, Kitajima JP. Comparison of the genomes of two Xanthomonas pathogens with differing host specificities. Nature. 2002; 417:459– 463. [PubMed: 12024217] 22. Shi D, Yu X, Cabrera-Luque J, Chen TY, Roth L, Morizono H, Allewell NM, Tuchman M. A single mutation in the active site swaps the substrate specificity of N-acetyl-L-ornithine transcarbamylase and N-succinyl-L-ornithine transcarbamylase. Protein Sci. 2007; 16:1689–1699. [PubMed: 17600144] 23. Shi D, Morizono H, Cabrera-Luque J, Yu X, Roth L, Malamy MH, Allewell NM, Tuchman M. Structure and catalytic mechanism of a novel N-succinyl-L-ornithine transcarbamylase in arginine biosynthesis of Bacteroides fragilis. J Biol Chem. 2006; 281:20623–20631. [PubMed: 16704984] 24. Pastra-Landis SC, Foote J, Kantrowitz ER. An improved colorimetric assay for aspartate and ornithine transcarbamylases. Anal Biochem. 1981; 118:358–363. [PubMed: 7337232] 25. Shi D, Yu X, Roth L, Morizono H, Hathout Y, Allewell NM, Tuchman M. Expression, purification, crystallization and preliminary X-ray crystallographic studies of a novel acetylcitrulline deacetylase from Xanthomonas campestris. Acta Crystallogr Sect F Struct Biol Cryst Commun. 2005; 61:676–679. 26. Otwinowski Z, Minor W. Processing of X-ray Diffraction Data Collected in Oscillation Mode. Methods in Enzymology. 1997; 276:307–326. 27. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr. 1994; 50:760–763. [PubMed: 15299374] 28. Jones TA, Zou JY, Cowan SW, Kjeldgaard M. Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr A. 1991; 47(Pt 2):110–119. [PubMed: 2025413] 29. Brunger AT, Adams PD, Clore GM, DeLano WL, Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges M, Pannu NS, Read RJ, Rice LM, Simonson T, Warren GL. Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr. 1998; 54:905–921. [PubMed: 9757107] 30. Brunger AT. Free R value: a novel statistical quantity for assessing the accuracy of crystal structures. Nature. 1992; 355:472–475. [PubMed: 18481394] 31. Laskowski RA, MacArthur MW, Moss DS, Thornton JM. PROCHECK: a program to check the stereochemical quality of protein structures. J Appl Crystallogr. 1993; 26:283–291. 32. Golemi D, Maveyraud L, Vakulenko S, Samama JP, Mobashery S. Critical involvement of a carbamylated lysine in catalytic function of class D beta-lactamases. Proc Natl Acad Sci U S A. 2001; 98:14280–14285. [PubMed: 11724923] 33. Lapko VN, Smith DL, Smith JB. In vivo carbamylation and acetylation of water-soluble human lens alphaB-crystallin lysine 92. Protein Sci. 2001; 10:1130–1136. [PubMed: 11369851] 34. Shi D, Morizono H, Ha Y, Aoyagi M, Tuchman M, Allewell NM. 1.85-A resolution crystal structure of human ornithine transcarbamoylase complexed with N-phosphonacetyl-L-ornithine. Catalytic mechanism and correlation with inherited deficiency. J Biol Chem. 1998; 273:34247– 34254. [PubMed: 9852088] 35. Langley DB, Templeton MD, Fields BA, Mitchell RE, Collyer CA. Mechanism of inactivation of ornithine transcarbamoylase by Ndelta -(N'-Sulfodiaminophosphinyl)-L-ornithine, a true transition state analogue? Crystal structure and implications for catalytic mechanism. J Biol Chem. 2000; 275:20012–20019. [PubMed: 10747936] 36. Cha J, Vakulenko SB, Mobashery S. Characterization of the beta-lactam antibiotic sensor domain of the MecR1 signal sensor/transducer protein from methicillin-resistant Staphylococcus aureus. Biochemistry. 2007; 46:7822–7831. [PubMed: 17550272] Li et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 37. O'Leary MH, Jaworski RJ, Hartman FC. C nuclear magnetic resonance study of the CO(2) activation of ribulosebisphosphate carboxylase from Rhodospirillum rubrum. Proc Natl Acad Sci U S A. 1979; 76:673–675. [PubMed: 16592618] 38. Golemi-Kotra D, Cha JY, Meroueh SO, Vakulenko SB, Mobashery S. Resistance to beta-lactam antibiotics and its mediation by the sensor domain of the transmembrane BlaR signaling pathway in Staphylococcus aureus. J Biol Chem. 2003; 278:18419–18425. [PubMed: 12591921] 39. Schneider KD, Bethel CR, Distler AM, Hujer AM, Bonomo RA, Leonard DA. Mutation of the active site carboxy-lysine (K70) of OXA-1 beta-lactamase results in a deacylation-deficient enzyme. Biochemistry. 2009; 48:6136–6145. [PubMed: 19485421] 40. Huang CY, Hsu CC, Chen MC, Yang YS. Effect of metal binding and posttranslational lysine carboxylation on the activity of recombinant hydantoinase. J Biol Inorg Chem. 2009; 14:111–121. [PubMed: 18781344] 41. Zhang Q, Wang Y. High mobility group proteins and their post-translational modifications. Biochim Biophys Acta. 2008; 1784:1159–1166. [PubMed: 18513496] 42. Jabri E, Carr MB, Hausinger RP, Karplus PA. The crystal structure of urease from Klebsiella aerogenes. Science. 1995; 268:998–1004. [PubMed: 7754395] 43. Young PG, Smith CA, Metcalf P, Baker EN. Structures of Mycobacterium tuberculosis folylpolyglutamate synthase complexed with ADP and AMPPCP. Acta Crystallogr D Biol Crystallogr D. 2008; 64:745–753. 44. Gordon E, Flouret B, Chantalat L, van Heijenoort J, Mengin-Lecreulx D, Dideberg O. Crystal structure of UDP-N-acetylmuramoyl-L-alanyl-D-glutamate: meso-diaminopimelate ligase from Escherichia coli. J Biol Chem. 2001; 276:10999–11006. [PubMed: 11124264] 45. Taylor TC, Andersson I. Structure of a product complex of spinach ribulose-1,5-bisphosphate carboxylase/oxygenase. Biochemistry. 1997; 36:4041–4046. [PubMed: 9092835] 46. Jensen R. Activation of Rubisco controls CO(2) assimilation in light: a perspective on its discovery. Photosynth Res. 2004; 82:187–193. [PubMed: 16151874] 47. Falkowski PG. Photosynthesis: the paradox of carbon dioxide efflux. Curr Biol. 1997; 7:R637– R639. [PubMed: 9368746] 48. Roos A, Boron WF. Intracellular pH. Physiol Rev. 1981; 61:296–434. [PubMed: 7012859] 49. Smith KS, Ferry JG. Prokaryotic carbonic anhydrases. FEMS Microbiol Rev. 2000; 24:335–366. [PubMed: 10978542] 50. Bahn YS, Muhlschlegel FA. CO2 sensing in fungi and beyond. Curr Opin Microbiol. 2006; 9:572– 578. [PubMed: 17045514] 51. Esposito G, Jaiswal BS, Xie F, Krajnc-Franken MA, Robben TJ, Strik AM, Kuil C, Philipsen RL, van Duin M, Conti M, Gossen JA. Mice deficient for soluble adenylyl cyclase are infertile because of a severe sperm-motility defect. Proc Natl Acad Sci U S A. 2004; 101:2993–2998. [PubMed: 14976244] 52. Townsend PD, Holliday PM, Fenyk S, Hess KC, Gray MA, Hodgson DR, Cann MJ. Stimulation of mammalian G-protein-responsive adenylyl cyclases by carbon dioxide. J Biol Chem. 2009; 284:784–791. [PubMed: 19008230] 53. Kwon JY, Dahanukar A, Weiss LA, Carlson JR. The molecular basis of CO2 reception in Drosophila. Proc Natl Acad Sci U S A. 2007; 104:3574–3578. [PubMed: 17360684] 54. Jones WD, Cayirlioglu P, Kadow IG, Vosshall LB. Two chemosensory receptors together mediate carbon dioxide detection in Drosophila. Nature. 2007; 445:86–90. [PubMed: 17167414] 55. Xu Z, Liu Y, Yang Y, Jiang W, Arnold E, Ding J. Crystal structure of D-Hydantoinase from Burkholderia pickettii at a resolution of 2.7 Angstroms: insights into the molecular basis of enzyme thermostability. J Bacteriol. 2003; 185:4038–4049. [PubMed: 12837777] 56. Sun T, Nukaga M, Mayama K, Braswell EH, Knox JR. Comparison of beta-lactamases of classes A and D: 1.5-A crystallographic structure of the class D OXA-1 oxacillinase. Protein Sci. 2003; 12:82–91. [PubMed: 12493831] 57. Maveyraud L, Golemi D, Kotra LP, Tranier S, Vakulenko S, Mobashery S, Samama JP. Insights into class D beta-lactamases are revealed by the crystal structure of the OXA10 enzyme from Pseudomonas aeruginosa. Structure. 2000; 8:1289–1298. [PubMed: 11188693] Li et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 58. Benning MM, Kuo JM, Raushel FM, Holden HM. Three-dimensional structure of the binuclear metal center of phosphotriesterase. Biochemistry. 1995; 34:7973–7978. [PubMed: 7794910] 59. Thoden JB, Phillips GN Jr, Neal TM, Raushel FM, Holden HM. Molecular structure of dihydroorotase: a paradigm for catalysis through the use of a binuclear metal center. Biochemistry. 2001; 40:6989–6997. [PubMed: 11401542] 60. Abendroth J, Niefind K, Schomburg D. X-ray structure of a dihydropyrimidinase from Thermus sp. at 1.3 A resolution. J Mol Biol. 2002; 320:143–156. [PubMed: 12079340] 61. Yang H, Carr PD, McLoughlin SY, Liu JW, Horne I, Qiu X, Jeffries CM, Russell RJ, Oakeshott JG, Ollis DL. Evolution of an organophosphate-degrading enzyme: a comparison of natural and directed evolution. Protein Eng. 2003; 16:135–145. [PubMed: 12676982] 62. Bertrand JA, Auger G, Martin L, Fanchon E, Blanot D, Le Beller D, van Heijenoort J, Dideberg O. Determination of the MurD mechanism through crystallographic analysis of enzyme complexes. J Mol Biol. 1999; 289:579–590. [PubMed: 10356330] 63. Park IS, Hausinger RP. Requirement of carbon dioxide for in vitro assembly of the urease nickel metallocenter. Science. 1995; 267:1156–1158. [PubMed: 7855593] 64. Sulzenbacher G, Bignon C, Nishimura T, Tarling CA, Withers SG, Henrissat B, Bourne Y. Crystal structure of Thermotoga maritima alpha-L-fucosidase. Insights into the catalytic mechanism and the molecular basis for fucosidosis. J Biol Chem. 2004; 279:13119–13128. [PubMed: 14715651] 65. Eichman BF, O'Rourke EJ, Radicella JP, Ellenberger T. Crystal structures of 3-methyladenine DNA glycosylase MagIII and the recognition of alkylated bases. Embo J. 2003; 22:4898–4909. [PubMed: 14517230] 66. Burman JD, Stevenson CE, Sawers RG, Lawson DM. The crystal structure of Escherichia coli TdcF, a member of the highly conserved YjgF/YER057c/UK114 family. BMC Struct Biol. 2007; 7:30. [PubMed: 17506874] Li et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Stereo view of the structure and hydrogen bonding network surrounding residue 302. A, PALAO bound wild-type AOTCase, B, PALAO bound K302A AOTCase C, PALAO bound K302E AOTCase, D, PALAO bound K302R AOTCase. Contours of the electron density maps (2Fo-Fc) around PALAO, residue 302 and water molecules are shown as a brown cage at 1.0σ. The final refined positions of the ligands and surrounding protein residues are represented as colored sticks. The predicted hydrogen bonding interactions are in pink dashed lines. The water molecules are represented as pink balls. The carbon of PALAO, residue 302 and other protein residues are shown in pink, light blue and green sticks, respectively. Li et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. 13C NMR spectra of wild-type (upper panel) and K302A mutant (lower panel) AOTCase (1 mM). Experiments were performed in 100 mM Tris HCl, 50 mM NaCl, 7% D2O, pH 8.0, supplemented with 20 mM NaH13CO3. The position of the resonance attributed to carboxylated lysine in the enzyme is around 164 ppm. Li et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Chemical structure of carbamylated vs. carboxylated lysine. Li et al. Page 16 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 17 Table 1 Data collection and refinement statistics Dataset PALAO K302A K302E K302R Space group I213 I213 I213 I213 Resolution (Å) 2.2 1.9 1.85 2.2 Unit-cell parameters (Å) a = b = c =128.88 a = b = c =128.92 a = b = c =129.29 a = b = c =127.39 Measurements 219,475 305,757 390,128 246,817 Unique reflections 18,269 (1,832) a 28,236 (1,365) 30,622 (1,456) 17,635 (879) Redundancy 12.0 (11.8) 10.8(5.4) 12.8 (5.4) 14.0 (13.1) Completeness (%) 99.8 (100.0) 100.0 (100.0) 99.7 (95.1) 100.0 (100.0) <I/σ (I)> 15.0 (4.9) 16.4 (2.3) 19.8 (2.8) 8.7 (3.7) Rmerg b 7.4 (48.4) 6.5(64.9) 5.2 (55.3) 9.8 (79.1) Wilson B (Å2) 30.4 27.6 28.6 21.9 Refinement Resolution range (Å) 50.0-2.2 50-1.9 50-1.85 50-2.2 No. of protein atoms 2620 2613 2617 2619 No. of water atoms 90 219 193 146 No. of hetero atoms 24 24 24 24 Rmsd of bond lengths (Å) 0.006 0.005 0.005 0.005 Rmsd of bond angle (°) 1.1 1.2 1.2 1.2 Rwork (%)c 20.0 19.8 20.0 18.9 Rfree (%)d 24.3 23.2 23.2 22.2 Average B factor (Å2) 41.7 32.2 32.3 35.3 aFigures in brackets apply to the highest-resolution shell. bRmerg = ΣhΣi|I(h,i)-<I(h)>|/∑hΣiI(h,i), where I(h,i) is the intensity of the ith observation of reflection h, and < I(h)> is the average intensity of redundant measurements of reflection h. cRwork= Σh‖Fobs| – |Fcalc‖/Σh|Fobs|. dRfree = Σh‖Fobs| – |Fcalc‖/Σh|Fobs| for 5% of the reserved reflections. Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 18 Table 2 Interactions between carboxylated lysine and other residues at the active site of AOTCase Kcx302 Other residues Bound ligands PALAO CPa AORNb CP +ANORc SO4+ACITd OQ1 K252 NZ 2.6 2.6 2.7 2.6 2.6 OQ1 W1e 2.6 2.6 2.6 2.7 OQ2 S253 N 3.0 3.1 2.8 2.9 2.9 OQ2 H293 NE2 3.0 3.2 3.0 2.9 2.9 NZ W2f 3.1 2.9 3.0 3.0 aThe values were calculated based on PDB ID 3KZM. bThe values were calculated based on PDB ID 3KZN. cThe values were calculated based on PDB ID 3KZO. dThe values were calculated based on PDB ID 3KZK. eThis water molecule hydrogen bonds to N1 atom of PALAO, AORN or ANOR, and backbone O atom of Pro296 as well. fThis water molecule hydrogen bonds to OE1 atom of Glu92 from adjacent subunit as well. Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 19 Table 3 Specific activity of wild-type and mutant AOTCase in the presence of acids (0.5M). Compounds added Specific activity(µmol/min/mg) Wild-type K302A K302E K302R None 43.4 ± 0.4a 23.0 ± 0.5 7.1 ± 0.1 0.059±0.01 Formate 44.1 ± 1.2 26.4 ± 0.6 6.7 ± 0.2 0.093±0.01 Acetate 48.5 ± 1.1 21.2 ± 0.8 6.6 ± 0.5 0.104±0.03 aThe Mean ± S.D. are shown (n = 3). Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 20 Table 4 Protein structures with lysine carboxylation modification PDB ID Enzyme name Residue Organism source Funciton 2OGJ Dihydroorotase 175 A.tumefaciens Bridging two Zn(II) 2Z26 Dihydroorotase 102 E.coli Bridging two Zn(II) 3JZE Dihydroorotase 103 S.enterica Bridging two Zn(II) 2GWN Dihydroorotase 149 P. gingivalis Bridging two Zn(II) 3F4C Organophosphorus hydrolase 243 G. stearothermophilus Bridging two Co(II) 3ICJ Metal-dependent hydrolase 294 P. furiosus Bridging two Zn(II) 3GTX Organophosphorus hydrolase 243 D. radiodurans Bridging two Co(II) 2QPX Metal-dependent hydrolase 166 L. casei Bridging two Zn(II) 2FTW Dihydropyrimidinase 158 D. discoideum Bridging two Zn(II) 2FVK Dihydropyrimidinase 167 S. kluyveri Bridging two Zn(II) 3DC8 Dihydropyrimidinase 147 S. meliloti Bridging two Zn(II) 3GNH L-Lys/Arg carboxypeptidase 211 C. crescentus cb15 Bridging two Zn(II) 3DUG Arginine carboxypeptidase 182 Unidentified Bridging two Zn(II) 2VC7 Phosphotriesterase 137 S. solfataricus Bridging two Co(II) 2R1N Metallophosphotriesterases 169 A. tumefaciens Bridging two Co(II) 2OB3 Phosphotriesterase 169 B. diminuta Bridging two Zn(II) 3E74 Allantoinase 146 E. coli Bridging two Fe(III) 1EJX Urease 217 K. aerogenes Bridging two Ni(II) 1E9Z Urease 219 H. pylori Bridging two Ni(II) 4UBP Urease 220 B. pasteurii Bridging two Ni(II) 1ONW Isoaspartyl dipeptidase 162 E. coli Bridging two Zn(II) 1K1D D-hydanroinase 150 G. stearothermophilus Bridging two Zn(II) 1GKR L-hydanroinase 147 A. aurescens Bridging two Zn(II) 1GKP D-hydanroinase 150 Thermus sp. Bridging two Zn(II) 1NFG D-hydantoinase 148 R. pickettii Bridging two Zn(II) 2ICS Adenine deaminase 154 E. faecalis Bridging two Zn(II) 1RQB Transcarboxylase 184 P. freudenreichii Binding one Co(II) 2QF7 Pyruvate carboxylase 718 R. etli Binding one Zn(II) 3BG3 Pyruvate carboxylase 741 H. sapiens Binding one Mn(II) 2OEM Rubisco-like protein 173 G. kaustophilus Binding one Mg(II) 1WDD Rubisco 201 O. sativa Binding one Mg(II) 1GK8 Rubisco 201 C. reinhardtii Binding one Mg(II) 1BWV Rubisco 201 G. partita Binding one Mg(II) 2WTZ ATP-dependent MurE ligase 262 M. tuberculosis Binding one Mg(II) 2JFG MurD ligase 198 E. coli Catalytic role? 1E8C MurE ligase 224 E. coli Catalytic role? 1JBW Folypolyglutamate synthetase 185 L. casei Catalytic role? 1W78 FolC bifunctional protein 188 E. coli Binding one Mg(II) 3HBR OXA-48 β-lactamase 73 K. pneumoniae Catalytic role Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 21 PDB ID Enzyme name Residue Organism source Funciton 3ISG Class D β-lactamase 70 E. coli Catalytic role 2P9V AmpC beta-lactamase 315 E. coli Catalytic role 1K55 OXA-10 β-lactamase 70 P. aeruginosa Catalytic role 1K38 β-lactamase OXA-2 70 S. typhimurium Catalytic role 1XQL Alanine racemase 129 G. stearothermophilus Binding substrate? 1VFS Alanine racemase 129 S. lavendulae Binding substrate? 1RCQ Alanine racemase 122 P. aeruginosa Binding substrate? 2J6V UV damage endonuclease 229 T. thermophilus Catalytic role 1H01 Cell division protein kinase 2 33 H. sapiens Catalytic role? 2UYN Protein TdcF A58 E. coli Structural role? 1HL9 Fucosidase 338 T. maritime Structural role? 1PU6 3-methyladenine DNA glycosylase 205 H. pylori Structural role? Biochemistry. Author manuscript; available in PMC 2011 August 17.
3M4N
Crystal structure of N-acetyl-L-ornithine transcarbamylase K302A mutant complexed with PALAO
Reversible Post-Translational Carboxylation Modulates The Enzymatic Activity Of N-Acetyl-L-Ornithine Transcarbamylase† Yongdong Li1,2, Xiaolin Yu1, Jeremy Ho1, David Fushman3, Norma M. Allewell3, Mendel Tuchman1, and Dashuang Shi1,‡ 1Research Center for Genetic Medicine and Department of Integrative Systems Biology, Children’s National Medical Center, The George Washington University, Washington, DC 20010, USA. 2Key Laboratory of Organo-Pharmaceutical Chemistry, Jiangxi Province, Gannan Normal University, Ganzhou 341000, China. 3Department of Chemistry and Biochemistry, College of Chemical and Life Sciences, University of Maryland, College Park, MD 20742, USA. Abstract N-acetyl-L-ornithine transcarbamylase (AOTCase), rather than ornithine transcarbamylase (OTCase), is the essential carbamylase enzyme in the arginine biosynthesis of several plant and human pathogens. The specificity of this unique enzyme provides a potential target for controlling the spread of these pathogens. Recently, several crystal structures of AOTCase from Xanthomonas campestris (xc) have been determined. In these structures, an unexplained electron density at the tip of Lys302 side-chain was observed. Using 13C NMR spectroscopy, we show herein that Lys302 is post-translationally carboxylated. The structure of wild-type AOTCase complexed with the bisubstrate analogue, Nδ-(phosphonoacetyl)-Nα-acetyl-L-ornithine (PALAO), indicates that the carboxyl group on Lys302 forms a strong hydrogen bonding network with surrounding active site residues, Lys252, Ser253, His293, and Glu92 from the adjacent subunit either directly or via a water molecule. Furthermore, the carboxyl group is involved in binding N-acetyl-L-ornithine via a water molecule. Activity assays with the wild-type enzyme and several mutants demonstrate that the post translational modification of lysine 302 has an important role in catalysis. Post-translational modification of the ε-amino group of lysine residues in proteins is a common mechanism used by organisms to regulate protein functions including DNA-protein interactions, subcellular localization, transcriptional activity, and protein stability and activity (1). Lysine residues can be modified by the addition of functional groups to become acetylated, methylated, carbamylated or carboxylated. The role of histone lysine acetylation and methylation in affecting chromatin structure and gene expression has been well established for more than a decade (2). However, the biological roles for lysine carbamylation and carboxylation have rarely been investigated. †This work was supported by Public Health Service grants DK-47870 (MT) and DK-067935 (DS) from the National Institute of Diabetes, Digestive and Kidney Diseases. JH was supported by a Scholarship from the Doug and Lynn Parsons Family Foundation. The Cornell High Energy Synchrotron Source (CHESS) is supported by the National Science Foundation under award DMR 0225180 and the Macromolecular Diffraction Facility at CHESS (MacCHESS) is supported by award RR-01646 from the National Institutes of Health, through its National Center for Research Resources. ‡Corresponding author. dshi@cnmcresearch.org. Phone: 202-476-5817. Fax: 202-476-6014. SUPPORTING INFORMATION AVAILABLE Figure S1. Structure and hydrogen bonding network around residue 302 for previously determined AOTCase structures. This material is available free of charge via the Internet at http://pubs.acs.org. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 August 17. Published in final edited form as: Biochemistry. 2010 August 17; 49(32): 6887–6895. doi:10.1021/bi1007386. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript In vivo, lysine acetylation and methylation are usually carried out by acetyltransferase and methyltransferase enzymes, respectively (3). In addition, some proteins such as hemoglobin and human serum albumin can be acetylated non-enzymatically by chemicals such as aspirin, methyl acetyl phosphate, and other acetylating agents such as acetyl-CoA (4–8). Lysine can also be methylated by small chemicals in vitro, and this has routinely been used as a rescue method for protein crystallization (9). Lysine carbamylation and lysine carboxylation have only been achieved by using chemicals and no enzyme has yet been found to catalyze these modifications. Lysine carbamylation was one of the earliest post- translational modification of proteins to be elucidated when it was identified as a product of reversible denaturation-renaturation studies of proteins with urea (10,11). This carbamylation, which produces homocitrulline, has also been detected in uremic patients (12) and in patients with elevated plasma and/or urinary lysine levels (13). In contrast, lysine carboxylation is not as commonly reported, but has been identified in a number of proteins via crystal structure determinations. In most of these proteins, the carboxyl groups of modified lysines are involved in bridging two metal ions that play a structural role in the active site. In several other proteins, however, a direct role for a carboxylated lysine in the catalytic mechanism has been reported (14–16). N-acetyl-L-ornithine transcarbamylase (AOTCase, EC 2.1.3.9) was recently discovered to be part of a novel arginine biosynthesis pathway in plant pathogens of the Xanthomonadaceae family such as Xylella and Xanthomonas (17–19). These pathogens attack a variety of economically important crops including citrus fruits, cotton, tomatoes, and rice (20,21). Genome sequence analyses showed that an AOTCase-like gene is also present in some human pathogens such as Stenotrophomonas maltophilia and members of the genus Bacteroides (22). In the case of Bacteroides fragilis, this gene was later confirmed to encode another novel transcarbamylase, N-succinyl-L-ornithine transcarbamylase (SOTCase, EC 2.1.3.11) (23). Crystal structures of both AOTCase and SOTCase bound with substrate or substrate analogues have recently been determined (17,18,23). An extended density at the side-chain tip of Lys302 in AOTCase was observed suggesting a post- translational modification. Since Lys302 is located within the active site of AOTCase and is not present in SOTCase, it was proposed as one of three key signature residues to distinguish the two carbamylases (22). Here, we demonstrate that Lys302 is post- translationally modified by carboxylation and that this change affects the catalytic function of the enzyme. MATERIALS AND METHODS Materials All chemicals were purchased from Sigma Chemical Company unless otherwise specified. ANOR was purchased from Indofine Chemical Co., Inc. N-acetyl-L-citrulline was custom synthesized and purified by Chiral Quest Company. PALAO (>95% purity) was synthesized by IMI TAMI Institute of Research and Development Ltd. (19). xcAOTCase was prepared and purified as previously described (18). Mutants K302A (primer: 5’- CTGCGTCGCAACGTCGCGGCTACTGATGCGGTG-3’), K302E (primer:5’- CTGCGTCGCAACGTCGAGGCTACTGATGCGGTG-3’) and K302R (primer: 5’- CTGCGTCGCAACGTCAGGGCTACTGATGCGGTG-3’) were generated by site-directed mutagenesis using the “Quik Change” mutagenesis kit (Stratagene) according to the manufacturer’s protocol. The correct mutants were confirmed by DNA sequencing. Recombinant mutant proteins were expressed and purified in the same manner as the wild- type enzyme. Li et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Activity assay The modified colorimetric assay method, which detects the formation of the ureido group during the transcarbamylation reaction (24), was used to measure enzyme activity. CP and AORN concentration were kept constant as 4.0 mM. After an incubation of 5 minutes, the reaction was stopped by the addition of 1 ml of color reagent, as described previously (19). A set of tubes containing known amounts of N-acetyl-L-citrulline was included with each rack of enzyme assays to produce a standard curve for calculation of the enzyme specific activity. Mass spectrometric analysis In order to identify the post-translational modification, mass spectrometric analysis was carried out on a 4700 ABI TOF/TOF mass spectrometer (Applier Biosystems) based on the method described previously (25). In brief, 10 µg of native protein were digested overnight at 310 K using trypsin in 50 mM ammonium bicarbonate pH 7.4. After desalting using a C18 ZipTip micropipette tip, the resulting peptides were eluted in 10 µl of acetonitrile/0.1% TFA [70:30(v:v)]. The sample was mixed with matrix solution and spotted on a MALDI plate to be submitted to the mass spectrometric analysis. Chemical rescue experiments The assay in the presence of various selected chemical was conducted as described above. The stock solutions of small chemicals were titrated to the pH of the assay with KOH or HCl. 13C NMR experiments The wild-type and K302A mutant protein of AOTCase (~10 mg) was precipitated by degassed buffer (pH 4.5) containing 25 mM sodium acetate. After centrifugation, the precipitate was re-dissolved by adding a buffer containing 20 mM NaH13CO3, 100 mM Tris HCl (pH 8.0) and 50 mM NaCl. Before NMR experiments, 40 µl D2O was added to 500 µl protein sample. The 13C NMR spectra were collected on a Bruker Avance 600 spectrometer (operating at 14.1 T) equipped with a direct 13C-detection probe at 298 K. The experimental settings and processing parameters for the wild-type protein and K302A variant were identical. 512 transients were collected with 4K time domain points and a spectral width of 3019 Hz centered at 160 ppm. The spectra were processed using exponential multiplication with the line broadening factor set to 3Hz. The similarity of protein concentration in both samples was verified by 1H NMR (not shown). Crystallization, data collection and processing PALAO-bound wild-type and mutant AOTCase crystals were grown using the hanging-drop vapor diffusion method, with conditions similar to those used to produce native and ligand- complexed AOTCase crystals (18,23). 2.0 µl of ~10 mg/ml solution of AOTCase were mixed with 1.6 µl of reservoir solution and 0.4 µl PALAO solution (~0.01 M). The reservoir solution contained 20% (w/v) PEG 3350, 0.2 M lithium sulfate and 0.1 M bis-Tris, pH 6.0. Diffraction data for the PALAO-bound crystal were collected at 100 K at the F1 beam line of the Cornell High Energy Synchrotron Source. Data sets for the PALAO-bound mutant AOTCase crystals were obtained using a Rigaku anode x-ray generator in the Molecular Structure Section of the National Institute of Health. All data were processed using HKL2000 package (26) and reduced using the program TRUNCATE in the CCP4 suite (27). Data collection parameters are listed in Table 1. Li et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Structure solution and refinement Molecular replacement was used for phase determination of the PALAO-bound wild-type and mutant AOTCase structures. The coordinates of AOTCase (PDB code: 3KZO) after removal of ligands or water molecules were used for phase determination. Upon rigid-body refinement, electron density corresponding to the ligands could be clearly visualized. The ligands were built into the map using the graphics program O (28). Refinements were carried out using molecular annealing, energy minimization and restrained B factor refinement with the program CNS1.1 (29). During refinements, 5% of the reflections at various resolutions were randomly selected to set aside to calculate Rfree to monitor the progress of refinement (30). After every cycle of refinement, the model was manually adjusted using the program O (28). Water molecules were automatically assigned using the program WATERPICK of CNS. Model quality was checked using the program PROCHECK (31) to ensure good stereochemistry for all three models. The final refinement statistics are listed in Table 1. Figures 1 was drawn using the programs Pymol (http://www.pymol.org). Figure 3 was drawn using ChemDraw 8.0. The coordinates have been deposited with the RCSB PDB as entries 3M4J, 3M4N, 3M5C and 3M5D. RESULTS Lys302 in AOTCase is carboxylated To investigate the nature of the modification of Lys302 and how it affects catalytic activity, we revisited all AOTCase structures. In the PALAO-bound AOTCase structure, the electron density map clearly indicates that Lys302 is post-translationally modified (Figure 1A). The type of modification can include methylation, acetylation, carbamylation, and carboxylation. The shape of the electron density can been used to distinguish methyl groups from larger functional groups, but it is difficult to distinguish between acetyl, carbamyl, and carboxyl groups, all of which have three non-hydrogen atoms in a plane. Given the hydrogen bonding network with surrounding residues (Lys252, Ser253 and His293, Table 2), a carboxylated modification is the most likely choice for the modification of Lys302 in AOTCase. To exclude that the modification’s identity represents chemically stable moieties (methyl, acetyl, carbamyl), we analyzed trypsin digested fragments of purified AOTCase by TOF- TOF mass spectroscopy. As expected, only a peptide fragment with an unmodified Lys302 was observed, consistent with the lability of the carboxylic group in acidic solutions. At low pH, the carboxyl group is spontaneously released as carbon dioxide (14, 32), in contrast to other modified groups that are stably bound and can be observed by mass spectrometry analysis after proteolysis (33). The putative carboxyl group on the modified Lys302 forms direct hydrogen bonds with main-chain or side-chain nitrogen atoms of Lys252, Ser253 and His293 (Figure 1A and Table 2). Among these, Lys252 is involved in direct hydrogen bonding to the carboxyl group of the AORN moiety of PALAO, and His293 forms a strong hydrogen bond with the main-chain nitrogen atom of Leu295 in the conserved His293-Cys294-Leu295-Pro206 (HCLP) motif. The hydrogen bonding network between the carboxyl group of modified Lys302, His293 and the main-chain nitrogen atom of Leu295 is reminiscent of the similar hydrogen bonding network, Glu310-His302-Leu304 and Glu299-Leu272-Leu274, found in human and E. coli OTCase, respectively (34, 35). These three residues are conserved in all OTCase sequences, and the interactions between them are important for maintaining the HCLP motif in a specific conformation to orientate their main-chain oxygen atoms towards the active site. In all known transcarbamylase structures, a leucine residue corresponding to Leu295 is in an energetically unfavorable conformation and the peptide bond between this Li et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript leucine and Pro296 is in the cis conformation. In addition to the direct hydrogen bonding interaction above, the carboxyl group on the modified Lys302 interacts with the α-amino nitrogen atom of the AORN moiety of PALAO and Glu92 from the adjacent subunit via water molecules. When we revisited all previously determined AOTCase structures (see supplementary Figure S1) we found: (1) Lys302 was carboxylated in the absence of substrate binding, but substrate binding immobilizes the side-chain of Lys302 further by hydrogen bonding interaction via water molecules. (2) Water-mediated hydrogen bonding promotes interactions of carboxylated Lys302 with Glu92 from the adjacent subunit and the α-amino nitrogen atom of AORN. (3) Similarly to AOTCase, in the structure of SOTCase E92Z (Z = Ala, Ser, Pro, Val), mutant with N-succinyl-L-norvaline bound (22) the carboxylated Lys302 hydrogen bonds to the α-amino nitrogen atom and the succinyl carboxyl group of N- succinyl-L-norvaline via water molecules (Figure S1). To obtain direct, independent evidence for the carboxylation of Lys302, 13C NMR experiments were carried out with both wild-type protein and the K302A mutant. As observed for other proteins with carboxylated lysine (36,37), the strong 13C NMR signal at 164 ppm characteristic of a carboxyl group was clearly detectable in AOTCase wild-type protein labeled by 13C-bicarbonate, in contrast to the K302A mutant where the signal was weak (Figure 2). Since there are 17 other lysine residues in the protein, the weak signal seen for the K302A mutant might be caused by the adventitious carboxylation of another lysine with reduced pKa, as has been observed for the K392A mutant of the sensor domain of the BlaR protein (38). Functional and structural studies of Lys302 mutants To investigate the effect of lysine carboxylation on enzyme activity, Lys302 was mutated to alanine, glutamate or arginine. Each of these variants was expressed in E. coli and gave similar yields. Enzymatic assays demonstrated a significant decrease in enzymatic activity in all three mutants, reflecting the functional importance of Lys302 (Table 3). The level of enzymatic activity for the wild-type (WT) and three mutants was WT > K302A > K302E ≫ K302R. To determine the structural basis of these results, the WT and mutant enzymes bound with PALAO were crystallized and their structures were determined at 1.8–2.2 Å resolution. Only the K302R mutation had and appreciable effect on the structure of the protein. Since K302 is located near the AORN binding site, the mutations would weaken AORN binding to the active site. In the structure of the K302A mutant, three additional water molecules (labeled as w3, w4 and w5 in Figure 1B) replace the carboxylated lysine. The two water molecules (labeled w1 and w2 in Figure 1A–1D) that mediate the hydrogen bonding interaction of carboxylated Lys302 with PALAO and Glu92 from the adjacent subunit are also found in the K302A mutant structure. Furthermore, these water molecules maintain a similar hydrogen-bonding network to the wild-type enzyme. These results might explain why the K302A mutant retains significant catalytic activity (Table 3). To investigate whether adding short-chain carboxylic acids to the K302A mutant increases its activity as other enzymes (14, 15, 39, 40), the activity of the K302A mutant was measured in the presence of high formate and acetate concentration (0.5 M). Surprisingly, the activity of the K302A mutant was not significantly improved. The crystal structure of the K302A mutant soaking with the crystallization buffer in the presence of 0.5 M acetate was also determined (not shown) and it was observed that the same five water molecules were present in the cavity that replaced the side chain of the carboxylated lysine. This, the acetate’s inability to replace the water molecules in the crystal structure, is consistent with the unchanged activity assay results. Li et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The side-chain of Glu302 in the K302E mutant structure is well defined and anchored by hydrogen bonding interaction with the main-chain nitrogen atom of Arg298 and weakly hydrogen bonded to the main-chain nitrogen atom of Ser253 (Figure 1C). Two of three additional water molecules (w4 and w5) observed in the K302A mutant structure occupied the same position as the carboxyl oxygen atoms of Glu302 and form a similar hydrogen- bonding network. Relative to the PALAO-bound wild-type structure, there is only one more water molecule (w3) at the position of the carboxyl group of the carboxylated Lys302. This water molecule mediates a hydrogen bonding interaction between Glu302 and Lys252. Two common water molecules (w1 and w2) that interact with PALAO and Glu92 from the adjacent subunit, respectively, were also identified in the K302E structure. Our observation that the K302E mutant had lower enzymatic activity than that of the K302A mutant (Table 3) was surprising since the carboxyl group of the glutamate could conceivably function similarly to a carboxylated lysine. The explanation may be that, in the K302A mutant, the hydrogen bonding network is well maintained by water molecules in the cavity that replaces the carboxylated lysine. In particular, w3 is optimally located for strong hydrogen bonding to w1 (2.7 Å), which in turn binds AORN. The distances between w1 and the carboxyl oxygen of carboxylated Lys302 in all wild-type crystal structures are within 2.4–2.7 Å, but the distance between w1 and w2 in the K302E structure is significantly greater (3.2 Å). The weaker hydrogen bonding interaction may be a reason for lower enzymatic activity of the K302E mutant. In contrast to the K302A and K302E structures, the K302R structure shows a much larger reduction in enzyme activity relative to the wild-type enzyme. The electron density for the side-chain of Arg302 is weak and the temperature factor of its side-chain is 54.4 Å2, significantly higher than those of carboxylated Lys302 (44.7 Å2) and Glu302 (33.4 Å2), implying greater flexibility. Furthermore, the side-chain of Arg302 is oriented differently from the carboxyl group of carboxylated Lys302 and pushes the nearby residues His180, Pro181 and Lys182 outwards about 1.0 Å (Figure 1D). However, the water molecules involved in hydrogen bonding to the α-amino nitrogen atom of PALAO (w1) and the side- chains of Lys252 (w3) and Glu92 (w2) from the adjacent subunit are conserved. Consistent with the K302E structure, the distance between w1 and w2 is even greater (3.4 Å) than in the WT structure and the hydrogen bonding interaction between w2 and w3 is no longer observed. Thus, the almost undetectable enzymatic activity of the K302R mutant probably results from the changes at its active site, including the weakened hydrogen bonding network involved in substrate binding. DISCUSSION Several lines of evidence clearly indicate that Lys302 in AOTCase is carboxylated. First, the extra electron density indicates that the side-chain of Lys302 is modified. Second, the hydrogen bonding environment of Lys302 for hydrogen bonding interactions is compatible with a carboxyl group, but not for a positively charged lysine side-chain. Third, the modification is labile at low pH, since mass spectroscopy of samples prepared at low pH indicated that Lys302 was no longer modified. Fourth, the clear presence of the indicative 13C NMR signal at 164 ppm for wild-type protein and its absence in the K302A mutant confirms carboxylation of Lys302. It is well known that lysine carboxylation is non-enzymatic and reversible, while other post- translational modifications such as methylation, acetylation, and carbamylation are irreversible and detectable by mass spectroscopy. Furthermore, lysine methylation and acetylation usually require an enzyme-catalyzed reaction in vivo (41). Therefore, it is unlikely that such lysine modifications will be observed in recombinant proteins Li et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript overexpressed in a foreign host (e.g. E. coli). Lysine methylation can be achieved by using special chemicals in vitro, but these chemicals are not present in vivo. Lysine carbamylation and carboxylation use completely different mechanisms to form functionally different groups (Figure 3). Carbamylation can be achieved by cyanate produced from myeloperoxidase-catalyzed oxidation of thiocyanate, an anion abundant in plasma and increased in smokers, or from urea in the plasma. Lysine carboxylation, on the other hand, occurs readily in aqueous solution in the presence of carbon dioxide at a basic pH (32,42). Even though carbamylation and carboxylation use very different mechanisms, the two are confused in the literatures. Lysine carbamylation (or carbamoylation) is referred to in several publications (15,32,42–44), when the actual reaction is in fact carboxylation. The activity of the K302A mutant is almost half of that of the wild-type enzyme raising the question of why AOTCase retains a lysine in this position. Perhaps this lysine was maintained through evolution to distinguish AOTCase from SOTCase which uses N- succinyl-L-ornithine (SORN) rather than AORN (22), and OTCase which uses L-ornithine. An alternative explanation may be found in the very low activity of the K302R mutant. The side-chain of arginine has a positive charge while carboxylated lysine has a negative charge. The side chain of unmodified lysine is usually located in a similar position as that of arginine, as observed in the structure of UV damage endonuclease (14). It would be expected that the activity of AOTCase with an uncarboxylated lysine would be as low as the K302R mutant’s. It could further be surmised that, the respective organisms need to use carboxylation as a switch to turn “on” or “off” the arginine biosynthetic pathway. It has been well known that rubulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) in plant cells uses the carboxylation on Lys201 as a switch to turn the enzyme “on” during the day and “off” at night by removing the carboxyl group (45,46). Carbon dioxide and bicarbonate have been found to play an important biological role in modulating several biological processes including photosynthetic carbon fixation (47), pH homeostasis (48), carbon metabolism (49), activation of virulence in pathogenic organisms (50), sperm maturation (51), stimulation of mammalian G-protein-responsive adenylyl cyclase (52), and as an alarmone in Drosophila (53,54). Whether or not carboxylation of a key lysine in their related proteins is used as an underlying regulatory mechanism should be investigated further. There are 197 structures with carboxylated lysine residue (modified residue indicated as Kcx) in the Protein Data Bank (PDB). If structures with 90% identity are counted only once, there are still 52 unique structures remaining in this pool (Table 4). These proteins include hydantoinase (40,55), folylpolyglutamate synthase (43), UV damage endonuclease (14), OXA10, OXA-1 class D β-lactamase (38,56,57), urease (42), phosphotriesterase (58), dihydroorotase (59), dihydropyrimidinase (60), organophosphate hydrolase (61) and MurE and MurD ligases (44,62). In most of these proteins, the carboxylated lysine bridges two metal ions, similar to the role of glutamate or aspartate in proteins with two metal-binding sites (26 structures among 52). However, the urease apoenzyme can be activated in vitro only in the presence of carbon dioxide prior to nickel binding (63), suggesting that the carboxylated lysine may have other structural roles beyond binding metals. In some proteins such as β-lactamase, UV damage endonuclease, Rubisco, MurD and MurE ligase and BlaR signal transducer protein, a carboxylated lysine plays an essential catalytic role. More interestingly, in three structures (PDB ID: 1HL9, 1PU6 and 2UYN for fucosiadase, 3- methyladenine DNA glycosylase and TdcF protein of unknown function, respectively), the carboxylated lysines are located near the surface of proteins, presumably playing primarily a structure stabilizing role (64–66). Since the carboxyl group is labile at acidic pH, but easily formed in the presence of carbon dioxide at basic pH, the number of proteins with lysine carboxylation must be underestimated. Furthermore, the carboxylated lysine must be fixed in place by metal ions (either one or two) or hydrogen bonding with other protein residues (at least one). Therefore, any detection method involving denaturing the proteins will result Li et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript in release of the carboxyl group. With current technology, 13C NMR (38) and crystallography are the only methods that can detect this modification. However, these methods are not amenable to high-throughput investigations. The majority (49 out of 52 structures in the PDB) of known lysine carboxylation modifications were found to be located at or near the active site, probably because these sites receive the most attention. Revisiting the structures in PDB with more attention to surface lysines might reveal more structures with carboxylated lysines. In conclusion, we have shown that Lys302 in AOTCase is post-translationally modified by carboxylation and that this modification may be functionally important for enzymatic activity. Lysine carboxylation is likely to be a more common event than currently appreciated and may play a critical role in enzymatic activity and protein stability. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Abbreviations ACIT N-acteyl-L-citrulline ANOR N-acetyl-L-norvaline AORN N-acetyl-L-Ornithine AOTCase N-acetyl-L-ornithine transcarbamylase ATCase aspartate transcarbamlyase OTCase ornithine transcarbamylase CP carbamyl phosphate ORN L-ornithine PALAO Nδ-(phosphonacetyl)-Nα-acetyl-L-ornithine SORN N-succinyl-L-ornithine WT wild-type xc Xanthomonas campestris Acknowledgments We thank Dr. David Davies for facilitating our use of the diffraction equipment in the Molecular Structure Section of the National Institute of Health and Dr. Fred Dyda for help in data collection and processing, and Dr. Yui-Fai Lam in the University of Maryland for help in setting up NMR measurements. REFERENCES 1. Close P, Creppe C, Gillard M, Ladang A, Chapelle JP, Nguyen L, Chariot A. The emerging role of lysine acetylation of non-nuclear proteins. Cell Mol Life Sci. 2010; 67:1255–1264. [PubMed: 20082207] 2. Geiman TM, Robertson KD. Chromatin remodeling, histone modifications, and DNA methylation- how does it all fit together? J Cell Biochem. 2002; 87:117–125. [PubMed: 12244565] 3. An W. Histone acetylation and methylation: combinatorial players for transcriptional regulation. Subcell Biochem. 2007; 41:351–369. [PubMed: 17484136] Li et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 4. Yang F, Bian C, Zhu L, Zhao G, Huang Z, Huang M. Effect of human serum albumin on drug metabolism: structural evidence of esterase activity of human serum albumin. J Struct Biol. 2007; 157:348–355. [PubMed: 17067818] 5. Xu AS, Labotka RJ, London RE. Acetylation of human hemoglobin by methyl acetylphosphate. Evidence of broad regio-selectivity revealed by NMR studies. J Biol Chem. 1999; 274:26629– 26632. [PubMed: 10480863] 6. Ueno H, Pospischil MA, Manning JM. Methyl acetyl phosphate as a covalent probe for anion- binding sites in human and bovine hemoglobins. J Biol Chem. 1989; 264:12344–12351. [PubMed: 2745446] 7. Stavropoulos P, Nagy V, Blobel G, Hoelz A. Molecular basis for the autoregulation of the protein acetyl transferase Rtt109. Proc Natl Acad Sci U S A. 2008; 105:12236–12241. [PubMed: 18719104] 8. Bewley MC, Graziano V, Jiang J, Matz E, Studier FW, Pegg AE, Coleman CS, Flanagan JM. Structures of wild-type and mutant human spermidine/spermine N1-acetyltransferase, a potential therapeutic drug target. Proc Natl Acad Sci U S A. 2006; 103:2063–2068. [PubMed: 16455797] 9. Walter TS, Meier C, Assenberg R, Au KF, Ren J, Verma A, Nettleship JE, Owens RJ, Stuart DI, Grimes JM. Lysine methylation as a routine rescue strategy for protein crystallization. Structure. 2006; 14:1617–1622. [PubMed: 17098187] 10. Stark GRSW, Moore S. Reactions of the cyanante present in aqueous urea with amino acids and proteins. J Biol Chem. 1960; 235:3177–3181. 11. Bobb D, Hofstee BH. Gel isoelectric focusing for following the successive carbamylations of amino groups in chymotrypsinogen A. Anal Biochem. 1971; 40:209–217. [PubMed: 5550146] 12. Kraus LM, Kraus AP Jr. Carbamoylation of amino acids and proteins in uremia. Kidney Int Suppl. 2001; 78:S102–S107. [PubMed: 11168993] 13. Al-Dirbashi OY, Al-Hassnan ZN, Rashed MS. Determination of homocitrulline in urine of patients with HHH syndrome by liquid chromatography tandem mass spectrometry. Anal Bioanal Chem. 2006; 386:2013–2017. [PubMed: 17053917] 14. Meulenbroek EM, Paspaleva K, Thomassen EA, Abrahams JP, Goosen N, Pannu NS. Involvement of a carboxylated lysine in UV damage endonuclease. Protein Sci. 2009; 18:549–558. [PubMed: 19241382] 15. Dementin S, Bouhss A, Auger G, Parquet C, Mengin-Lecreulx D, Dideberg O, van Heijenoort J, Blanot D. Evidence of a functional requirement for a carbamoylated lysine residue in MurD, MurE and MurF synthetases as established by chemical rescue experiments. Eur J Biochem. 2001; 268:5800–5807. [PubMed: 11722566] 16. Cha J, Mobashery S. Lysine N(zeta)-decarboxylation in the BlaR1 protein from Staphylococcus aureus at the root of its function as an antibiotic sensor. J Am Chem Soc. 2007; 129:3834–3835. [PubMed: 17343387] 17. Shi D, Yu X, Roth L, Morizono H, Tuchman M, Allewell NM. Structures of N-acetylornithine transcarbamoylase from Xanthomonas campestris complexed with substrates and substrate analogs imply mechanisms for substrate binding and catalysis. Proteins. 2006; 64:532–542. [PubMed: 16741992] 18. Shi D, Morizono H, Yu X, Roth L, Caldovic L, Allewell NM, Malamy MH, Tuchman M. Crystal structure of N-acetylornithine transcarbamylase from Xanthomonas campestris: a novel enzyme in a new arginine biosynthetic pathway found in several eubacteria. J Biol Chem. 2005; 280:14366– 14369. [PubMed: 15731101] 19. Morizono H, Cabrera-Luque J, Shi D, Gallegos R, Yamaguchi S, Yu X, Allewell NM, Malamy MH, Tuchman M. Acetylornithine transcarbamylase: a novel enzyme in arginine biosynthesis. J Bacteriol. 2006; 188:2974–2982. [PubMed: 16585758] 20. da Silva FR, Vettore AL, Kemper EL, Leite A, Arruda P. Fastidian gum: the Xylella fastidiosa exopolysaccharide possibly involved in bacterial pathogenicity. FEMS Microbiol Lett. 2001; 203:165–171. [PubMed: 11583843] 21. da Silva AC, Ferro JA, Reinach FC, Farah CS, Furlan LR, Quaggio RB, Monteiro-Vitorello CB, Van Sluys MA, Almeida NF, Alves LM, do Amaral AM, Bertolini MC, Camargo LE, Camarotte G, Cannavan F, Cardozo J, Chambergo F, Ciapina LP, Cicarelli RM, Coutinho LL, Cursino-Santos Li et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript JR, El-Dorry H, Faria JB, Ferreira AJ, Ferreira RC, Ferro MI, Formighieri EF, Franco MC, Greggio CC, Gruber A, Katsuyama AM, Kishi LT, Leite RP, Lemos EG, Lemos MV, Locali EC, Machado MA, Madeira AM, Martinez-Rossi NM, Martins EC, Meidanis J, Menck CF, Miyaki CY, Moon DH, Moreira LM, Novo MT, Okura VK, Oliveira MC, Oliveira VR, Pereira HA, Rossi A, Sena JA, Silva C, de Souza RF, Spinola LA, Takita MA, Tamura RE, Teixeira EC, Tezza RI, Trindade dos Santos M, Truffi D, Tsai SM, White FF, Setubal JC, Kitajima JP. Comparison of the genomes of two Xanthomonas pathogens with differing host specificities. Nature. 2002; 417:459– 463. [PubMed: 12024217] 22. Shi D, Yu X, Cabrera-Luque J, Chen TY, Roth L, Morizono H, Allewell NM, Tuchman M. A single mutation in the active site swaps the substrate specificity of N-acetyl-L-ornithine transcarbamylase and N-succinyl-L-ornithine transcarbamylase. Protein Sci. 2007; 16:1689–1699. [PubMed: 17600144] 23. Shi D, Morizono H, Cabrera-Luque J, Yu X, Roth L, Malamy MH, Allewell NM, Tuchman M. Structure and catalytic mechanism of a novel N-succinyl-L-ornithine transcarbamylase in arginine biosynthesis of Bacteroides fragilis. J Biol Chem. 2006; 281:20623–20631. [PubMed: 16704984] 24. Pastra-Landis SC, Foote J, Kantrowitz ER. An improved colorimetric assay for aspartate and ornithine transcarbamylases. Anal Biochem. 1981; 118:358–363. [PubMed: 7337232] 25. Shi D, Yu X, Roth L, Morizono H, Hathout Y, Allewell NM, Tuchman M. Expression, purification, crystallization and preliminary X-ray crystallographic studies of a novel acetylcitrulline deacetylase from Xanthomonas campestris. Acta Crystallogr Sect F Struct Biol Cryst Commun. 2005; 61:676–679. 26. Otwinowski Z, Minor W. Processing of X-ray Diffraction Data Collected in Oscillation Mode. Methods in Enzymology. 1997; 276:307–326. 27. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr. 1994; 50:760–763. [PubMed: 15299374] 28. Jones TA, Zou JY, Cowan SW, Kjeldgaard M. Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr A. 1991; 47(Pt 2):110–119. [PubMed: 2025413] 29. Brunger AT, Adams PD, Clore GM, DeLano WL, Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges M, Pannu NS, Read RJ, Rice LM, Simonson T, Warren GL. Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr. 1998; 54:905–921. [PubMed: 9757107] 30. Brunger AT. Free R value: a novel statistical quantity for assessing the accuracy of crystal structures. Nature. 1992; 355:472–475. [PubMed: 18481394] 31. Laskowski RA, MacArthur MW, Moss DS, Thornton JM. PROCHECK: a program to check the stereochemical quality of protein structures. J Appl Crystallogr. 1993; 26:283–291. 32. Golemi D, Maveyraud L, Vakulenko S, Samama JP, Mobashery S. Critical involvement of a carbamylated lysine in catalytic function of class D beta-lactamases. Proc Natl Acad Sci U S A. 2001; 98:14280–14285. [PubMed: 11724923] 33. Lapko VN, Smith DL, Smith JB. In vivo carbamylation and acetylation of water-soluble human lens alphaB-crystallin lysine 92. Protein Sci. 2001; 10:1130–1136. [PubMed: 11369851] 34. Shi D, Morizono H, Ha Y, Aoyagi M, Tuchman M, Allewell NM. 1.85-A resolution crystal structure of human ornithine transcarbamoylase complexed with N-phosphonacetyl-L-ornithine. Catalytic mechanism and correlation with inherited deficiency. J Biol Chem. 1998; 273:34247– 34254. [PubMed: 9852088] 35. Langley DB, Templeton MD, Fields BA, Mitchell RE, Collyer CA. Mechanism of inactivation of ornithine transcarbamoylase by Ndelta -(N'-Sulfodiaminophosphinyl)-L-ornithine, a true transition state analogue? Crystal structure and implications for catalytic mechanism. J Biol Chem. 2000; 275:20012–20019. [PubMed: 10747936] 36. Cha J, Vakulenko SB, Mobashery S. Characterization of the beta-lactam antibiotic sensor domain of the MecR1 signal sensor/transducer protein from methicillin-resistant Staphylococcus aureus. Biochemistry. 2007; 46:7822–7831. [PubMed: 17550272] Li et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 37. O'Leary MH, Jaworski RJ, Hartman FC. C nuclear magnetic resonance study of the CO(2) activation of ribulosebisphosphate carboxylase from Rhodospirillum rubrum. Proc Natl Acad Sci U S A. 1979; 76:673–675. [PubMed: 16592618] 38. Golemi-Kotra D, Cha JY, Meroueh SO, Vakulenko SB, Mobashery S. Resistance to beta-lactam antibiotics and its mediation by the sensor domain of the transmembrane BlaR signaling pathway in Staphylococcus aureus. J Biol Chem. 2003; 278:18419–18425. [PubMed: 12591921] 39. Schneider KD, Bethel CR, Distler AM, Hujer AM, Bonomo RA, Leonard DA. Mutation of the active site carboxy-lysine (K70) of OXA-1 beta-lactamase results in a deacylation-deficient enzyme. Biochemistry. 2009; 48:6136–6145. [PubMed: 19485421] 40. Huang CY, Hsu CC, Chen MC, Yang YS. Effect of metal binding and posttranslational lysine carboxylation on the activity of recombinant hydantoinase. J Biol Inorg Chem. 2009; 14:111–121. [PubMed: 18781344] 41. Zhang Q, Wang Y. High mobility group proteins and their post-translational modifications. Biochim Biophys Acta. 2008; 1784:1159–1166. [PubMed: 18513496] 42. Jabri E, Carr MB, Hausinger RP, Karplus PA. The crystal structure of urease from Klebsiella aerogenes. Science. 1995; 268:998–1004. [PubMed: 7754395] 43. Young PG, Smith CA, Metcalf P, Baker EN. Structures of Mycobacterium tuberculosis folylpolyglutamate synthase complexed with ADP and AMPPCP. Acta Crystallogr D Biol Crystallogr D. 2008; 64:745–753. 44. Gordon E, Flouret B, Chantalat L, van Heijenoort J, Mengin-Lecreulx D, Dideberg O. Crystal structure of UDP-N-acetylmuramoyl-L-alanyl-D-glutamate: meso-diaminopimelate ligase from Escherichia coli. J Biol Chem. 2001; 276:10999–11006. [PubMed: 11124264] 45. Taylor TC, Andersson I. Structure of a product complex of spinach ribulose-1,5-bisphosphate carboxylase/oxygenase. Biochemistry. 1997; 36:4041–4046. [PubMed: 9092835] 46. Jensen R. Activation of Rubisco controls CO(2) assimilation in light: a perspective on its discovery. Photosynth Res. 2004; 82:187–193. [PubMed: 16151874] 47. Falkowski PG. Photosynthesis: the paradox of carbon dioxide efflux. Curr Biol. 1997; 7:R637– R639. [PubMed: 9368746] 48. Roos A, Boron WF. Intracellular pH. Physiol Rev. 1981; 61:296–434. [PubMed: 7012859] 49. Smith KS, Ferry JG. Prokaryotic carbonic anhydrases. FEMS Microbiol Rev. 2000; 24:335–366. [PubMed: 10978542] 50. Bahn YS, Muhlschlegel FA. CO2 sensing in fungi and beyond. Curr Opin Microbiol. 2006; 9:572– 578. [PubMed: 17045514] 51. Esposito G, Jaiswal BS, Xie F, Krajnc-Franken MA, Robben TJ, Strik AM, Kuil C, Philipsen RL, van Duin M, Conti M, Gossen JA. Mice deficient for soluble adenylyl cyclase are infertile because of a severe sperm-motility defect. Proc Natl Acad Sci U S A. 2004; 101:2993–2998. [PubMed: 14976244] 52. Townsend PD, Holliday PM, Fenyk S, Hess KC, Gray MA, Hodgson DR, Cann MJ. Stimulation of mammalian G-protein-responsive adenylyl cyclases by carbon dioxide. J Biol Chem. 2009; 284:784–791. [PubMed: 19008230] 53. Kwon JY, Dahanukar A, Weiss LA, Carlson JR. The molecular basis of CO2 reception in Drosophila. Proc Natl Acad Sci U S A. 2007; 104:3574–3578. [PubMed: 17360684] 54. Jones WD, Cayirlioglu P, Kadow IG, Vosshall LB. Two chemosensory receptors together mediate carbon dioxide detection in Drosophila. Nature. 2007; 445:86–90. [PubMed: 17167414] 55. Xu Z, Liu Y, Yang Y, Jiang W, Arnold E, Ding J. Crystal structure of D-Hydantoinase from Burkholderia pickettii at a resolution of 2.7 Angstroms: insights into the molecular basis of enzyme thermostability. J Bacteriol. 2003; 185:4038–4049. [PubMed: 12837777] 56. Sun T, Nukaga M, Mayama K, Braswell EH, Knox JR. Comparison of beta-lactamases of classes A and D: 1.5-A crystallographic structure of the class D OXA-1 oxacillinase. Protein Sci. 2003; 12:82–91. [PubMed: 12493831] 57. Maveyraud L, Golemi D, Kotra LP, Tranier S, Vakulenko S, Mobashery S, Samama JP. Insights into class D beta-lactamases are revealed by the crystal structure of the OXA10 enzyme from Pseudomonas aeruginosa. Structure. 2000; 8:1289–1298. [PubMed: 11188693] Li et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 58. Benning MM, Kuo JM, Raushel FM, Holden HM. Three-dimensional structure of the binuclear metal center of phosphotriesterase. Biochemistry. 1995; 34:7973–7978. [PubMed: 7794910] 59. Thoden JB, Phillips GN Jr, Neal TM, Raushel FM, Holden HM. Molecular structure of dihydroorotase: a paradigm for catalysis through the use of a binuclear metal center. Biochemistry. 2001; 40:6989–6997. [PubMed: 11401542] 60. Abendroth J, Niefind K, Schomburg D. X-ray structure of a dihydropyrimidinase from Thermus sp. at 1.3 A resolution. J Mol Biol. 2002; 320:143–156. [PubMed: 12079340] 61. Yang H, Carr PD, McLoughlin SY, Liu JW, Horne I, Qiu X, Jeffries CM, Russell RJ, Oakeshott JG, Ollis DL. Evolution of an organophosphate-degrading enzyme: a comparison of natural and directed evolution. Protein Eng. 2003; 16:135–145. [PubMed: 12676982] 62. Bertrand JA, Auger G, Martin L, Fanchon E, Blanot D, Le Beller D, van Heijenoort J, Dideberg O. Determination of the MurD mechanism through crystallographic analysis of enzyme complexes. J Mol Biol. 1999; 289:579–590. [PubMed: 10356330] 63. Park IS, Hausinger RP. Requirement of carbon dioxide for in vitro assembly of the urease nickel metallocenter. Science. 1995; 267:1156–1158. [PubMed: 7855593] 64. Sulzenbacher G, Bignon C, Nishimura T, Tarling CA, Withers SG, Henrissat B, Bourne Y. Crystal structure of Thermotoga maritima alpha-L-fucosidase. Insights into the catalytic mechanism and the molecular basis for fucosidosis. J Biol Chem. 2004; 279:13119–13128. [PubMed: 14715651] 65. Eichman BF, O'Rourke EJ, Radicella JP, Ellenberger T. Crystal structures of 3-methyladenine DNA glycosylase MagIII and the recognition of alkylated bases. Embo J. 2003; 22:4898–4909. [PubMed: 14517230] 66. Burman JD, Stevenson CE, Sawers RG, Lawson DM. The crystal structure of Escherichia coli TdcF, a member of the highly conserved YjgF/YER057c/UK114 family. BMC Struct Biol. 2007; 7:30. [PubMed: 17506874] Li et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Stereo view of the structure and hydrogen bonding network surrounding residue 302. A, PALAO bound wild-type AOTCase, B, PALAO bound K302A AOTCase C, PALAO bound K302E AOTCase, D, PALAO bound K302R AOTCase. Contours of the electron density maps (2Fo-Fc) around PALAO, residue 302 and water molecules are shown as a brown cage at 1.0σ. The final refined positions of the ligands and surrounding protein residues are represented as colored sticks. The predicted hydrogen bonding interactions are in pink dashed lines. The water molecules are represented as pink balls. The carbon of PALAO, residue 302 and other protein residues are shown in pink, light blue and green sticks, respectively. Li et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. 13C NMR spectra of wild-type (upper panel) and K302A mutant (lower panel) AOTCase (1 mM). Experiments were performed in 100 mM Tris HCl, 50 mM NaCl, 7% D2O, pH 8.0, supplemented with 20 mM NaH13CO3. The position of the resonance attributed to carboxylated lysine in the enzyme is around 164 ppm. Li et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Chemical structure of carbamylated vs. carboxylated lysine. Li et al. Page 16 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 17 Table 1 Data collection and refinement statistics Dataset PALAO K302A K302E K302R Space group I213 I213 I213 I213 Resolution (Å) 2.2 1.9 1.85 2.2 Unit-cell parameters (Å) a = b = c =128.88 a = b = c =128.92 a = b = c =129.29 a = b = c =127.39 Measurements 219,475 305,757 390,128 246,817 Unique reflections 18,269 (1,832) a 28,236 (1,365) 30,622 (1,456) 17,635 (879) Redundancy 12.0 (11.8) 10.8(5.4) 12.8 (5.4) 14.0 (13.1) Completeness (%) 99.8 (100.0) 100.0 (100.0) 99.7 (95.1) 100.0 (100.0) <I/σ (I)> 15.0 (4.9) 16.4 (2.3) 19.8 (2.8) 8.7 (3.7) Rmerg b 7.4 (48.4) 6.5(64.9) 5.2 (55.3) 9.8 (79.1) Wilson B (Å2) 30.4 27.6 28.6 21.9 Refinement Resolution range (Å) 50.0-2.2 50-1.9 50-1.85 50-2.2 No. of protein atoms 2620 2613 2617 2619 No. of water atoms 90 219 193 146 No. of hetero atoms 24 24 24 24 Rmsd of bond lengths (Å) 0.006 0.005 0.005 0.005 Rmsd of bond angle (°) 1.1 1.2 1.2 1.2 Rwork (%)c 20.0 19.8 20.0 18.9 Rfree (%)d 24.3 23.2 23.2 22.2 Average B factor (Å2) 41.7 32.2 32.3 35.3 aFigures in brackets apply to the highest-resolution shell. bRmerg = ΣhΣi|I(h,i)-<I(h)>|/∑hΣiI(h,i), where I(h,i) is the intensity of the ith observation of reflection h, and < I(h)> is the average intensity of redundant measurements of reflection h. cRwork= Σh‖Fobs| – |Fcalc‖/Σh|Fobs|. dRfree = Σh‖Fobs| – |Fcalc‖/Σh|Fobs| for 5% of the reserved reflections. Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 18 Table 2 Interactions between carboxylated lysine and other residues at the active site of AOTCase Kcx302 Other residues Bound ligands PALAO CPa AORNb CP +ANORc SO4+ACITd OQ1 K252 NZ 2.6 2.6 2.7 2.6 2.6 OQ1 W1e 2.6 2.6 2.6 2.7 OQ2 S253 N 3.0 3.1 2.8 2.9 2.9 OQ2 H293 NE2 3.0 3.2 3.0 2.9 2.9 NZ W2f 3.1 2.9 3.0 3.0 aThe values were calculated based on PDB ID 3KZM. bThe values were calculated based on PDB ID 3KZN. cThe values were calculated based on PDB ID 3KZO. dThe values were calculated based on PDB ID 3KZK. eThis water molecule hydrogen bonds to N1 atom of PALAO, AORN or ANOR, and backbone O atom of Pro296 as well. fThis water molecule hydrogen bonds to OE1 atom of Glu92 from adjacent subunit as well. Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 19 Table 3 Specific activity of wild-type and mutant AOTCase in the presence of acids (0.5M). Compounds added Specific activity(µmol/min/mg) Wild-type K302A K302E K302R None 43.4 ± 0.4a 23.0 ± 0.5 7.1 ± 0.1 0.059±0.01 Formate 44.1 ± 1.2 26.4 ± 0.6 6.7 ± 0.2 0.093±0.01 Acetate 48.5 ± 1.1 21.2 ± 0.8 6.6 ± 0.5 0.104±0.03 aThe Mean ± S.D. are shown (n = 3). Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 20 Table 4 Protein structures with lysine carboxylation modification PDB ID Enzyme name Residue Organism source Funciton 2OGJ Dihydroorotase 175 A.tumefaciens Bridging two Zn(II) 2Z26 Dihydroorotase 102 E.coli Bridging two Zn(II) 3JZE Dihydroorotase 103 S.enterica Bridging two Zn(II) 2GWN Dihydroorotase 149 P. gingivalis Bridging two Zn(II) 3F4C Organophosphorus hydrolase 243 G. stearothermophilus Bridging two Co(II) 3ICJ Metal-dependent hydrolase 294 P. furiosus Bridging two Zn(II) 3GTX Organophosphorus hydrolase 243 D. radiodurans Bridging two Co(II) 2QPX Metal-dependent hydrolase 166 L. casei Bridging two Zn(II) 2FTW Dihydropyrimidinase 158 D. discoideum Bridging two Zn(II) 2FVK Dihydropyrimidinase 167 S. kluyveri Bridging two Zn(II) 3DC8 Dihydropyrimidinase 147 S. meliloti Bridging two Zn(II) 3GNH L-Lys/Arg carboxypeptidase 211 C. crescentus cb15 Bridging two Zn(II) 3DUG Arginine carboxypeptidase 182 Unidentified Bridging two Zn(II) 2VC7 Phosphotriesterase 137 S. solfataricus Bridging two Co(II) 2R1N Metallophosphotriesterases 169 A. tumefaciens Bridging two Co(II) 2OB3 Phosphotriesterase 169 B. diminuta Bridging two Zn(II) 3E74 Allantoinase 146 E. coli Bridging two Fe(III) 1EJX Urease 217 K. aerogenes Bridging two Ni(II) 1E9Z Urease 219 H. pylori Bridging two Ni(II) 4UBP Urease 220 B. pasteurii Bridging two Ni(II) 1ONW Isoaspartyl dipeptidase 162 E. coli Bridging two Zn(II) 1K1D D-hydanroinase 150 G. stearothermophilus Bridging two Zn(II) 1GKR L-hydanroinase 147 A. aurescens Bridging two Zn(II) 1GKP D-hydanroinase 150 Thermus sp. Bridging two Zn(II) 1NFG D-hydantoinase 148 R. pickettii Bridging two Zn(II) 2ICS Adenine deaminase 154 E. faecalis Bridging two Zn(II) 1RQB Transcarboxylase 184 P. freudenreichii Binding one Co(II) 2QF7 Pyruvate carboxylase 718 R. etli Binding one Zn(II) 3BG3 Pyruvate carboxylase 741 H. sapiens Binding one Mn(II) 2OEM Rubisco-like protein 173 G. kaustophilus Binding one Mg(II) 1WDD Rubisco 201 O. sativa Binding one Mg(II) 1GK8 Rubisco 201 C. reinhardtii Binding one Mg(II) 1BWV Rubisco 201 G. partita Binding one Mg(II) 2WTZ ATP-dependent MurE ligase 262 M. tuberculosis Binding one Mg(II) 2JFG MurD ligase 198 E. coli Catalytic role? 1E8C MurE ligase 224 E. coli Catalytic role? 1JBW Folypolyglutamate synthetase 185 L. casei Catalytic role? 1W78 FolC bifunctional protein 188 E. coli Binding one Mg(II) 3HBR OXA-48 β-lactamase 73 K. pneumoniae Catalytic role Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 21 PDB ID Enzyme name Residue Organism source Funciton 3ISG Class D β-lactamase 70 E. coli Catalytic role 2P9V AmpC beta-lactamase 315 E. coli Catalytic role 1K55 OXA-10 β-lactamase 70 P. aeruginosa Catalytic role 1K38 β-lactamase OXA-2 70 S. typhimurium Catalytic role 1XQL Alanine racemase 129 G. stearothermophilus Binding substrate? 1VFS Alanine racemase 129 S. lavendulae Binding substrate? 1RCQ Alanine racemase 122 P. aeruginosa Binding substrate? 2J6V UV damage endonuclease 229 T. thermophilus Catalytic role 1H01 Cell division protein kinase 2 33 H. sapiens Catalytic role? 2UYN Protein TdcF A58 E. coli Structural role? 1HL9 Fucosidase 338 T. maritime Structural role? 1PU6 3-methyladenine DNA glycosylase 205 H. pylori Structural role? Biochemistry. Author manuscript; available in PMC 2011 August 17.
3M4O
RNA polymerase II elongation complex B
X-ray structure and mechanism of RNA polymerase II stalled at an antineoplastic monofunctional platinum-DNA adduct Dong Wanga,b,1, Guangyu Zhuc, Xuhui Huangd, and Stephen J. Lippardc,1 aDepartment of Structural Biology, Stanford University School of Medicine, Stanford, CA 94305; bSkaggs School of Pharmacy and Pharmaceutical Sciences, University of California, San Diego, La Jolla, CA 92093; cDepartment of Chemistry, Massachusetts Institute of Technology, Cambridge, MA 02139; and dDepartment of Chemistry, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, P.R. China Contributed by Stephen J. Lippard, March 3, 2010 (sent for review February 9, 2010) DNA is a major target of anticancer drugs. The resulting adducts interfere with key cellular processes, such as transcription, to trigger downstream events responsible for drug activity. cis- Diammine(pyridine)chloroplatinum(II), cDPCP or pyriplatin, is a monofunctional platinum(II) analogue of the widely used antican- cer drug cisplatin having significant anticancer properties with a different spectrum of activity. Its novel structure-activity properties hold promise for overcoming drug resistance and improving the spectrum of treatable cancers over those responsive to cisplatin. However, the detailed molecular mechanism by which cells process DNA modified by pyriplatin and related monofunctional complexes is not at all understood. Here we report the structure of a transcri- bing RNA polymerase II (pol II) complex stalled at a site-specific monofunctional pyriplatin-DNA adduct in the active site. The re- sults reveal a molecular mechanism of pol II transcription inhibition and drug action that is dramatically different from transcription in- hibition by cisplatin and UV-induced 1,2-intrastrand cross-links. Our findings provide insight into structure-activity relationships that may apply to the entire family of monofunctional DNA-damaging agents and pave the way for rational improvement of monofunc- tional platinum anticancer drugs. anticancer ∣chemotherapy ∣DNA damage ∣pyriplatin ∣transcription T he DNA template for transcription is not only the site of in- born errors of metabolism and of continuous attack by harm- ful environmental agents, but it also represents a major target for cancer therapy. Platinum-based anticancer drugs such as cisplatin, cis-diamminedichloroplatinum(II), are widely used and among the most effective antineoplastic treatments (1, 2). Platinum-based drugs typically form bifunctional intra- or inter- strand DNA cross-links by covalent bonding to the N7 positions of two guanosine residues, triggering a variety of cellular processes, including transcription inhibition with attendant apoptosis (1, 2). However, resistance and side effects can require with- drawal of these drugs before they can effect a cure in certain types of cancer (3). In the effort to find new compounds that circumvent resis- tance to conventional bifunctional platinum-based drugs, a class of monofunctional platinum compounds were synthesized and screened for anticancer activity (4–6). In contrast to other inactive monofunctional platinum(II) compounds such as ½PtðdienÞClþ and ½PtðNH3Þ3Clþ, cis-diammine(pyridine)chloro- platinum(II) [cDPCP or “pyriplatin” (Fig. 1)] and related com- plexes display significant anticancer properties and a different spectrum of activity compared to conventional platinum-based drugs. These features render them attractive candidates for treat- ing cisplatin-refractory patients if the potency could be raised to or beyond the level of that of cisplatin (4, 5, 7). Pyriplatin exhibits unique chemical and biological properties, forming monofunc- tional DNA adducts (Fig. 1 and Fig. S1) that can inhibit transcrip- tion and better elude DNA repair (7). The x-ray crystal structure of pyriplatin bound to a DNA duplex reveals substantially different features than those of DNA adducts formed by conven- tional, bifunctional platinum-based drugs. The overall DNA duplex is much less distorted, with the pyridine ligand of the cis-fPtðNH3Þ2ðpyÞg2þ moiety directed toward the 50-end of the platinated strand. A hydrogen bond forms between the NH3 ligand trans to pyridine and O6 of the platinated guanosine residue (7). The detailed molecular mechanism by which cells process DNA modified by monofunctional complexes such as pyriplatin is not understood. Several important questions remain unan- swered. By what process do monofunctional adducts block pol II transcription? Does the mechanism differ from that of tran- scription inhibition by 1,2- and 1,3-intrastrand cross-links that comprise the major adducts of cisplatin? Why do pyriplatin and its homologues, which violate the classical structure-activity relationships (SARs) for active, bifunctional platinum drugs (8), show such promise by comparison to related monofunctional complexes like ½PtðNH3Þ3Clþ? Would knowledge of the struc- ture of pyriplatin-modified DNA at its site(s) of biological action inform the design of more potent analogues? In the present work we take a combined biochemical and x-ray structural approach to investigate the molecular mechanism of pol II transcription inhibition by a site-specific monofunctional platinum(II)-DNA adduct of pyriplatin. An unprecedented mo- lecular mechanism for pol II transcription inhibition is revealed, providing insight into structure-activity relationships that may ap- ply to the entire family of monofunctional DNA-damaging agents, whether or not they contain platinum. Results A Different Configuration of a Pyriplatin-DNA Adduct Accommodated in the Pol II Active Site. To understand how a monofunctional pyriplatin-DNA adduct is accommodated in the active site of the transcribing pol II elongation complex, we designed and pre- pared a DNA template containing a site-specific DNA lesion of this complex, as described previously (7). A transcribing pol II complex was then assembled in which the pyriplatin-DNA lesion occupies the active (þ1) site (Complex B, Table 1). The crystal structure of this complex reveals that the platinated nucleotide is captured as a pol II complex in the post-translocation state, in which the addition site is empty and ready for NTP loading (Dashed Ring, Fig. 2A and Fig. S2). Fig. 2A reveals that the Author contributions: D.W. and S.J.L. designed research; D.W., G.Z., and X.H. performed research; D.W., G.Z., X.H., and S.J.L. analyzed data; and D.W., X.H., and S.J.L. wrote the paper. The authors declare no conflict of interest. Data deposition: The atomic coordinates have been deposited in the Protein Data Bank, www.pdb.org (PDB ID codes 3M4O and 3M3Y). 1To whom correspondence may be addressed. E-mail: dongwang@ucsd.edu or lippard@ mit.edu. This article contains supporting information online at www.pnas.org/cgi/content/full/ 1002565107/DCSupplemental. 9584–9589 ∣PNAS ∣May 25, 2010 ∣vol. 107 ∣no. 21 www.pnas.org/cgi/doi/10.1073/pnas.1002565107 positioning of the pyriplatin-damaged guanosine residue is lo- cated above the bridge helix. This structure requires rotation of the cis-fPtðNH3Þ2ðpyÞg2þ moiety and its bound guanosine re- sidue into a different configuration compared to that adopted in the pyriplatin-duplex DNA structure, in order to avoid a steric clash with bridge helix (7). Fig. 2B depicts this comparison. The rotation is energetically facilitated by the formation of hydro- gen bonds between the ammine ligands on platinum with the phosphodiester moiety of the backbone between positions þ1 and þ2, with concomitant loss of a hydrogen bond between O6 of the platinated guanosine residue and an ammine ligand. An additional feature is that the pyridine group of the cis- fPtðNH3Þ2ðpyÞg2þ fragment, which points downstream toward the 50-direction of the template DNA, forms van der Waals inter- actions with bridge helix residues Val 829 and Ala 832. The purine base of the guanosine residue at position þ1 is displaced toward the major groove of the RNA–DNA duplex by comparison with structures having an undamaged base at this site in the post-trans- location state (9–11). Transcription Elongation Inhibited by a Pyriplatin–DNA Adduct. Be- cause transcription inhibition is an important component in the mechanism of action of platinum anticancer drugs (12–20), we investigated the effect of a site-specific pyriplatin–DNA ad- duct on the kinetics of pol II transcription elongation. We per- formed an extension assay using platinated (Complex A, Table 1) and unplatinated (Complex A0, control, Table 1) pol II transcribing complexes having a 9mer RNA as primer. These complexes were then incubated with a mixture of ATP, CTP, and GTP. The RNA transcripts in A could be elongated from the 9mer to the 11mer, stopping at a position corresponding to the Pt– DNA lesion site observed in the pol II complex of the damaged template DNA, whereas RNA transcripts in A0 were extended much farther downstream on the undamaged template control DNA (Fig. 3A). In order to avoid the possibility of misincorpora- tion-induced transcription inhibition in this assay, we carried out a similar extension assay using an RNA containing a 30-end CMP matched against the damaged base (pol II complex C, 11mer) (Table 1). A single matching GTP was incubated with this pol II complex to test whether the enzyme could bypass the Pt– DNA lesion. Consistent with the results of the previous assay, RNA transcripts could not be extended beyond an 11mer in the pol II complex with the damaged DNA template, whereas RNA transcripts were efficiently extended farther downstream along the undamaged DNA template (Fig. 3B). Similar extension assay results were obtained using a chain-terminated GTP analo- gue 30-dGTP or an RNA primer of different length (complex B, 10mer) (Table 1) (Fig. 3 C and D). Finally, to investigate whether the presence of the damaged base affects the rate of NTP incor- poration in a single round, we used complex B (10mer) and com- plex C (11mer), incubating with CTP and 30-dGTP, respectively. For CTP incorporation, RNA transcripts could be efficiently ex- tended from the 10mer to the 11mer using both damaged and nondamaged templates at a comparable rate (Fig. 3E), whereas no obvious extension of RNA transcripts from the 11mer to a 12mer was observed on the damaged DNA template (Fig. 3C). UTP failed to incorporate at the damaged template under the same conditions (Fig. S3A). No obvious intrinsic cleavage was observed for a pol II complex containing the 11mer RNA and Pt-damaged DNA template in the presence of 20 mM Mg2þ ion, suggesting that most of complex C (11mer) is not in the back- tracked state (Fig. S3B) (21–23). X-ray Structure of Pol II stalled at a Pyriplatin–DNA Adduct. To under- stand the nature of the pol II complex stalled at the pyriplatin- induced Pt–DNA adduct, we solved the x-ray crystal structure Fig. 1. Scheme depicting the formation of a monofunctional platinum-DNA adduct by pyriplatin on double-stranded duplex DNA. The structure of the pyriplatin-damaged DNA duplex used coordinates from the PDB (code 3CO3). The damaged and nondamaged DNA strands are shown in cyan and green, respectively. The pyridine ligand and two ammine groups of the cis-fPtðNH3Þ2ðpyÞg2þ moiety are depicted in magenta and blue, respec- tively. The platinum atom and nitrogen atoms of the cis-fPtðNH3Þ2ðpyÞg2þ moiety are highlighted in yellow and as a blue ball, respectively. The termini of the DNA strands are labeled. A B +1 -1 +1 -1 5’ 3’ 3’ 5’ 5’ 3’ Non-template DNA Bridge Helix Bridge Helix Addition Site Addition Site 5’ V829 A832 RNA RNA Template DNA Template DNA 3’ Fig. 2. Structure of a pol II transcribing complex encountering a site-specific pyriplatin-dG adduct in DNA. (A) A site-specific pyriplatin-DNA adduct is ac- commodated in the pol II active site. The view is a standard one, from the “Rpb2 side,” as described elsewhere (9–11, 39). The RNA transcript, template DNA strand, and nontemplate DNA strand are depicted in red, cyan, and green, respectively. Parts of the bridge helix (Rpb1 825–848) are shown in gray. The pyriplatin-damaged guanosine is colored magenta. The platinum atom of the cis-fPtðNH3Þ2ðpyÞg2þ moiety is denoted as a yellow ball and the two ammine groups are in blue. The dashed oval represents the empty nucleotide addition site in the post-translocation state. The positions of the RNA strand are labeled. (B) cis-fPtðNH3Þ2ðpyÞg2þ-dG in the pol II active site adopts a different configuration in comparison with its conformation in the structure of pyriplatin-modified duplex DNA. The superimposed geometry of the cis-fPtðNH3Þ2ðpyÞg2þ-guanosine unit from the DNA duplex structure (3CO3) is shown in light blue. Side chains of Val 829 and Ala 832 are depicted in orange. The remainder of the figure is the same as in A. Wang et al. PNAS ∣ May 25, 2010 ∣ vol. 107 ∣ no. 21 ∣ 9585 BIOCHEMISTRY of the enzyme in complex with a platinated DNA using an RNA- containing CTP matched against the damaged guanosine residue. In this structure, pol II is in pre-translocation state, with the newly added CMP still occupying the addition site without transloca- tion. The platinated guanosine residue forms Watson–Crick base pairs with the newly added CMP (Fig. 4 A and B and Fig. S4). The cis-fPtðNH3Þ2ðpyÞg2þ moiety is surrounded by the bridge helix at the bottom, part of the Rpb2 fork region (528–534) on the left side, and the sugar-phosphate backbone connecting template DNA positions þ1 and þ2 on the right side (Fig. 4B). Interest- ingly, upon CMP incorporation, the cis-fPtðNH3Þ2ðpyÞg2þ moiety adopts a different conformation. The pyridine group of this unit now faces toward 30-direction of template DNA (Fig. 4 A and B). The ammine group trans to pyridine is directed toward the bridge helix and forms hydrogen bonds with main chain atoms of Ala 828 and the side chain of Thr 831 (Fig. 4B). The residues in the bridge helix are highly conserved from yeast to humans. Because Thr 831 and Ala 828 are absolutely conserved between S. cerevisiae and humans, the interactions we observe in the S. cerevisiae pol II structure will also occur in human pol II. These structural results provide important insights into the transcription stalling process at a monofunctional pyriplatin– DNA adduct. The adduct adopts a significantly different confor- mation within the pol II active site compared to that in duplex DNA (7). The present structural and biochemical evidence reveals that pol II stalls after efficient incorporation of CTP against the damaged guanosine residue. The conformation of the pyriplatin–DNA adduct changes significantly upon incorpora- tion of CTP. The modified guanosine rotates into the pol II active site and serves as a template for base pairing with the matched substrate, and the cis-fPtðNH3Þ2ðpyÞg2þ moiety is now directed toward 30-end of the platinated DNA. Table 1. RNA and DNA scaffold of pol II transcribing complexes Complex A: (Damaged template 29mer with 9mer RNA) RNA: 5′ AUGGAGAGG 3′ DNA: 3′ CTACCTCTCCTG*CCCCACCAATACCCATC 5′ DNA: 5′ GTGGTTATGGGTAG 3′ Complex A′: (Nondamaged template 29mer with 9mer RNA) RNA: 5′ AUGGAGAGG 3′ DNA: 3′ CTACCTCTCCTGCCCCACCAATACCCATC 5′ DNA: 5′ GTGGTTATGGGTAG 3′ Complex B: (Damaged template 29mer with 10mer RNA) RNA: 5′ AUGGAGAGGA 3′ DNA: 3′ CTACCTCTCCTG*CCCCACCAATACCCATC 5′ DNA: 5′ GTGGTTATGGGTAG 3′ Complex B′: (Nondamaged template 29mer with 10mer RNA) RNA: 5′ AUGGAGAGGA 3′ DNA: 3′ CTACCTCTCCTGCCCCACCAATACCCATC 5′ DNA: 5′ GTGGTTATGGGTAG 3′ Complex C: (Damaged template 29mer with 11mer RNA) RNA: 5′ AUGGAGAGGAC3′ DNA: 3′ CTACCTCTCCTG*CCCCACCAATACCCATC 5′ DNA: 5′ GTGGTTATGGGTAG 3′ Complex C′: (Non-damaged Template 29mer with 11mer RNA) RNA: 5′ AUGGAGAGGAC3′ DNA: 3′ CTACCTCTCCTGCCCCACCAATACCCATC 5′ DNA: 5′ GTGGTTATGGGTAG 3′ G*: cDPCP-dG. Fig. 3. Pol II transcription elongation blocked by a site-specific pyriplatin- DNA adduct. (A) In vitro transcription with preformed pol II elongation com- plexes A and A0 incubated with a mixture of ATP, CTP, and GTP (25 μM each). Time points were taken after 0, 0.5, 1, 2, 3, 4, 8, 16, 32, or 64 min incubation. The RNA transcripts in lanes 1–10 were taken from reactions of the pol II com- plex with a nondamaged DNA template, whereas the RNA transcripts in lanes 11–20 were taken from reactions of the pol II complex with a site-specifically damaged DNA template. The stalled RNA transcript is indicated by a black arrow (Right), and the extended RNA transcript is visible to the left. The length and sequences of RNA transcripts are given at the left margin of the gel. (B) In vitro transcription with preformed pol II elongation complexes C and C’ incubated with 25 μM GTP. The remainder of gel is the same as in A. (C) In vitro transcription with preformed pol II elongation complexes C and C’ incubated with 25 μM of 30-dGTP. The rest of gel is same as in A. (D) In vitro transcription with preformed pol II elongation complexes B and B’ incubated with a mixture of 25 μM CTP and 30-dGTP. Time points were taken after 0, 0.5, 1, 2, 4, 8, 16, 32, 64 min of incubation. The rest of gel is the same as in A. (E) In vitro transcription with preformed pol II elongation complexes B and B’ incubated with 25 μM CTP. The remainder of the gel is the same as in A. 9586 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1002565107 Wang et al. The result is that the RNA transcript fails to extend beyond the site of damage, subsequent translocation and nucleotide addition being strongly inhibited. Several factors contribute to such translocation inhibition, including (i) stabilization of the initial pre-translocation state by interaction between the platinated guanosine and pol II residues (Fig. 4B); (ii) a high translocation energy barrier; and (iii) an unfavorable subsequent post-translo- cation state induced by the DNA lesion. Hydrogen bonding inter- actions between an ammine group of the cis-fPtðNH3Þ2ðpyÞg2þ moiety with bridge helix partially help to stabilize the initial pre- translocation state (Fig. 4B). To address the factors ii and iii, we modeled the pyriplatin-damaged guanosine residue at the −1 po- sition to mimic the state following translocation of the pyriplatin- modified guanosine from the þ1 to −1 position. The structure clearly reveals that the cis-fPtðNH3Þ2ðpyÞg2þ moiety serves as a strong steric block, narrowing the space between the DNA nucleotide base (−1) and the bridge helix and preventing the downstream undamaged nucleoside base on the DNA template strand from rotating into the canonical þ1 position (Fig. 5A). Moreover, the fact that the cis-fPtðNH3Þ2ðpyÞg2þ moiety at the −1 position sterically clashes with the downstream nucleotide base at the þ1 position suggests that this final state is unfavorable (Fig. 5 A and B). In summary, our results indicate that pyriplatin– DNA adducts inhibit pol II transcription elongation by prevent- ing subsequent translocation and nucleotide addition beyond the site of damage. Discussion Insights into Structure-Activity Relationships (SARs) for the Monofunc- tional Platinum Drug Family. The original SARs pertaining to bifunctional platinum compounds such as cisplatin (8) were for- mulated to explain why anticancer activity appeared to require neutral, cis-[PtA2X2] compositions, in which A is an amine ligand and X is a monoanionic leaving group. These rules are clearly violated by cationic, monofunctional platinum compounds such as pyriplatin (4, 5). Other monofunctional platinum complexes, including ½PtðdienÞClþ, ½PtðNH3Þ3Clþ, and trans-½PtðNH3Þ2 ðpyÞClþ, are inactive and do not arrest pol II transcription, whereas the cis-fPtðNH3Þ2ðpyÞg2þ unit bound to guanosine blocks pol II transcription and has significant anticancer proper- ties in mice when administered as pyriplatin (4, 5, 8, 24–32). The present structure of pol II in complex with DNA site- specifically modified by pyriplatin provides unique insight into SARs to be expected for monofunctional platinum drug candi- dates. We constructed models of potential stalled transcription complexes containing DNA modified by the following three representative units, fPtðNH3Þ3g2þ, trans-fPtðNH3Þ2ðpyÞg2þ, and cis-fPtðNH3Þ2ðpyÞg2þ bound to guanosine in DNA and posi- tioned in either the −1 or þ1 site of pol II, in order to mimic the A B +1 -1 +1 -1 Bridge Helix Bridge Helix Rpb2 528-534 +1 5’ -1 3’ 5’ 3’ T831 A828 5’ 3’ 5’ 3’ Non-template DNA 5’ 3’ +2 RNA Template DNA Template DNA RNA 3.9 Å 3.9 Å Fig. 4. Structure of pol II transcribing complex stalled at a site-specific pyriplatin-DNA adduct after CMP incorporation. (A) The newly incorporated matched CMP is highlighted in yellow. Other colors are as in Fig. 2. Interac- tions of the damaged nucleotide and pol II residues are highlighted in (B). The view is taken roughly from an ∼90 degree clockwise rotation along the RNA/DNA helix axis from A. Nitrogen and oxygen atoms are depicted in blue and red, respectively. Hydrogen bonds between ammine group of the cis-fPtðNH3Þ2ðpyÞg2þ moiety and bridge helix residues are shown as black dashed lines. The loop of Rpb2 828–834 is shown in green. X A +1 -1 -2 X +1 -1 -2 RNA Template DNA Bridge Helix Non-template DNA Addition Site 3’ 5’ 5’ 5’ 3’ 3’ +1 -1 -1 +1 +2 -2 Bridge Helix Addition Site 3’ 3’ 5’ Template DNA RNA 5’ 3’ B Fig. 5. Pol II translocation following CMP incorporation is inhibited by a site- specific pyriplatin-DNA adduct. (A) The cis-fPtðNH3Þ2ðpyÞg2þ-guanosine unit is superimposed with a nucleoside in −1 position shown in magenta and as a surface view. In the latter, the nitrogen and oxygen atoms are highlighted in blue and red, respectively. CMP at the 30-end of RNA chain is highlighted in yellow. The bridge helix is shown in gray as a surface view. The nucleosides at the þ1 and þ2 position of the template DNA are drawn in wheat and orange, respectively. The rotation of the downstream nucleoside base during trans- location, from the þ2 position to the þ1 position, is blocked by the cis- fPtðNH3Þ2ðpyÞg2þ moiety, as indicated. Other colors are as in Fig. 2. (B) The cis-fPtðNH3Þ2ðpyÞg2þ moiety of pyriplatin-dG adduct modeled at −1 posi- tion clashes with the base at the þ1 position. Colors are as in A, and the view as in Fig. 4B. Wang et al. PNAS ∣ May 25, 2010 ∣ vol. 107 ∣ no. 21 ∣ 9587 BIOCHEMISTRY post- and pre-translocation states, respectively (Fig. S1). For each modeled structure, we rotated the platinum unit about the Pt-N7 bond by 360° and computed van der Waals energies arising from contacts between platinum ligands and the rest of the pol II complex (Figs. S5–S9). The fPtðNH3Þ3g2þ and trans-fPtðNH3Þ2ðpyÞg2þ moieties could be readily accommo- dated within the pol II active site over wide energy minima. The lack of a significant steric clash for these two groups, in either the −1 or þ1 position of the pol II transcribing complex, indicates the absence of a barrier to transcriptional bypass (Figs. S6–S9). This finding agrees with experiment. In contrast, the energy bar- rier is prohibitively high for cis-fPtðNH3Þ2ðpyÞg2þ platinated DNA modeled at −1 position, which is consistent with its ability to block pol II bypass and the failure of pol II to reach the sub- sequent post-translocation state (Figs. S5 and S8). The presence of a pyridine or other bulky group in the cis configuration is important for restricting the rotation range of the cis- fPtðNH3Þ2ðpyÞg2þ moiety and thus rendering it a strong steric block to translocation. For a fPtðNH3Þ3g2þ or trans-fPtðNH3Þ2 ðpyÞg2þ adduct at the −1 position, such a steric clash can be avoided by rotation about the Pt-N7 bond, facilitating subsequent pol II translocation. These results are fully consistent with previous biochemical studies revealing that the latter two DNA adducts are inactive and fail to block transcription (5, 7, 12, 26–33). A Unique Molecular Mechanism of Pol II Transcription Inhibition. The stalling mechanism of monofunctional platinum drugs of the pyriplatin family is dramatically different from transcription inhi- bition by cisplatin and UV-induced 1,2-intrastrand cross-links. For the latter two DNA-modifications, a translocation barrier prevents delivery of damaged bases to the active site and/or leads to misincorporation of NTPs against the damage site, respectively (19, 34). Monofunctional platinum-damaged residues, on the other hand, can cross over the bridge helix and be accommodated in the pol II active site. For Pt–dG adducts, the correct CMP nucleotide can be efficiently incorporated against the damaged guanosine. It is blockage of the subsequent translocation from this position after incorporation of the cytosine nucleotide that leads to inhibition of the RNA polymerase, but only when a bulky pyridine ligand is present in the cis coordination site. In conclusion, we report here the structure of a pol II transcri- bing complex stalled at a site-specific monofunctional DNA adduct, revealing a unique mechanism of transcription inhibition by this kind of genome damage. The results establish a basis for SARs that govern the anticancer drug potential of monofunc- tional platinum-based DNA-damaging agents. Specific inter- actions between pol II active site residues and the platinum ligands are revealed, providing a structural framework for rational design of more potent monofunctional pyriplatin analo- gues. Because the spectrum of activity of pyriplatin is dramatically different from that of cisplatin against an extensive panel of can- cer cell lines but with reduced potency (7), this information will be valuable for increasing the anticancer drug potential of this family of compounds based on pol II stalling with concomitant induction of apoptosis. Methods Preparation of Pol II Transcribing Complexes. Ten-subunit S. cerevisiae pol II was purified as described (35). RNA oligonucleotides were purchased from Dharmacon and DNA oligonucleotides were obtained from IDT. cis- ½PtðNH3Þ2ðpyÞClCl was prepared by Ryan Todd at MIT. The site-specifically platinated template DNA was obtained as described (7). Pol II transcribing complexes were assembled with the use of synthetic oli- gonucleotides (10). Briefly, DNA and RNA oligonucleotides were annealed and mixed with pol II in 20 mM Tris (pH 7.5), 40 mM KCl, and 5 mM DTT. The final mixture contained 2 μM pol II, 10 μM site-specific pyriplatin- damaged template DNA strand, and 20 μM nontemplate DNA and RNA oli- gonucleotides. The mixture was kept for 1 h at room temperature, and excess oligonucleotides were removed by ultrafiltration. Crystals were obtained from solutions containing 390 mM ðNH4Þ2HPO4∕NaH2PO4, pH 5.9–6.3, 50 mM dioxane, 10 mM DTT, and 9–11% PEG6000. Crystals of transcribing complexes were transferred in a stepwise manner to cryobuffer as described (10, 11). For the structure of the pol II complex with CTP incorporation, 10 mM CTP was added to the cryobuffer (10, 11). Crystal Structure Determination and Analysis. Diffraction data were collected on beam line 11-1 at the Stanford Synchrotron Radiation Laboratory. Data were processed in DENZO and SCALEPACK (HKL2000) (36). Model building was performed with the program Coot (37), and refinement was done with REFMAC with TLS (CCP4i) (Table S1). In the structure of pol II complex with a CTP incorporation against damaged guanosine residue, we also observed additional weaker density within the second channel in comparison to the nucleoside residue at the þ1 position, which may correspond to nonspecific binding of a second CTP molecule through the soaking process. All structure models in the figures were superimposed with nucleoside residues near the active site using PYMOL (38). Transcription Assay. Transcription assays were performed essentially as de- scribed (11). In a typical reaction, 32P-labeled RNA oligonucleotide (10 pmol) was annealed with template DNA 29mer (20 pmol, damaged or nondamaged template) and nontemplate DNA 14mer (20 pmol) in elongation buffer (20 mM Tris-HCl, pH 7.5, 40 mM KCl, 0.5 mM MgCl2) in a final volume of 20 μL. An aliquot of the annealed RNA/DNA hybrid was incubated with a five- fold excess of pol II (final concentration of pol II 1.1 μM, of RNA oligonucleo- tide 0.22 μM, and of DNA oligonucleotides 0.44 μM) for 10 min at room tem- perature. Equal volumes of the NTP mixture solution were added (final concentrations 25 μM) and the mixture was then incubated for 0, 0.5, 1, 2, 3, 4, 8, 16, 32, or 64 min at room temperature before addition of stop solu- tion (final concentrations 5 M urea, 44.5 mM Tris-HCl, 44.5 mM boric acid, 26 mM EDTA, pH 8.0, Xylene Cyanol and Bromophenol Blue dyes). RNA pro- ducts were analyzed by PAGE in the presence of urea. Visualization and quantification of products were performed with the use of a PhosphorIma- ger (Molecular Dynamics). Computer Modeling Analysis. Three representative platinum units, fPtðNH3Þ3g2þ, trans-fPtðNH3Þ2ðpyÞg2þ, and cis-fPtðNH3Þ2ðpyÞg2þ bound to guanosine in DNA and positioned in either the −1 or þ1 site of pol II were modeled to mimic the post- and pre-translocation states, respectively. The vdW interaction energies between the three ligands at different orientations and the rest of the pol II complex were systematically computed and taken as direct indicators of steric effects. The structure of the cis-fPtðNH3Þ2ðpyÞg2þ fragment on DNA in pol II is available from the current study. Initial configurations for the other two units, fPtðNH3Þ3g2þ and trans-fPtðNH3Þ2ðpyÞg2þ, were obtained by modeling. Briefly, the same configuration of pol II, DNA, and RNA as found in the struc- ture containing cis-fPtðNH3Þ2ðpyÞg2þ was used for these two complexes. The geometry of the fPtðNH3Þ3g2þ moiety was taken from a previous structure where it binds to a B-DNA dodecamer (PDB ID: 5BNA) (39). Docking was achieved by aligning the damaged guanosine base of the two structures. Finally, the trans-ammine group in fPtðNH3Þ3g2þ was replaced with a pyridine ligand, and the Pt-N bond length was appropriately adjusted to obtain the structure for trans-fPtðNH3Þ2ðpyÞg2þ. The same procedure was used to generate structures at both þ1 and −1 positions. The vdW energies were computed for different configurations generated by rotating about the Pt-N7 bond from −180° to 180° for each platinum modi- fication (see Figs. S5–S7). The rotation angle (φ) was defined to be positive when rotating in the anticlockwise direction. In the configuration with φ ¼ 0°, the plane composed of two Pt-N bonds of the ligand which are per- pendicular to the Pt-N7 bond was set to be parallel to the damaged guano- sine base. We noticed that, for fPtðNH3Þ3g2þ and trans-fPtðNH3Þ2ðpyÞg2þ, two trans ammine groups were accommodated at slightly different configura- tions, with φ ¼ 0° due to the different local environment, which leads to slightly different energies between conformations with φ and φ  180°. Because the purpose of our modeling study is to identify major steric clashes instead of accurately computing free energy changes associated with rota- tion of the ligand, which requires extensive conformational sampling, we performed a simple average of the two energies (E1ðφÞ and E2ðφ  180°Þ) based on their Boltzmann weights (T 298 K), eq 1, ¯E ¼ ðe−βE1E1 þ e−βE2E2Þ∕ðe−βE1 þ e−βE2Þ [1] to get a better estimate of vdW energy profiles. 9588 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1002565107 Wang et al. The GROMACS simulation package was used to compute vdW energies between the ligands and the pol II complex (40). A 20-Å cutoff was adopted for computing the vdW interactions. The AMBER03 force field was used for the pol II complex including protein, RNA, and DNA (41). The vdW force field (Leonard–Jones potential) parameters for ligands were generated from the AMTECHAMBER module of the AMBER 9 package (42) using the general AMBER force field (GAFF) (43) developed for rational drug design. Since the Pt atom is not in direct contact with the pol II complex and does not con- tribute significantly to any steric effects, we excluded it from our vdW energy calculations. ACKNOWLEDGMENTS. This research was supported by the National Institute of General Medical Sciences (NIH Pathway to Independence Award GM085136 to D.W. and GM49985 to R.D. Kornberg) and by the National Cancer Institute (Grant CA034992 to S.J.L.). Portions of the research were carried out at the Stanford Synchrotron Radiation Laboratory, a national user facility operated by Stanford University on behalf of the U.S. Department of Energy, Office of Basic Energy Sciences. The SSRL Structural Molecular Biology Program is sup- ported by the Department of Energy, Office of Biological and Environmental Research, and by the National Institutes of Health, National Center for Research Resources, Biomedical Technology Program, and the National Insti- tute of General Medical Sciences. 1. Jamieson ER, Lippard SJ (1999) Structure, recognition, and processing of cisplatin–DNA adducts. Chem Rev 99:2467–2498. 2. Wang D, Lippard SJ (2005) Cellular processing of platinum anticancer drugs. Nat Rev Drug Discov 4:307–320. 3. Roy S, et al. (2008) Phenanthroline derivatives with improved selectivity as DNA- targeting anticancer or antimicrobial drugs. ChemMedChem 3:1427–1434. 4. Hollis LS, Amundsen AR, Stern EW (1989) Chemical and biological properties of a new series of cis-diammineplatinum(II) antitumor agents containing three nitrogen donors: cis-½PtðNH3Þ2ðN-donorÞClþ. J Med Chem 32:128–136. 5. Hollis LS, et al. (1991) Mechanistic studies of a novel class of trisubstituted platinum(II) antitumor agents. Cancer Res 51:1866–1875. 6. Bierbach U, Sabat M, Farrell N (2000) Inversion of the cis geometry requirement for cytotoxicity in structurally novel platinum(II) complexes containing the bidentate N,O-donor pyridin-2-yl-acetate. Inorg Chem 39:1882–1890. 7. Lovejoy KS, et al. (2008) cis-Diammine(pyridine)chloroplatinum(II), a monofunctional platinum(II) antitumor agent: Uptake, structure, function, and prospects. Proc Natl Acad Sci USA 105:8902–8907. 8. Cleare MJ, Hoeschele JD (1973) Studies on the antitumor activity of group VIII transi- tion metal complexes. Part I. Platinum (II) complexes. Bioinorg Chem 2:187–210. 9. Westover KD, Bushnell DA, Kornberg RD (2004) Structural basis of transcription: separation of RNA from DNA by RNA polymerase II. Science 303:1014–1016. 10. Westover KD, Bushnell DA, Kornberg RD (2004) Structural basis of transcription: nu- cleotide selection by rotation in the RNA polymerase II active center. Cell 119:481–489. 11. Wang D, et al. (2006) Structural basis of transcription: Role of the trigger loop in sub- strate specificity and catalysis. Cell 127:941–954. 12. Corda Y, et al. (1993) Spectrum of DNA–platinum adduct recognition by prokaryotic and eukaryotic DNA-dependent RNA polymerases. Biochemistry 32:8582–8588. 13. Mello JA, Lippard SJ, Essigmann JM (1995) DNA adducts of cis-diamminedichloropla- tinum(II) and its trans isomer inhibit RNA polymerase II differentially in vivo. Biochem- istry 34:14783–14791. 14. Sandman KE, Marla SS, Zlokarnik G, Lippard SJ (1999) Rapid fluorescence-based repor- ter-gene assays to evaluate the cytotoxicity and antitumor drug potential of platinum complexes. Chem Biol 6:541–551. 15. Lee KB, Wang D, Lippard SJ, Sharp PA (2002) Transcription-coupled and DNA damage- dependent ubiquitination of RNA polymerase II in vitro. Proc Natl Acad Sci USA 99:4239–4244. 16. Tornaletti S, Patrick SM, Turchi JJ, Hanawalt PC (2003) Behavior of T7 RNA polymerase and mammalian RNA polymerase II at site-specific cisplatin adducts in the template DNA. J Biol Chem 278:35791–35797. 17. Tremeau-Bravard A, Riedl T, Egly JM, Dahmus ME (2004) Fate of RNA polymerase II stalled at a cisplatin lesion. J Biol Chem 279:7751–7759. 18. Jung Y, Lippard SJ (2006) RNA polymerase II blockage by cisplatin-damaged DNA. Stability and polyubiquitylation of stalled polymerase. J Biol Chem 281:1361–1370. 19. Damsma GE, et al. (2007) Mechanism of transcriptional stalling at cisplatin-damaged DNA. Nat Struct Mol Biol 14:1127–1133. 20. Jung Y, Lippard SJ (2007) Direct cellular responses to platinum-induced DNA damage. Chem Rev 107:1387–1407. 21. Surratt CK, Milan SC, Chamberlin MJ (1991) Spontaneous cleavage of RNA in ternary complexes of Escherichia coli RNA polymerase and its significance for the mechanism of transcription. Proc Natl Acad Sci USA 88:7983–7987. 22. Orlova M, et al. (1995) Intrinsic transcript cleavage activity of RNA polymerase. Proc Natl Acad Sci USA 92:4596–4600. 23. Wang D, et al. (2009) Structural basis of transcription: backtracked RNA polymerase II at 3.4 angstrom resolution. Science 324:1203–1206. 24. Cohen GL, Bauer WR, Barton JK, Lippard SJ (1979) Binding of cis- and trans- dichlorodiammineplatinum(II) to DNA: evidence for unwinding and shortening of the double helix. Science 203:1014–1016. 25. Lecointe P, Macquet JP, Butour JL (1979) Correlation between the toxicity of platinum drugs to L1210 leukemia cells and their mutagenic properties. Biochem Biophys Res Commun 90:209–213. 26. Macquet JP, Butour JL (1983) Platinum-amine compounds: importance of the labile and inert ligands for their pharmacological activities toward L1210 leukemia cells. J Natl Cancer Inst 70:899–905. 27. Calvert AH (1986) Clinical applications of platinum metal complexes. Biochemical Mechanisms of Platinum Antitumor Drugs (IRI Press, Washington). 28. Balcarová Z, et al. (1998) DNA interactions of a novel platinum drug, cis-½PtClðNH3Þ2 ðN7-acyclovirÞþ. Mol Pharmacol 53:846–855. 29. Brabec V, Leng M (1993) DNA interstrand cross-links of trans-diamminedichloroplati- num(II) are preferentially formed between guanine and complementary cytosine residues. Proc Natl Acad Sci USA 90:5345–5349. 30. Lemaire MA, Schwartz A, Rahmouni AR, Leng M (1991) Interstrand cross-links are preferentially formed at the d(GC) sites in the reaction between cis-diamminedichlor- oplatinum (II) and DNA. Proc Natl Acad Sci USA 88:1982–1985. 31. Brabec V, Boudny V (1994) Monofunctional and interstrand DNA adducts of platinum (II) complexes. Met Based Drugs 1:195–200. 32. Brabec V, Boudný V, Balcarová Z (1994) Monofunctional adducts of platinum(II) pro- duce in DNA a sequence-dependent local denaturation. Biochemistry 33:1316–1322. 33. Novakova O, et al. (2009) Energetics, conformation, and recognition of DNA duplexes modified by methylated analogues of ½PtClðdienÞþ. Chemistry 15:6211–6221. 34. Brueckner F, Hennecke U, Carell T, Cramer P (2007) CPD damage recognition by tran- scribing RNA polymerase II. Science 315:859–862. 35. Cramer P, Bushnell DA, Kornberg RD (2001) Structural basis of transcription: RNA polymerase II at 2.8 angstrom resolution. Science 292:1863–1876. 36. Otwinowski Z, Minor W (1997) Processing of x-ray diffraction data collected in oscilla- tion mode. Method Enzymol 276:307–326. 37. Emsley P, Cowtan K (2004) Coot: Model-building tools for molecular graphics. Acta Cryst D60:2126–2132. 38. DeLano WL (2002) The PyMOL Molecular Graphics System (DeLano Scientific, Palo Alto, CA). 39. Wing RM, Pjura P, Drew HR, Dickerson RE (1984) The primary mode of binding of cisplatin to a B-DNA dodecamer: C-G-C-G-A-A-T-T-C-G-C-G. EMBO J 3:1201–1206. 40. Lindahl E, Hess B, van der Spoel D (2001) GROMACS 3.0: A package for molecular simulation and trajectory analysis. J Mol Model 7:306–317. 41. Duan Y, et al. (2003) A point-charge force field for molecular mechanics simulations of proteins based on condensed-phase quantum mechanical calculations. J Comput Chem 24:1999–2012. 42. Wang JM, Wang W, Kollman PA, Case DA (2006) Automatic atom type and bond type perception in molecular mechanical calculations. J Mol Graphics Modell 25:247–260. 43. Wang J, et al. (2004) Development and testing of a general Amber force field. J Comput Chem 25:1157–1174. Wang et al. PNAS ∣ May 25, 2010 ∣ vol. 107 ∣ no. 21 ∣ 9589 BIOCHEMISTRY
3M4U
Crystal Structure of Trypanosoma brucei Protein Tyrosine Phosphatase TbPTP1
The Trypanosoma brucei Life Cycle Switch TbPTP1 Is Structurally Conserved and Dephosphorylates the Nucleolar Protein NOPP44/46* Received for publication,January 28, 2010, and in revised form, April 19, 2010 Published, JBC Papers in Press,May 5, 2010, DOI 10.1074/jbc.M110.108860 Seemay Chou‡, Bryan C. Jensen§, Marilyn Parsons§¶, Tom Alber‡, and Christoph Grundner§¶1 From the ‡Department of Molecular and Cell Biology and QB3 Institute, University of California, Berkeley, California 94720-3200, the §Seattle Biomedical Research Institute, Seattle, Washington 98109-5219, and the ¶Department of Global Health, University of Washington, Seattle, Washington 98195-5065 Trypanosoma brucei adapts to changing environments as it cycles through arrested and proliferating stages in the human and tsetse fly hosts. Changes in protein tyrosine phosphoryla- tion of several proteins, including NOPP44/46, accompany T. brucei development. Moreover, inactivation of T. brucei pro- tein-tyrosine phosphatase 1 (TbPTP1) triggers differentiation of bloodstream stumpy forms into tsetse procyclic forms through unknown downstream effects. Here, we link these events by showing that NOPP44/46 is a major substrate of TbPTP1. TbPTP1 substrate-trapping mutants selectively enrich NOPP44/46 from procyclic stage cell lysates, and TbPTP1 effi- ciently and selectively dephosphorylates NOPP44/46 in vitro. To provide insights into the mechanism of NOPP44/46 recog- nition, we determined the crystal structure of TbPTP1. The TbPTP1 structure, the first of a kinetoplastid protein-tyrosine phosphatase (PTP), emphasizes the conservation of the PTP fold, extending to one of the most diverged eukaryotes. The structure reveals surfaces that may mediate substrate specificity and affords a template for the design of selective inhibitors to interfere with T. brucei transmission. Trypanosoma brucei causes human African trypanosomiasis or African sleeping sickness, which is marked by debilitating neurologic symptoms ranging from sensory impairment to the characteristic aberrant sleeping patterns that progress to coma. If untreated, human African trypanosomiasis is fatal. With 30,000 deaths a year and 60 million people living at risk (1), human African trypanosomiasis is a major disease burden in sub-Saharan Africa. Current drugs are ineffective and toxic, and drug resistance is becoming a growing hurdle for treatment (2). T. brucei alternates between human and tsetse fly hosts, requiring extensive and rapid physiologic adaptations. In humans, the major T. brucei population consists of the extra- cellular, proliferative slender form in the bloodstream, which irreversibly differentiates into the G1-arrested stumpy form poised for transmission to the tsetse fly. Taken up by the tsetse fly, the stumpy form differentiates into the proliferative procy- clic form in the insect midgut. Eventually, the tsetse salivary gland becomes populated with metacyclic forms, which infect the human host (3). This differentiation cycle requires survival in a diverse set of environments and forms the basis for infec- tivity and transmission. The molecular signals, regulators, and effectors underlying this complex sequence of events are not well understood but could provide novel targets for therapeutic interference. Dis- tinct patterns of protein tyrosine phosphorylation accompany and often precede stage progression (4), suggesting that tyro- sine phosphorylation is a key mechanism of developmental reg- ulation. Studies of the T. brucei dual specificity kinases also pro- vide evidence that tyrosine phosphorylation regulates the trypanosome life cycle (5–7). Moreover, NOPP44/46,2 a nucle- olar RNA-binding protein required for ribosome biogenesis (8), exhibits dramatic changes in tyrosine phosphorylation in con- cert with the T. brucei life cycle transitions (9). NOPP44/46 is tyrosine-phosphorylated in both proliferating procyclic and non-proliferating stumpy forms, but not in proliferating slen- der forms, indicating a complex interplay between life cycle and cell cycle in modulating tyrosine phosphorylation. Recently, TbPTP1, a PTP with sequence similarity to classi- cal human PTPs, was identified as a central molecular switch for the stumpy-to-procyclic progression (10). TbPTP1 activity arrests stumpy bloodstream forms, suggesting a model in which TbPTP1 inactivation in the fly midgut releases the arrest and triggers development into the procyclic form (10). Thus, TbPTP1 might function downstream of the recently described proteins associated with differentiation (PAD) transporters, which represent the first known step in the pathway that allows the differentiation signals citrate or cis-aconitate to trigger developmental changes (11). However, the substrates and downstream effects of TbPTP1 remain unknown. By sequence comparison, TbPTP1 is similar to human clas- sical PTPs such as the prototypical PTP1B. TbPTP1 has an ortholog in Trypanosoma cruzi and Leishmania major and con- tains six regions that appear to be specific to trypanosomatids (10). The Kinetoplastida, including T. brucei, constitute some * Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth Grant R01 AI31077 (to M. P.). The atomic coordinates and structure factors (code 3M4U) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformat- ics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). 1 To whom correspondence should be addressed: 307 Westlake Ave. N, Ste. 500, Seattle, WA 98109-5219. Tel.: 206-256-7295; Fax: 206-256-7229; E-mail: christoph.grundner@sbri.org. 2 The abbreviations used are: NOPP44/46, nucleolar phosphoprotein 44/46; TbPTP1, T. brucei protein-tyrosine phosphatase 1; PTP, protein-tyrosine phosphatase; r.m.s., root mean square; CHES, 2-(cyclohexylamino)ethane- sulfonic acid. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 29, pp. 22075–22081, July 16, 2010 © 2010 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. JULY 16, 2010•VOLUME 285•NUMBER 29 JOURNAL OF BIOLOGICAL CHEMISTRY 22075 of the most diverged eukaryotes. The evolutionary distance and unique sequences raise the question whether the overall struc- tural conservation of the PTP fold is maintained in these distant eukaryotes. Here, we identify NOPP44/46 as a major substrate of the life cycle switch TbPTP1. We also describe the TbPTP1 crystal structure, revealing strong conformational similarity to other eukaryotic PTPs and surface characteristics that help ratio- nalize NOPP44/46 binding. Trypanosome-specific sequence motifs follow the canonical PTP fold, and all major functional elements are structurally conserved. These data establish the structural correlates of kinetoplastid PTPs within the PTP fam- ily and provide a new link in the signaling pathway controlling the stumpy-to-procyclic transition. EXPERIMENTAL PROCEDURES Cloning, Protein Expression, and Purification—The full- length TbPTP1 (systematic ID Tb10.70.0070) gene was ampli- fied from genomic T. brucei DNA (kindly provided by Dr. Christian Klotz) and cloned into the pET28b expression vector in-frame with the N-terminal six-histidine tag. Point mutants were generated according to the QuikChange protocol (Strat- agene). BL21 (DE3)-CodonPlus cells were transformed, and protein expression was induced at A600 of 0.6 by adding 100 M isopropyl-1-thio--D-galactopyranoside. After 20 h of induc- tion at 20 °C, cells were harvested, resuspended in 20 mM Tris, pH 7.5, 100 mM NaCl, and lysed by sonication. The lysate was centrifuged for 1 h at 20,000  g, and the supernatant was loaded on a metal-chelating affinity column. Fractions contain- ing TbPTP1 were identified by measuring the hydrolysis of p-nitrophenyl phosphate (12). Fractions were pooled, loaded on a gel filtration column, and eluted in 20 mM Tris, pH 7.5, 100 mM NaCl. Recombinant TbPTP1 was concentrated to 10 mg/ml. In Vitro Dephosphorylation—NOPP44/46 (Genbank acces- sion number HM44803) was amplified from T. brucei strain 29.13 (13) genomic DNA and cloned into pLEW-MHTAP (14) for expression in procyclic form T. brucei 29.13. Expression of the tagged protein was induced with tetracycline for 24 h, and the protein was purified using a modified tandem affinity puri- fication protocol with 1 mM sodium orthovanadate in the lysis buffer (15, 16). The purified preparation was treated with 10 mM dithiothreitol for 15 min to inactivate the sodium orthovanadate. Dephosphorylation reactions were carried out at room tempera- ture for 15 min in 20 mM Tris, pH 7.5, 100 mM NaCl buffer, with varying amounts of TbPTP1. For dephosphorylation of NOPP44/46 with other PTPs, PTP input was normalized to the activity of 100 nM TbPTP1 at a saturating concentration of the non-cognate substrate p-nitrophenyl phosphate. Reactions were stopped by the addition of SDS-PAGE loading buffer, separated by SDS-PAGE, and detected by Western blot using the 4G10 anti- Tyr(P) antibody. Blots were stripped and reprobed with mono- clonal anti-NOPP44/46 1D2 (9). Substrate Trapping—TbPTP1 resin was prepared by cou- pling TbPTP1 to NHS-activated SepharoseTM 4 fast flow (GE Healthcare) according to the manufacturer’s protocol. Wild- type or the D199A mutant TbPTP1 was coupled at a concen- tration of 1 mg/ml followed by an incubation in 0.1 M Tris blocking buffer. Lysates were prepared from procyclic form T. brucei grown for 16 h in medium containing 1.5 M sodium orthovanadate. Cells were extracted in lysis buffer (50 mM Tris- HCl, pH 7.5, 150 mM NaCl, 2 mM EGTA, 1% Triton X-100) containing 1 mM sodium orthovanadate, complete protease inhibitor mixture (Roche Applied Science), and 5 mM iodoac- etamide to inhibit endogenous PTP activity. Iodoacetamide and orthovanadate were inactivated by the addition of 10 mM dithiothreitol. To capture substrates of TbPTP1, 10 l of wild- type or D199A TbPTP1 resin was incubated for 2 h at 4 °C with 500 l of lysate corresponding to 0.5  109 T. brucei cells. The resin was washed five times in high salt buffer (20 mM HEPES, pH 8.0, 2 M NaCl, 15% glycerol, 0.5% Nonidet P-40) followed by five washes alternating in guanidine-HCl buffer (20 mM HEPES, pH 8.0, 200 mM guanidine-HCl, 15% glycerol, 0.5% Nonidet P-40) and low salt buffer (20 mM HEPES, pH 8.0, 300 mM NaCl, 15% glycerol). The resin was boiled in reducing SDS-PAGE buffer for 15 min, run on a 12% Tris-glycine gel, and transferred to nitrocellulose membrane for Western analysis with 4G10 anti-Tyr(P) (GE Healthcare) and anti-NOPP44/46 antibody (9). To control for input of recombinant wild-type and trapping TbPTP1, TbPTP1 was released from the column material by boiling in SDS sample buffer prior to trapping and analyzed by SDS-PAGE. Crystallization, Structure Determination, and Structure Analysis—Initial crystals were obtained by sitting drop vapor diffusion trials at 18 °C from a 1:1 mixture of TbPTP1 at 10 mg/ml and 10% polyethylene glycol 3000, 100 mM CHES, pH 9.5. Diffraction quality crystals were obtained by hanging drop vapor diffusion after introducing mutations E138A, E139A, and E140A, predicted to reduce the surface entropy of TbPTP1 (17), the addition of 10 mM tris(2-carboxyethyl)phosphine, and in- drop trypsin cleavage of the His6 tag using a trypsin:TbPTP1 ratio of 1:1,000 (w/w). Crystals were immersed in mother liquor containing 10% glycerol, mounted, and flash frozen in liquid nitrogen. Diffraction data were collected at the Lawrence Berkeley National Laboratory Advanced Light Source Beamline 8.3.1. Data were reduced using the HKL2000 program suite (18). Phases were obtained by molecular replacement using MolRep (19) and the search model PTPN3 (Protein Data Bank (PDB) accession 2B49) modified by CHAINSAW (20). After auto- mated model building in PHENIX (21), the final model was built by alternating manual model building using Coot (22) and maximum likelihood refinement using PHENIX. The Rfree was determined using a random 5% of the data. The structure was validated using MOLProbity (23). Images were generated in PyMOL, and structure comparisons were performed using the DALI server and PDBsum. The crystal structure was deposited in the Protein Data Bank under accession number 3M4U. RESULTS TbPTP1 Binds NOPP44/46—To identify cellular substrates of TbPTP1, we generated a substrate-trapping mutant by replacing the general acid Asp199 with Ala. This mutant is cat- alytically inactive but retains substrate binding, thus allowing for stable trapping and isolation of substrates (24). Based on previous studies suggesting that TbPTP1 is inactive and tyro- TbPTP1 Structure and Substrate 22076 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 29•JULY 16, 2010 sine phosphorylation most pronounced in procyclic forms (4, 10), we used cell lysates from T. brucei procyclic forms for trap- ping experiments. Tyrosine phosphorylation was preserved throughout cell lysis by the addition of iodoacetamide and sodium orthovanadate to inhibit endogenous PTPs. The wild- type and D199A TbPTP1 variants were covalently linked to NHS-Sepharose beads and incubated with T. brucei lysate. After high salt, guanidinium hydrochloride, and detergent washes, bound proteins were eluted by boiling in SDS-PAGE loading buffer. Western blotting of eluates using anti-Tyr(P) antibody showed the enrichment of several tyrosine-phosphor- ylated proteins by the D199A mutant relative to the wild-type TbPTP1 (Fig. 1A). Two major bands migrated at a molecular mass of 45 and 70 kDa, and a minor band migrated at 40 kDa. The enrichment of these bands was selective as other Tyr(P) proteins apparent in the total lysate were not trapped by TbPTP1. Because the molecular mass of the 45-kDa species corre- sponded to that of NOPP44/46, a protein known to be tyrosine- phosphorylated in procyclic forms in vivo (4, 9), we explored the possibility that NOPP44/46 was a trapped substrate of TbPTP1. Experimental replicates of trapping eluates were probed with anti-NOPP44/46, resulting in signal at the position identical to the band detected with anti-Tyr(P) antibodies (Fig. 1B). Some phospho-independent binding of NOPP44/46 to wild-type TbPTP1 was also observed. The specific enrichment of phosphorylated NOPP44/46 with the trapping TbPTP1 identifies NOPP44/46 as a potential in vivo substrate of TbPTP1. TbPTP1 Dephosphorylates NOPP44/46 in Vitro—To con- firm the observed interaction between TbPTP1 and NOPP44/ 46, we tested dephosphorylation of NOPP44/46 by TbPTP1 in vitro. Although the pH optimum of TbPTP1 is 6 (10), this pref- erence is unlikely to reflect physiologic function rather than the generally higher nucleophilicity of cysteine residues at low pH. In vitro dephosphorylation assays were therefore performed at pH 7.5, more similar to the pH at which TbPTP1 likely func- tions. Phosphorylated NOPP44/46 was obtained by overex- pressing a C-terminally TAP-tagged version of the full-length protein in procyclic forms. Rapid dephosphorylation of NOPP44/46 was observed at the lowest TbPTP1 concentration tested (3.7 nM) and was complete at TbPTP1 concentrations at or above 33 nM (Fig. 2A). To test whether dephosphorylation of NOPP44/46 by TbPTP1 is specific, we tested the activity of a panel of unrelated microbial PTPs on NOPP44/46. Of the five tested PTPs, only TbPTP1 and Yersinia YopH dephosphory- lated NOPP44/46; no activity was observed using Mycobacte- rium tuberculosis PtpA and PtpB, Staphylococcus aureus SaPtpA, or Listeria monocytogenes lmo1935 (Fig. 2B). With YopH known to have broad substrate specificity (25), these data show complete, efficient, and selective dephosphorylation of NOPP44/46 by TbPTP1 in vitro. NOPP44/46 Is Phosphorylated on Tyr181—To define the phosphorylation site(s) on NOPP44/46 recognized by TbPTP1, we assessed the phosphorylation state of NOPP44/46 variants in which each of the five Tyr residues was replaced individually with Phe. Y181F completely abrogated tyrosine phosphoryla- tion of NOPP44/46, as shown by anti-Tyr(P) Western analysis (Fig. 3). None of the other mutations reduced the level of NOPP44/46 Tyr phosphorylation, indicating that Tyr181 is the only phosphorylated Tyr and the target of TbPTP1. The NOPP44/46 phosphorylation site is located in a 40-residue acidic loop encompassing residues 167–207. The sequence of this segment, 167DAGDEDDNDDDDEAYDEDDSDDDDDD- DDDDDDDDDDDDDDE207, indicates that phosphorylation adds additional negative charges to a nearly uninterrupted acidic sequence. This acidic region containing the target Tyr is found only in T. brucei homologs. TbPTP1 Has a Classical PTP Fold—To explore the basis for recognition of the unusual substrate target sequence and the architecture of this diverged kinetoplastid PTP, we determined the crystal structure of TbPTP1 at 2.4 Å resolution (Table 1). The asymmetric unit contains two TbPTP1 molecules with a root mean square (r.m.s.) deviation of all atoms of 0.3 Å. The structure comprises residues Ser6–Thr301 in chain A and residues Met1–Leu298 in chain B (Fig. 4), as well as one phos- phate per TbPTP1 and 192 water molecules. No clear elec- tron density was visible for residues Leu66–Gln73 and Ala160–Ala162 of chain A and residues Lys67–Arg75, Ala140, and Gln147–His151 of chain B. The protein contains four changes. The catalytic cysteine is changed to alanine, possibly through desulfurization by the phosphine tris(2-carboxyethyl)- FIGURE 1. Substrate trapping identifies NOPP44/46 as a major TbPTP1 substrate. A, anti-Tyr(P) (-pTyr) Western blot after TbPTP1 substrate trap- ping. The trapping mutant (trap) selectively enriches three major tyrosine- phosphorylated proteins. WT, wild type. B, anti-NOPP44/46 Western blot of proteins bound to wild type and trapping mutant. The Western blot in B is an experimental replicate of A and identifies the 45-kDa band as NOPP44/46. Some phospho-independent binding is also apparent. C, Coomassie Blue staining showing equivalent amounts of TbPTP1 wild type and mutant elute from the resin prior to trapping. TbPTP1 Structure and Substrate JULY 16, 2010•VOLUME 285•NUMBER 29 JOURNAL OF BIOLOGICAL CHEMISTRY 22077 phosphine present at high concentrations in the protein drop, and Glu138-140 were mutated to alanine to improve the crystallization properties of the protein (17). The overall fold of TbPTP1 resembles that of other classical PTPs, with an extended, twisted -sheet at the center and -helices surrounding it (Fig. 4A). The catalytic loop, or P-loop, is situated at the center of the active site and comprises the invariant PTP signature motif Cys-Xaa5-Arg. A phosphate binds in the position similar to that of Tyr(P) substrate phos- phate in PTP-peptide substrate structures (26) (Fig. 4B). The active site cavity is further delineated by the Tyr(P) loop that deepens the cavity to 9 Å, thus excluding Ser(P) and Thr(P) resi- dues. The WPD loop containing the general acid Asp199 assumes a closed conformation, similar to that seen in other PTP structures with small ligands bound (27). TbPTP1 contains 9 of 10 PTP sequence motifs (28) in the same spatial organization as human PTPs. The six trypanosome-specific sequence motifs of TbPTP1 follow the canon- ical PTP fold and do not give rise to new structural features. Con- sistent with a role in substrate recognition or regulation, the trypanosome-specific sequences predominantly map to the surface of TbPTP1 (Fig. 5A). The phos- phate engages in the typical inter- actions with the P-loop, hydrogen- bonding with six main chain amides and the invariant Arg235 side chain. The closest structural homologs of TbPTP1 found by the DALI server are the prototypical human PTP1B and PTPRO (Glepp1), with a C r.m.s. deviation of 2.1 and 2.2 Å, respec- tively. The superposition of TbPTP1 with PTPRO (PDB ID 2G59) shows the overall large similarity, with major differences only at the termini and surface loops (Fig. 5B). The TbPTP1 loop from 62 to 79, although mostly invisible in the structure, has shifted at the base when compared with the equivalent PTPRO loop and contains a single-residue insertion when com- pared with PTPRO and up to seven residues when compared with other human PTPs. The TbPTP1 loop 138–154 contain- ing a PEST sequence shows the most divergence from the PTPRO structure. TbPTP1 forms weak interactions between the two molecules in the asymmetric unit in the crystal (data not shown). The interactions comprise two symmetric salt bridges between Lys123 and Glu201, hydrogen bonds between Gly127 and Glu201, as well as 44 non-bonded interactions resulting in an interface of 500 Å2. PTPs such as PTP form dimers in the crystal and in solution (27). However, TbPTP1 migrates as a monomer in size exclusion chromatography (data not shown), suggesting that these contacts do not reflect a physiologic state but are a result of crystal packing. Although the three-dimensional organization of the PTP active site is highly similar in all PTPs across families, PTP surface properties vary widely and produce large diversity (27). The TbPTP1 electrostatic surface shows distinct and continuous electronegative and positive areas (Fig. 6). The active site shows moderately electropositive potential, with the closed WPD loop burying additional electropositive regions of the phosphate binding pocket. A continuous elec- tropositive stretch runs across the active site and along one side of the molecule. This stretch includes the side chains of Arg15, -23, -30, -50, -125, -175, -276, and Lys113 and -131 (Fig. 6A). The electrostatic surface of PTP1B with a closed WPD loop FIGURE 2. TbPTP1 efficiently and selectively dephosphorylates NOPP44/46 in vitro. A TAP-tagged allele of NOPP44/46 was expressed in procyclic forms and affinity-purified for dephosphorylation reactions. A, TbPTP1 dephosphorylates NOPP44/46 in a dose-dependent manner. -pTyr, anti-Tyr(P) antibody. B, TbPTP1, but not unrelated phosphatases from M. tuberculosis (PtpA and PtpB), S. aureus (SaPtpA), and L. monocytogenes (lmo1935), dephosphorylates NOPP44/46. FIGURE 3. NOPP44/46 is phosphorylated on Tyr181. A, schematic of NOPP44/46 indicating domain organization and position of the five tyrosine residues. U, unique region; J, junction; A, acidic region; R, RGG repeat region. B, all five NOPP44/46 tyrosines were individually changed to Phe, and the phosphorylation of NOPP44/46 was detected by anti-Tyr(P) antibody (-pTyr) (upper panel) and anti-NOPP44/46 control Western (lower panel). WT, wild type. TbPTP1 Structure and Substrate 22078 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 29•JULY 16, 2010 also shows extended electropositive areas (Fig. 6B). These surfaces likely complement the sequence of the highly elec- tronegative PTP1B substrate, the insulin receptor. PTP1B binds the triply phosphorylated peptide sequence pYETD- pYpY, which also concentrates a large number of charges around the PTP1B active site. In contrast, PTPRO shows few electropositive areas outside of the direct active site vicinity and a predicted second Tyr(P) binding site (Fig. 6B, right panel). DISCUSSION T. brucei requires stringent control of developmental pro- grams to successfully infect its human and tsetse fly hosts, and a molecular hallmark of life cycle transitions in T. brucei is the coordinated change in tyrosine phosphorylation. Recently, the tyrosine phosphatase TbPTP1 was identified as a key regulator of the trypanosome life cycle. TbPTP1 has an ortholog in T. cruzi with 61.3% sequence identity. Although an intracellular parasite, T. cruzi shares the bloodstream-to- insect route of transmission controlled by TbPTP1 in T. bru- cei, suggesting that the function of the two PTPs may be conserved. Despite the evolutionary distance between humans and trypanosomes, our crystal structure of TbPTP1 shows a high degree of structural conservation of the conventional PTP fold. The 24% sequence identity of TbPTP1 to human PTP1B translates into a C r.m.s. deviation of 2.1 Å. The conser- vation of the PTP fold thus extends not only to bacteria but also to distant eukaryotes and underscores the utility and evolutionarysuccessofthisscaffoldfortyrosinedephosphory- lation. TbPTP1 shares 9 of 10 signature motifs with the human PTPs and has six additional trypanosome-specific motifs that may play roles in func- tional regulation or substrate rec- ognition. The folding of these motifs suggests that although not giving rise to new structural ele- ments, their position mostly on the surface is consistent with a role in substrate recognition and/or regulation. The TbPTP1 structure also pro- vides the basis for inhibitor design. The strong similarity to human PTPs highlights the need for struc- tural information to guide the design of selective TbPTP1 inhibi- tors as tools and potential therapeu- tics. TbPTP1 prevents premature differentiation of stumpy blood- stream forms to procyclic forms, which lack immune evasion mecha- nisms that allow survival in the mammalian host. Thus, inhibition of TbPTP1 would reduce the pool of tsetse-infective parasites within the mammalian host, potentially atten- uating transmission, an approach that has gained acceptance for the reduction of malaria (29). Blocking transmission could be particularly advantageous for controlling animal trypanosomiasis, which affects live- stock and remains a major hurdle to economic development in sub- FIGURE 4. Overall structure of TbPTP1. A, TbPTP1 shares the canonical PTP fold. The catalytic motifs P-loop and WPD loop are highlighted in orange and yellow, respectively. No electron density for residues 65–74 was visible in chain A (dotted line). N-term, N terminus; C-term, C terminus. B, 2Fo  Fc electron density map of the active site showing phosphate (center), contoured at 1.0 . FIGURE 5. TbPTP1 has a similar fold to human PTPs. A, trypanosome-specific sequence motifs (green) map on the surface outside of the active site (P-loop in orange). B, superposition of the C chain in ribbon representation showing overall strong similarity to human PTPRO. TABLE 1 Data collection and refinement statistics for TbPTP1 Parentheses denote values for the highest resolution shell. Data collection Crystal symmetry P212121 Unit cell a, b, c (Å) 76.63, 77.38, 117,2 , ,  (°) 90, 90, 90 Resolution (Å) 2.4 Rmerge (%) 10 Completeness ( %) 99.46 (97) Multiplicity 5 (4.9) I/I 54 (2.5) Refinement statistics Resolution (Å) 47-2.4 Reflections 28,040 Rwork/Rfree (%) 20/26 r.m.s.  bonds (Å) 0.008 r.m.s.  angles (°) 1.065 Average B-factor (Å2) 42.4 Main chain dihedral angles Most favored (%) 96.6 Allowed (%) 3.2 TbPTP1 Structure and Substrate JULY 16, 2010•VOLUME 285•NUMBER 29 JOURNAL OF BIOLOGICAL CHEMISTRY 22079 Saharan Africa. Moreover, T. brucei rhodesiense infects both humans and animals, providing a parasite reservoir for human infection. Although tyrosine phosphorylation is emerging as a key reg- ulator of the trypanosome life cycle, little is known about the molecular pathways that lead to downstream developmental changes. Identification of TbPTP1 substrates is essential to understanding the mechanisms by which TbPTP1 regulates T. brucei differentiation. Among the phosphoproteins selec- tively enriched using a TbPTP1 trapping mutant, we identified NOPP44/46 and yet unidentified 70- and 40-kDa phospho- proteins as substrates of this PTP. Other substrate phosphopro- teins might be associated with the insoluble fraction and not detected by our methods. The functional interaction of TbPTP1 and NOPP44/46 is supported by efficient and selective in vitro dephosphorylation. Furthermore, the phosphorylation pattern of NOPP44/46 during developmental stages, unlike that of other major tyrosine-phosphorylated species, matches the proposed activity profile of TbPTP1 in slender and procy- clic forms (4, 9). The presence of phosphorylated NOPP44/46 in stumpy forms may reflect decreasing TbPTP1 activity or changes in the activity of the cognate kinase(s) in combination with a large increase in NOPP44/46 protein levels observed in stumpy forms (9). The TbPTP1 crystal structure allows rationalizing NOPP44/46 substrate binding. The distinct, continuous elec- tropositive area running across the TbPTP1 surface and the active site might be a footprint of electronegative regions of its substrate(s), such as the acidic stretch harboring the NOPP44/46 Tyr(P). Phosphorylation sites are usually found in flexible loop regions, suggesting that linear sequences rather than conformational sites serve as dephosphorylation sub- strates. This is consistent with the NOPP44/46 dephosphory- lation site, which is predicted to be highly disordered. More- over, the kinetics of substrate peptide turnover by PTPs are often approaching the limits of diffusion, suggesting that the selectivity and binding determinants of peptides are contained within the primary peptide sequence (30). A phosphoproteomic study of tyrosine-phosphorylated pro- teins in T. brucei procyclic forms identified 34 phosphoproteins (31). NOPP44/46 Tyr181, however, was not identified, likely due to experimental limitations of that study. By immunofluores- cence, phosphotyrosine proteins mostly associated with the cytoskeleton and the nucleolus, which is also the site of NOPP44/46 (32). A physiological interaction between TbPTP1 and NOPP44/46 would require cellular co-localization, and consistent with this tenet, TbPTP1 was found to associate with the cytoskeletal and nuclear fractions (10). However, it remains possible that NOPP44/46 is dephosphorylated outside of the nucleus as a recent study suggests that a pool of NOPP44/46 is exported out of the nucleus via exportin 1 (33). TbPTP1 is a molecular switch for the stumpy-to-procyclic transition as both genetic and pharmacological inhibition of the phosphatase lead to spontaneous differentiation of committed stumpy forms to procyclic forms in vitro (10). Because trypano- somatids do not use transcriptional control to regulate ex- pression of protein-coding genes (with a few exceptions), the key substrates of TbPTP1 that mediate this effect are likely to modulate mRNA stability, translation, or protein turnover. The identification of a known ribosome biogenesis protein, NOPP44/46 (8), as a potential in vivo substrate of TbPTP1 points toward a possible role in translational control. For exam- ple, the tyrosine phosphorylation state of NOPP44/46 may modulate ribosome biogenesis, which would in turn affect translational capacity. This possibility will be examined in future studies. Alternatively, NOPP44/46 may have uncharac- terized cellular functions in addition to its essential role in ribo- some biogenesis that could play specific roles in differentiation. In vivo, NOPP44/46 acts as a structural scaffold for several nucleolar proteins (34, 35), and its phosphorylation state may modulate these interactions to promote specific cellular changes. Studies in other organisms suggest the existence of functional links between proteins involved in ribosome biogen- esis and development, highlighting complex yet conserved modes of developmental regulation. In zebrafish, Pescadillo, an essential gene required for nucleolar assembly and 60 S biogen- esis, was originally discovered in a screen for regulators of embryonic development (36–39). In yeast, the Pescadillo ortholog, Yph1p, is found in two distinct multiprotein com- plexes with different functions in ribosome biogenesis and DNA replication (40). Interestingly, depletion of NOPP44/46 (44) and Yph1p both lead to cell cycle arrest with defective S-phase progression (40, 41). Furthermore, the ribosomal bio- genesis factors such as nucleolin and nucleophosmin play roles in the cytosol or nucleus distinct from their function in ribo- some biogenesis (42, 43). As other substrates of TbPTP1 are identified and tools for the study of TbPTP1 in vivo are refined, we will be better able to determine how these processes act together or apart to influence trypanosomatid differentiation and potentially provide insight into a novel signaling mecha- nism conserved in eukaryotic development. FIGURE 6. Electrostatic surface potential of TbPTP1. A, left, the TbPTP1 sur- face shows distinct electronegative (red) and positive (blue) regions, with a continuous electropositive stretch across the active site. The entry to the active site is indicated by the circle. Right, schematic representation in the sameorientationasleftpanel.B,electrostaticsurfacerepresentationofPTP1B (1SUG, left) and PTPRO (2G59, right) in the same orientation as TbPTP1. PTP1B also binds a highly negatively charged substrate and shows large electropos- itive areas. TbPTP1 Structure and Substrate 22080 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 29•JULY 16, 2010 Acknowledgments—We thank Christine L. Gee for help with model building and refinement, the staff at the Advanced Light Source Beamline 8.3.1 for help with data collection, and Carolina Vega and Charles Kifer for technical assistance. REFERENCES 1. Fe`vre, E. M., Wissmann, B. V., Welburn, S. C., and Lutumba, P. (2008) PLoS. Negl. Trop. Dis. 2, e333 2. Brun, R., Blum, J., Chappuis, F., and Burri, C. (2010) Lancet 375, 148–159 3. Fenn, K., and Matthews, K. R. (2007) Curr. Opin. Microbiol. 10, 539–546 4. Parsons, M., Valentine, M., Deans, J., Schieven, G. L., and Ledbetter, J. A. (1991) Mol. Biochem. Parasitol. 45, 241–248 5. García-Salcedo, J. A., Nolan, D. P., Gijo´n, P., Go´mez-Rodriguez, J., and Pays, E. (2002) Mol. Microbiol. 45, 307–319 6. Li, Z., and Wang, C. C. (2006) Eukaryot. Cell 5, 1026–1035 7. Mu¨ller, I. B., Domenicali-Pfister, D., Roditi, I., and Vassella, E. (2002) Mol. Biol. Cell 13, 3787–3799 8. Jensen, B. C., Brekken, D. L., Randall, A. C., Kifer, C. T., and Parsons, M. (2005) Eukaryot. Cell 4, 30–35 9. Parsons, M., Ledbetter, J. A., Schieven, G. L., Nel, A. E., and Kanner, S. B. (1994) Mol. Biochem. Parasitol. 63, 69–78 10. Szo¨or, B., Wilson, J., McElhinney, H., Tabernero, L., and Matthews, K. R. (2006) J. Cell Biol. 175, 293–303 11. Dean, S., Marchetti, R., Kirk, K., and Matthews, K. R. (2009) Nature 459, 213–217 12. Montalibet, J., Skorey, K. I., and Kennedy, B. P. (2005) Methods 35, 2–8 13. Wirtz, E., Leal, S., Ochatt, C., and Cross, G. A. (1999) Mol. Biochem. Para- sitol. 99, 89–101 14. Jensen, B. C., Kifer, C. T., Brekken, D. L., Randall, A. C., Wang, Q., Drees, B. L., and Parsons, M. (2007) Mol. Biochem. Parasitol. 151, 28–40 15. Panigrahi, A. K., Schnaufer, A., Carmean, N., Igo, R. P., Jr., Gygi, S. P., Ernst, N. L., Palazzo, S. S., Weston, D. S., Aebersold, R., Salavati, R., and Stuart, K. D. (2001) Mol. Cell. Biol. 21, 6833–6840 16. Rigaut, G., Shevchenko, A., Rutz, B., Wilm, M., Mann, M., and Se´raphin, B. (1999) Nat. Biotechnol. 17, 1030–1032 17. Goldschmidt, L., Cooper, D. R., Derewenda, Z. S., and Eisenberg, D. (2007) Protein Sci. 16, 1569–1576 18. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307–326 19. Collaborative Computational Project Number 4 (1994) Acta Crystallogr. D Biol. Crystallogr. 50, 760–763 20. Stein, N. (2008) J. Appl. Crystallogr. 41, 641–643 21. Adams, P. D., Grosse-Kunstleve, R. W., Hung, L. W., Ioerger, T. R., McCoy, A. J., Moriarty, N. W., Read, R. J., Sacchettini, J. C., Sauter, N. K., and Terwilliger, T. C. (2002) Acta Crystallogr. D Biol. Crystallogr. 58, 1948–1954 22. Emsley, P., and Cowtan, K. (2004) Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 23. Chen, V. B., Arendall, W. B., 3rd, Headd, J. J., Keedy, D. A., Immormino, R. M., Kapral, G. J., Murray, L. W., Richardson, J. S., and Richardson, D. C. (2010) Acta Crystallogr. D Biol. Crystallogr. 66, 12–21 24. Flint, A. J., Tiganis, T., Barford, D., and Tonks, N. K. (1997) Proc. Natl. Acad. Sci. U.S.A. 94, 1680–1685 25. Zhang, Z. Y., Clemens, J. C., Schubert, H. L., Stuckey, J. A., Fischer, M. W., Hume, D. M., Saper, M. A., and Dixon, J. E. (1992) J. Biol. Chem. 267, 23759–23766 26. Salmeen, A., Andersen, J. N., Myers, M. P., Tonks, N. K., and Barford, D. (2000) Mol. Cell 6, 1401–1412 27. Barr, A. J., Ugochukwu, E., Lee, W. H., King, O. N., Filippakopoulos, P., Alfano, I., Savitsky, P., Burgess-Brown, N. A., Mu¨ller, S., and Knapp, S. (2009) Cell 136, 352–363 28. Andersen, J. N., Mortensen, O. H., Peters, G. H., Drake, P. G., Iversen, L. F., Olsen, O. H., Jansen, P. G., Andersen, H. S., Tonks, N. K., and Møller, N. P. (2001) Mol. Cell. Biol. 21, 7117–7136 29. White, N. J. (2008) Malar. J. 7, Suppl. 1, S8 30. Zhang, Z. Y. (2002) Annu. Rev. Pharmacol. Toxicol. 42, 209–234 31. Nett, I. R., Davidson, L., Lamont, D., and Ferguson, M. A. (2009) Eukaryot. Cell 8, 617–626 32. Das, A., Peterson, G. C., Kanner, S. B., Frevert, U., and Parsons, M. (1996) J. Biol. Chem. 271, 15675–15681 33. Hellman, K., Prohaska, K., and Williams, N. (2007) Eukaryot. Cell 6, 2206–2213 34. Park, J. H., Brekken, D. L., Randall, A. C., and Parsons, M. (2002) Mol. Biochem. Parasitol. 119, 97–106 35. Park, J. H., Jensen, B. C., Kifer, C. T., and Parsons, M. (2001) J. Cell Sci. 114, 173–185 36. Allende, M. L., Amsterdam, A., Becker, T., Kawakami, K., Gaiano, N., and Hopkins, N. (1996) Genes Dev. 10, 3141–3155 37. Kinoshita, Y., Jarell, A. D., Flaman, J. M., Foltz, G., Schuster, J., Sopher, B. L., Irvin, D. K., Kanning, K., Kornblum, H. I., Nelson, P. S., Hieter, P., and Morrison, R. S. (2001) J. Biol. Chem. 276, 6656–6665 38. Lerch-Gaggl, A., Haque, J., Li, J., Ning, G., Traktman, P., and Duncan, S. A. (2002) J. Biol. Chem. 277, 45347–45355 39. Maiorana, A., Tu, X., Cheng, G., and Baserga, R. (2004) Oncogene 23, 7116–7124 40. Du, Y. C., and Stillman, B. (2002) Cell 109, 835–848 41. Das, A., Park, J. H., Hagen, C. B., and Parsons, M. (1998) J. Cell Sci. 111, 2615–2623 42. Inder, K. L., Lau, C., Loo, D., Chaudhary, N., Goodall, A., Martin, S., Jones, A., van der Hoeven, D., Parton, R. G., Hill, M. M., and Hancock, J. F. (2009) J. Biol. Chem. 284, 28410–28419 43. Okuda, M., Horn, H. F., Tarapore, P., Tokuyama, Y., Smulian, A. G., Chan, P. K., Knudsen, E. S., Hofmann, I. A., Snyder, J. D., Bove, K. E., and Fuka- sawa, K. (2000) Cell 103, 127–140 44. Worthen, C., Jensen, B. C., and Parsons, M. (2010) PLoS Negl. Trop. Dis. e678 TbPTP1 Structure and Substrate JULY 16, 2010•VOLUME 285•NUMBER 29 JOURNAL OF BIOLOGICAL CHEMISTRY 22081
3M4W
Structural basis for the negative regulation of bacterial stress response by RseB
PROTEIN STRUCTURE REPORT Structural basis for the negative regulation of bacterial stress response by RseB Dong Young Kim, Eunju Kwon, JongKeun Choi, Hye-Yeon Hwang, and Kyeong Kyu Kim* Department of Molecular Cell Biology, Sungkyunkwan University School of Medicine, Suwon 440-746, Korea Received 29 December 2009; Revised 13 March 2010; Accepted 16 March 2010 DOI: 10.1002/pro.393 Published online 29 March 2010 proteinscience.org Abstract: The rE-dependent stress response in bacterial cells is initiated by the DegS- and RseP-regulated intramembrane proteolysis of a membrane-spanning antisigma factor, RseA. RseB binds to RseA and inhibits its sequential cleavage, thereby functioning as a negative modulator of this response. In the crystal structure of the periplasmic domain of RseA bound to RseB, the DegS cleavage site of RseA is unstructured, however, its P1 residue is buried in the hydrophobic pocket of RseB, which suggests that RseB binding blocks the access of DegS to the cleavage site. Keywords: RseA; RseB; RseP; stress response; sigma factor; crystal Introduction Regulated intramembrane proteolysis (RIP) is a con- trol mechanism underlying transmembrane signal transfer and performs a key role in the initiation of the essential signal transduction pathways in diverse organisms.1 For example, the Notch signal- ing pathway, which is critical for a variety of cell– cell communications in multicellular organisms, is controlled by the RIP of the Notch receptor by ADAM-family metalloprotease and gamma- secretase.1 In Gram-negative bacteria, the sequen- tial cleavage of RseA, a membrane-spanning anti-rE factor, modulates the initiation of the envelope-stress response.2,3 RseA forms a tight complex with rE using its N-terminal cytoplasmic domain, thereby inhibiting the transcription of rE-dependent genes. Under stress conditions that include the misfolding of periplasmic proteins, two membrane proteases, DegS and RseP, sequentially degrade RseA to liber- ate rE (Supporting Information Fig. S1). DegS, which is activated when its PDZ domain is bound to the C-terminal peptide of unfolded outer membrane porins (OMPs), cleaves the C-terminal periplasmic domain of RseA.4 Subsequently, RseP cleavage within the membrane domain of RseA releases the cytoplasmic domain of RseA (associated with the rE) from the membrane. In the final step, the cyto- plasmic domain of RseA is degraded such that the released rE can interact with RNA polymerase.5 RseB also participates in the regulation of rE-dependent envelope-stress response by inhibiting Additional Supporting Information may be found in the online version of this article. Dong Young Kim’s current address is Department of Pharmaceutical Chemistry, University of California, San Francisco, 600 16th Street, San Francisco, CA 94107, USA. Grant sponsor: 21C Frontier Functional Proteomics Program; Grant number: FPR08B2-270; Grant sponsor: Korea Healthcare technology R&D Project; Grant number: A092006; Grant sponsor: Ubiquitome Research Program; Grant number: M105 33010001-05N3301-00100; Grant sponsor: National Research Laboratory Program; Grant number: NRL-2006-02287. *Correspondence to: Kyeong Kyu Kim, Department of Molecular Cell Biology, Sungkyunkwan University School of Medicine, Suwon 440-746, Korea. E-mail: kkim@med.skku.ac.kr 1258 PROTEIN SCIENCE 2010 VOL 19:1258—1263 Published by Wiley-Blackwell. V C 2010 The Protein Society the intramembrane proteolysis of RseA.6 RseB has been previously demonstrated to suppress the pro- teolytic activity of DegS for RseA,6,7 independently of the activation mechanism of DegS.8 Accordingly, either the deletion of the rseB gene or the release of RseB from RseA results in a more rapid degradation of RseA and increased activity of rE.6,9 In this study, we attempted to characterize the regulatory role of RseB in the proteolytic cleavage of RseA and determined the crystal structure of RseB in complex with RseAperi (the periplasmic domain of RseA, resi- dues 121–216) at a resolution of 2.3 A˚ by molecular replacement using apo-RseB (PDB ID: 2P4B) as a template (Table I). Results and Discussion Structure determination The RseAperiRseB structure was determined at a resolution of 2.3 A˚ by molecular replacement using the large domain (residues 26–200) of E.coli RseB (PDB ID: 2P4B) as a template.10 Although the asym- metric unit of the crystal harbors four RseAperiRseB complexes (Supporting Information Fig. S2), the di- meric structure of the complex has been known from size exclusion and SAXS (Small Angle X-ray Scatter- ing) data.11 Each complex (Com1–Com 4 in Support- ing Information Fig. S2) is composed of one RseB monomer and one RseAperi monomer. Com1 is in contact with two other complexes, Com2 and Com3. The Com1:Com2 interaction, which is mediated by hydrogen bonds between relatively well-conserved residues, buries the 1190 A˚ 2 surface area of each complex. RseB:RseB, RseAperi:RseAperi, and RseAper- i:RseB interfaces contribute to this burial of surface area by 889 A˚ 2, 57 A˚ 2, and 244 A˚ 2, respectively. The Com1:Com3 interaction results in the burial of 370 A˚ 2, which is primarily a RseB:RseB contact and involves a zinc ion that was added for the purposes of crystallization (Supporting Information Fig. S3). Moreover, the dimeric interaction between the N-ter- minal regions of the two RseAs in the Com1:Com2 dimer and their proximity to the transmembrane region demonstrate the involvement of the trans- membrane domain of RseA in dimeric contacts [Fig. 1(a) and Supporting Information Fig. S4]. There is no biologically relevant higher-order oligo- meric form that can be generated by symmetry operations. Accordingly, the Com1:Com2 dimer (or Com3:Com4 dimer) was considered biologically relevant [Fig. 1(a) and Supporting Information Fig. S2]. In this manuscript, we used Com1 and Com1:Com2 to describe the monomeric and dimeric RseAperiRseB complexes, respectively. The RMSD between Com1:Com2 and Com3:Com4 complexes was 0.43 A˚ for 626 Ca atoms. Overall structure We were able to model most residues in RseB, with the exception of the first N-terminal residue, resi- dues 240–246, and three C-terminal residues. The loop connecting the large (RseB25–209) and small (RseB217–315) domains was disordered in the apo- RseB model10,12; however, it was well ordered in the RseAperiRseB complex due to the interaction with the bound RseAperi, which results in the stabilization of the loop (Fig. 1). By way of contrast, the 96-resi- due periplasmic domain of RseA (residues 121–216) was largely unstructured, and only two regions involved in RseB binding, RseA132–151 and RseA169– 190, were modeled (Fig. 1 and Supporting Informa- tion Fig. S5). RseB evidenced similar structures in their apo- (Chain A of 2P4B or Chain A of 2V43) and RseAperi-bound states, with an RMSD of 0.85 A˚ for 276 Ca atoms (2P4B) or 1.61 A˚ for 261 Ca atoms (2V43), thereby indicating that its overall conforma- tion is largely maintained upon RseA binding. The major local conformational change was found in two b-strands (b5 and b6; residues 88–104) in the small domain of RseB (2P4B), which binds directly to the C-terminus of RseA132–151 [Fig. 1 and Supporting Information Fig. S6(a)]. The conformational changes are more drastic when the RseA-bound RseB was compared with another crystal structure of apo- RseB (PDB ID: 2V43) where four b-strands (b3–b6; residues 68–104) exhibit large conformational Table I. Data Collection and Refinement Statistics RseAperiRseB Data collection Space group P212121 Cell dimensions a, b, c (A˚ ) 87.05, 119.58, 150.67 a, b, c () 90.00, 90.00, 90.00 Resolution (A˚ ) 30.00–2.30 (2.38–2.30)a Rsym or Rmerge 6.0 (20.8) I/rI 20.0 (3.5) Completeness (%) 93.0 (81.5) Redundancy 4.6 Refinement Resolution (A˚ ) 20.00–2.30 No. reflections, working/free 62725/3320 Rwork/Rfree 23.9/27.1 No. atoms Protein 10108 Zn2þ 6 Water 421 B-factors Protein 52.1 Zn 82.3 Water 50.0 R.m.s. deviations Bond lengths (A˚ ) 0.008 Bond angles () 1.482 Ramachandran plot Most favored (%) 87.2 Additionally allowed (%) 12.7 Generously allowed (%) 0.2 Disallowed (%) 0.0 a Values in parentheses are for highest-resolution shell. Kim et al. PROTEIN SCIENCE VOL 19:1258—1263 1259 changes after binding to the C-terminus of RseA [Fig. 1 and Supporting Information Fig. S6(b)]. Interaction between RseAperi and RseB RseA binds to a broad area of the RseB groove that is formed between the large and small domains of RseB (Figs. 1 and 2). RseA132–151 mostly forms a random coil rather than a regular secondary struc- ture and interacts with residues in the large domain of RseB. The residues in RseA132–151 form hydropho- bic interactions with the hydrophobic residues or ali- phatic carbons of bulky residues of RseB, with the exception of Lys144, which forms a salt-bridge with Glu181 of RseB [Fig. 2(a)]. The results of the histi- dine pull-down assay verified that an RseAperi mu- tant featuring Ala substitutions at Gly143, Lys144, and Pro147 was still capable of binding to RseB (data not shown), thereby indicating that the electro- static interaction attributable to Lys144 is not crit- ically important to the association between RseB and RseA132–151. Consistent with this finding, Lys144 is not conserved among RseA homologues in Gram-negative bacteria (Fig. 3). RseA169–190 exhibits a helical conformation and binds principally to the small domain of RseB. The charged residues in RseA169–190 are well-conserved in the RseA homologues and are important to RseB binding (Figs. 2 and 3). Most notably, Arg172, Asp179, Glu181, and Arg184/Arg185 in RseA169–190 form salt bridges with Glu293, Arg239, Arg282, and Asp109 of RseB, respectively [Figs. 2(a,b)]. It has been demonstrated that RseA169–185 is the minimum fragment necessary for RseB binding8,10; addition- ally, the mutation of the conserved Arg residues in this fragment (Arg172, Arg184, and Arg185) abol- ishes RseB binding activity10,13 (Fig. 3). Therefore, RseA132–151 does not appear to be the primary deter- minant in RseB binding, but it may perform other additional functions, such as recruiting RseB or sterically inhibiting the access of proteases. Two RseAperiRseB complexes in a dimer (Com1:Com2) are also stabilized via intercomplex interactions [Fig. 1(a) and Supporting Information Fig. S2]. The Val135, Phe136, and Thr138 residues of RseA in Com1 are in contact with Ile50, Asn51/Thr179/ Gln182, and Arg169/Arg184 of RseB in Com2, respectively, and vice versa [Figs. 1(a),2(b)]. Structural implication of the binding of RseA to RseB The DegS cleavage site (Val148-Ser149; P1-P10) at the C-terminal end of RseA132–151, is located within the RseB groove in the RseAperiRseB complex [Figs. 1,2(c)]. Val148 is buried in the hydrophobic pocket formed by Phe100 and Leu102 of RseB and Leu182 of RseA169–190. Ser149, which is located near the he- lix in RseA169–190, forms a hydrogen bond with Gln178 of RseA. As a result, the DegS cleavage site is almost completely hidden by RseB and RseA169– 190, such that DegS access is restricted in the RseA- periRseB complex, and probably also in the RseAR- seB complex (Fig. 1). From this perspective, it has been theorized that the binding of RseA132–151 to RseB contributes to locating the cleavage site deep inside of the RseB groove, thereby rendering it re- sistant to DegS cleavage. This mechanism is consist- ent with the model proposed in Ref. 8. In the proteolytic cascade of RseA, the cleavage by RseP requires prior periplasmic cleavage by DegS and the release of RseA149–216. 5 It was reported Figure 1. Structure of the RseAperiRseB complex. (a) Dimer model of the RseAperiRseB complex. Each RseAperi is depicted in magenta or yellow, and RseB is depicted in green or slate. Ribbon diagram (b) and surface model (c) of the monomeric RseAperiRseB complex. The regions important for their binding, RseB25–209, RseB210–216, RseB217–315, RseA132–151, and RseA169–190, are colored slate, purple, red, green, and yellow, respectively. The DegS cleavage site is shown in blue. 1260 PROTEINSCIENCE.ORG Crystal Structure of RseB in Complex with RseA recently that the interaction of the newly exposed C-terminal residue of RseA1–148, Val 148, with the second PDZ domain of RseP is critically important for the cleavage.14 It is expected that the interaction between RseB and RseA1–148 is not very strong due to the lower binding affinity of RseA121–173 for RseB detected in previous biochemical studies.10 These findings suggest that RseB in complex with RseA149–216 will dissociate from RseA1–148 after DegS cleavage and that RseB is unlikely to Figure 2. RseAperiRseB interaction. (a) Schematic drawing of RseAperiRseB interaction. RseA residues located in random coils and helices are shown as orange circles and green pentagons, respectively. RseB residues in the same complex and from the other RseAperiRseB complex are marked as white and cyan boxes, respectively. Charge interactions, hydrogen bonds, and hydrophobic contacts are shown as red, blue, and black lines, respectively. (b) Binding interface between RseA and RseB. RseB, RseA132–151, and RseA169–190 are colored purple, green, and yellow, respectively. Dotted lines indicate charge interactions and the involved residues are depicted by stick models. (c) The DegS cleavage site (Val148-Ser149) bound to the RseB groove in the RseAperiRseB complex is drawn in a ribbon model with the same color scheme as in Fig. 1(b). The residues near the cleavage site are drawn as stick models and labeled. The cleavage site is indicated by a black arrow and labeled. Figure 3. Multiple sequence alignment of the periplasmic domain of RseAs from Gram-negative bacteria. Identical and similar residues are boxed in blue and yellow, respectively. Species abbreviations are as follows: Ec, Escherichia coli; Sf, Shigella flexneri; Se, Salmonella enterica; Yp, Yersinia pestis; Eca, Erwinia carotovora; Vc, Vibrio cholerae; So, Shewanella oneidensis; Hi, Haemophilus influenzae; Ms, Mannheimia succiniciproducens. Kim et al. PROTEIN SCIENCE VOL 19:1258—1263 1261 reassociate with RseA1–148. Therefore, RseP will interact with the C-terminal end of the DegS-cleaved RseA (RseA1–148). Materials and Methods Protein expression and purification E.coli RseAperi (periplasmic domain containing resi- dues 121–216) and RseB (residues 24–318) were expressed separately in E. coli BL21(DE3) as previ- ously described.10,11 His-Trx-RseAperi- and RseB- expressing cells were harvested and mixed at a ratio of 1:3 (wet weight) to ensure the formation of the complex. The cells were then sonicated in buffer A (20 mM Tris-HCl pH 7.5 and 0.1M NaCl). The RseAperiRseB complex was then purified by nickel- affinity chromatography and size exclusion chroma- tography. The cleared lysates were loaded onto a metal-chelating column (GE Healthcare, Princeton, NJ) and the proteins were eluted with 50–500 mM imidazole gradient. The fractions containing His- Trx-RseAperiRseB were pooled and dialyzed twice against buffer A. The His-Trx tag was removed with thrombin at room temperature, and RseAperiRseB was purified further using a Superdex-200 column (GE Healthcare, Princeton, NJ) pre-equilibrated with buffer A, then concentrated to 15 mg/mL. Crystallization and data collection The crystallization of the RseAperiRseB complex was performed using the microbatch method at 14C. The crystallization drop was prepared by mix- ing 1 lL protein solution (8–10 mg/mL) and 1 lL crystallization reagent (28% PEG550MME, 10 mM ZnSO4, and 100 mM MES pH 6.5) under a layer of Al’s oil (Hampton Research, Aliso Viejo, CA). The crystals in a drop were flash-frozen in a cold nitro- gen stream at 100 K without the addition of a cryo- protectant, and the diffraction data were collected at PLS-BL4A (Beam line 4A, Pohang Light Source, South Korea; wavelength 1.0000 A˚ ). The diffraction images were recorded to an ADSC Quantum 210 CCD detector. The diffraction data were indexed and integrated using HKL2000 and scaled using SCALEPACK.15 Structure determination The RseAperiRseB structure was determined by molecular replacement using PHASER16 with the large domain (residues 26–200) of E.coli RseB10 (PDB 2P4B) as a template. Three large domains were initially identified, and the small domains were added to the model. The fourth RseB was generated by a noncrystallographic symmetry operation. Sev- eral cycles of rigid body, positional, simulated annealing and B-factor refinements, and model rebuilding were conducted at a resolution of 2.3 A˚ using the CNS and COOT programs.17,18 The RseA model was placed on the additional electron density. The RseAperiRseB structure was refined further using REFMAC.19 The final refinement with sol- vents resulted in R and Rfree values of 23.8% and 27.0%, respectively. The data collection and refine- ment statistics are summarized in Table I. The fig- ures were drawn using PyMOL.20 The protein–pro- tein interface was calculated using Protorp.21 The coordinates and structure factors for the RseAperi RseB complex have been deposited under accession code 3M4W. Summary In this study, we characterized the inhibitory mecha- nisms of RseB in the regulated proteolysis of RseA. The C-terminal helix in RseA169–190 is a major con- tributor to the formation of a stable complex with RseB, whereas the N-terminal region of the periplas- mic domain of RseA is necessary for the burial of the DegS cleavage site within an inaccessible pocket in the RseARseB complex. In the regulation of the envelope-stress response, RseB functions by blocking the access of DegS protease rather than converting RseA into a compact structure that is resistant to proteolysis, as the random coil structure around the cleavage site is maintained in the complex. Accord- ingly, in this study we explain why the release of RseB is a prerequisite for the degradation of RseA and the activation of the rE-dependent envelope stress response at the atomic level. In the crystal structure of the RseARseB complex, RseB is unlikely to interfere with the further degradation of the periplasmically cleaved RseA. Therefore, the newly exposed Val148 will be readily recognized by RseP for subsequent cleavage, which results in the activation of the envelope-stress response. References 1. Brown MS, Ye J, Rawson RB, Goldstein JL (2000) Regulated intramembrane proteolysis: a control mecha- nism conserved from bacteria to humans. Cell 100: 391–398. 2. Alba BM, Leeds JA, Onufryk C, Lu CZ, Gross CA (2002) DegS and YaeL participate sequentially in the cleavage of RseA to activate the sigma(E)-dependent extracyto- plasmic stress response. Genes Dev 16:2156–2168. 3. Kanehara K, Ito K, Akiyama Y (2002) YaeL (EcfE) acti- vates the sigma(E) pathway of stress response through a site-2 cleavage of anti-sigma(E), RseA. Genes Dev 16: 2147–2155. 4. Walsh NP, Alba BM, Bose B, Gross CA, Sauer RT (2003) OMP peptide signals initiate the envelope-stress response by activating DegS protease via relief of inhi- bition mediated by its PDZ domain. Cell 113:61–71. 5. Flynn JM, Levchenko I, Sauer RT, Baker TA (2004) Modulating substrate choice: the SspB adaptor delivers a regulator of the extracytoplasmic-stress response to the AAAþ protease ClpXP for degradation. Genes Dev 18:2292–2301. 6. Grigorova IL, Chaba R, Zhong HJ, Alba BM, Rhodius V, Herman C, Gross CA (2004) Fine-tuning of the 1262 PROTEINSCIENCE.ORG Crystal Structure of RseB in Complex with RseA Escherichia coli sigmaE envelope stress response relies on multiple mechanisms to inhibit signal-independent proteolysis of the transmembrane anti-sigma factor, RseA. Genes Dev 18:2686–2697. 7. Chaba R, Grigorova IL, Flynn JM, Baker TA, Gross CA (2007) Design principles of the proteolytic cascade gov- erning the sigmaE-mediated envelope stress response in Escherichia coli: keys to graded, buffered, and rapid signal transduction. Genes Dev 21:124–136. 8. Cezairliyan BO, Sauer RT (2007) Inhibition of regu- lated proteolysis by RseB. Proc Natl Acad Sci USA 104: 3771–3776. 9. Collinet B, Yuzawa H, Chen T, Herrera C, Missiakas D (2000) RseB binding to the periplasmic domain of RseA modulates the RseA:sigmaE interaction in the cyto- plasm and the availability of sigmaE RNA polymerase. J Biol Chem 275:33898–33904. 10. Kim DY, Jin KS, Kwon E, Ree M, Kim KK (2007) Crys- tal structure of RseB and a model of its binding mode to RseA. Proc Natl Acad Sci USA 104:8779–8784. 11. Jin KS, Kim DY, Rho Y, Le VB, Kwon E, Kim KK, Ree M (2008) Solution structures of RseA and its complex with RseB. J Synchrotron Radiat 15:219–222. 12. Wollmann P, Zeth K (2007) The structure of RseB: a sensor in periplasmic stress response of E. coli. J Mol Biol 372:927–941. 13. Ahuja N, Korkin D, Chaba R, Cezairliyan BO, Sauer RT, Kim KK, Gross CA (2009) Analyzing the interac- tion of RseA and RseB, the two negative regulators of the sigmaE envelope stress response, using a combined bioinformatic and experimental strategy. J Biol Chem 284:5403–5413. 14. Li X, Wang B, Feng L, Kang H, Qi Y, Wang J, Shi Y (2009) Cleavage of RseA by RseP requires a carboxyl- terminal hydrophobic amino acid following DegS cleav- age. Proc Natl Acad Sci USA 106:14837–14842. 15. Otwinowski Z, Minor W (1997) Processing of X-ray dif- fraction data collected in oscillation mode. Methods Enzymol 276:307–326. 16. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ (2007) Phaser crystallo- graphic software. J Appl Crystallogr 40:658–674. 17. Brunger AT, Adams PD, Clore GM, DeLano WL, Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges M, Pannu NS, Read RJ, Rice LM, Simonson T, Warren GL (1998) Crystallography & NMR system: a new soft- ware suite for macromolecular structure determination. Acta Crystallogr D54:905–921. 18. Emsley P, Cowtan K (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr D60:2126–2132. 19. Murshudov GN, Vagin AA, Dodson EJ (1997) Refine- ment of macromolecular structures by the maximum- likelihood method. Acta Crystallogr D53:240–255. 20. DeLano WL (2006) The PyMOL molecular graphics system. CA: DeLano Scientific LLC. 21. Reynolds C, Damerell D, Jones S (2009) Protorp: a pro- tein-protein interaction analysis tool. Bioinformatics 25:413–414. Kim et al. PROTEIN SCIENCE VOL 19:1258—1263 1263
3M53
SET7/9 in complex with TAF10 peptide and AdoHcy
SET7/9 Catalytic Mutants Reveal the Role of Active Site Water Molecules in Lysine Multiple Methylation*□ S Received for publication,February 15, 2010, and in revised form, July 14, 2010 Published, JBC Papers in Press,August 1, 2010, DOI 10.1074/jbc.M110.114587 Paul A. Del Rizzo‡, Jean-Franc¸ois Couture‡§1, Lynnette M. A. Dirk¶, Bethany S. Strunk‡, Marijo S. Roiko‡, Joseph S. Brunzelle, Robert L. Houtz¶, and Raymond C. Trievel‡2 From the ‡Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109, the §Ottawa Institute of Systems Biology, Department of Biochemistry, Microbiology, and Immunology, University of Ottawa, Ottawa, Ontario K1H 8M5, Canada, the ¶Department of Horticulture, Plant Physiology/Biochemistry/Molecular Biology Program, University of Kentucky, Lexington, Kentucky 40546, and the Department of Molecular Pharmacology and Biological Chemistry, Northwestern University Center for Synchrotron Research, Life Sciences Collaborative Access Team, Argonne, Illinois 60439 SET domain lysine methyltransferases (KMTs) methylate specific lysine residues in histone and non-histone substrates. These enzymes also display product specificity by catalyzing dis- tinct degrees of methylation of the lysine -amino group. To elucidate the molecular mechanism underlying this specificity, we have characterized the Y245A and Y305F mutants of the human KMT SET7/9 (also known as KMT7) that alter its prod- uct specificity from a monomethyltransferase to a di- and a tri- methyltransferase, respectively. Crystal structures of these mutants in complex with peptides bearing unmodified, mono-, di-, and trimethylated lysines illustrate the roles of active site water molecules in aligning the lysine -amino group for methyl transfer with S-adenosylmethionine. Displacement or dissocia- tion of these solvent molecules enlarges the diameter of the active site, accommodating the increasing size of the methylated -amino group during successive methyl transfer reactions. Together, these results furnish new insights into the roles of active site water molecules in modulating lysine multiple meth- ylation by SET domain KMTs and provide the first molecular snapshots of the mono-, di-, and trimethyl transfer reactions catalyzed by these enzymes. SET domain enzymes represent a family of S-adenosylmethi- onine (AdoMet)3-dependent methyltransferases that catalyze the site-specific methylation of protein lysyl residues in a host of proteins, including histones, transcription factors, chroma- tin-modifying enzymes, ribosomal subunits, and other sub- strates (1–3). In many instances, these modifications serve to recruit effector proteins that recognize methyl-lysyl residues in a sequence-dependent fashion (4). In addition, SET domain KMTs exhibit product specificity, defined as their ability to cat- alyze mono-, di-, or trimethylation of the lysine -amino group. This specificity is biologically relevant because many methyl- lysine-binding proteins can discriminate among different degrees of lysine methylation (4). Thus, both the site and degree of lysine methylation are critical to recognition by effector proteins. Structural and functional studies have identified a Phe/Tyr switch in the active site of SET domain KMTs that governs their respective product specificities (5, 6). According to this model, KMTs that possess a tyrosine in the Phe/Tyr switch site are limited to catalyzing lysine monomethylation, whereas en- zymes that possess a phenylalanine or another hydrophobic residue in this position display di- or trimethyltransferase activ- ity. Mutational analysis of various SET domain KMTs, includ- ing DIM-5 (KMT1) (6), SET7/9 (6), G9A (KMT1C) (5), and SET8 (also known as PR-SET7 and KMT5A) (7, 8), has demon- strated that substitutions in the Phe/Tyr switch result in pre- dictable changes in product specificity. Several models have been proposed to explain the mechanism by which the Phe/Tyr switch site governs this specificity, including variations in the diameter of the active site due to the size of Phe/Tyr switch residue and steric hindrance by the tyrosine hydroxyl group (6, 9–11). However, our recent studies of the Phe/Tyr switch mutant Y334F in the human histone H4 Lys-20 (H4K20) meth- yltransferase SET8 indicate that the Phe/Tyr switch regulates product specificity via a more subtle mechanism (8). Specifi- cally, the switch modulates the binding of an active site water molecule that in turn regulates the transition from mono- methylation to multiple methylation. Among the KMTs that have been structurally characterized, SET7/9 has emerged as an archetypal model for studying the catalytic mechanism and product specificity of the SET domain family. Although initially isolated as a histone H3 Lys-4 (H3K4)-specific methyltransferase, this KMT has been shown to regulate the functions of numerous non-histone substrates through site-specific methylation (12–21). Early structural and functional studies of SET7/9 identified two active site mutants, * Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth Grant R01 GM073839 (to R. C. T.) and National Institutes of Health Admin- istrative Supplement GM073839-04S1 (to R. C. T.) funded through the American Recovery and Reinvestment Act. This work was also supported by Department of Energy Grant DE-FG02-92ER20075 (to R. L. H.). □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Table 1. The atomic coordinates and structure factors (codes 3M53, 3M54, 3M55, 3M56, 3M57,3M58,3M59,and3M5A)havebeendepositedintheProteinDataBank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). 1 Supported by a Canadian Institutes of Health Research postdoctoral fellowship. 2 To whom correspondence should be addressed: Dept. of Biological Chem- istry, University of Michigan Medical School, 1150 West Medical Center Dr., 5301 MSRB III, Ann Arbor, MI 48109. Tel.: 734-647-0889; Fax: 734-763-4581; E-mail: rtrievel@umich.edu. 3 The abbreviations used are: AdoMet, S-adenosylmethionine; AdoHcy, S-ad- enosylhomocysteine; KMT, lysine methyltransferase; ITC, isothermal titra- tion calorimetry; CH–O, carbon-oxygen hydrogen bond; BisTris, 2-[bis(2- hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 41, pp. 31849–31858, October 8, 2010 © 2010 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31849 Y245A and Y305F, which change its product specificity. The Phe/Tyr switch mutant Y305F alters SET7/9 product specificity from a mono- to dimethyltransferase (6), whereas the Y245A substitution converts the enzyme into a trimethyltransferase with weak monomethyltransferase activity (11). These mutants have been the subjects of numerous molecular modeling simu- lations that have led to various models to explain their distinct product specificities (22–26). However, the lack of structural data for the SET7/9 Y245A and Y305F mutants in complex with cognate methylated peptides has hindered our understanding of the mechanisms that define the respective product specific- ities of these mutants. Moreover, these structures would yield a framework for visualizing the mono-, di-, and trimethylation reactions catalyzed by SET domain KMTs. To gain insight into the molecular basis of their product specificities, we have determined high resolution crystal structures of the SET7/9 Y245A and Y305F mutants in com- plex with peptides of the TATA box-binding protein-associ- ated factor TAF10 bearing the Lys-189 methylation site in unmodified (K189), monomethylated (K189me1), dimethyl- ated (K189me2), and trimethylated (K189me3) states. The structures and accompanying biochemical data support a model whereby changes in the occupancy or position of water molecules in the active site are critical in establishing the prod- uct specificities of the SET7/9 Y245A and Y305F mutants. Together, our results provide new insights into the mechanisms that govern SET domain product specificity and provide step- wise snapshots of the lysine mono-, di-, and trimethyl transfer reactions catalyzed by KMTs. EXPERIMENTAL PROCEDURES Cloning, Expression, and Purification of the SET7/9 Mutants— The Y245A and Y305F mutants were introduced into the pHIS2 SET7/9 expression vector encoding residues 110–366 (27) via QuikChange site-directed mutagenesis (Stratagene) and were verified by dideoxy DNA sequencing. The plasmids encoding wild type (WT) SET7/9 and the Y245A and Y305F mutants were transformed into Rosetta2 DE3 cells (Novagen) and were expressed as described previously (27, 28). In the course of characterizing WT SET7/9, we observed that the enzyme co- purified with AdoMet or another contaminant that resulted in technical difficulties in the isothermal titration calorimetry (ITC) experiments and co-crystallization trials with the TAF10 peptides. To overcome this problem, a denaturation and refold- ing step was inserted in the purification scheme. The denatur- ation and refolding protocol involved adding 6 M guanidine HCl, 50 mM sodium phosphate, pH 7.5, and 500 mM NaCl to the protein while it was immobilized on a nickel-Sepharose column (GE Healthcare). The column was washed with this buffer, fol- lowed by a solution of 6 M urea, 50 mM sodium phosphate, pH 7.5, 150 mM NaCl, and 10 mM 2-mercaptoethanol to remove the cofactor from the denatured enzyme. A reverse gradient from 6 to 0 M urea was then performed in the same buffer to refold the protein, which was subsequently eluted from the column using a linear gradient of 0–500 mM imidazole in 50 mM sodium phosphate, pH 7.5, 500 NaCl, and 10 mM 2-mercaptoethanol. The refolded protein was digested with tobacco etch virus pro- tease (29) during dialysis against 20 mM Tris, pH 7.5, 150 mM NaCl, and 5 mM 2-mercaptoethanol and then purified using a Superdex 200 gel filtration column (GE Healthcare). Protein concentration was determined by its absorbance at 280 nm. Synthetic Peptides—The TAF10 peptides bearing K189, K189me1, K189me2, and K189me3 (sequence, acetyl- SKSK189DRKYTL-amide) and a biotinylated TAF10 peptide (sequence, acetyl-SKSK189DRKYTLT(K-EZLink-S-S-biotin)- amide) were synthesized and purified by New England Peptide, Inc. Peptide concentrations were measured using the absorb- ance of their tyrosine residue at 274 nm. Crystallization and Data Collection—Crystals were pro- duced by hanging drop vapor diffusion by mixing the crystalli- zation solution in a 1:1 ratio with 10–20 mg/ml SET7/9, 1 mM S-adenosylhomocysteine (AdoHcy), and 1.0–3.0 mM unmodi- fied or methylated TAF10-K189 peptide in 20 mM Tris, pH 8.0, 100 mM NaCl, and 2 mM tris(2-carboxyethyl)phosphine. Crys- tals were obtained at 20 °C in either 1.8–2.0 M (NH4)2SO4, with 0.1 M BisTris, pH 6.2–6.6, or in 0.95–1.1 M sodium citrate with 100 mM imidazole pH 8.0–8.4. In both crystallization condi- tions, the final pH values were between pH 8.0 and 9.0. Crystals in the (NH4)2SO4 condition were typically flash-frozen in the mother liquor containing 25–30% glycerol, and the crystals in the citrate condition were frozen in 1.6 M sodium citrate. Data were collected at the Advanced Photon Source beamlines 21-IDG (LS-CAT) and 23-IDD (GM/CA-CAT). Images were indexed, integrated, and scaled using HKL2000 (30). Structures of the mutants were solved by molecular replacement using MOLREP (31) with the coordinates of a previously reported SET7/9 ternary complex used as the search model (Protein Data Bank code 2F69). Successive rounds of model building and refinement were carried out using Coot (32) and REFMAC (33), respectively. The geometry of the models were verified by Mol- Probity (34). Simulated annealing omit maps were calculated using CNS (35) with the peptide and cofactor removed to elim- inate model bias in the active site. Structural figures were ren- dered using PyMOL (Schro¨dinger, LLC.). Fluorescent Methyltransferase Assay—A coupled fluorescent methyltransferase assay was used to measure the kinetic parameters of WT SET7/9 and the Y245A and Y305F mutants as reported previously, with the exception that 50–150 nM enzyme, 100 M AdoMet, and varying concentrations of TAF10 peptide substrate were used (27, 36). Assays were performed in triplicate, and a homocysteine calibration curve was used to calculate the initial velocities. Kinetic parameters were calcu- lated by plotting the velocities versus peptide concentration and by fitting the Michaelis-Menten equation to the data via non- linear regression using Prism 5.0 (GraphPad). In cases where the Km value was beyond the measurable range of the assay, the kcat/Km value was determined as described previously (7). Calorimetry Experiments—ITC was performed at 20 °C using a MicroCal VP-ITC calorimeter (GE Healthcare) with 0.12 mM protein and 4 mM AdoHcy in 20 mM sodium phosphate, pH 7, and 100 mM NaCl with 1.5 mM peptide as the injectant. Data were processed, and equilibrium dissociation constants (KD) and curve fitting errors were calculated from the binding iso- therms using Origin 7.0 (OriginLab Corp.). WT SET7/9 and the Y245A and Y305F mutants displayed ligand:protein binding stoichiometries (N values) between 0.8 and 1.0, demonstrating Lysine Methylation by SET7/9 Mutants 31850 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 that WT SET7/9 and its mutants were properly refolded due to their ability to bind peptides in an 1:1 molar ratio. TLC Product Analysis—Methyltransferase assays were per- formed in triplicate at 37 °C with the biotinylated TAF10 pep- tide (0.5 mM) for 30 min in the presence of WT SET7/9 (3 pmol), Y305F mutant (6 pmol), or the Y245A mutant (100 pmol). Assays contained 100 mM HEPES, pH 8.0, 50 mM NaCl, 1.0 mM [methyl-3H]AdoMet (diluted to a specific activity of 0.2 Ci/nmol with purified AdoMet (37)), 15 M Sulfolobus solfa- taricus AdoHcy hydrolase (36), and 2 units of adenosine deami- nase (Roche Applied Science) in a final volume of 20 l. The reactions were terminated by addition of an equal volume of 200 mM MES, pH 5.1, prior to addition of a 2-fold molar excess of immobilized avidin resin (UltraLink; Pierce). Biotinylated peptides were allowed to bind at room temperature for 30 min, and the resin was then collected by centrifugation (9000  g). The resin was washed three times with 300 mM NaCl, and the peptide was eluted overnight from the avidin resin by cleavage of the disulfide bond in the linker of the peptide using 10 mM tris(2-carboxyethyl)phosphine (Thermo Scientific). The resin was incubated with additional 10 mM tris(2-carboxyethyl)phos- phine the following day until the radiolabel was essentially removed from the resin. The recovered peptides were hydro- lyzed by incubation with 6 M HCl at 110 °C for 24 h. Subsequent steps in measuring the radiolabel incorporated into the mono-, di-, and trimethyl-lysine products were performed as reported previously (8). RESULTS Functional Analysis of the SET7/9 Y305F Mutant—Prior studies of SET7/9 by Zhang et al. (6) reported that mutation of the Phe/Tyr switch residue Tyr-305 to a phenylalanine alters its product specificity from a mono- to dimethyltransferase. We verified these findings by demon- strating that WT SET7/9 mono- methylated the TAF10-K189 pep- tide, whereas the Y305F mutant mono- and dimethylated this sub- strate, as demonstrated by mass spectrometry (data not shown). We next examined whether the Y305F substitution altered the affinity of SET7/9 for the TAF10-K189 peptides using ITC (Fig. 1). A comparison of the KD values revealed that SET7/9 Y305F bound the TAF10-K189 and TAF10-K189me1 peptides 4- and 6-fold more tightly, respec- tively, than the WT enzyme, whereas this mutant displayed a substantially diminished affinity for the TAF10-K189me2 pep- tide (Table 1). Although the WT enzyme and the Y305F mutant exhibited discernable differences in their affinities for the unmodified and monomethylated peptides, these variations are modest and cannot account for their distinct product specific- ities, suggesting that a kinetic effect during methylation may be responsible. To investigate this possibility, we characterized the kinetic parameters of WT SET7/9 and the Y305F mutant using the TAF10 peptides as substrates. Both enzymes methylated the unmodified peptide with comparable kcat and Km values (Table 2). In analyzing the kinetic parameters for the methylation of the monomethylated peptide by SET7/9 Y305F, we found that this substrate displayed an elevated Km value that was beyond the measurable range of the assay due to its limited solubility. In this case, we measured the catalytic efficiency (kcat/Km) for the methylation of this peptide and found that it was methylated 15-fold less efficiently than the unmodified peptide by SET7/9 Y305F. Given the fact that the Y305F mutant exhibited a higher binding affinity for the TAF10-K189me1 peptide than the WT enzyme (Table 1), the kinetic data suggest that a step in the reaction pathway following substrate binding limits the catalytic efficiency of this mutant. We next examined whether the Y305F mutant dimethylated the TAF10-K189 peptide via a processive or a distributive mechanism. In a processive mechanism, the methyl-lysine substrate would remain bound to the enzyme during successive methyl transfer reactions; thus, the concentration of an inter- mediate, such as monomethyl-lysine, cannot exceed the en- zyme concentration during the assay. In a distributive mecha- nism, the intermediates are released into solution where they accumulate prior to the next round of methylation, resulting in an intermediate concentration that is greater than that of the enzyme. Using a radiometric TLC assay and a biotinylated TAF10 peptide, we quantified the amounts of monomethylated products generated by the WT SET7/9 and the Y305F mutant FIGURE 1. ITC analysis of WT SET7/9 and the Y245A and Y305F mutants. Representative titrations and binding isotherms for the calorimetry experiments are shown for the unmodified TAF10-K189 peptide titrated into WT SET7/9 (left plot), SET7/9 Y305F (middle plot), and SET7/9 Y245A (right plot). In each panel, the ITC titration experiment (upper panel) is shown with the binding isotherms (lower panel) fit to a single binding site model. TABLE 1 Analysis of the binding affinity of WT SET7/9 and its catalytic mutants for unmodified and methylated TAF10 peptides SET7/9 TAF10 peptide KD a M WT K189 4.9  0.20 WT K189me1 4.0  0.36 Y305F K189 1.3  0.10 Y305F K189me1 0.62  0.065 Y305Fb K189me2 70 Y245A K189 4.0  0.25 Y245A K189me1 3.3  0.10 Y245A K189me2 5.8  0.22 Y245A K189me3 11  0.28 a Curve fitting errors were calculated from the binding isotherms. b An estimate of the affinity is reported due to weak peptide binding. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31851 (Table 3). The data demonstrate that comparable amounts of monomethyl-lysine were generated when the quantity of enzyme usedistakenintoaccount,inagreementwiththeirsimilar turnover numbers for the TAF10-K189 peptide (Table 2). The Y305F mutant also produced small but measurable quantities of radiolabeled dimethyl-lysine product that were substantially smaller than the amount of monomethyl-lysine generated. Therefore, the TLC data are consistent with a distributive mechanism for dimethylation by the Y305F mutant because the amount of monomethyl-lysine produced exceeded the quantity of enzyme used in the assay. Structures of WT SET7/9 and the Y305F Mutant in Complex with Unmodified and Methylated TAF10 Peptides—To deter- mine the mechanism by which the Y305F substitution alters the product specificity of SET7/9, we determined the crystal struc- tures of this mutant bound to AdoHcy and TAF10-K189, TAF10-K189me1, and TAF10-K189me2 peptides and com- pared these to the structures of the WT SET7/9AdoHcy TAF10-K189 complex (supplemental Table 1). The structures of these complexes were determined to 1.85 Å or higher reso- lution, permitting unambiguous modeling of the K189 side chains in the active site of the enzyme based on simulated annealing omit maps (Fig. 2). The ternary complexes of the WT and the Y305F mutant superimpose with overall root mean square differences of less than 0.3 Å for all aligned atoms, indi- cating that neither the Y305F mutation nor the binding of the various TAF10-K189 peptides results in substantial changes in its overall structure. An inspection of the active sites of the SET7/9 WT and Y305F complexes illustrates the binding modes of the unmod- ified and methylated forms of K189 in the TAF10 peptides (Fig. 2, A–D). The K189 side chain binds in an extended all trans conformation in a deep pocket, termed the lysine binding chan- nel, that is composed of Tyr-245, Thr-266, Leu-267, Ser-268, Tyr-305, Tyr-335, and Tyr-337 (not shown for clarity) (Fig. 2A). These residues interact with the aliphatic portion of the K189 side chain primarily through van der Waals contacts. The lysine binding channel connects to the AdoMet-binding site on the opposite face of the catalytic domain via an oxygen-lined methyl transfer pore (38). During catalysis, the methyl group of the cofactor is positioned within the methyl transfer pore for the SN2 reaction with the -amino group of the lysine or methyl-lysine substrate (see below). To lower the activation barrier for this reaction, the lysine -amine nucleophile is aligned for methyl transfer through a hydrogen bond network within the active site. In the WT enzyme, the K189 -amino group hydrogen bonds to the hydroxyl group of Tyr-245 as well as to two water molecules (Fig. 2A). One of the water molecules (termed water 1), is coor- dinated in a solvent pocket, through hydrogen bonds to the carbonyl oxygens of Gly-292 and Ala-295 and to the hydroxyl group of the Phe/Tyr switch residue Tyr-305. This solvent pocket is structurally conserved in SET domain KMTs and has an important role in defining product specificity through the adjacent Phe/Tyr switch residue, as shown in our prior studies of the human H4K20 methyltransferase SET8 (8). The other water molecule is bound within the methyl transfer pore between the lysine substrate and the thioether sulfur atom of AdoHcy through hydrogen bonds to Tyr-245, Asn-265, and His-293 in SET7/9 and the TAF10-K189 -amino group. This water is not observed in other structures of SET7/9 ternary complexes and may represent the approximate position that the AdoMet methyl group occupies in the methyl transfer pore in the Michaelis complex. In structures of the Y305F ternary complexes, the K189, K189me1, and K189me2 side chains also adopt extended trans side chain geometries within the lysine binding channel that are stabilized via hydrogen bonding to Tyr-245 in the enzyme (Fig. 2, B–D). The orientations of the K189me1 and K189me2 side chains are further maintained through carbon-oxygen (CH–O) hydrogen bonding between the methyl groups and oxygen atoms within the vicinity of the methyl transfer pore, as reported previously in other SET domain KMT structures (8, 10, 38). A superimposition of the SET7/9 WT and Y305F com- plexes underscores the similarity of the lysyl binding conforma- tions (Fig. 2E). However, there are notable differences in the hydrogen bond patterns and occupancy of water 1 within the solvent pocket in the Y305F mutant compared with the WT enzyme. Specifically, the Y305F substitution results in the loss of one hydrogen bond to water 1 in the structures of the TAF10- K189 and TAF10-K189me1 complexes (Fig. 2, B and C). In con- trast, water 1 is absent in TAF10-K189me2 complex, and the vacated solvent pocket is occupied by one of the methyl groups TABLE 2 Kinetic parameters of WT SET7/9 and the Y305F and Y245A mutants Enzyme TAF10 peptide substrate Km a kcat a kcat/Km a M min1 M1 min1  103 WT K189 160  17 17  0.62 110  17 Y305F K189 88  5.0 17  0.30 190  11 Y305Fb K189me1 11  0.50 Y245A K189 200  35 0.53  0.04 2.6  0.47 Y245A K189me1 210  23 5.9  0.23 28  3.3 Y245A K189me2 400  29 6.5  0.16 15  1.2 a Curve fitting errors were calculated from the hyperbolic fits for the Michaelis-Menten equation. b Km and kcat were not determined because the Km value was beyond the measurable range; therefore, the kcat/Km value is reported. TABLE 3 Product analysis of WT SET7/9 and the Y305F and Y245A mutants Enzyme Quantity of enzyme Measured product Amount of product formeda nmol nmol WT 0.003 Kme1 0.65  0.07 Y305F 0.006 Kme1 1.5  0.49 Kme2 0.033  0.009 Y245A 0.100 Kme1 0.80  0.22 Kme2 0.39  0.021 Kme3 0.076  0.019 a Standard deviation was calculated from triplicate measurements. Lysine Methylation by SET7/9 Mutants 31852 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 of the dimethyl -amine (Fig. 2D). This methyl group forms a 3.5-Å CH–O hydrogen bond to the carbonyl oxygen of Ala-295, further stabilizing the binding of the dimethyl-lysine side chain. A homologous dimethyl-lysine-binding mode and CH–O hydrogen bond was observed in our prior structural studies of the SET8 Y334F Phe/Tyr switch mutant that confers an analo- gous change in product specificity from a mono- to a dimeth- yltransferase (8). A structural alignment of the active sites of the SET7/9 Y305F and SET8 Y334F mutants bound to cognate dimethylated peptides illustrates that the coordinates of the dimethyl-lysyl side chains are virtually superimposable, with one methyl group oriented toward the methyltransfer pore and the second positioned within the vacant solvent pocket (Fig. 2F). Taken together, the structures of the SET7/9 Y305F com- plexes and the similarities in the dimethyl-lysine conforma- tions in the SET7/9 Y305F and SET8 Y334F mutants imply that the Phe/Tyr switch governs product specificity through a con- served mechanism whereby it indirectly influences the binding modes of the methyl-lysine side chain by modulating the affin- ity of the water molecule (water 1) bound in the solvent pocket. Biochemical Characterization of the SET7/9 Y245A Mutant—Previ- ous studies by Xiao et al. (11) reported that the Y245A mutation yields an unusual change in the product specificity of SET7/9, converting the enzyme to a trimeth- yltransferase with weak monometh- yltransferase activity. We deter- mined that the SET7/9 Y245A could mono-, di-, and trimethylate the TAF10-K189 peptide by mass spec- trometry (data not shown) and TLC (Table 3), confirming the earlier studies of Xiao et al. (11). ITC analysis revealed that the Y245A mutant displayed comparable KD values for the unmodified and methylated TAF10-K189 peptides (Fig. 1), although its affinity for the trimethylated peptide was modestly diminished in comparison with the other peptides (Table 1). The ITC data demonstrate that the Y245A mutant bound the unmodified, mono-, and dimethylated sub- strates with equivalent affinities, suggesting that a kinetic effect or a structural alteration in the active site may be responsible for its diminished activity toward un- modified substrates. To gain further insight into its peculiar product specificity, we characterized the kinetic properties of the SET7/9 Y245A mutant. Steady state analysis demonstrated that this mutant displayed similar Km values for the unmodified, mono- and dimethylated TAF10 peptides (Table 2). However, the turnover number for the TAF10-K189 peptide was diminished over 10-fold versus the methylated peptides and was reduced 30-fold versus the WT enzyme, in agreement with the weak monomethyltransferase activity reported by Xiao et al. (11). In addition, we investigated whether this mutant catalyzes lysine trimethylation via a pro- cessive or distributive mechanism as described for SET7/9 Y305F. The TLC data illustrate that the mono- and dimethyl- lysine intermediates accumulated at quantities greater than that of the enzyme used in the assay, indicating that SET7/9 Y245A obeys a distributive mechanism, analogous to the Y305F mutant (Table 3). Structures of SET7/9 Y245A Bound to Unmodified and Meth- ylated TAF10 Peptides—To elucidate the mechanism underly- ing its unusual product specificity, we determined the crystal structures of SET7/9 Y245A in complex with AdoHcy and unmodified, mono-, di-, and trimethylated TAF10 peptides (supplemental Table 1). These complexes superimpose with the structure of the WT SET7/9AdoHcyTAF10-K189 com- FIGURE 2. Crystal structures of WT SET7/9 and the Y305F mutant in complex with AdoHcy and unmodi- fied and methylated TAF10 peptides. A, active site of WT SET7/9 bound to AdoHcy and TAF10-K189 and the active site of SET7/9 Y305F bound to AdoHcy and TAF10-K189 (B), TAF10-K189me1 (C), and TAF10-K189me2 peptides (D). A–D, the K189 residue in the TAF10 peptide is depicted with yellow carbon atoms with the corresponding Fo – Fc simulated-annealing omit maps (contoured at 2.5 ) shown. AdoHcy is colored with green carbon atoms, and SET7/9 carbon atoms are colored gray, with the exception of Tyr-305/Y305F (magenta) and Tyr-245 (cyan). Gly-264 to Ser-268 are shown, and the main chain atoms of Gly-292 to Ala-295 are depicted for clarity. Red dashed lines indicate conventional hydrogen bonds, and blue dashed lines indicate CH–O hydrogen bonds. Cyan dashed lines indicate weak CH–O hydrogen bonds between 3.5 and 3.7 Å in length. The water molecule in the solvent binding pocket is numbered 1. E, overlay of the active sites of the WT enzyme (light blue carbon atoms) and the Y305F mutant (green, yellow, and pink carbon atoms correspond to K189, K189me1, and K189me2, respectively). Red dashed lines indicate coordination of the waters in the struc- turesoftheY305Fmutant,andthelightbluedashedlineindicateshydrogenbondingthatonlyoccursintheWT enzyme. F, superimposition of the active sites of the SET7/9 Y305FAdoHcyTAF10-K189me2 and SET8 Y334FAdoHcyH4K20me2 (Protein Data Bank code 3F9X) complexes shown with green and yellow carbon atoms, respectively. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31853 plex with root mean squared differences of less than 0.4 Å for all aligned atoms, indicating that the Y245A mutant does not per- turb the overall structure of the enzyme. Simulated annealing omit maps illustrate that K189 side chains are bound within the lysine binding channel through hydrogen bonds and van der Waals contacts (Fig. 3, A–D), although the interactions and binding modes are distinct from those in the complexes of WT SET7/9 and the Y305F mutant (Fig. 2, A–D). In the unmodified TAF10 peptide complex, the K189 -amino group forms a weak hydrogen bond with the carbonyl oxygen of Gly-264 (Fig. 3A), whereas the -amino groups of K189me1 and K189me2 hydro- gen bond to the hydroxyl group of Tyr-305 in the mono- and dimethylated peptide complexes (Fig. 3, B and C). The confor- mations of the K189me1 and K189me2 side chains are further stabilized by water-mediated hydrogen bonding and through CH–O hydrogen bonding to their methyl groups. In the TAF10-K189me3 peptide complex, the trimethyl-lysine side chain is coordinated exclusively through direct and water-me- diated CH–O hydrogen bonds to its methyl groups because the quaternary -ammonium cation cannot engage in hydrogen bonding (Fig. 3D). A structural alignment of the four SET7/9 Y245A complexes illustrates distinct binding modes for the unmodified versus the methylated K189 side chains, highlighting the selectivity of this mutant for methylated substrates. The side chains of K189me1, K189me2, and K189me3 roughly overlay with their respective -amino groups superimposed and adopt slightly kinked con- formations (Fig. 3E), as opposed to the extended trans geome- try of the unmodified and methylated lysines in the complexes of the WT enzyme and the Y305F mutant (Fig. 2, A–D). Con- versely, the unmodified K189 side chain does not superimpose with its methylated counterparts and is oriented in an alterna- tive configuration due to its hydrogen bonding to Gly-264 (Fig. 3, A and E). An overlay of the structures of the WT enzyme and Y245A mutant bound to the unmodified TAF10 peptide illus- trates that the side chains of K189 do not superimpose and that the K189 -amino group appears to be misaligned with AdoHcy in the Y245A complex (Fig. 3F). This suboptimal alignment may explain the diminished kcat value of SET7/9 Y245A mutant toward substrates with unmodified lysines (Table 2). A comparison of the structures of the SET7/9 Y245A and Y305F complexes yields a molecular explanation for the differ- ent product specificities of these two mutants. In the SET7/9 Y305F complexes, Tyr-245 aligns the K189 -amino group for methyl transfer through hydrogen bonding to its hydroxyl group (Fig. 2, B–D). Conversely, in the Y245A mutant, the K189me1 and K189me2 -amino groups are oriented through hydrogen bonding to Tyr-305 (Fig. 3, B and C). These distinct hydrogen bond patterns impart differences in the conforma- tions of the lysyl side chains due to the relative orientations of Tyr-245 and Tyr-305 in the lysine binding channel. Specifically, the kinked conformation adopted by the K189me1 and K189me2 side chains in the Y245A complexes (Fig. 3, B and C) may contribute to the differences in the turnover numbers of this mutant versus those of the WT enzyme and the Y305F mutant (Table 2). In addition, the dimethyl -amino group of the K189me2 side chain binds in distinct orientations in the Y245A and Y305F mutants due to their hydrogen bonding to FIGURE 3. Crystal structures of the SET7/9 Y245A mutant in complex with AdoHcy and unmodified and methylated TAF10 peptides. Active site of SET7/9 Y245A bound to AdoHcy and TAF10-K189 (A), TAF10-K189me1 (B), TAF10-K189me2 (C), and TAF10-K189me3 peptides (D). A–D, Fo – Fc simulated- annealing omit maps (contoured at 2.5 ) for the unmodified and methylated K189 side chains are illustrated. The residues and hydrogen bonds in each com- plexarecoloredasinFig.2.Thewatermoleculesinthelysinebindingchannelsof the Y245A complexes are numbered 1–4, as described in the text. E, superimpo- sition of the active sites of the Y245A complexes bound to the four methylated statesofTAF10-K189(unmodified,mono-,di-,andtrimethylatedshowningreen, yellow, pink, and blue, respectively). F, overlay of WT SET7/9 (blue carbons) and SET7/9 Y245A (green carbons) in complex with TAF10-K189. Waters correspond- ing to the WT and Y245A structures are colored cyan and green, respectively. G, alignment of the active sites of the Y245A (yellow carbons) and Y305F (green carbons) mutants in complex with the TAF10-K189me2 peptide. Hydrogen bonds from the Y305F structure are shown as green dashed lines, and waters and hydrogen bonds in the Y245A structure are shown in yellow and orange, respectively. Lysine Methylation by SET7/9 Mutants 31854 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 Tyr-305 and Tyr-245, respectively (Fig. 3G). In the Y305F mutant, hydrogen bonds to the dimethyl -amino group cou- pled with steric constraints in the lysine binding channel pre- vent the K189me2 side chain from undergoing a conforma- tional change that is conducive to trimethylation (Fig. 2D), consistent with its dimethyltransferase activity. However, in the Y245A mutant, the alanine substitution enlarges the diameter of the lysine binding channel, accommodating trimethyl-lysine (Fig. 3D). In addition, the larger diameter would permit the dimethyl-lysine substrate to undergo the conformational reor- ganization necessary to align the -amino group in a productive geometry for trimethylation. A major difference in the active site of the Y245A mutant versus the other SET7/9 structures is the presence of several water molecules bound in the cavity generated by the Y245A mutation. In the structure of the Y245A mutant bound to TAF10-K189, three water molecules (waters 2–4) occupy this cavity and are arranged in a triangular geometry (Fig. 3A). In addition, water 1 shifts 1.6 Å from its position in the solvent pocket toward water 2 to which it forms a hydrogen bond (Fig. 3, A and E). The shift in water 1 was unexpected given its con- served orientation in the solvent pocket of the SET7/9 WT and Y305F complexes (Fig. 2, A–C) as well as in the structures of other SET domain KMTs (8). This displacement is presumably related to the alternative conformation of the K189 side chain whose -amino group is too distant (4.3 Å) to form a productive hydrogen bond to water 1. Conversely, in the Y245A complexes bound to TAF10-K189me1 and TAF10-K189me2, water 1 remains tightly bound in the solvent pocket through hydrogen bonds to Tyr-305 hydroxyl group, the carbonyl oxygens of Gly- 292 and Ala-295, and the K189 -amino group (Fig. 3, B, C, and E), analogous to its binding in the WT enzyme (Fig. 2A). How- ever, in the TAF10-K189me3 complex, one of the methyl groups of the trimethyl -ammonium cation is oriented into the solvent pocket (Fig. 3D), similar to the dimethyl-lysine binding mode observed in the Y305F mutant (Fig. 2D). The binding of the methyl group in the solvent pocket displaces water 1 by 3.2 Å relative to its position in the TAF10-K189me1 complex (Fig. 3E), thereby avoiding a steric clash with the trimethylated -ammonium group. Variations in the occupancy of water 2 are also seen in the different Y245A structures. Water 2 is bound in similar orientations in the active site of the unmodified and monomethylated peptide complexes but is absent in the di- and trimethylated peptide complexes due to the binding of a methyl group in this position (Fig. 3, A–E). In summary, the changes in the positions or occupancies of waters 1 and 2 correlate with the binding modes of the unmodified and methylated K189 within the active site of the Y245A mutant. Catalytic Models of Lysine Multiple Methylation by SET7/9 Y245A, and Y305F—The structures of the SET7/9 complexes reported here offer a prime opportunity to generate stepwise models for lysine mono-, di-, and trimethylation by a SET domain KMT. We modeled the AdoMet-bound Michaelis complexes by superimposing the SET7/9 product complexes with the previously reported structure of the SET7/9-AdoMet binary complex (Fig. 4) (39). The conformations of the mono- and dimethyl -amino groups in the Michaelis complexes were inferred from the coordinates of the corresponding dimethyl- and trimethyl-lysine products, respectively. In addition, we modeled the -amino group in a deprotonated state with its hydrogen atoms oriented toward the hydrogen bond acceptors that align the lysyl side chain for methylation. As a basis for this comparison, we first modeled the monomethylation reaction catalyzed by WT SET7/9 (Fig. 4A). In the substrate ternary complex, the lysine -amine is aligned with the methyl group and sulfonium cation of AdoMet through a hydrogen bond to the Tyr-245 hydroxyl group and water 1 in the solvent pocket. The values of the reaction distance and angle are 2.8 Å and 153°, respectively, in approximate agreement with the linear geome- try of a SN2 methyl transfer reaction calculated in other mod- eled substrate complexes (8, 10). In the product complex, the monomethyl-lysine side chain is bound in an extended confor- mation with its methyl group oriented within the methyl trans- fer pore, thereby obstructing AdoMet binding. Furthermore, water 1 remains tightly coordinated in the solvent pocket through four hydrogen bonds to Gly-292, Ala-295, Tyr-305, and the monomethyl -amino group. These interactions hinder the dissociation of water 1 and the related rearrangement of the monomethyl-lysine side chain required for a second methyl transfer reaction, explaining why the WT enzyme cannot cata- lyze di- and trimethylation. These findings concur with the FIGURE 4. Catalytic models of the methyl transfer reactions catalyzed by WT SET7/9 and the Y305F mutant. A, monomethylation of TAF10-K189 by the WT enzyme. The reaction scheme depicts the modeled substrate ternary complex (left) and the product complex (right) for the transfer of the methyl group from AdoMet (green carbon atoms) to K189 in TAF10 (yellow carbons), yielding AdoHcy and K189me1. The red arrow indicates the direction of the nucleophilic attack of the deprotonated -amino group on the AdoMet methyl group. The transferred methyl group is colored green, and the white atoms represent the hydrogens of the -amino group. Hydrogen bonds and residues in the enzyme active site are illustrated as in Fig. 2. The reaction distance and angle are labeled in red. B and C, models of the Y305F mutant for the first methyl transfer reaction with TAF10-K189 (B) and second methyl transfer reaction with TAF10-K189me1 (C). Color schemes are the same as in A. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31855 model for SET7/9 product specificity reported in previous structural and functional studies (6, 11). Similar reaction geometry is observed in the model for the monomethyl transfer reaction catalyzed by SET7/9 Y305F. Hydrogen bonds from the Tyr-245 hydroxyl group and water 1 align the lysine -amino group with the AdoMet methyl group at a distance of 2.1 Å and an angle of 160°, equivalent to those measured in the Michaelis complex of the WT enzyme (Fig. 4, A and B). In the product complex, monomethyl-lysine adopts an extended trans configuration analogous to that in the WT enzyme. For dimethylation to occur, the monomethyl-lysine must undergo a conformational change in which its methyl group is rotated out of the methyl transfer path with AdoMet. The structure of the Y305F mutant bound to the dimethylated TAF10 peptide (Fig. 2D) implies that this rearrangement occurs through the dissociation of water 1 due to the loss of the Tyr- 305 hydrogen bond in the solvent pocket. The dissociation of water 1 would enable the monomethyl-lysine side chain to adopt an alternative conformation through a rotation about its C–N bond, projecting the methyl group into the solvent pocket (Fig. 4C). This rotation reorients the methyl group out of the methyl transfer path while realigning the monomethyl - amino group for a second methylation reaction through a direct hydrogen bond to the Tyr-245 hydroxyl group and a CH–O hydrogen bond between its methyl group and Ala-295. The modeled reaction geometry for monomethyl-lysine substrate complex of the Y305F mutant (2.7 Å, 160°) is equivalent to that of the first methyl transfer reaction in SET7/9 Y305F. These geometries concur with our previous models for mono- and dimethylation catalyzed by SET8 Y334F (8), illustrating that the orientation of a methyl group into the solvent pocket is a con- served feature of SET domain KMTs that catalyze multiple methylation. In addition, we modeled the methyl transfer reactions cata- lyzed by SET7/9 Y245A (Fig. 5). In the model of the lysine sub- strate complex, the -amino group is aligned for methyl transfer by a hydrogen bond to the carbonyl oxygen of Gly-264, result- ing in a short reaction distance (2.3 Å) and a suboptimal reac- tion angle (141°) with the methyl group of AdoMet (Fig. 5A). This misalignment appears to be a direct consequence of the Y245A mutation that abolishes hydrogen bonding to the - amino group, illustrating that the suboptimal orientation of the -amine likely contributes to the diminished activity of this mutant toward unmodified substrates (11). Conversely, in the modeled monomethyl-lysine substrate complex for SET7/9 Y245A, Tyr-305 hydrogen bonds to the -amino group, aligning it for methyl transfer (Fig. 5B). In addition, CH–O hydrogen bonds to the monomethyl-lysine methyl group and the dissoci- ation of water 2 from the active site also contribute to reposition- ing the -amino group for dimethylation. Collectively, these inter- actions orient the -amine in a reaction angle of 165° that is more conducive to methyl transfer. However, the reaction distance for dimethylation is 0.6 Å longer than that in the corresponding Y305F model because Tyr-305 is positioned further from AdoMet than Tyr-245 (Figs. 4C and 5B). In the third methyl transfer reaction catalyzed by SET7/9 Y245A, the lone pair of electrons of the dimethyl-lysine - amino group acts as the nucleophile and thus cannot engage in hydrogen bonding. The structure of the trimethyl-lysine prod- uct complex (Fig. 3D) implies that the dimethyl -amine is aligned via CH–O hydrogen bonds to its methyl groups, as shown in the model of the Michaelis complex for this reaction (Fig. 5C). These CH–O hydrogen bonds restrain the orienta- tion of the -amino group and position one of the methyl groups into the solvent pocket, displacing water 1 as discussed earlier (Fig. 3, D and E). These interactions cumulatively align the -amino group and AdoMet methyl group with a reaction distance of 3.0 Å and angle of 162° (Fig. 5C). Taken together, the models of the substrate complexes for SET7/9 Y245A suggest that CH–O hydrogen bonds play an increasingly important role in aligning the methylated -amino group in successive rounds of methyl transfer. DISCUSSION The structural and functional characterization of the SET7/9 Y245A and Y305F mutants presented here yields new insights into the mechanism underlying the product specificity of SET domain KMTs. Importantly, it resolves a general paradox concerning this specificity. How does the active site constrain the motion of the lysine -amino group to align it for methyl transfer with AdoMet, while providing adequate volume to accommodate the mono-, di-, and tri- methylated lysine side chain generated during multiple methyl transfer reactions? The structures of the Y305F and Y245A mutants resolve this paradox, illustrating that alter- ations in the positions or occupancies of water molecules within their active sites generate the space required to FIGURE 5. Models for the methyl transfer reactions catalyzed by the SET7/9 Y245A mutant. Models of the Y245A mutant for the first methyl transfer reaction with TAF10-K189 (A), the second methyl transfer reaction with TAF10-K189me1 (B), and the third methyl transfer reaction with TAF10- K189me2 (C). Residues and hydrogen bonds are depicted as in Fig. 3. Lysine Methylation by SET7/9 Mutants 31856 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 accommodate the multiply methylated -amine produced during successive catalytic cycles. Minor perturbations in the side chains of certain active site residues, such as Tyr- 305, are also observed in alignments of the WT enzyme and the Y245A and Y305F complexes, although these changes are modest compared with the displacement or dissociation of the water molecules in the active site. These findings suggest that the waters function as transient place holders that facil- itate the SN2 methyl transfer reaction. During monomethy- lation, they function to constrain the movement of the lysine -amino group by mediating hydrogen bonds between the substrate and enzyme, thereby promoting the linear align- ment with the methyl group and sulfonium cation of AdoMet (Fig. 4, A and B). During di- and trimethylation, the water molecules either relocate within the lysine binding channel or dissociate from the enzyme, yielding the space required to rotate the methyl group away from the methyl transfer pore and to realign the -amine in productive geom- etry for the next methyl transfer reaction (Figs. 4C and 5, B and C). These findings agree with our prior analysis of the SET8 Phe/Tyr switch mutant in which we demonstrated that the Y334F substitution attenuates hydrogen bonding to the water molecule bound in the solvent pocket, promoting its dissociation and the conformational changes necessary for lysine dimethylation (8). Indeed, there is a nearly identical alignment of the dimethyl-lysine side chains in the structures of SET7/9Y305FandSET8Y334Fcomplexes,despitethedifferences in the orientations of the Phe-305 and Phe-334 side chains in each structure (Fig. 2F). Finally, both the SET8 Y334F (8) and SET7/9 Y305F mutants (Table 2) displayed diminished catalytic efficien- cies for lysine dimethylation versus monomethylation. These dif- ferences may reflect the kinetics of the reorganization within the active site, including the dissociation of the water molecule from the solvent pocket and the concomitant realignment of the monomethyl-lysine into a productive geometry for dimethylation. In addition to their place-holding role, the active site waters may also facilitate the deprotonation of the lysine -amino group between methyl transfer reactions. For methylation to occur, the -amino group must be deprotonated to function as the nucleophile in the SN2 methyl transfer reaction with AdoMet (Figs. 4 and 5). Although the pKa value of the lysine -amine in solution is 10.5, molecular dynamics simulations by Zhang and Bruice (25, 26) indicate that this value diminishes to 8.2 upon formation of the SET7/9 Michaelis complex due to the proximity of the AdoMet sulfonium cation and the low dielectric constant of the active site. Furthermore, their simu- lations show that a chain of water molecules facilitates the dep- rotonation of the -amino group prior to methyl transfer, trans- ferring the proton to bulk solvent. Although these water molecule chains are not evident in our crystal structures, the Y305F and Y245A complexes suggest another potential mech- anism for deprotonation. In the dimethyl-lysine complexes of the Y305F and Y245A mutants, the dissociation of water 1 and 2, respectively, from the lysine binding channel requires that the solvent-mediated hydrogen bond to the -amino group is broken (Figs. 2D and 3C). It is conceivable that these waters dissociate from the active site as hydronium ions, promoting the realignment and deprotonation of the methyl -amino group for the next methyl transfer reaction. A comparison of the SET7/9 Y305F and SET8 Y334F com- plexes yields insights into the mechanism by which the Phe/Tyr switch influences water binding within the solvent pocket. The phenylalanine substitution in the Phe/Tyr switch results in the loss of a single hydrogen bond to the water molecule (water 1) in the solvent pocket compared with the four hydrogen bonds that coordinate the solvent molecule in WT SET7/9 (Fig. 2, A and B) and SET8 (7, 8). Although this attenuation in hydrogen bonding may appear insignificant, this difference is nonetheless impor- tant for at least two reasons. First, theoretical calculations indi- cate that, on average, water molecules form 3.5 hydrogen bonds in solutions (40, 41). This value is greater than the num- ber of hydrogen bonds coordinating water 1 in the solvent pocket in SET7/9 Y305F (Fig. 2, B and C) as well as in SET8 Y334F and other di- and trimethyltransferases that possess a hydrophobic residue in the Phe/Tyr switch site (8). From the perspective of the water molecule, the greater hydrogen bond- ing potential in solution would tend to thermodynamically favor its dissociation from the solvent pocket in SET domain KMTs that lack a tyrosine in the Phe/Tyr switch position. Sec- ond, the ordered binding of water molecules observed in the active sites of SET domain ternary complexes represents an unfavorable entropy compared with their diffusion in bulk sol- vent. In WT SET7/9 (Fig. 2A) and SET8 (7, 8), this entropic penalty can be partially offset through the favorable enthalpy of binding associated with the four hydrogen bonds that coordi- nate the water within the solvent pocket. It is conceivable that the loss of the tyrosine-mediated hydrogen bond in the Phe/Tyr switch shifts the equilibrium in favor of dissociation of the water molecule from the solvent pocket, thereby facilitating dimethylation in SET7/9 Y305F, SET8 Y334F, and other di- and trimethyltransferases. The structures of the SET7/9 Y245A and Y305F complexes illustrate the interactions that align the lysine -amino group during the methyl transfer reactions in each enzyme. In the WT enzyme and the Y305F mutant, hydrogen bonding to the hydroxyl group of Tyr-245 appears to be critical in properly aligning the -amine for methyl transfer (Fig. 4). Tyr-245 is conserved in the sequences of many SET domain KMTs (8, 42), and substitutions of this residue generally impair or abolish activity, indicating its importance in catalysis (8, 43). However, SET7/9 appears to be an exception to this rule, as the Y245A mutant is not only active but is capable of catalyzing lysine trimethylation. In this mutant, Tyr-305 appears to assume the role of Tyr-245 by hydrogen bonding to the monomethylated -amino group to align it for methyl transfer with AdoMet, as illustrated in the modeled substrate complex for the dimethy- lation reaction (Fig. 5B). Conversely, in the model for trimethy- lation, the Tyr-305 hydroxyl group does not hydrogen bond to the -amine but instead participates in a CH–O hydrogen bond with one of the methyl groups to assist in aligning the dimethy- lated -amine for the methyl transfer reaction (Fig. 5C). Addi- tional structural and functional studies of the SET domain tri- methyltransferases will aid in further illuminating the roles of CH–O hydrogen bonds in facilitating lysine multiple methylation. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31857 Acknowledgments—We acknowledge S. Schiebold for assistance in protein expression, purification, and crystallization and S. Anderson and R. Sanishvili for their assistance with x-ray data collection. We also thank S. Bulfer and S. Horowitz for reading the manuscript and providing useful comments. This work utilized the Protein Structure Facility of the Michigan Diabetes Research and Training Center, Uni- versity of Michigan, supported by National Institutes of Health Grant DK020572, NIDDK. Use of the Advanced Photon Source was sup- ported by the United States Department of Energy, Basic Energy Sci- ences, Office of Science, under Contract DE-AC02-06CH11357. GM/CA CAT has been funded in whole or in part by National Institutes of Health NCI Grant Y1-CO-1020 and NIGMS Grant Y1-GM-1104. Use of the LS-CAT Sector 21 was supported by Michi- gan Economic Development Corporation and the Michigan Technol- ogy Tri-Corridor Grant 085P1000817 for the support of this research program. REFERENCES 1. Huang, J., and Berger, S. L. (2008) Curr. Opin. Genet. Dev. 18, 152–158 2. Morgunkova, A., and Barlev, N. A. (2006) Cell Cycle 5, 1308–1312 3. Yang, X. D., Lamb, A., and Chen, L. F. (2009) Epigenetics 4, 429–433 4. Taverna, S. D., Li, H., Ruthenburg, A. J., Allis, C. D., and Patel, D. J. (2007) Nat. Struct. Mol. Biol. 14, 1025–1040 5. Collins, R. E., Tachibana, M., Tamaru, H., Smith, K. M., Jia, D., Zhang, X., Selker, E. U., Shinkai, Y., and Cheng, X. (2005) J. Biol. Chem. 280, 5563–5570 6. Zhang, X., Yang, Z., Khan, S. I., Horton, J. R., Tamaru, H., Selker, E. U., and Cheng, X. (2003) Mol. Cell 12, 177–185 7. Couture, J. F., Collazo, E., Brunzelle, J. S., and Trievel, R. C. (2005) Genes Dev. 19, 1455–1465 8. Couture, J. F., Dirk, L. M., Brunzelle, J. S., Houtz, R. L., and Trievel, R. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 20659–20664 9. Qian, C., Wang, X., Manzur, K., Sachchidanand, Farooq, A., Zeng, L., Wang, R., and Zhou, M. M. (2006) J. Mol. Biol. 359, 86–96 10. Trievel, R. C., Flynn, E. M., Houtz, R. L., and Hurley, J. H. (2003) Nat. Struct. Biol. 10, 545–552 11. Xiao, B., Jing, C., Wilson, J. R., Walker, P. A., Vasisht, N., Kelly, G., Howell, S., Taylor, I. A., Blackburn, G. M., and Gamblin, S. J. (2003) Nature 421, 652–656 12. Chuikov, S., Kurash, J. K., Wilson, J. R., Xiao, B., Justin, N., Ivanov, G. S., McKinney, K., Tempst, P., Prives, C., Gamblin, S. J., Barlev, N. A., and Reinberg, D. (2004) Nature 432, 353–360 13. Ea, C. K., and Baltimore, D. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 18972–18977 14. Este`ve, P. O., Chin, H. G., Benner, J., Feehery, G. R., Samaranayake, M., Horwitz, G. A., Jacobsen, S. E., and Pradhan, S. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 5076–5081 15. Kouskouti, A., Scheer, E., Staub, A., Tora, L., and Talianidis, I. (2004) Mol. Cell 14, 175–182 16. Masatsugu, T., and Yamamoto, K. (2009) Biochem. Biophys. Res. Commun. 381, 22–26 17. Munro, S., Khaire, N., Inche, A., Carr, S., and La Thangue, N. B. (2010) Oncogene 29, 2357–2367 18. Pagans, S., Kauder, S. E., Kaehlcke, K., Sakane, N., Schroeder, S., Dorm- eyer, W., Trievel, R. C., Verdin, E., Schnolzer, M., and Ott, M. (2010) Cell Host Microbe 7, 234–244 19. Subramanian, K., Jia, D., Kapoor-Vazirani, P., Powell, D. R., Collins, R. E., Sharma, D., Peng, J., Cheng, X., and Vertino, P. M. (2008) Mol. Cell 30, 336–347 20. Wang, J., Hevi, S., Kurash, J. K., Lei, H., Gay, F., Bajko, J., Su, H., Sun, W., Chang, H., Xu, G., Gaudet, F., Li, E., and Chen, T. (2009) Nat. Genet. 41, 125–129 21. Yang, X. D., Huang, B., Li, M., Lamb, A., Kelleher, N. L., and Chen, L. F. (2009) EMBO J. 28, 1055–1066 22. Guo, H. B., and Guo, H. (2007) Proc. Natl. Acad. Sci. U.S.A. 104, 8797–8802 23. Hu, P., Wang, S., and Zhang, Y. (2008) J. Am. Chem. Soc. 130, 3806–3813 24. Hu, P., and Zhang, Y. (2006) J. Am. Chem. Soc. 128, 1272–1278 25. Zhang, X., and Bruice, T. C. (2007) Biochemistry 46, 14838–14844 26. Zhang, X., and Bruice, T. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 5728–5732 27. Couture, J. F., Collazo, E., Hauk, G., and Trievel, R. C. (2006) Nat. Struct. Mol. Biol. 13, 140–146 28. Trievel, R. C., Beach, B. M., Dirk, L. M., Houtz, R. L., and Hurley, J. H. (2002) Cell 111, 91–103 29. Kapust, R. B., To¨zse´r, J., Fox, J. D., Anderson, D. E., Cherry, S., Copeland, T. D., and Waugh, D. S. (2001) Protein Eng. 14, 993–1000 30. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307–326 31. Vagin, A., and Teplyakov, A. (2000) Acta Crystallogr. D Biol. Crystallogr. 56, 1622–1624 32. Emsley, P., and Cowtan, K. (2004) Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 33. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Acta Crystallogr. D Biol. Crystallogr. 53, 240–255 34. Davis, I. W., Leaver-Fay, A., Chen, V. B., Block, J. N., Kapral, G. J., Wang, X., Murray, L. W., Arendall, W. B., 3rd, Snoeyink, J., Richardson, J. S., and Richardson, D. C. (2007) Nucleic Acids Res. 35, W375–W383 35. Bru¨nger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. D Biol. Crystallogr. 54, 905–921 36. Collazo, E., Couture, J. F., Bulfer, S., and Trievel, R. C. (2005) Anal. Bio- chem. 342, 86–92 37. Chirpich, T. P., Zappia, V., Costilow, R. N., and Barker, H. A. (1970) J. Biol. Chem. 245, 1778–1789 38. Couture, J. F., Hauk, G., Thompson, M. J., Blackburn, G. M., and Trievel, R. C. (2006) J. Biol. Chem. 281, 19280–19287 39. Kwon, T., Chang, J. H., Kwak, E., Lee, C. W., Joachimiak, A., Kim, Y. C., Lee, J., and Cho, Y. (2003) EMBO J. 22, 292–303 40. Chandra, A., and Chowdhuri, S. (2002) J. Phys. Chem. B 106, 6779–6783 41. Guardia, E., Marti, J., Garcia-Tarres, L., and Laria, D. (2005) J. Mol. Liq. 117, 63–67 42. Dillon, S. C., Zhang, X., Trievel, R. C., and Cheng, X. (2005) Genome Biol. 6, 227 43. Zhang, X., Tamaru, H., Khan, S. I., Horton, J. R., Keefe, L. J., Selker, E. U., and Cheng, X. (2002) Cell 111, 117–127 Lysine Methylation by SET7/9 Mutants 31858 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010
3M54
SET7/9 Y305F in complex with TAF10 peptide and AdoHcy
SET7/9 Catalytic Mutants Reveal the Role of Active Site Water Molecules in Lysine Multiple Methylation*□ S Received for publication,February 15, 2010, and in revised form, July 14, 2010 Published, JBC Papers in Press,August 1, 2010, DOI 10.1074/jbc.M110.114587 Paul A. Del Rizzo‡, Jean-Franc¸ois Couture‡§1, Lynnette M. A. Dirk¶, Bethany S. Strunk‡, Marijo S. Roiko‡, Joseph S. Brunzelle, Robert L. Houtz¶, and Raymond C. Trievel‡2 From the ‡Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109, the §Ottawa Institute of Systems Biology, Department of Biochemistry, Microbiology, and Immunology, University of Ottawa, Ottawa, Ontario K1H 8M5, Canada, the ¶Department of Horticulture, Plant Physiology/Biochemistry/Molecular Biology Program, University of Kentucky, Lexington, Kentucky 40546, and the Department of Molecular Pharmacology and Biological Chemistry, Northwestern University Center for Synchrotron Research, Life Sciences Collaborative Access Team, Argonne, Illinois 60439 SET domain lysine methyltransferases (KMTs) methylate specific lysine residues in histone and non-histone substrates. These enzymes also display product specificity by catalyzing dis- tinct degrees of methylation of the lysine -amino group. To elucidate the molecular mechanism underlying this specificity, we have characterized the Y245A and Y305F mutants of the human KMT SET7/9 (also known as KMT7) that alter its prod- uct specificity from a monomethyltransferase to a di- and a tri- methyltransferase, respectively. Crystal structures of these mutants in complex with peptides bearing unmodified, mono-, di-, and trimethylated lysines illustrate the roles of active site water molecules in aligning the lysine -amino group for methyl transfer with S-adenosylmethionine. Displacement or dissocia- tion of these solvent molecules enlarges the diameter of the active site, accommodating the increasing size of the methylated -amino group during successive methyl transfer reactions. Together, these results furnish new insights into the roles of active site water molecules in modulating lysine multiple meth- ylation by SET domain KMTs and provide the first molecular snapshots of the mono-, di-, and trimethyl transfer reactions catalyzed by these enzymes. SET domain enzymes represent a family of S-adenosylmethi- onine (AdoMet)3-dependent methyltransferases that catalyze the site-specific methylation of protein lysyl residues in a host of proteins, including histones, transcription factors, chroma- tin-modifying enzymes, ribosomal subunits, and other sub- strates (1–3). In many instances, these modifications serve to recruit effector proteins that recognize methyl-lysyl residues in a sequence-dependent fashion (4). In addition, SET domain KMTs exhibit product specificity, defined as their ability to cat- alyze mono-, di-, or trimethylation of the lysine -amino group. This specificity is biologically relevant because many methyl- lysine-binding proteins can discriminate among different degrees of lysine methylation (4). Thus, both the site and degree of lysine methylation are critical to recognition by effector proteins. Structural and functional studies have identified a Phe/Tyr switch in the active site of SET domain KMTs that governs their respective product specificities (5, 6). According to this model, KMTs that possess a tyrosine in the Phe/Tyr switch site are limited to catalyzing lysine monomethylation, whereas en- zymes that possess a phenylalanine or another hydrophobic residue in this position display di- or trimethyltransferase activ- ity. Mutational analysis of various SET domain KMTs, includ- ing DIM-5 (KMT1) (6), SET7/9 (6), G9A (KMT1C) (5), and SET8 (also known as PR-SET7 and KMT5A) (7, 8), has demon- strated that substitutions in the Phe/Tyr switch result in pre- dictable changes in product specificity. Several models have been proposed to explain the mechanism by which the Phe/Tyr switch site governs this specificity, including variations in the diameter of the active site due to the size of Phe/Tyr switch residue and steric hindrance by the tyrosine hydroxyl group (6, 9–11). However, our recent studies of the Phe/Tyr switch mutant Y334F in the human histone H4 Lys-20 (H4K20) meth- yltransferase SET8 indicate that the Phe/Tyr switch regulates product specificity via a more subtle mechanism (8). Specifi- cally, the switch modulates the binding of an active site water molecule that in turn regulates the transition from mono- methylation to multiple methylation. Among the KMTs that have been structurally characterized, SET7/9 has emerged as an archetypal model for studying the catalytic mechanism and product specificity of the SET domain family. Although initially isolated as a histone H3 Lys-4 (H3K4)-specific methyltransferase, this KMT has been shown to regulate the functions of numerous non-histone substrates through site-specific methylation (12–21). Early structural and functional studies of SET7/9 identified two active site mutants, * Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth Grant R01 GM073839 (to R. C. T.) and National Institutes of Health Admin- istrative Supplement GM073839-04S1 (to R. C. T.) funded through the American Recovery and Reinvestment Act. This work was also supported by Department of Energy Grant DE-FG02-92ER20075 (to R. L. H.). □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Table 1. The atomic coordinates and structure factors (codes 3M53, 3M54, 3M55, 3M56, 3M57,3M58,3M59,and3M5A)havebeendepositedintheProteinDataBank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). 1 Supported by a Canadian Institutes of Health Research postdoctoral fellowship. 2 To whom correspondence should be addressed: Dept. of Biological Chem- istry, University of Michigan Medical School, 1150 West Medical Center Dr., 5301 MSRB III, Ann Arbor, MI 48109. Tel.: 734-647-0889; Fax: 734-763-4581; E-mail: rtrievel@umich.edu. 3 The abbreviations used are: AdoMet, S-adenosylmethionine; AdoHcy, S-ad- enosylhomocysteine; KMT, lysine methyltransferase; ITC, isothermal titra- tion calorimetry; CH–O, carbon-oxygen hydrogen bond; BisTris, 2-[bis(2- hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 41, pp. 31849–31858, October 8, 2010 © 2010 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31849 Y245A and Y305F, which change its product specificity. The Phe/Tyr switch mutant Y305F alters SET7/9 product specificity from a mono- to dimethyltransferase (6), whereas the Y245A substitution converts the enzyme into a trimethyltransferase with weak monomethyltransferase activity (11). These mutants have been the subjects of numerous molecular modeling simu- lations that have led to various models to explain their distinct product specificities (22–26). However, the lack of structural data for the SET7/9 Y245A and Y305F mutants in complex with cognate methylated peptides has hindered our understanding of the mechanisms that define the respective product specific- ities of these mutants. Moreover, these structures would yield a framework for visualizing the mono-, di-, and trimethylation reactions catalyzed by SET domain KMTs. To gain insight into the molecular basis of their product specificities, we have determined high resolution crystal structures of the SET7/9 Y245A and Y305F mutants in com- plex with peptides of the TATA box-binding protein-associ- ated factor TAF10 bearing the Lys-189 methylation site in unmodified (K189), monomethylated (K189me1), dimethyl- ated (K189me2), and trimethylated (K189me3) states. The structures and accompanying biochemical data support a model whereby changes in the occupancy or position of water molecules in the active site are critical in establishing the prod- uct specificities of the SET7/9 Y245A and Y305F mutants. Together, our results provide new insights into the mechanisms that govern SET domain product specificity and provide step- wise snapshots of the lysine mono-, di-, and trimethyl transfer reactions catalyzed by KMTs. EXPERIMENTAL PROCEDURES Cloning, Expression, and Purification of the SET7/9 Mutants— The Y245A and Y305F mutants were introduced into the pHIS2 SET7/9 expression vector encoding residues 110–366 (27) via QuikChange site-directed mutagenesis (Stratagene) and were verified by dideoxy DNA sequencing. The plasmids encoding wild type (WT) SET7/9 and the Y245A and Y305F mutants were transformed into Rosetta2 DE3 cells (Novagen) and were expressed as described previously (27, 28). In the course of characterizing WT SET7/9, we observed that the enzyme co- purified with AdoMet or another contaminant that resulted in technical difficulties in the isothermal titration calorimetry (ITC) experiments and co-crystallization trials with the TAF10 peptides. To overcome this problem, a denaturation and refold- ing step was inserted in the purification scheme. The denatur- ation and refolding protocol involved adding 6 M guanidine HCl, 50 mM sodium phosphate, pH 7.5, and 500 mM NaCl to the protein while it was immobilized on a nickel-Sepharose column (GE Healthcare). The column was washed with this buffer, fol- lowed by a solution of 6 M urea, 50 mM sodium phosphate, pH 7.5, 150 mM NaCl, and 10 mM 2-mercaptoethanol to remove the cofactor from the denatured enzyme. A reverse gradient from 6 to 0 M urea was then performed in the same buffer to refold the protein, which was subsequently eluted from the column using a linear gradient of 0–500 mM imidazole in 50 mM sodium phosphate, pH 7.5, 500 NaCl, and 10 mM 2-mercaptoethanol. The refolded protein was digested with tobacco etch virus pro- tease (29) during dialysis against 20 mM Tris, pH 7.5, 150 mM NaCl, and 5 mM 2-mercaptoethanol and then purified using a Superdex 200 gel filtration column (GE Healthcare). Protein concentration was determined by its absorbance at 280 nm. Synthetic Peptides—The TAF10 peptides bearing K189, K189me1, K189me2, and K189me3 (sequence, acetyl- SKSK189DRKYTL-amide) and a biotinylated TAF10 peptide (sequence, acetyl-SKSK189DRKYTLT(K-EZLink-S-S-biotin)- amide) were synthesized and purified by New England Peptide, Inc. Peptide concentrations were measured using the absorb- ance of their tyrosine residue at 274 nm. Crystallization and Data Collection—Crystals were pro- duced by hanging drop vapor diffusion by mixing the crystalli- zation solution in a 1:1 ratio with 10–20 mg/ml SET7/9, 1 mM S-adenosylhomocysteine (AdoHcy), and 1.0–3.0 mM unmodi- fied or methylated TAF10-K189 peptide in 20 mM Tris, pH 8.0, 100 mM NaCl, and 2 mM tris(2-carboxyethyl)phosphine. Crys- tals were obtained at 20 °C in either 1.8–2.0 M (NH4)2SO4, with 0.1 M BisTris, pH 6.2–6.6, or in 0.95–1.1 M sodium citrate with 100 mM imidazole pH 8.0–8.4. In both crystallization condi- tions, the final pH values were between pH 8.0 and 9.0. Crystals in the (NH4)2SO4 condition were typically flash-frozen in the mother liquor containing 25–30% glycerol, and the crystals in the citrate condition were frozen in 1.6 M sodium citrate. Data were collected at the Advanced Photon Source beamlines 21-IDG (LS-CAT) and 23-IDD (GM/CA-CAT). Images were indexed, integrated, and scaled using HKL2000 (30). Structures of the mutants were solved by molecular replacement using MOLREP (31) with the coordinates of a previously reported SET7/9 ternary complex used as the search model (Protein Data Bank code 2F69). Successive rounds of model building and refinement were carried out using Coot (32) and REFMAC (33), respectively. The geometry of the models were verified by Mol- Probity (34). Simulated annealing omit maps were calculated using CNS (35) with the peptide and cofactor removed to elim- inate model bias in the active site. Structural figures were ren- dered using PyMOL (Schro¨dinger, LLC.). Fluorescent Methyltransferase Assay—A coupled fluorescent methyltransferase assay was used to measure the kinetic parameters of WT SET7/9 and the Y245A and Y305F mutants as reported previously, with the exception that 50–150 nM enzyme, 100 M AdoMet, and varying concentrations of TAF10 peptide substrate were used (27, 36). Assays were performed in triplicate, and a homocysteine calibration curve was used to calculate the initial velocities. Kinetic parameters were calcu- lated by plotting the velocities versus peptide concentration and by fitting the Michaelis-Menten equation to the data via non- linear regression using Prism 5.0 (GraphPad). In cases where the Km value was beyond the measurable range of the assay, the kcat/Km value was determined as described previously (7). Calorimetry Experiments—ITC was performed at 20 °C using a MicroCal VP-ITC calorimeter (GE Healthcare) with 0.12 mM protein and 4 mM AdoHcy in 20 mM sodium phosphate, pH 7, and 100 mM NaCl with 1.5 mM peptide as the injectant. Data were processed, and equilibrium dissociation constants (KD) and curve fitting errors were calculated from the binding iso- therms using Origin 7.0 (OriginLab Corp.). WT SET7/9 and the Y245A and Y305F mutants displayed ligand:protein binding stoichiometries (N values) between 0.8 and 1.0, demonstrating Lysine Methylation by SET7/9 Mutants 31850 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 that WT SET7/9 and its mutants were properly refolded due to their ability to bind peptides in an 1:1 molar ratio. TLC Product Analysis—Methyltransferase assays were per- formed in triplicate at 37 °C with the biotinylated TAF10 pep- tide (0.5 mM) for 30 min in the presence of WT SET7/9 (3 pmol), Y305F mutant (6 pmol), or the Y245A mutant (100 pmol). Assays contained 100 mM HEPES, pH 8.0, 50 mM NaCl, 1.0 mM [methyl-3H]AdoMet (diluted to a specific activity of 0.2 Ci/nmol with purified AdoMet (37)), 15 M Sulfolobus solfa- taricus AdoHcy hydrolase (36), and 2 units of adenosine deami- nase (Roche Applied Science) in a final volume of 20 l. The reactions were terminated by addition of an equal volume of 200 mM MES, pH 5.1, prior to addition of a 2-fold molar excess of immobilized avidin resin (UltraLink; Pierce). Biotinylated peptides were allowed to bind at room temperature for 30 min, and the resin was then collected by centrifugation (9000  g). The resin was washed three times with 300 mM NaCl, and the peptide was eluted overnight from the avidin resin by cleavage of the disulfide bond in the linker of the peptide using 10 mM tris(2-carboxyethyl)phosphine (Thermo Scientific). The resin was incubated with additional 10 mM tris(2-carboxyethyl)phos- phine the following day until the radiolabel was essentially removed from the resin. The recovered peptides were hydro- lyzed by incubation with 6 M HCl at 110 °C for 24 h. Subsequent steps in measuring the radiolabel incorporated into the mono-, di-, and trimethyl-lysine products were performed as reported previously (8). RESULTS Functional Analysis of the SET7/9 Y305F Mutant—Prior studies of SET7/9 by Zhang et al. (6) reported that mutation of the Phe/Tyr switch residue Tyr-305 to a phenylalanine alters its product specificity from a mono- to dimethyltransferase. We verified these findings by demon- strating that WT SET7/9 mono- methylated the TAF10-K189 pep- tide, whereas the Y305F mutant mono- and dimethylated this sub- strate, as demonstrated by mass spectrometry (data not shown). We next examined whether the Y305F substitution altered the affinity of SET7/9 for the TAF10-K189 peptides using ITC (Fig. 1). A comparison of the KD values revealed that SET7/9 Y305F bound the TAF10-K189 and TAF10-K189me1 peptides 4- and 6-fold more tightly, respec- tively, than the WT enzyme, whereas this mutant displayed a substantially diminished affinity for the TAF10-K189me2 pep- tide (Table 1). Although the WT enzyme and the Y305F mutant exhibited discernable differences in their affinities for the unmodified and monomethylated peptides, these variations are modest and cannot account for their distinct product specific- ities, suggesting that a kinetic effect during methylation may be responsible. To investigate this possibility, we characterized the kinetic parameters of WT SET7/9 and the Y305F mutant using the TAF10 peptides as substrates. Both enzymes methylated the unmodified peptide with comparable kcat and Km values (Table 2). In analyzing the kinetic parameters for the methylation of the monomethylated peptide by SET7/9 Y305F, we found that this substrate displayed an elevated Km value that was beyond the measurable range of the assay due to its limited solubility. In this case, we measured the catalytic efficiency (kcat/Km) for the methylation of this peptide and found that it was methylated 15-fold less efficiently than the unmodified peptide by SET7/9 Y305F. Given the fact that the Y305F mutant exhibited a higher binding affinity for the TAF10-K189me1 peptide than the WT enzyme (Table 1), the kinetic data suggest that a step in the reaction pathway following substrate binding limits the catalytic efficiency of this mutant. We next examined whether the Y305F mutant dimethylated the TAF10-K189 peptide via a processive or a distributive mechanism. In a processive mechanism, the methyl-lysine substrate would remain bound to the enzyme during successive methyl transfer reactions; thus, the concentration of an inter- mediate, such as monomethyl-lysine, cannot exceed the en- zyme concentration during the assay. In a distributive mecha- nism, the intermediates are released into solution where they accumulate prior to the next round of methylation, resulting in an intermediate concentration that is greater than that of the enzyme. Using a radiometric TLC assay and a biotinylated TAF10 peptide, we quantified the amounts of monomethylated products generated by the WT SET7/9 and the Y305F mutant FIGURE 1. ITC analysis of WT SET7/9 and the Y245A and Y305F mutants. Representative titrations and binding isotherms for the calorimetry experiments are shown for the unmodified TAF10-K189 peptide titrated into WT SET7/9 (left plot), SET7/9 Y305F (middle plot), and SET7/9 Y245A (right plot). In each panel, the ITC titration experiment (upper panel) is shown with the binding isotherms (lower panel) fit to a single binding site model. TABLE 1 Analysis of the binding affinity of WT SET7/9 and its catalytic mutants for unmodified and methylated TAF10 peptides SET7/9 TAF10 peptide KD a M WT K189 4.9  0.20 WT K189me1 4.0  0.36 Y305F K189 1.3  0.10 Y305F K189me1 0.62  0.065 Y305Fb K189me2 70 Y245A K189 4.0  0.25 Y245A K189me1 3.3  0.10 Y245A K189me2 5.8  0.22 Y245A K189me3 11  0.28 a Curve fitting errors were calculated from the binding isotherms. b An estimate of the affinity is reported due to weak peptide binding. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31851 (Table 3). The data demonstrate that comparable amounts of monomethyl-lysine were generated when the quantity of enzyme usedistakenintoaccount,inagreementwiththeirsimilar turnover numbers for the TAF10-K189 peptide (Table 2). The Y305F mutant also produced small but measurable quantities of radiolabeled dimethyl-lysine product that were substantially smaller than the amount of monomethyl-lysine generated. Therefore, the TLC data are consistent with a distributive mechanism for dimethylation by the Y305F mutant because the amount of monomethyl-lysine produced exceeded the quantity of enzyme used in the assay. Structures of WT SET7/9 and the Y305F Mutant in Complex with Unmodified and Methylated TAF10 Peptides—To deter- mine the mechanism by which the Y305F substitution alters the product specificity of SET7/9, we determined the crystal struc- tures of this mutant bound to AdoHcy and TAF10-K189, TAF10-K189me1, and TAF10-K189me2 peptides and com- pared these to the structures of the WT SET7/9AdoHcy TAF10-K189 complex (supplemental Table 1). The structures of these complexes were determined to 1.85 Å or higher reso- lution, permitting unambiguous modeling of the K189 side chains in the active site of the enzyme based on simulated annealing omit maps (Fig. 2). The ternary complexes of the WT and the Y305F mutant superimpose with overall root mean square differences of less than 0.3 Å for all aligned atoms, indi- cating that neither the Y305F mutation nor the binding of the various TAF10-K189 peptides results in substantial changes in its overall structure. An inspection of the active sites of the SET7/9 WT and Y305F complexes illustrates the binding modes of the unmod- ified and methylated forms of K189 in the TAF10 peptides (Fig. 2, A–D). The K189 side chain binds in an extended all trans conformation in a deep pocket, termed the lysine binding chan- nel, that is composed of Tyr-245, Thr-266, Leu-267, Ser-268, Tyr-305, Tyr-335, and Tyr-337 (not shown for clarity) (Fig. 2A). These residues interact with the aliphatic portion of the K189 side chain primarily through van der Waals contacts. The lysine binding channel connects to the AdoMet-binding site on the opposite face of the catalytic domain via an oxygen-lined methyl transfer pore (38). During catalysis, the methyl group of the cofactor is positioned within the methyl transfer pore for the SN2 reaction with the -amino group of the lysine or methyl-lysine substrate (see below). To lower the activation barrier for this reaction, the lysine -amine nucleophile is aligned for methyl transfer through a hydrogen bond network within the active site. In the WT enzyme, the K189 -amino group hydrogen bonds to the hydroxyl group of Tyr-245 as well as to two water molecules (Fig. 2A). One of the water molecules (termed water 1), is coor- dinated in a solvent pocket, through hydrogen bonds to the carbonyl oxygens of Gly-292 and Ala-295 and to the hydroxyl group of the Phe/Tyr switch residue Tyr-305. This solvent pocket is structurally conserved in SET domain KMTs and has an important role in defining product specificity through the adjacent Phe/Tyr switch residue, as shown in our prior studies of the human H4K20 methyltransferase SET8 (8). The other water molecule is bound within the methyl transfer pore between the lysine substrate and the thioether sulfur atom of AdoHcy through hydrogen bonds to Tyr-245, Asn-265, and His-293 in SET7/9 and the TAF10-K189 -amino group. This water is not observed in other structures of SET7/9 ternary complexes and may represent the approximate position that the AdoMet methyl group occupies in the methyl transfer pore in the Michaelis complex. In structures of the Y305F ternary complexes, the K189, K189me1, and K189me2 side chains also adopt extended trans side chain geometries within the lysine binding channel that are stabilized via hydrogen bonding to Tyr-245 in the enzyme (Fig. 2, B–D). The orientations of the K189me1 and K189me2 side chains are further maintained through carbon-oxygen (CH–O) hydrogen bonding between the methyl groups and oxygen atoms within the vicinity of the methyl transfer pore, as reported previously in other SET domain KMT structures (8, 10, 38). A superimposition of the SET7/9 WT and Y305F com- plexes underscores the similarity of the lysyl binding conforma- tions (Fig. 2E). However, there are notable differences in the hydrogen bond patterns and occupancy of water 1 within the solvent pocket in the Y305F mutant compared with the WT enzyme. Specifically, the Y305F substitution results in the loss of one hydrogen bond to water 1 in the structures of the TAF10- K189 and TAF10-K189me1 complexes (Fig. 2, B and C). In con- trast, water 1 is absent in TAF10-K189me2 complex, and the vacated solvent pocket is occupied by one of the methyl groups TABLE 2 Kinetic parameters of WT SET7/9 and the Y305F and Y245A mutants Enzyme TAF10 peptide substrate Km a kcat a kcat/Km a M min1 M1 min1  103 WT K189 160  17 17  0.62 110  17 Y305F K189 88  5.0 17  0.30 190  11 Y305Fb K189me1 11  0.50 Y245A K189 200  35 0.53  0.04 2.6  0.47 Y245A K189me1 210  23 5.9  0.23 28  3.3 Y245A K189me2 400  29 6.5  0.16 15  1.2 a Curve fitting errors were calculated from the hyperbolic fits for the Michaelis-Menten equation. b Km and kcat were not determined because the Km value was beyond the measurable range; therefore, the kcat/Km value is reported. TABLE 3 Product analysis of WT SET7/9 and the Y305F and Y245A mutants Enzyme Quantity of enzyme Measured product Amount of product formeda nmol nmol WT 0.003 Kme1 0.65  0.07 Y305F 0.006 Kme1 1.5  0.49 Kme2 0.033  0.009 Y245A 0.100 Kme1 0.80  0.22 Kme2 0.39  0.021 Kme3 0.076  0.019 a Standard deviation was calculated from triplicate measurements. Lysine Methylation by SET7/9 Mutants 31852 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 of the dimethyl -amine (Fig. 2D). This methyl group forms a 3.5-Å CH–O hydrogen bond to the carbonyl oxygen of Ala-295, further stabilizing the binding of the dimethyl-lysine side chain. A homologous dimethyl-lysine-binding mode and CH–O hydrogen bond was observed in our prior structural studies of the SET8 Y334F Phe/Tyr switch mutant that confers an analo- gous change in product specificity from a mono- to a dimeth- yltransferase (8). A structural alignment of the active sites of the SET7/9 Y305F and SET8 Y334F mutants bound to cognate dimethylated peptides illustrates that the coordinates of the dimethyl-lysyl side chains are virtually superimposable, with one methyl group oriented toward the methyltransfer pore and the second positioned within the vacant solvent pocket (Fig. 2F). Taken together, the structures of the SET7/9 Y305F com- plexes and the similarities in the dimethyl-lysine conforma- tions in the SET7/9 Y305F and SET8 Y334F mutants imply that the Phe/Tyr switch governs product specificity through a con- served mechanism whereby it indirectly influences the binding modes of the methyl-lysine side chain by modulating the affin- ity of the water molecule (water 1) bound in the solvent pocket. Biochemical Characterization of the SET7/9 Y245A Mutant—Previ- ous studies by Xiao et al. (11) reported that the Y245A mutation yields an unusual change in the product specificity of SET7/9, converting the enzyme to a trimeth- yltransferase with weak monometh- yltransferase activity. We deter- mined that the SET7/9 Y245A could mono-, di-, and trimethylate the TAF10-K189 peptide by mass spec- trometry (data not shown) and TLC (Table 3), confirming the earlier studies of Xiao et al. (11). ITC analysis revealed that the Y245A mutant displayed comparable KD values for the unmodified and methylated TAF10-K189 peptides (Fig. 1), although its affinity for the trimethylated peptide was modestly diminished in comparison with the other peptides (Table 1). The ITC data demonstrate that the Y245A mutant bound the unmodified, mono-, and dimethylated sub- strates with equivalent affinities, suggesting that a kinetic effect or a structural alteration in the active site may be responsible for its diminished activity toward un- modified substrates. To gain further insight into its peculiar product specificity, we characterized the kinetic properties of the SET7/9 Y245A mutant. Steady state analysis demonstrated that this mutant displayed similar Km values for the unmodified, mono- and dimethylated TAF10 peptides (Table 2). However, the turnover number for the TAF10-K189 peptide was diminished over 10-fold versus the methylated peptides and was reduced 30-fold versus the WT enzyme, in agreement with the weak monomethyltransferase activity reported by Xiao et al. (11). In addition, we investigated whether this mutant catalyzes lysine trimethylation via a pro- cessive or distributive mechanism as described for SET7/9 Y305F. The TLC data illustrate that the mono- and dimethyl- lysine intermediates accumulated at quantities greater than that of the enzyme used in the assay, indicating that SET7/9 Y245A obeys a distributive mechanism, analogous to the Y305F mutant (Table 3). Structures of SET7/9 Y245A Bound to Unmodified and Meth- ylated TAF10 Peptides—To elucidate the mechanism underly- ing its unusual product specificity, we determined the crystal structures of SET7/9 Y245A in complex with AdoHcy and unmodified, mono-, di-, and trimethylated TAF10 peptides (supplemental Table 1). These complexes superimpose with the structure of the WT SET7/9AdoHcyTAF10-K189 com- FIGURE 2. Crystal structures of WT SET7/9 and the Y305F mutant in complex with AdoHcy and unmodi- fied and methylated TAF10 peptides. A, active site of WT SET7/9 bound to AdoHcy and TAF10-K189 and the active site of SET7/9 Y305F bound to AdoHcy and TAF10-K189 (B), TAF10-K189me1 (C), and TAF10-K189me2 peptides (D). A–D, the K189 residue in the TAF10 peptide is depicted with yellow carbon atoms with the corresponding Fo – Fc simulated-annealing omit maps (contoured at 2.5 ) shown. AdoHcy is colored with green carbon atoms, and SET7/9 carbon atoms are colored gray, with the exception of Tyr-305/Y305F (magenta) and Tyr-245 (cyan). Gly-264 to Ser-268 are shown, and the main chain atoms of Gly-292 to Ala-295 are depicted for clarity. Red dashed lines indicate conventional hydrogen bonds, and blue dashed lines indicate CH–O hydrogen bonds. Cyan dashed lines indicate weak CH–O hydrogen bonds between 3.5 and 3.7 Å in length. The water molecule in the solvent binding pocket is numbered 1. E, overlay of the active sites of the WT enzyme (light blue carbon atoms) and the Y305F mutant (green, yellow, and pink carbon atoms correspond to K189, K189me1, and K189me2, respectively). Red dashed lines indicate coordination of the waters in the struc- turesoftheY305Fmutant,andthelightbluedashedlineindicateshydrogenbondingthatonlyoccursintheWT enzyme. F, superimposition of the active sites of the SET7/9 Y305FAdoHcyTAF10-K189me2 and SET8 Y334FAdoHcyH4K20me2 (Protein Data Bank code 3F9X) complexes shown with green and yellow carbon atoms, respectively. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31853 plex with root mean squared differences of less than 0.4 Å for all aligned atoms, indicating that the Y245A mutant does not per- turb the overall structure of the enzyme. Simulated annealing omit maps illustrate that K189 side chains are bound within the lysine binding channel through hydrogen bonds and van der Waals contacts (Fig. 3, A–D), although the interactions and binding modes are distinct from those in the complexes of WT SET7/9 and the Y305F mutant (Fig. 2, A–D). In the unmodified TAF10 peptide complex, the K189 -amino group forms a weak hydrogen bond with the carbonyl oxygen of Gly-264 (Fig. 3A), whereas the -amino groups of K189me1 and K189me2 hydro- gen bond to the hydroxyl group of Tyr-305 in the mono- and dimethylated peptide complexes (Fig. 3, B and C). The confor- mations of the K189me1 and K189me2 side chains are further stabilized by water-mediated hydrogen bonding and through CH–O hydrogen bonding to their methyl groups. In the TAF10-K189me3 peptide complex, the trimethyl-lysine side chain is coordinated exclusively through direct and water-me- diated CH–O hydrogen bonds to its methyl groups because the quaternary -ammonium cation cannot engage in hydrogen bonding (Fig. 3D). A structural alignment of the four SET7/9 Y245A complexes illustrates distinct binding modes for the unmodified versus the methylated K189 side chains, highlighting the selectivity of this mutant for methylated substrates. The side chains of K189me1, K189me2, and K189me3 roughly overlay with their respective -amino groups superimposed and adopt slightly kinked con- formations (Fig. 3E), as opposed to the extended trans geome- try of the unmodified and methylated lysines in the complexes of the WT enzyme and the Y305F mutant (Fig. 2, A–D). Con- versely, the unmodified K189 side chain does not superimpose with its methylated counterparts and is oriented in an alterna- tive configuration due to its hydrogen bonding to Gly-264 (Fig. 3, A and E). An overlay of the structures of the WT enzyme and Y245A mutant bound to the unmodified TAF10 peptide illus- trates that the side chains of K189 do not superimpose and that the K189 -amino group appears to be misaligned with AdoHcy in the Y245A complex (Fig. 3F). This suboptimal alignment may explain the diminished kcat value of SET7/9 Y245A mutant toward substrates with unmodified lysines (Table 2). A comparison of the structures of the SET7/9 Y245A and Y305F complexes yields a molecular explanation for the differ- ent product specificities of these two mutants. In the SET7/9 Y305F complexes, Tyr-245 aligns the K189 -amino group for methyl transfer through hydrogen bonding to its hydroxyl group (Fig. 2, B–D). Conversely, in the Y245A mutant, the K189me1 and K189me2 -amino groups are oriented through hydrogen bonding to Tyr-305 (Fig. 3, B and C). These distinct hydrogen bond patterns impart differences in the conforma- tions of the lysyl side chains due to the relative orientations of Tyr-245 and Tyr-305 in the lysine binding channel. Specifically, the kinked conformation adopted by the K189me1 and K189me2 side chains in the Y245A complexes (Fig. 3, B and C) may contribute to the differences in the turnover numbers of this mutant versus those of the WT enzyme and the Y305F mutant (Table 2). In addition, the dimethyl -amino group of the K189me2 side chain binds in distinct orientations in the Y245A and Y305F mutants due to their hydrogen bonding to FIGURE 3. Crystal structures of the SET7/9 Y245A mutant in complex with AdoHcy and unmodified and methylated TAF10 peptides. Active site of SET7/9 Y245A bound to AdoHcy and TAF10-K189 (A), TAF10-K189me1 (B), TAF10-K189me2 (C), and TAF10-K189me3 peptides (D). A–D, Fo – Fc simulated- annealing omit maps (contoured at 2.5 ) for the unmodified and methylated K189 side chains are illustrated. The residues and hydrogen bonds in each com- plexarecoloredasinFig.2.Thewatermoleculesinthelysinebindingchannelsof the Y245A complexes are numbered 1–4, as described in the text. E, superimpo- sition of the active sites of the Y245A complexes bound to the four methylated statesofTAF10-K189(unmodified,mono-,di-,andtrimethylatedshowningreen, yellow, pink, and blue, respectively). F, overlay of WT SET7/9 (blue carbons) and SET7/9 Y245A (green carbons) in complex with TAF10-K189. Waters correspond- ing to the WT and Y245A structures are colored cyan and green, respectively. G, alignment of the active sites of the Y245A (yellow carbons) and Y305F (green carbons) mutants in complex with the TAF10-K189me2 peptide. Hydrogen bonds from the Y305F structure are shown as green dashed lines, and waters and hydrogen bonds in the Y245A structure are shown in yellow and orange, respectively. Lysine Methylation by SET7/9 Mutants 31854 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 Tyr-305 and Tyr-245, respectively (Fig. 3G). In the Y305F mutant, hydrogen bonds to the dimethyl -amino group cou- pled with steric constraints in the lysine binding channel pre- vent the K189me2 side chain from undergoing a conforma- tional change that is conducive to trimethylation (Fig. 2D), consistent with its dimethyltransferase activity. However, in the Y245A mutant, the alanine substitution enlarges the diameter of the lysine binding channel, accommodating trimethyl-lysine (Fig. 3D). In addition, the larger diameter would permit the dimethyl-lysine substrate to undergo the conformational reor- ganization necessary to align the -amino group in a productive geometry for trimethylation. A major difference in the active site of the Y245A mutant versus the other SET7/9 structures is the presence of several water molecules bound in the cavity generated by the Y245A mutation. In the structure of the Y245A mutant bound to TAF10-K189, three water molecules (waters 2–4) occupy this cavity and are arranged in a triangular geometry (Fig. 3A). In addition, water 1 shifts 1.6 Å from its position in the solvent pocket toward water 2 to which it forms a hydrogen bond (Fig. 3, A and E). The shift in water 1 was unexpected given its con- served orientation in the solvent pocket of the SET7/9 WT and Y305F complexes (Fig. 2, A–C) as well as in the structures of other SET domain KMTs (8). This displacement is presumably related to the alternative conformation of the K189 side chain whose -amino group is too distant (4.3 Å) to form a productive hydrogen bond to water 1. Conversely, in the Y245A complexes bound to TAF10-K189me1 and TAF10-K189me2, water 1 remains tightly bound in the solvent pocket through hydrogen bonds to Tyr-305 hydroxyl group, the carbonyl oxygens of Gly- 292 and Ala-295, and the K189 -amino group (Fig. 3, B, C, and E), analogous to its binding in the WT enzyme (Fig. 2A). How- ever, in the TAF10-K189me3 complex, one of the methyl groups of the trimethyl -ammonium cation is oriented into the solvent pocket (Fig. 3D), similar to the dimethyl-lysine binding mode observed in the Y305F mutant (Fig. 2D). The binding of the methyl group in the solvent pocket displaces water 1 by 3.2 Å relative to its position in the TAF10-K189me1 complex (Fig. 3E), thereby avoiding a steric clash with the trimethylated -ammonium group. Variations in the occupancy of water 2 are also seen in the different Y245A structures. Water 2 is bound in similar orientations in the active site of the unmodified and monomethylated peptide complexes but is absent in the di- and trimethylated peptide complexes due to the binding of a methyl group in this position (Fig. 3, A–E). In summary, the changes in the positions or occupancies of waters 1 and 2 correlate with the binding modes of the unmodified and methylated K189 within the active site of the Y245A mutant. Catalytic Models of Lysine Multiple Methylation by SET7/9 Y245A, and Y305F—The structures of the SET7/9 complexes reported here offer a prime opportunity to generate stepwise models for lysine mono-, di-, and trimethylation by a SET domain KMT. We modeled the AdoMet-bound Michaelis complexes by superimposing the SET7/9 product complexes with the previously reported structure of the SET7/9-AdoMet binary complex (Fig. 4) (39). The conformations of the mono- and dimethyl -amino groups in the Michaelis complexes were inferred from the coordinates of the corresponding dimethyl- and trimethyl-lysine products, respectively. In addition, we modeled the -amino group in a deprotonated state with its hydrogen atoms oriented toward the hydrogen bond acceptors that align the lysyl side chain for methylation. As a basis for this comparison, we first modeled the monomethylation reaction catalyzed by WT SET7/9 (Fig. 4A). In the substrate ternary complex, the lysine -amine is aligned with the methyl group and sulfonium cation of AdoMet through a hydrogen bond to the Tyr-245 hydroxyl group and water 1 in the solvent pocket. The values of the reaction distance and angle are 2.8 Å and 153°, respectively, in approximate agreement with the linear geome- try of a SN2 methyl transfer reaction calculated in other mod- eled substrate complexes (8, 10). In the product complex, the monomethyl-lysine side chain is bound in an extended confor- mation with its methyl group oriented within the methyl trans- fer pore, thereby obstructing AdoMet binding. Furthermore, water 1 remains tightly coordinated in the solvent pocket through four hydrogen bonds to Gly-292, Ala-295, Tyr-305, and the monomethyl -amino group. These interactions hinder the dissociation of water 1 and the related rearrangement of the monomethyl-lysine side chain required for a second methyl transfer reaction, explaining why the WT enzyme cannot cata- lyze di- and trimethylation. These findings concur with the FIGURE 4. Catalytic models of the methyl transfer reactions catalyzed by WT SET7/9 and the Y305F mutant. A, monomethylation of TAF10-K189 by the WT enzyme. The reaction scheme depicts the modeled substrate ternary complex (left) and the product complex (right) for the transfer of the methyl group from AdoMet (green carbon atoms) to K189 in TAF10 (yellow carbons), yielding AdoHcy and K189me1. The red arrow indicates the direction of the nucleophilic attack of the deprotonated -amino group on the AdoMet methyl group. The transferred methyl group is colored green, and the white atoms represent the hydrogens of the -amino group. Hydrogen bonds and residues in the enzyme active site are illustrated as in Fig. 2. The reaction distance and angle are labeled in red. B and C, models of the Y305F mutant for the first methyl transfer reaction with TAF10-K189 (B) and second methyl transfer reaction with TAF10-K189me1 (C). Color schemes are the same as in A. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31855 model for SET7/9 product specificity reported in previous structural and functional studies (6, 11). Similar reaction geometry is observed in the model for the monomethyl transfer reaction catalyzed by SET7/9 Y305F. Hydrogen bonds from the Tyr-245 hydroxyl group and water 1 align the lysine -amino group with the AdoMet methyl group at a distance of 2.1 Å and an angle of 160°, equivalent to those measured in the Michaelis complex of the WT enzyme (Fig. 4, A and B). In the product complex, monomethyl-lysine adopts an extended trans configuration analogous to that in the WT enzyme. For dimethylation to occur, the monomethyl-lysine must undergo a conformational change in which its methyl group is rotated out of the methyl transfer path with AdoMet. The structure of the Y305F mutant bound to the dimethylated TAF10 peptide (Fig. 2D) implies that this rearrangement occurs through the dissociation of water 1 due to the loss of the Tyr- 305 hydrogen bond in the solvent pocket. The dissociation of water 1 would enable the monomethyl-lysine side chain to adopt an alternative conformation through a rotation about its C–N bond, projecting the methyl group into the solvent pocket (Fig. 4C). This rotation reorients the methyl group out of the methyl transfer path while realigning the monomethyl - amino group for a second methylation reaction through a direct hydrogen bond to the Tyr-245 hydroxyl group and a CH–O hydrogen bond between its methyl group and Ala-295. The modeled reaction geometry for monomethyl-lysine substrate complex of the Y305F mutant (2.7 Å, 160°) is equivalent to that of the first methyl transfer reaction in SET7/9 Y305F. These geometries concur with our previous models for mono- and dimethylation catalyzed by SET8 Y334F (8), illustrating that the orientation of a methyl group into the solvent pocket is a con- served feature of SET domain KMTs that catalyze multiple methylation. In addition, we modeled the methyl transfer reactions cata- lyzed by SET7/9 Y245A (Fig. 5). In the model of the lysine sub- strate complex, the -amino group is aligned for methyl transfer by a hydrogen bond to the carbonyl oxygen of Gly-264, result- ing in a short reaction distance (2.3 Å) and a suboptimal reac- tion angle (141°) with the methyl group of AdoMet (Fig. 5A). This misalignment appears to be a direct consequence of the Y245A mutation that abolishes hydrogen bonding to the - amino group, illustrating that the suboptimal orientation of the -amine likely contributes to the diminished activity of this mutant toward unmodified substrates (11). Conversely, in the modeled monomethyl-lysine substrate complex for SET7/9 Y245A, Tyr-305 hydrogen bonds to the -amino group, aligning it for methyl transfer (Fig. 5B). In addition, CH–O hydrogen bonds to the monomethyl-lysine methyl group and the dissoci- ation of water 2 from the active site also contribute to reposition- ing the -amino group for dimethylation. Collectively, these inter- actions orient the -amine in a reaction angle of 165° that is more conducive to methyl transfer. However, the reaction distance for dimethylation is 0.6 Å longer than that in the corresponding Y305F model because Tyr-305 is positioned further from AdoMet than Tyr-245 (Figs. 4C and 5B). In the third methyl transfer reaction catalyzed by SET7/9 Y245A, the lone pair of electrons of the dimethyl-lysine - amino group acts as the nucleophile and thus cannot engage in hydrogen bonding. The structure of the trimethyl-lysine prod- uct complex (Fig. 3D) implies that the dimethyl -amine is aligned via CH–O hydrogen bonds to its methyl groups, as shown in the model of the Michaelis complex for this reaction (Fig. 5C). These CH–O hydrogen bonds restrain the orienta- tion of the -amino group and position one of the methyl groups into the solvent pocket, displacing water 1 as discussed earlier (Fig. 3, D and E). These interactions cumulatively align the -amino group and AdoMet methyl group with a reaction distance of 3.0 Å and angle of 162° (Fig. 5C). Taken together, the models of the substrate complexes for SET7/9 Y245A suggest that CH–O hydrogen bonds play an increasingly important role in aligning the methylated -amino group in successive rounds of methyl transfer. DISCUSSION The structural and functional characterization of the SET7/9 Y245A and Y305F mutants presented here yields new insights into the mechanism underlying the product specificity of SET domain KMTs. Importantly, it resolves a general paradox concerning this specificity. How does the active site constrain the motion of the lysine -amino group to align it for methyl transfer with AdoMet, while providing adequate volume to accommodate the mono-, di-, and tri- methylated lysine side chain generated during multiple methyl transfer reactions? The structures of the Y305F and Y245A mutants resolve this paradox, illustrating that alter- ations in the positions or occupancies of water molecules within their active sites generate the space required to FIGURE 5. Models for the methyl transfer reactions catalyzed by the SET7/9 Y245A mutant. Models of the Y245A mutant for the first methyl transfer reaction with TAF10-K189 (A), the second methyl transfer reaction with TAF10-K189me1 (B), and the third methyl transfer reaction with TAF10- K189me2 (C). Residues and hydrogen bonds are depicted as in Fig. 3. Lysine Methylation by SET7/9 Mutants 31856 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 accommodate the multiply methylated -amine produced during successive catalytic cycles. Minor perturbations in the side chains of certain active site residues, such as Tyr- 305, are also observed in alignments of the WT enzyme and the Y245A and Y305F complexes, although these changes are modest compared with the displacement or dissociation of the water molecules in the active site. These findings suggest that the waters function as transient place holders that facil- itate the SN2 methyl transfer reaction. During monomethy- lation, they function to constrain the movement of the lysine -amino group by mediating hydrogen bonds between the substrate and enzyme, thereby promoting the linear align- ment with the methyl group and sulfonium cation of AdoMet (Fig. 4, A and B). During di- and trimethylation, the water molecules either relocate within the lysine binding channel or dissociate from the enzyme, yielding the space required to rotate the methyl group away from the methyl transfer pore and to realign the -amine in productive geom- etry for the next methyl transfer reaction (Figs. 4C and 5, B and C). These findings agree with our prior analysis of the SET8 Phe/Tyr switch mutant in which we demonstrated that the Y334F substitution attenuates hydrogen bonding to the water molecule bound in the solvent pocket, promoting its dissociation and the conformational changes necessary for lysine dimethylation (8). Indeed, there is a nearly identical alignment of the dimethyl-lysine side chains in the structures of SET7/9Y305FandSET8Y334Fcomplexes,despitethedifferences in the orientations of the Phe-305 and Phe-334 side chains in each structure (Fig. 2F). Finally, both the SET8 Y334F (8) and SET7/9 Y305F mutants (Table 2) displayed diminished catalytic efficien- cies for lysine dimethylation versus monomethylation. These dif- ferences may reflect the kinetics of the reorganization within the active site, including the dissociation of the water molecule from the solvent pocket and the concomitant realignment of the monomethyl-lysine into a productive geometry for dimethylation. In addition to their place-holding role, the active site waters may also facilitate the deprotonation of the lysine -amino group between methyl transfer reactions. For methylation to occur, the -amino group must be deprotonated to function as the nucleophile in the SN2 methyl transfer reaction with AdoMet (Figs. 4 and 5). Although the pKa value of the lysine -amine in solution is 10.5, molecular dynamics simulations by Zhang and Bruice (25, 26) indicate that this value diminishes to 8.2 upon formation of the SET7/9 Michaelis complex due to the proximity of the AdoMet sulfonium cation and the low dielectric constant of the active site. Furthermore, their simu- lations show that a chain of water molecules facilitates the dep- rotonation of the -amino group prior to methyl transfer, trans- ferring the proton to bulk solvent. Although these water molecule chains are not evident in our crystal structures, the Y305F and Y245A complexes suggest another potential mech- anism for deprotonation. In the dimethyl-lysine complexes of the Y305F and Y245A mutants, the dissociation of water 1 and 2, respectively, from the lysine binding channel requires that the solvent-mediated hydrogen bond to the -amino group is broken (Figs. 2D and 3C). It is conceivable that these waters dissociate from the active site as hydronium ions, promoting the realignment and deprotonation of the methyl -amino group for the next methyl transfer reaction. A comparison of the SET7/9 Y305F and SET8 Y334F com- plexes yields insights into the mechanism by which the Phe/Tyr switch influences water binding within the solvent pocket. The phenylalanine substitution in the Phe/Tyr switch results in the loss of a single hydrogen bond to the water molecule (water 1) in the solvent pocket compared with the four hydrogen bonds that coordinate the solvent molecule in WT SET7/9 (Fig. 2, A and B) and SET8 (7, 8). Although this attenuation in hydrogen bonding may appear insignificant, this difference is nonetheless impor- tant for at least two reasons. First, theoretical calculations indi- cate that, on average, water molecules form 3.5 hydrogen bonds in solutions (40, 41). This value is greater than the num- ber of hydrogen bonds coordinating water 1 in the solvent pocket in SET7/9 Y305F (Fig. 2, B and C) as well as in SET8 Y334F and other di- and trimethyltransferases that possess a hydrophobic residue in the Phe/Tyr switch site (8). From the perspective of the water molecule, the greater hydrogen bond- ing potential in solution would tend to thermodynamically favor its dissociation from the solvent pocket in SET domain KMTs that lack a tyrosine in the Phe/Tyr switch position. Sec- ond, the ordered binding of water molecules observed in the active sites of SET domain ternary complexes represents an unfavorable entropy compared with their diffusion in bulk sol- vent. In WT SET7/9 (Fig. 2A) and SET8 (7, 8), this entropic penalty can be partially offset through the favorable enthalpy of binding associated with the four hydrogen bonds that coordi- nate the water within the solvent pocket. It is conceivable that the loss of the tyrosine-mediated hydrogen bond in the Phe/Tyr switch shifts the equilibrium in favor of dissociation of the water molecule from the solvent pocket, thereby facilitating dimethylation in SET7/9 Y305F, SET8 Y334F, and other di- and trimethyltransferases. The structures of the SET7/9 Y245A and Y305F complexes illustrate the interactions that align the lysine -amino group during the methyl transfer reactions in each enzyme. In the WT enzyme and the Y305F mutant, hydrogen bonding to the hydroxyl group of Tyr-245 appears to be critical in properly aligning the -amine for methyl transfer (Fig. 4). Tyr-245 is conserved in the sequences of many SET domain KMTs (8, 42), and substitutions of this residue generally impair or abolish activity, indicating its importance in catalysis (8, 43). However, SET7/9 appears to be an exception to this rule, as the Y245A mutant is not only active but is capable of catalyzing lysine trimethylation. In this mutant, Tyr-305 appears to assume the role of Tyr-245 by hydrogen bonding to the monomethylated -amino group to align it for methyl transfer with AdoMet, as illustrated in the modeled substrate complex for the dimethy- lation reaction (Fig. 5B). Conversely, in the model for trimethy- lation, the Tyr-305 hydroxyl group does not hydrogen bond to the -amine but instead participates in a CH–O hydrogen bond with one of the methyl groups to assist in aligning the dimethy- lated -amine for the methyl transfer reaction (Fig. 5C). Addi- tional structural and functional studies of the SET domain tri- methyltransferases will aid in further illuminating the roles of CH–O hydrogen bonds in facilitating lysine multiple methylation. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31857 Acknowledgments—We acknowledge S. Schiebold for assistance in protein expression, purification, and crystallization and S. Anderson and R. Sanishvili for their assistance with x-ray data collection. We also thank S. Bulfer and S. Horowitz for reading the manuscript and providing useful comments. This work utilized the Protein Structure Facility of the Michigan Diabetes Research and Training Center, Uni- versity of Michigan, supported by National Institutes of Health Grant DK020572, NIDDK. Use of the Advanced Photon Source was sup- ported by the United States Department of Energy, Basic Energy Sci- ences, Office of Science, under Contract DE-AC02-06CH11357. GM/CA CAT has been funded in whole or in part by National Institutes of Health NCI Grant Y1-CO-1020 and NIGMS Grant Y1-GM-1104. Use of the LS-CAT Sector 21 was supported by Michi- gan Economic Development Corporation and the Michigan Technol- ogy Tri-Corridor Grant 085P1000817 for the support of this research program. REFERENCES 1. Huang, J., and Berger, S. L. (2008) Curr. Opin. Genet. Dev. 18, 152–158 2. Morgunkova, A., and Barlev, N. A. (2006) Cell Cycle 5, 1308–1312 3. Yang, X. D., Lamb, A., and Chen, L. F. (2009) Epigenetics 4, 429–433 4. Taverna, S. D., Li, H., Ruthenburg, A. J., Allis, C. D., and Patel, D. J. (2007) Nat. Struct. Mol. Biol. 14, 1025–1040 5. Collins, R. E., Tachibana, M., Tamaru, H., Smith, K. M., Jia, D., Zhang, X., Selker, E. U., Shinkai, Y., and Cheng, X. (2005) J. Biol. Chem. 280, 5563–5570 6. Zhang, X., Yang, Z., Khan, S. I., Horton, J. R., Tamaru, H., Selker, E. U., and Cheng, X. (2003) Mol. Cell 12, 177–185 7. Couture, J. F., Collazo, E., Brunzelle, J. S., and Trievel, R. C. (2005) Genes Dev. 19, 1455–1465 8. Couture, J. F., Dirk, L. M., Brunzelle, J. S., Houtz, R. L., and Trievel, R. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 20659–20664 9. Qian, C., Wang, X., Manzur, K., Sachchidanand, Farooq, A., Zeng, L., Wang, R., and Zhou, M. M. (2006) J. Mol. Biol. 359, 86–96 10. Trievel, R. C., Flynn, E. M., Houtz, R. L., and Hurley, J. H. (2003) Nat. Struct. Biol. 10, 545–552 11. Xiao, B., Jing, C., Wilson, J. R., Walker, P. A., Vasisht, N., Kelly, G., Howell, S., Taylor, I. A., Blackburn, G. M., and Gamblin, S. J. (2003) Nature 421, 652–656 12. Chuikov, S., Kurash, J. K., Wilson, J. R., Xiao, B., Justin, N., Ivanov, G. S., McKinney, K., Tempst, P., Prives, C., Gamblin, S. J., Barlev, N. A., and Reinberg, D. (2004) Nature 432, 353–360 13. Ea, C. K., and Baltimore, D. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 18972–18977 14. Este`ve, P. O., Chin, H. G., Benner, J., Feehery, G. R., Samaranayake, M., Horwitz, G. A., Jacobsen, S. E., and Pradhan, S. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 5076–5081 15. Kouskouti, A., Scheer, E., Staub, A., Tora, L., and Talianidis, I. (2004) Mol. Cell 14, 175–182 16. Masatsugu, T., and Yamamoto, K. (2009) Biochem. Biophys. Res. Commun. 381, 22–26 17. Munro, S., Khaire, N., Inche, A., Carr, S., and La Thangue, N. B. (2010) Oncogene 29, 2357–2367 18. Pagans, S., Kauder, S. E., Kaehlcke, K., Sakane, N., Schroeder, S., Dorm- eyer, W., Trievel, R. C., Verdin, E., Schnolzer, M., and Ott, M. (2010) Cell Host Microbe 7, 234–244 19. Subramanian, K., Jia, D., Kapoor-Vazirani, P., Powell, D. R., Collins, R. E., Sharma, D., Peng, J., Cheng, X., and Vertino, P. M. (2008) Mol. Cell 30, 336–347 20. Wang, J., Hevi, S., Kurash, J. K., Lei, H., Gay, F., Bajko, J., Su, H., Sun, W., Chang, H., Xu, G., Gaudet, F., Li, E., and Chen, T. (2009) Nat. Genet. 41, 125–129 21. Yang, X. D., Huang, B., Li, M., Lamb, A., Kelleher, N. L., and Chen, L. F. (2009) EMBO J. 28, 1055–1066 22. Guo, H. B., and Guo, H. (2007) Proc. Natl. Acad. Sci. U.S.A. 104, 8797–8802 23. Hu, P., Wang, S., and Zhang, Y. (2008) J. Am. Chem. Soc. 130, 3806–3813 24. Hu, P., and Zhang, Y. (2006) J. Am. Chem. Soc. 128, 1272–1278 25. Zhang, X., and Bruice, T. C. (2007) Biochemistry 46, 14838–14844 26. Zhang, X., and Bruice, T. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 5728–5732 27. Couture, J. F., Collazo, E., Hauk, G., and Trievel, R. C. (2006) Nat. Struct. Mol. Biol. 13, 140–146 28. Trievel, R. C., Beach, B. M., Dirk, L. M., Houtz, R. L., and Hurley, J. H. (2002) Cell 111, 91–103 29. Kapust, R. B., To¨zse´r, J., Fox, J. D., Anderson, D. E., Cherry, S., Copeland, T. D., and Waugh, D. S. (2001) Protein Eng. 14, 993–1000 30. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307–326 31. Vagin, A., and Teplyakov, A. (2000) Acta Crystallogr. D Biol. Crystallogr. 56, 1622–1624 32. Emsley, P., and Cowtan, K. (2004) Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 33. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Acta Crystallogr. D Biol. Crystallogr. 53, 240–255 34. Davis, I. W., Leaver-Fay, A., Chen, V. B., Block, J. N., Kapral, G. J., Wang, X., Murray, L. W., Arendall, W. B., 3rd, Snoeyink, J., Richardson, J. S., and Richardson, D. C. (2007) Nucleic Acids Res. 35, W375–W383 35. Bru¨nger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. D Biol. Crystallogr. 54, 905–921 36. Collazo, E., Couture, J. F., Bulfer, S., and Trievel, R. C. (2005) Anal. Bio- chem. 342, 86–92 37. Chirpich, T. P., Zappia, V., Costilow, R. N., and Barker, H. A. (1970) J. Biol. Chem. 245, 1778–1789 38. Couture, J. F., Hauk, G., Thompson, M. J., Blackburn, G. M., and Trievel, R. C. (2006) J. Biol. Chem. 281, 19280–19287 39. Kwon, T., Chang, J. H., Kwak, E., Lee, C. W., Joachimiak, A., Kim, Y. C., Lee, J., and Cho, Y. (2003) EMBO J. 22, 292–303 40. Chandra, A., and Chowdhuri, S. (2002) J. Phys. Chem. B 106, 6779–6783 41. Guardia, E., Marti, J., Garcia-Tarres, L., and Laria, D. (2005) J. Mol. Liq. 117, 63–67 42. Dillon, S. C., Zhang, X., Trievel, R. C., and Cheng, X. (2005) Genome Biol. 6, 227 43. Zhang, X., Tamaru, H., Khan, S. I., Horton, J. R., Keefe, L. J., Selker, E. U., and Cheng, X. (2002) Cell 111, 117–127 Lysine Methylation by SET7/9 Mutants 31858 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010
3M55
SET7/9 Y305F in complex with TAF10-K189me1 peptide and AdoHcy
SET7/9 Catalytic Mutants Reveal the Role of Active Site Water Molecules in Lysine Multiple Methylation*□ S Received for publication,February 15, 2010, and in revised form, July 14, 2010 Published, JBC Papers in Press,August 1, 2010, DOI 10.1074/jbc.M110.114587 Paul A. Del Rizzo‡, Jean-Franc¸ois Couture‡§1, Lynnette M. A. Dirk¶, Bethany S. Strunk‡, Marijo S. Roiko‡, Joseph S. Brunzelle, Robert L. Houtz¶, and Raymond C. Trievel‡2 From the ‡Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109, the §Ottawa Institute of Systems Biology, Department of Biochemistry, Microbiology, and Immunology, University of Ottawa, Ottawa, Ontario K1H 8M5, Canada, the ¶Department of Horticulture, Plant Physiology/Biochemistry/Molecular Biology Program, University of Kentucky, Lexington, Kentucky 40546, and the Department of Molecular Pharmacology and Biological Chemistry, Northwestern University Center for Synchrotron Research, Life Sciences Collaborative Access Team, Argonne, Illinois 60439 SET domain lysine methyltransferases (KMTs) methylate specific lysine residues in histone and non-histone substrates. These enzymes also display product specificity by catalyzing dis- tinct degrees of methylation of the lysine -amino group. To elucidate the molecular mechanism underlying this specificity, we have characterized the Y245A and Y305F mutants of the human KMT SET7/9 (also known as KMT7) that alter its prod- uct specificity from a monomethyltransferase to a di- and a tri- methyltransferase, respectively. Crystal structures of these mutants in complex with peptides bearing unmodified, mono-, di-, and trimethylated lysines illustrate the roles of active site water molecules in aligning the lysine -amino group for methyl transfer with S-adenosylmethionine. Displacement or dissocia- tion of these solvent molecules enlarges the diameter of the active site, accommodating the increasing size of the methylated -amino group during successive methyl transfer reactions. Together, these results furnish new insights into the roles of active site water molecules in modulating lysine multiple meth- ylation by SET domain KMTs and provide the first molecular snapshots of the mono-, di-, and trimethyl transfer reactions catalyzed by these enzymes. SET domain enzymes represent a family of S-adenosylmethi- onine (AdoMet)3-dependent methyltransferases that catalyze the site-specific methylation of protein lysyl residues in a host of proteins, including histones, transcription factors, chroma- tin-modifying enzymes, ribosomal subunits, and other sub- strates (1–3). In many instances, these modifications serve to recruit effector proteins that recognize methyl-lysyl residues in a sequence-dependent fashion (4). In addition, SET domain KMTs exhibit product specificity, defined as their ability to cat- alyze mono-, di-, or trimethylation of the lysine -amino group. This specificity is biologically relevant because many methyl- lysine-binding proteins can discriminate among different degrees of lysine methylation (4). Thus, both the site and degree of lysine methylation are critical to recognition by effector proteins. Structural and functional studies have identified a Phe/Tyr switch in the active site of SET domain KMTs that governs their respective product specificities (5, 6). According to this model, KMTs that possess a tyrosine in the Phe/Tyr switch site are limited to catalyzing lysine monomethylation, whereas en- zymes that possess a phenylalanine or another hydrophobic residue in this position display di- or trimethyltransferase activ- ity. Mutational analysis of various SET domain KMTs, includ- ing DIM-5 (KMT1) (6), SET7/9 (6), G9A (KMT1C) (5), and SET8 (also known as PR-SET7 and KMT5A) (7, 8), has demon- strated that substitutions in the Phe/Tyr switch result in pre- dictable changes in product specificity. Several models have been proposed to explain the mechanism by which the Phe/Tyr switch site governs this specificity, including variations in the diameter of the active site due to the size of Phe/Tyr switch residue and steric hindrance by the tyrosine hydroxyl group (6, 9–11). However, our recent studies of the Phe/Tyr switch mutant Y334F in the human histone H4 Lys-20 (H4K20) meth- yltransferase SET8 indicate that the Phe/Tyr switch regulates product specificity via a more subtle mechanism (8). Specifi- cally, the switch modulates the binding of an active site water molecule that in turn regulates the transition from mono- methylation to multiple methylation. Among the KMTs that have been structurally characterized, SET7/9 has emerged as an archetypal model for studying the catalytic mechanism and product specificity of the SET domain family. Although initially isolated as a histone H3 Lys-4 (H3K4)-specific methyltransferase, this KMT has been shown to regulate the functions of numerous non-histone substrates through site-specific methylation (12–21). Early structural and functional studies of SET7/9 identified two active site mutants, * Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth Grant R01 GM073839 (to R. C. T.) and National Institutes of Health Admin- istrative Supplement GM073839-04S1 (to R. C. T.) funded through the American Recovery and Reinvestment Act. This work was also supported by Department of Energy Grant DE-FG02-92ER20075 (to R. L. H.). □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Table 1. The atomic coordinates and structure factors (codes 3M53, 3M54, 3M55, 3M56, 3M57,3M58,3M59,and3M5A)havebeendepositedintheProteinDataBank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). 1 Supported by a Canadian Institutes of Health Research postdoctoral fellowship. 2 To whom correspondence should be addressed: Dept. of Biological Chem- istry, University of Michigan Medical School, 1150 West Medical Center Dr., 5301 MSRB III, Ann Arbor, MI 48109. Tel.: 734-647-0889; Fax: 734-763-4581; E-mail: rtrievel@umich.edu. 3 The abbreviations used are: AdoMet, S-adenosylmethionine; AdoHcy, S-ad- enosylhomocysteine; KMT, lysine methyltransferase; ITC, isothermal titra- tion calorimetry; CH–O, carbon-oxygen hydrogen bond; BisTris, 2-[bis(2- hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 41, pp. 31849–31858, October 8, 2010 © 2010 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31849 Y245A and Y305F, which change its product specificity. The Phe/Tyr switch mutant Y305F alters SET7/9 product specificity from a mono- to dimethyltransferase (6), whereas the Y245A substitution converts the enzyme into a trimethyltransferase with weak monomethyltransferase activity (11). These mutants have been the subjects of numerous molecular modeling simu- lations that have led to various models to explain their distinct product specificities (22–26). However, the lack of structural data for the SET7/9 Y245A and Y305F mutants in complex with cognate methylated peptides has hindered our understanding of the mechanisms that define the respective product specific- ities of these mutants. Moreover, these structures would yield a framework for visualizing the mono-, di-, and trimethylation reactions catalyzed by SET domain KMTs. To gain insight into the molecular basis of their product specificities, we have determined high resolution crystal structures of the SET7/9 Y245A and Y305F mutants in com- plex with peptides of the TATA box-binding protein-associ- ated factor TAF10 bearing the Lys-189 methylation site in unmodified (K189), monomethylated (K189me1), dimethyl- ated (K189me2), and trimethylated (K189me3) states. The structures and accompanying biochemical data support a model whereby changes in the occupancy or position of water molecules in the active site are critical in establishing the prod- uct specificities of the SET7/9 Y245A and Y305F mutants. Together, our results provide new insights into the mechanisms that govern SET domain product specificity and provide step- wise snapshots of the lysine mono-, di-, and trimethyl transfer reactions catalyzed by KMTs. EXPERIMENTAL PROCEDURES Cloning, Expression, and Purification of the SET7/9 Mutants— The Y245A and Y305F mutants were introduced into the pHIS2 SET7/9 expression vector encoding residues 110–366 (27) via QuikChange site-directed mutagenesis (Stratagene) and were verified by dideoxy DNA sequencing. The plasmids encoding wild type (WT) SET7/9 and the Y245A and Y305F mutants were transformed into Rosetta2 DE3 cells (Novagen) and were expressed as described previously (27, 28). In the course of characterizing WT SET7/9, we observed that the enzyme co- purified with AdoMet or another contaminant that resulted in technical difficulties in the isothermal titration calorimetry (ITC) experiments and co-crystallization trials with the TAF10 peptides. To overcome this problem, a denaturation and refold- ing step was inserted in the purification scheme. The denatur- ation and refolding protocol involved adding 6 M guanidine HCl, 50 mM sodium phosphate, pH 7.5, and 500 mM NaCl to the protein while it was immobilized on a nickel-Sepharose column (GE Healthcare). The column was washed with this buffer, fol- lowed by a solution of 6 M urea, 50 mM sodium phosphate, pH 7.5, 150 mM NaCl, and 10 mM 2-mercaptoethanol to remove the cofactor from the denatured enzyme. A reverse gradient from 6 to 0 M urea was then performed in the same buffer to refold the protein, which was subsequently eluted from the column using a linear gradient of 0–500 mM imidazole in 50 mM sodium phosphate, pH 7.5, 500 NaCl, and 10 mM 2-mercaptoethanol. The refolded protein was digested with tobacco etch virus pro- tease (29) during dialysis against 20 mM Tris, pH 7.5, 150 mM NaCl, and 5 mM 2-mercaptoethanol and then purified using a Superdex 200 gel filtration column (GE Healthcare). Protein concentration was determined by its absorbance at 280 nm. Synthetic Peptides—The TAF10 peptides bearing K189, K189me1, K189me2, and K189me3 (sequence, acetyl- SKSK189DRKYTL-amide) and a biotinylated TAF10 peptide (sequence, acetyl-SKSK189DRKYTLT(K-EZLink-S-S-biotin)- amide) were synthesized and purified by New England Peptide, Inc. Peptide concentrations were measured using the absorb- ance of their tyrosine residue at 274 nm. Crystallization and Data Collection—Crystals were pro- duced by hanging drop vapor diffusion by mixing the crystalli- zation solution in a 1:1 ratio with 10–20 mg/ml SET7/9, 1 mM S-adenosylhomocysteine (AdoHcy), and 1.0–3.0 mM unmodi- fied or methylated TAF10-K189 peptide in 20 mM Tris, pH 8.0, 100 mM NaCl, and 2 mM tris(2-carboxyethyl)phosphine. Crys- tals were obtained at 20 °C in either 1.8–2.0 M (NH4)2SO4, with 0.1 M BisTris, pH 6.2–6.6, or in 0.95–1.1 M sodium citrate with 100 mM imidazole pH 8.0–8.4. In both crystallization condi- tions, the final pH values were between pH 8.0 and 9.0. Crystals in the (NH4)2SO4 condition were typically flash-frozen in the mother liquor containing 25–30% glycerol, and the crystals in the citrate condition were frozen in 1.6 M sodium citrate. Data were collected at the Advanced Photon Source beamlines 21-IDG (LS-CAT) and 23-IDD (GM/CA-CAT). Images were indexed, integrated, and scaled using HKL2000 (30). Structures of the mutants were solved by molecular replacement using MOLREP (31) with the coordinates of a previously reported SET7/9 ternary complex used as the search model (Protein Data Bank code 2F69). Successive rounds of model building and refinement were carried out using Coot (32) and REFMAC (33), respectively. The geometry of the models were verified by Mol- Probity (34). Simulated annealing omit maps were calculated using CNS (35) with the peptide and cofactor removed to elim- inate model bias in the active site. Structural figures were ren- dered using PyMOL (Schro¨dinger, LLC.). Fluorescent Methyltransferase Assay—A coupled fluorescent methyltransferase assay was used to measure the kinetic parameters of WT SET7/9 and the Y245A and Y305F mutants as reported previously, with the exception that 50–150 nM enzyme, 100 M AdoMet, and varying concentrations of TAF10 peptide substrate were used (27, 36). Assays were performed in triplicate, and a homocysteine calibration curve was used to calculate the initial velocities. Kinetic parameters were calcu- lated by plotting the velocities versus peptide concentration and by fitting the Michaelis-Menten equation to the data via non- linear regression using Prism 5.0 (GraphPad). In cases where the Km value was beyond the measurable range of the assay, the kcat/Km value was determined as described previously (7). Calorimetry Experiments—ITC was performed at 20 °C using a MicroCal VP-ITC calorimeter (GE Healthcare) with 0.12 mM protein and 4 mM AdoHcy in 20 mM sodium phosphate, pH 7, and 100 mM NaCl with 1.5 mM peptide as the injectant. Data were processed, and equilibrium dissociation constants (KD) and curve fitting errors were calculated from the binding iso- therms using Origin 7.0 (OriginLab Corp.). WT SET7/9 and the Y245A and Y305F mutants displayed ligand:protein binding stoichiometries (N values) between 0.8 and 1.0, demonstrating Lysine Methylation by SET7/9 Mutants 31850 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 that WT SET7/9 and its mutants were properly refolded due to their ability to bind peptides in an 1:1 molar ratio. TLC Product Analysis—Methyltransferase assays were per- formed in triplicate at 37 °C with the biotinylated TAF10 pep- tide (0.5 mM) for 30 min in the presence of WT SET7/9 (3 pmol), Y305F mutant (6 pmol), or the Y245A mutant (100 pmol). Assays contained 100 mM HEPES, pH 8.0, 50 mM NaCl, 1.0 mM [methyl-3H]AdoMet (diluted to a specific activity of 0.2 Ci/nmol with purified AdoMet (37)), 15 M Sulfolobus solfa- taricus AdoHcy hydrolase (36), and 2 units of adenosine deami- nase (Roche Applied Science) in a final volume of 20 l. The reactions were terminated by addition of an equal volume of 200 mM MES, pH 5.1, prior to addition of a 2-fold molar excess of immobilized avidin resin (UltraLink; Pierce). Biotinylated peptides were allowed to bind at room temperature for 30 min, and the resin was then collected by centrifugation (9000  g). The resin was washed three times with 300 mM NaCl, and the peptide was eluted overnight from the avidin resin by cleavage of the disulfide bond in the linker of the peptide using 10 mM tris(2-carboxyethyl)phosphine (Thermo Scientific). The resin was incubated with additional 10 mM tris(2-carboxyethyl)phos- phine the following day until the radiolabel was essentially removed from the resin. The recovered peptides were hydro- lyzed by incubation with 6 M HCl at 110 °C for 24 h. Subsequent steps in measuring the radiolabel incorporated into the mono-, di-, and trimethyl-lysine products were performed as reported previously (8). RESULTS Functional Analysis of the SET7/9 Y305F Mutant—Prior studies of SET7/9 by Zhang et al. (6) reported that mutation of the Phe/Tyr switch residue Tyr-305 to a phenylalanine alters its product specificity from a mono- to dimethyltransferase. We verified these findings by demon- strating that WT SET7/9 mono- methylated the TAF10-K189 pep- tide, whereas the Y305F mutant mono- and dimethylated this sub- strate, as demonstrated by mass spectrometry (data not shown). We next examined whether the Y305F substitution altered the affinity of SET7/9 for the TAF10-K189 peptides using ITC (Fig. 1). A comparison of the KD values revealed that SET7/9 Y305F bound the TAF10-K189 and TAF10-K189me1 peptides 4- and 6-fold more tightly, respec- tively, than the WT enzyme, whereas this mutant displayed a substantially diminished affinity for the TAF10-K189me2 pep- tide (Table 1). Although the WT enzyme and the Y305F mutant exhibited discernable differences in their affinities for the unmodified and monomethylated peptides, these variations are modest and cannot account for their distinct product specific- ities, suggesting that a kinetic effect during methylation may be responsible. To investigate this possibility, we characterized the kinetic parameters of WT SET7/9 and the Y305F mutant using the TAF10 peptides as substrates. Both enzymes methylated the unmodified peptide with comparable kcat and Km values (Table 2). In analyzing the kinetic parameters for the methylation of the monomethylated peptide by SET7/9 Y305F, we found that this substrate displayed an elevated Km value that was beyond the measurable range of the assay due to its limited solubility. In this case, we measured the catalytic efficiency (kcat/Km) for the methylation of this peptide and found that it was methylated 15-fold less efficiently than the unmodified peptide by SET7/9 Y305F. Given the fact that the Y305F mutant exhibited a higher binding affinity for the TAF10-K189me1 peptide than the WT enzyme (Table 1), the kinetic data suggest that a step in the reaction pathway following substrate binding limits the catalytic efficiency of this mutant. We next examined whether the Y305F mutant dimethylated the TAF10-K189 peptide via a processive or a distributive mechanism. In a processive mechanism, the methyl-lysine substrate would remain bound to the enzyme during successive methyl transfer reactions; thus, the concentration of an inter- mediate, such as monomethyl-lysine, cannot exceed the en- zyme concentration during the assay. In a distributive mecha- nism, the intermediates are released into solution where they accumulate prior to the next round of methylation, resulting in an intermediate concentration that is greater than that of the enzyme. Using a radiometric TLC assay and a biotinylated TAF10 peptide, we quantified the amounts of monomethylated products generated by the WT SET7/9 and the Y305F mutant FIGURE 1. ITC analysis of WT SET7/9 and the Y245A and Y305F mutants. Representative titrations and binding isotherms for the calorimetry experiments are shown for the unmodified TAF10-K189 peptide titrated into WT SET7/9 (left plot), SET7/9 Y305F (middle plot), and SET7/9 Y245A (right plot). In each panel, the ITC titration experiment (upper panel) is shown with the binding isotherms (lower panel) fit to a single binding site model. TABLE 1 Analysis of the binding affinity of WT SET7/9 and its catalytic mutants for unmodified and methylated TAF10 peptides SET7/9 TAF10 peptide KD a M WT K189 4.9  0.20 WT K189me1 4.0  0.36 Y305F K189 1.3  0.10 Y305F K189me1 0.62  0.065 Y305Fb K189me2 70 Y245A K189 4.0  0.25 Y245A K189me1 3.3  0.10 Y245A K189me2 5.8  0.22 Y245A K189me3 11  0.28 a Curve fitting errors were calculated from the binding isotherms. b An estimate of the affinity is reported due to weak peptide binding. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31851 (Table 3). The data demonstrate that comparable amounts of monomethyl-lysine were generated when the quantity of enzyme usedistakenintoaccount,inagreementwiththeirsimilar turnover numbers for the TAF10-K189 peptide (Table 2). The Y305F mutant also produced small but measurable quantities of radiolabeled dimethyl-lysine product that were substantially smaller than the amount of monomethyl-lysine generated. Therefore, the TLC data are consistent with a distributive mechanism for dimethylation by the Y305F mutant because the amount of monomethyl-lysine produced exceeded the quantity of enzyme used in the assay. Structures of WT SET7/9 and the Y305F Mutant in Complex with Unmodified and Methylated TAF10 Peptides—To deter- mine the mechanism by which the Y305F substitution alters the product specificity of SET7/9, we determined the crystal struc- tures of this mutant bound to AdoHcy and TAF10-K189, TAF10-K189me1, and TAF10-K189me2 peptides and com- pared these to the structures of the WT SET7/9AdoHcy TAF10-K189 complex (supplemental Table 1). The structures of these complexes were determined to 1.85 Å or higher reso- lution, permitting unambiguous modeling of the K189 side chains in the active site of the enzyme based on simulated annealing omit maps (Fig. 2). The ternary complexes of the WT and the Y305F mutant superimpose with overall root mean square differences of less than 0.3 Å for all aligned atoms, indi- cating that neither the Y305F mutation nor the binding of the various TAF10-K189 peptides results in substantial changes in its overall structure. An inspection of the active sites of the SET7/9 WT and Y305F complexes illustrates the binding modes of the unmod- ified and methylated forms of K189 in the TAF10 peptides (Fig. 2, A–D). The K189 side chain binds in an extended all trans conformation in a deep pocket, termed the lysine binding chan- nel, that is composed of Tyr-245, Thr-266, Leu-267, Ser-268, Tyr-305, Tyr-335, and Tyr-337 (not shown for clarity) (Fig. 2A). These residues interact with the aliphatic portion of the K189 side chain primarily through van der Waals contacts. The lysine binding channel connects to the AdoMet-binding site on the opposite face of the catalytic domain via an oxygen-lined methyl transfer pore (38). During catalysis, the methyl group of the cofactor is positioned within the methyl transfer pore for the SN2 reaction with the -amino group of the lysine or methyl-lysine substrate (see below). To lower the activation barrier for this reaction, the lysine -amine nucleophile is aligned for methyl transfer through a hydrogen bond network within the active site. In the WT enzyme, the K189 -amino group hydrogen bonds to the hydroxyl group of Tyr-245 as well as to two water molecules (Fig. 2A). One of the water molecules (termed water 1), is coor- dinated in a solvent pocket, through hydrogen bonds to the carbonyl oxygens of Gly-292 and Ala-295 and to the hydroxyl group of the Phe/Tyr switch residue Tyr-305. This solvent pocket is structurally conserved in SET domain KMTs and has an important role in defining product specificity through the adjacent Phe/Tyr switch residue, as shown in our prior studies of the human H4K20 methyltransferase SET8 (8). The other water molecule is bound within the methyl transfer pore between the lysine substrate and the thioether sulfur atom of AdoHcy through hydrogen bonds to Tyr-245, Asn-265, and His-293 in SET7/9 and the TAF10-K189 -amino group. This water is not observed in other structures of SET7/9 ternary complexes and may represent the approximate position that the AdoMet methyl group occupies in the methyl transfer pore in the Michaelis complex. In structures of the Y305F ternary complexes, the K189, K189me1, and K189me2 side chains also adopt extended trans side chain geometries within the lysine binding channel that are stabilized via hydrogen bonding to Tyr-245 in the enzyme (Fig. 2, B–D). The orientations of the K189me1 and K189me2 side chains are further maintained through carbon-oxygen (CH–O) hydrogen bonding between the methyl groups and oxygen atoms within the vicinity of the methyl transfer pore, as reported previously in other SET domain KMT structures (8, 10, 38). A superimposition of the SET7/9 WT and Y305F com- plexes underscores the similarity of the lysyl binding conforma- tions (Fig. 2E). However, there are notable differences in the hydrogen bond patterns and occupancy of water 1 within the solvent pocket in the Y305F mutant compared with the WT enzyme. Specifically, the Y305F substitution results in the loss of one hydrogen bond to water 1 in the structures of the TAF10- K189 and TAF10-K189me1 complexes (Fig. 2, B and C). In con- trast, water 1 is absent in TAF10-K189me2 complex, and the vacated solvent pocket is occupied by one of the methyl groups TABLE 2 Kinetic parameters of WT SET7/9 and the Y305F and Y245A mutants Enzyme TAF10 peptide substrate Km a kcat a kcat/Km a M min1 M1 min1  103 WT K189 160  17 17  0.62 110  17 Y305F K189 88  5.0 17  0.30 190  11 Y305Fb K189me1 11  0.50 Y245A K189 200  35 0.53  0.04 2.6  0.47 Y245A K189me1 210  23 5.9  0.23 28  3.3 Y245A K189me2 400  29 6.5  0.16 15  1.2 a Curve fitting errors were calculated from the hyperbolic fits for the Michaelis-Menten equation. b Km and kcat were not determined because the Km value was beyond the measurable range; therefore, the kcat/Km value is reported. TABLE 3 Product analysis of WT SET7/9 and the Y305F and Y245A mutants Enzyme Quantity of enzyme Measured product Amount of product formeda nmol nmol WT 0.003 Kme1 0.65  0.07 Y305F 0.006 Kme1 1.5  0.49 Kme2 0.033  0.009 Y245A 0.100 Kme1 0.80  0.22 Kme2 0.39  0.021 Kme3 0.076  0.019 a Standard deviation was calculated from triplicate measurements. Lysine Methylation by SET7/9 Mutants 31852 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 of the dimethyl -amine (Fig. 2D). This methyl group forms a 3.5-Å CH–O hydrogen bond to the carbonyl oxygen of Ala-295, further stabilizing the binding of the dimethyl-lysine side chain. A homologous dimethyl-lysine-binding mode and CH–O hydrogen bond was observed in our prior structural studies of the SET8 Y334F Phe/Tyr switch mutant that confers an analo- gous change in product specificity from a mono- to a dimeth- yltransferase (8). A structural alignment of the active sites of the SET7/9 Y305F and SET8 Y334F mutants bound to cognate dimethylated peptides illustrates that the coordinates of the dimethyl-lysyl side chains are virtually superimposable, with one methyl group oriented toward the methyltransfer pore and the second positioned within the vacant solvent pocket (Fig. 2F). Taken together, the structures of the SET7/9 Y305F com- plexes and the similarities in the dimethyl-lysine conforma- tions in the SET7/9 Y305F and SET8 Y334F mutants imply that the Phe/Tyr switch governs product specificity through a con- served mechanism whereby it indirectly influences the binding modes of the methyl-lysine side chain by modulating the affin- ity of the water molecule (water 1) bound in the solvent pocket. Biochemical Characterization of the SET7/9 Y245A Mutant—Previ- ous studies by Xiao et al. (11) reported that the Y245A mutation yields an unusual change in the product specificity of SET7/9, converting the enzyme to a trimeth- yltransferase with weak monometh- yltransferase activity. We deter- mined that the SET7/9 Y245A could mono-, di-, and trimethylate the TAF10-K189 peptide by mass spec- trometry (data not shown) and TLC (Table 3), confirming the earlier studies of Xiao et al. (11). ITC analysis revealed that the Y245A mutant displayed comparable KD values for the unmodified and methylated TAF10-K189 peptides (Fig. 1), although its affinity for the trimethylated peptide was modestly diminished in comparison with the other peptides (Table 1). The ITC data demonstrate that the Y245A mutant bound the unmodified, mono-, and dimethylated sub- strates with equivalent affinities, suggesting that a kinetic effect or a structural alteration in the active site may be responsible for its diminished activity toward un- modified substrates. To gain further insight into its peculiar product specificity, we characterized the kinetic properties of the SET7/9 Y245A mutant. Steady state analysis demonstrated that this mutant displayed similar Km values for the unmodified, mono- and dimethylated TAF10 peptides (Table 2). However, the turnover number for the TAF10-K189 peptide was diminished over 10-fold versus the methylated peptides and was reduced 30-fold versus the WT enzyme, in agreement with the weak monomethyltransferase activity reported by Xiao et al. (11). In addition, we investigated whether this mutant catalyzes lysine trimethylation via a pro- cessive or distributive mechanism as described for SET7/9 Y305F. The TLC data illustrate that the mono- and dimethyl- lysine intermediates accumulated at quantities greater than that of the enzyme used in the assay, indicating that SET7/9 Y245A obeys a distributive mechanism, analogous to the Y305F mutant (Table 3). Structures of SET7/9 Y245A Bound to Unmodified and Meth- ylated TAF10 Peptides—To elucidate the mechanism underly- ing its unusual product specificity, we determined the crystal structures of SET7/9 Y245A in complex with AdoHcy and unmodified, mono-, di-, and trimethylated TAF10 peptides (supplemental Table 1). These complexes superimpose with the structure of the WT SET7/9AdoHcyTAF10-K189 com- FIGURE 2. Crystal structures of WT SET7/9 and the Y305F mutant in complex with AdoHcy and unmodi- fied and methylated TAF10 peptides. A, active site of WT SET7/9 bound to AdoHcy and TAF10-K189 and the active site of SET7/9 Y305F bound to AdoHcy and TAF10-K189 (B), TAF10-K189me1 (C), and TAF10-K189me2 peptides (D). A–D, the K189 residue in the TAF10 peptide is depicted with yellow carbon atoms with the corresponding Fo – Fc simulated-annealing omit maps (contoured at 2.5 ) shown. AdoHcy is colored with green carbon atoms, and SET7/9 carbon atoms are colored gray, with the exception of Tyr-305/Y305F (magenta) and Tyr-245 (cyan). Gly-264 to Ser-268 are shown, and the main chain atoms of Gly-292 to Ala-295 are depicted for clarity. Red dashed lines indicate conventional hydrogen bonds, and blue dashed lines indicate CH–O hydrogen bonds. Cyan dashed lines indicate weak CH–O hydrogen bonds between 3.5 and 3.7 Å in length. The water molecule in the solvent binding pocket is numbered 1. E, overlay of the active sites of the WT enzyme (light blue carbon atoms) and the Y305F mutant (green, yellow, and pink carbon atoms correspond to K189, K189me1, and K189me2, respectively). Red dashed lines indicate coordination of the waters in the struc- turesoftheY305Fmutant,andthelightbluedashedlineindicateshydrogenbondingthatonlyoccursintheWT enzyme. F, superimposition of the active sites of the SET7/9 Y305FAdoHcyTAF10-K189me2 and SET8 Y334FAdoHcyH4K20me2 (Protein Data Bank code 3F9X) complexes shown with green and yellow carbon atoms, respectively. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31853 plex with root mean squared differences of less than 0.4 Å for all aligned atoms, indicating that the Y245A mutant does not per- turb the overall structure of the enzyme. Simulated annealing omit maps illustrate that K189 side chains are bound within the lysine binding channel through hydrogen bonds and van der Waals contacts (Fig. 3, A–D), although the interactions and binding modes are distinct from those in the complexes of WT SET7/9 and the Y305F mutant (Fig. 2, A–D). In the unmodified TAF10 peptide complex, the K189 -amino group forms a weak hydrogen bond with the carbonyl oxygen of Gly-264 (Fig. 3A), whereas the -amino groups of K189me1 and K189me2 hydro- gen bond to the hydroxyl group of Tyr-305 in the mono- and dimethylated peptide complexes (Fig. 3, B and C). The confor- mations of the K189me1 and K189me2 side chains are further stabilized by water-mediated hydrogen bonding and through CH–O hydrogen bonding to their methyl groups. In the TAF10-K189me3 peptide complex, the trimethyl-lysine side chain is coordinated exclusively through direct and water-me- diated CH–O hydrogen bonds to its methyl groups because the quaternary -ammonium cation cannot engage in hydrogen bonding (Fig. 3D). A structural alignment of the four SET7/9 Y245A complexes illustrates distinct binding modes for the unmodified versus the methylated K189 side chains, highlighting the selectivity of this mutant for methylated substrates. The side chains of K189me1, K189me2, and K189me3 roughly overlay with their respective -amino groups superimposed and adopt slightly kinked con- formations (Fig. 3E), as opposed to the extended trans geome- try of the unmodified and methylated lysines in the complexes of the WT enzyme and the Y305F mutant (Fig. 2, A–D). Con- versely, the unmodified K189 side chain does not superimpose with its methylated counterparts and is oriented in an alterna- tive configuration due to its hydrogen bonding to Gly-264 (Fig. 3, A and E). An overlay of the structures of the WT enzyme and Y245A mutant bound to the unmodified TAF10 peptide illus- trates that the side chains of K189 do not superimpose and that the K189 -amino group appears to be misaligned with AdoHcy in the Y245A complex (Fig. 3F). This suboptimal alignment may explain the diminished kcat value of SET7/9 Y245A mutant toward substrates with unmodified lysines (Table 2). A comparison of the structures of the SET7/9 Y245A and Y305F complexes yields a molecular explanation for the differ- ent product specificities of these two mutants. In the SET7/9 Y305F complexes, Tyr-245 aligns the K189 -amino group for methyl transfer through hydrogen bonding to its hydroxyl group (Fig. 2, B–D). Conversely, in the Y245A mutant, the K189me1 and K189me2 -amino groups are oriented through hydrogen bonding to Tyr-305 (Fig. 3, B and C). These distinct hydrogen bond patterns impart differences in the conforma- tions of the lysyl side chains due to the relative orientations of Tyr-245 and Tyr-305 in the lysine binding channel. Specifically, the kinked conformation adopted by the K189me1 and K189me2 side chains in the Y245A complexes (Fig. 3, B and C) may contribute to the differences in the turnover numbers of this mutant versus those of the WT enzyme and the Y305F mutant (Table 2). In addition, the dimethyl -amino group of the K189me2 side chain binds in distinct orientations in the Y245A and Y305F mutants due to their hydrogen bonding to FIGURE 3. Crystal structures of the SET7/9 Y245A mutant in complex with AdoHcy and unmodified and methylated TAF10 peptides. Active site of SET7/9 Y245A bound to AdoHcy and TAF10-K189 (A), TAF10-K189me1 (B), TAF10-K189me2 (C), and TAF10-K189me3 peptides (D). A–D, Fo – Fc simulated- annealing omit maps (contoured at 2.5 ) for the unmodified and methylated K189 side chains are illustrated. The residues and hydrogen bonds in each com- plexarecoloredasinFig.2.Thewatermoleculesinthelysinebindingchannelsof the Y245A complexes are numbered 1–4, as described in the text. E, superimpo- sition of the active sites of the Y245A complexes bound to the four methylated statesofTAF10-K189(unmodified,mono-,di-,andtrimethylatedshowningreen, yellow, pink, and blue, respectively). F, overlay of WT SET7/9 (blue carbons) and SET7/9 Y245A (green carbons) in complex with TAF10-K189. Waters correspond- ing to the WT and Y245A structures are colored cyan and green, respectively. G, alignment of the active sites of the Y245A (yellow carbons) and Y305F (green carbons) mutants in complex with the TAF10-K189me2 peptide. Hydrogen bonds from the Y305F structure are shown as green dashed lines, and waters and hydrogen bonds in the Y245A structure are shown in yellow and orange, respectively. Lysine Methylation by SET7/9 Mutants 31854 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 Tyr-305 and Tyr-245, respectively (Fig. 3G). In the Y305F mutant, hydrogen bonds to the dimethyl -amino group cou- pled with steric constraints in the lysine binding channel pre- vent the K189me2 side chain from undergoing a conforma- tional change that is conducive to trimethylation (Fig. 2D), consistent with its dimethyltransferase activity. However, in the Y245A mutant, the alanine substitution enlarges the diameter of the lysine binding channel, accommodating trimethyl-lysine (Fig. 3D). In addition, the larger diameter would permit the dimethyl-lysine substrate to undergo the conformational reor- ganization necessary to align the -amino group in a productive geometry for trimethylation. A major difference in the active site of the Y245A mutant versus the other SET7/9 structures is the presence of several water molecules bound in the cavity generated by the Y245A mutation. In the structure of the Y245A mutant bound to TAF10-K189, three water molecules (waters 2–4) occupy this cavity and are arranged in a triangular geometry (Fig. 3A). In addition, water 1 shifts 1.6 Å from its position in the solvent pocket toward water 2 to which it forms a hydrogen bond (Fig. 3, A and E). The shift in water 1 was unexpected given its con- served orientation in the solvent pocket of the SET7/9 WT and Y305F complexes (Fig. 2, A–C) as well as in the structures of other SET domain KMTs (8). This displacement is presumably related to the alternative conformation of the K189 side chain whose -amino group is too distant (4.3 Å) to form a productive hydrogen bond to water 1. Conversely, in the Y245A complexes bound to TAF10-K189me1 and TAF10-K189me2, water 1 remains tightly bound in the solvent pocket through hydrogen bonds to Tyr-305 hydroxyl group, the carbonyl oxygens of Gly- 292 and Ala-295, and the K189 -amino group (Fig. 3, B, C, and E), analogous to its binding in the WT enzyme (Fig. 2A). How- ever, in the TAF10-K189me3 complex, one of the methyl groups of the trimethyl -ammonium cation is oriented into the solvent pocket (Fig. 3D), similar to the dimethyl-lysine binding mode observed in the Y305F mutant (Fig. 2D). The binding of the methyl group in the solvent pocket displaces water 1 by 3.2 Å relative to its position in the TAF10-K189me1 complex (Fig. 3E), thereby avoiding a steric clash with the trimethylated -ammonium group. Variations in the occupancy of water 2 are also seen in the different Y245A structures. Water 2 is bound in similar orientations in the active site of the unmodified and monomethylated peptide complexes but is absent in the di- and trimethylated peptide complexes due to the binding of a methyl group in this position (Fig. 3, A–E). In summary, the changes in the positions or occupancies of waters 1 and 2 correlate with the binding modes of the unmodified and methylated K189 within the active site of the Y245A mutant. Catalytic Models of Lysine Multiple Methylation by SET7/9 Y245A, and Y305F—The structures of the SET7/9 complexes reported here offer a prime opportunity to generate stepwise models for lysine mono-, di-, and trimethylation by a SET domain KMT. We modeled the AdoMet-bound Michaelis complexes by superimposing the SET7/9 product complexes with the previously reported structure of the SET7/9-AdoMet binary complex (Fig. 4) (39). The conformations of the mono- and dimethyl -amino groups in the Michaelis complexes were inferred from the coordinates of the corresponding dimethyl- and trimethyl-lysine products, respectively. In addition, we modeled the -amino group in a deprotonated state with its hydrogen atoms oriented toward the hydrogen bond acceptors that align the lysyl side chain for methylation. As a basis for this comparison, we first modeled the monomethylation reaction catalyzed by WT SET7/9 (Fig. 4A). In the substrate ternary complex, the lysine -amine is aligned with the methyl group and sulfonium cation of AdoMet through a hydrogen bond to the Tyr-245 hydroxyl group and water 1 in the solvent pocket. The values of the reaction distance and angle are 2.8 Å and 153°, respectively, in approximate agreement with the linear geome- try of a SN2 methyl transfer reaction calculated in other mod- eled substrate complexes (8, 10). In the product complex, the monomethyl-lysine side chain is bound in an extended confor- mation with its methyl group oriented within the methyl trans- fer pore, thereby obstructing AdoMet binding. Furthermore, water 1 remains tightly coordinated in the solvent pocket through four hydrogen bonds to Gly-292, Ala-295, Tyr-305, and the monomethyl -amino group. These interactions hinder the dissociation of water 1 and the related rearrangement of the monomethyl-lysine side chain required for a second methyl transfer reaction, explaining why the WT enzyme cannot cata- lyze di- and trimethylation. These findings concur with the FIGURE 4. Catalytic models of the methyl transfer reactions catalyzed by WT SET7/9 and the Y305F mutant. A, monomethylation of TAF10-K189 by the WT enzyme. The reaction scheme depicts the modeled substrate ternary complex (left) and the product complex (right) for the transfer of the methyl group from AdoMet (green carbon atoms) to K189 in TAF10 (yellow carbons), yielding AdoHcy and K189me1. The red arrow indicates the direction of the nucleophilic attack of the deprotonated -amino group on the AdoMet methyl group. The transferred methyl group is colored green, and the white atoms represent the hydrogens of the -amino group. Hydrogen bonds and residues in the enzyme active site are illustrated as in Fig. 2. The reaction distance and angle are labeled in red. B and C, models of the Y305F mutant for the first methyl transfer reaction with TAF10-K189 (B) and second methyl transfer reaction with TAF10-K189me1 (C). Color schemes are the same as in A. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31855 model for SET7/9 product specificity reported in previous structural and functional studies (6, 11). Similar reaction geometry is observed in the model for the monomethyl transfer reaction catalyzed by SET7/9 Y305F. Hydrogen bonds from the Tyr-245 hydroxyl group and water 1 align the lysine -amino group with the AdoMet methyl group at a distance of 2.1 Å and an angle of 160°, equivalent to those measured in the Michaelis complex of the WT enzyme (Fig. 4, A and B). In the product complex, monomethyl-lysine adopts an extended trans configuration analogous to that in the WT enzyme. For dimethylation to occur, the monomethyl-lysine must undergo a conformational change in which its methyl group is rotated out of the methyl transfer path with AdoMet. The structure of the Y305F mutant bound to the dimethylated TAF10 peptide (Fig. 2D) implies that this rearrangement occurs through the dissociation of water 1 due to the loss of the Tyr- 305 hydrogen bond in the solvent pocket. The dissociation of water 1 would enable the monomethyl-lysine side chain to adopt an alternative conformation through a rotation about its C–N bond, projecting the methyl group into the solvent pocket (Fig. 4C). This rotation reorients the methyl group out of the methyl transfer path while realigning the monomethyl - amino group for a second methylation reaction through a direct hydrogen bond to the Tyr-245 hydroxyl group and a CH–O hydrogen bond between its methyl group and Ala-295. The modeled reaction geometry for monomethyl-lysine substrate complex of the Y305F mutant (2.7 Å, 160°) is equivalent to that of the first methyl transfer reaction in SET7/9 Y305F. These geometries concur with our previous models for mono- and dimethylation catalyzed by SET8 Y334F (8), illustrating that the orientation of a methyl group into the solvent pocket is a con- served feature of SET domain KMTs that catalyze multiple methylation. In addition, we modeled the methyl transfer reactions cata- lyzed by SET7/9 Y245A (Fig. 5). In the model of the lysine sub- strate complex, the -amino group is aligned for methyl transfer by a hydrogen bond to the carbonyl oxygen of Gly-264, result- ing in a short reaction distance (2.3 Å) and a suboptimal reac- tion angle (141°) with the methyl group of AdoMet (Fig. 5A). This misalignment appears to be a direct consequence of the Y245A mutation that abolishes hydrogen bonding to the - amino group, illustrating that the suboptimal orientation of the -amine likely contributes to the diminished activity of this mutant toward unmodified substrates (11). Conversely, in the modeled monomethyl-lysine substrate complex for SET7/9 Y245A, Tyr-305 hydrogen bonds to the -amino group, aligning it for methyl transfer (Fig. 5B). In addition, CH–O hydrogen bonds to the monomethyl-lysine methyl group and the dissoci- ation of water 2 from the active site also contribute to reposition- ing the -amino group for dimethylation. Collectively, these inter- actions orient the -amine in a reaction angle of 165° that is more conducive to methyl transfer. However, the reaction distance for dimethylation is 0.6 Å longer than that in the corresponding Y305F model because Tyr-305 is positioned further from AdoMet than Tyr-245 (Figs. 4C and 5B). In the third methyl transfer reaction catalyzed by SET7/9 Y245A, the lone pair of electrons of the dimethyl-lysine - amino group acts as the nucleophile and thus cannot engage in hydrogen bonding. The structure of the trimethyl-lysine prod- uct complex (Fig. 3D) implies that the dimethyl -amine is aligned via CH–O hydrogen bonds to its methyl groups, as shown in the model of the Michaelis complex for this reaction (Fig. 5C). These CH–O hydrogen bonds restrain the orienta- tion of the -amino group and position one of the methyl groups into the solvent pocket, displacing water 1 as discussed earlier (Fig. 3, D and E). These interactions cumulatively align the -amino group and AdoMet methyl group with a reaction distance of 3.0 Å and angle of 162° (Fig. 5C). Taken together, the models of the substrate complexes for SET7/9 Y245A suggest that CH–O hydrogen bonds play an increasingly important role in aligning the methylated -amino group in successive rounds of methyl transfer. DISCUSSION The structural and functional characterization of the SET7/9 Y245A and Y305F mutants presented here yields new insights into the mechanism underlying the product specificity of SET domain KMTs. Importantly, it resolves a general paradox concerning this specificity. How does the active site constrain the motion of the lysine -amino group to align it for methyl transfer with AdoMet, while providing adequate volume to accommodate the mono-, di-, and tri- methylated lysine side chain generated during multiple methyl transfer reactions? The structures of the Y305F and Y245A mutants resolve this paradox, illustrating that alter- ations in the positions or occupancies of water molecules within their active sites generate the space required to FIGURE 5. Models for the methyl transfer reactions catalyzed by the SET7/9 Y245A mutant. Models of the Y245A mutant for the first methyl transfer reaction with TAF10-K189 (A), the second methyl transfer reaction with TAF10-K189me1 (B), and the third methyl transfer reaction with TAF10- K189me2 (C). Residues and hydrogen bonds are depicted as in Fig. 3. Lysine Methylation by SET7/9 Mutants 31856 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 accommodate the multiply methylated -amine produced during successive catalytic cycles. Minor perturbations in the side chains of certain active site residues, such as Tyr- 305, are also observed in alignments of the WT enzyme and the Y245A and Y305F complexes, although these changes are modest compared with the displacement or dissociation of the water molecules in the active site. These findings suggest that the waters function as transient place holders that facil- itate the SN2 methyl transfer reaction. During monomethy- lation, they function to constrain the movement of the lysine -amino group by mediating hydrogen bonds between the substrate and enzyme, thereby promoting the linear align- ment with the methyl group and sulfonium cation of AdoMet (Fig. 4, A and B). During di- and trimethylation, the water molecules either relocate within the lysine binding channel or dissociate from the enzyme, yielding the space required to rotate the methyl group away from the methyl transfer pore and to realign the -amine in productive geom- etry for the next methyl transfer reaction (Figs. 4C and 5, B and C). These findings agree with our prior analysis of the SET8 Phe/Tyr switch mutant in which we demonstrated that the Y334F substitution attenuates hydrogen bonding to the water molecule bound in the solvent pocket, promoting its dissociation and the conformational changes necessary for lysine dimethylation (8). Indeed, there is a nearly identical alignment of the dimethyl-lysine side chains in the structures of SET7/9Y305FandSET8Y334Fcomplexes,despitethedifferences in the orientations of the Phe-305 and Phe-334 side chains in each structure (Fig. 2F). Finally, both the SET8 Y334F (8) and SET7/9 Y305F mutants (Table 2) displayed diminished catalytic efficien- cies for lysine dimethylation versus monomethylation. These dif- ferences may reflect the kinetics of the reorganization within the active site, including the dissociation of the water molecule from the solvent pocket and the concomitant realignment of the monomethyl-lysine into a productive geometry for dimethylation. In addition to their place-holding role, the active site waters may also facilitate the deprotonation of the lysine -amino group between methyl transfer reactions. For methylation to occur, the -amino group must be deprotonated to function as the nucleophile in the SN2 methyl transfer reaction with AdoMet (Figs. 4 and 5). Although the pKa value of the lysine -amine in solution is 10.5, molecular dynamics simulations by Zhang and Bruice (25, 26) indicate that this value diminishes to 8.2 upon formation of the SET7/9 Michaelis complex due to the proximity of the AdoMet sulfonium cation and the low dielectric constant of the active site. Furthermore, their simu- lations show that a chain of water molecules facilitates the dep- rotonation of the -amino group prior to methyl transfer, trans- ferring the proton to bulk solvent. Although these water molecule chains are not evident in our crystal structures, the Y305F and Y245A complexes suggest another potential mech- anism for deprotonation. In the dimethyl-lysine complexes of the Y305F and Y245A mutants, the dissociation of water 1 and 2, respectively, from the lysine binding channel requires that the solvent-mediated hydrogen bond to the -amino group is broken (Figs. 2D and 3C). It is conceivable that these waters dissociate from the active site as hydronium ions, promoting the realignment and deprotonation of the methyl -amino group for the next methyl transfer reaction. A comparison of the SET7/9 Y305F and SET8 Y334F com- plexes yields insights into the mechanism by which the Phe/Tyr switch influences water binding within the solvent pocket. The phenylalanine substitution in the Phe/Tyr switch results in the loss of a single hydrogen bond to the water molecule (water 1) in the solvent pocket compared with the four hydrogen bonds that coordinate the solvent molecule in WT SET7/9 (Fig. 2, A and B) and SET8 (7, 8). Although this attenuation in hydrogen bonding may appear insignificant, this difference is nonetheless impor- tant for at least two reasons. First, theoretical calculations indi- cate that, on average, water molecules form 3.5 hydrogen bonds in solutions (40, 41). This value is greater than the num- ber of hydrogen bonds coordinating water 1 in the solvent pocket in SET7/9 Y305F (Fig. 2, B and C) as well as in SET8 Y334F and other di- and trimethyltransferases that possess a hydrophobic residue in the Phe/Tyr switch site (8). From the perspective of the water molecule, the greater hydrogen bond- ing potential in solution would tend to thermodynamically favor its dissociation from the solvent pocket in SET domain KMTs that lack a tyrosine in the Phe/Tyr switch position. Sec- ond, the ordered binding of water molecules observed in the active sites of SET domain ternary complexes represents an unfavorable entropy compared with their diffusion in bulk sol- vent. In WT SET7/9 (Fig. 2A) and SET8 (7, 8), this entropic penalty can be partially offset through the favorable enthalpy of binding associated with the four hydrogen bonds that coordi- nate the water within the solvent pocket. It is conceivable that the loss of the tyrosine-mediated hydrogen bond in the Phe/Tyr switch shifts the equilibrium in favor of dissociation of the water molecule from the solvent pocket, thereby facilitating dimethylation in SET7/9 Y305F, SET8 Y334F, and other di- and trimethyltransferases. The structures of the SET7/9 Y245A and Y305F complexes illustrate the interactions that align the lysine -amino group during the methyl transfer reactions in each enzyme. In the WT enzyme and the Y305F mutant, hydrogen bonding to the hydroxyl group of Tyr-245 appears to be critical in properly aligning the -amine for methyl transfer (Fig. 4). Tyr-245 is conserved in the sequences of many SET domain KMTs (8, 42), and substitutions of this residue generally impair or abolish activity, indicating its importance in catalysis (8, 43). However, SET7/9 appears to be an exception to this rule, as the Y245A mutant is not only active but is capable of catalyzing lysine trimethylation. In this mutant, Tyr-305 appears to assume the role of Tyr-245 by hydrogen bonding to the monomethylated -amino group to align it for methyl transfer with AdoMet, as illustrated in the modeled substrate complex for the dimethy- lation reaction (Fig. 5B). Conversely, in the model for trimethy- lation, the Tyr-305 hydroxyl group does not hydrogen bond to the -amine but instead participates in a CH–O hydrogen bond with one of the methyl groups to assist in aligning the dimethy- lated -amine for the methyl transfer reaction (Fig. 5C). Addi- tional structural and functional studies of the SET domain tri- methyltransferases will aid in further illuminating the roles of CH–O hydrogen bonds in facilitating lysine multiple methylation. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31857 Acknowledgments—We acknowledge S. Schiebold for assistance in protein expression, purification, and crystallization and S. Anderson and R. Sanishvili for their assistance with x-ray data collection. We also thank S. Bulfer and S. Horowitz for reading the manuscript and providing useful comments. This work utilized the Protein Structure Facility of the Michigan Diabetes Research and Training Center, Uni- versity of Michigan, supported by National Institutes of Health Grant DK020572, NIDDK. Use of the Advanced Photon Source was sup- ported by the United States Department of Energy, Basic Energy Sci- ences, Office of Science, under Contract DE-AC02-06CH11357. GM/CA CAT has been funded in whole or in part by National Institutes of Health NCI Grant Y1-CO-1020 and NIGMS Grant Y1-GM-1104. Use of the LS-CAT Sector 21 was supported by Michi- gan Economic Development Corporation and the Michigan Technol- ogy Tri-Corridor Grant 085P1000817 for the support of this research program. REFERENCES 1. Huang, J., and Berger, S. L. (2008) Curr. Opin. Genet. Dev. 18, 152–158 2. Morgunkova, A., and Barlev, N. A. (2006) Cell Cycle 5, 1308–1312 3. Yang, X. D., Lamb, A., and Chen, L. F. (2009) Epigenetics 4, 429–433 4. Taverna, S. D., Li, H., Ruthenburg, A. J., Allis, C. D., and Patel, D. J. (2007) Nat. Struct. Mol. Biol. 14, 1025–1040 5. Collins, R. E., Tachibana, M., Tamaru, H., Smith, K. M., Jia, D., Zhang, X., Selker, E. U., Shinkai, Y., and Cheng, X. (2005) J. Biol. Chem. 280, 5563–5570 6. Zhang, X., Yang, Z., Khan, S. I., Horton, J. R., Tamaru, H., Selker, E. U., and Cheng, X. (2003) Mol. Cell 12, 177–185 7. Couture, J. F., Collazo, E., Brunzelle, J. S., and Trievel, R. C. (2005) Genes Dev. 19, 1455–1465 8. Couture, J. F., Dirk, L. M., Brunzelle, J. S., Houtz, R. L., and Trievel, R. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 20659–20664 9. Qian, C., Wang, X., Manzur, K., Sachchidanand, Farooq, A., Zeng, L., Wang, R., and Zhou, M. M. (2006) J. Mol. Biol. 359, 86–96 10. Trievel, R. C., Flynn, E. M., Houtz, R. L., and Hurley, J. H. (2003) Nat. Struct. Biol. 10, 545–552 11. Xiao, B., Jing, C., Wilson, J. R., Walker, P. A., Vasisht, N., Kelly, G., Howell, S., Taylor, I. A., Blackburn, G. M., and Gamblin, S. J. (2003) Nature 421, 652–656 12. Chuikov, S., Kurash, J. K., Wilson, J. R., Xiao, B., Justin, N., Ivanov, G. S., McKinney, K., Tempst, P., Prives, C., Gamblin, S. J., Barlev, N. A., and Reinberg, D. (2004) Nature 432, 353–360 13. Ea, C. K., and Baltimore, D. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 18972–18977 14. Este`ve, P. O., Chin, H. G., Benner, J., Feehery, G. R., Samaranayake, M., Horwitz, G. A., Jacobsen, S. E., and Pradhan, S. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 5076–5081 15. Kouskouti, A., Scheer, E., Staub, A., Tora, L., and Talianidis, I. (2004) Mol. Cell 14, 175–182 16. Masatsugu, T., and Yamamoto, K. (2009) Biochem. Biophys. Res. Commun. 381, 22–26 17. Munro, S., Khaire, N., Inche, A., Carr, S., and La Thangue, N. B. (2010) Oncogene 29, 2357–2367 18. Pagans, S., Kauder, S. E., Kaehlcke, K., Sakane, N., Schroeder, S., Dorm- eyer, W., Trievel, R. C., Verdin, E., Schnolzer, M., and Ott, M. (2010) Cell Host Microbe 7, 234–244 19. Subramanian, K., Jia, D., Kapoor-Vazirani, P., Powell, D. R., Collins, R. E., Sharma, D., Peng, J., Cheng, X., and Vertino, P. M. (2008) Mol. Cell 30, 336–347 20. Wang, J., Hevi, S., Kurash, J. K., Lei, H., Gay, F., Bajko, J., Su, H., Sun, W., Chang, H., Xu, G., Gaudet, F., Li, E., and Chen, T. (2009) Nat. Genet. 41, 125–129 21. Yang, X. D., Huang, B., Li, M., Lamb, A., Kelleher, N. L., and Chen, L. F. (2009) EMBO J. 28, 1055–1066 22. Guo, H. B., and Guo, H. (2007) Proc. Natl. Acad. Sci. U.S.A. 104, 8797–8802 23. Hu, P., Wang, S., and Zhang, Y. (2008) J. Am. Chem. Soc. 130, 3806–3813 24. Hu, P., and Zhang, Y. (2006) J. Am. Chem. Soc. 128, 1272–1278 25. Zhang, X., and Bruice, T. C. (2007) Biochemistry 46, 14838–14844 26. Zhang, X., and Bruice, T. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 5728–5732 27. Couture, J. F., Collazo, E., Hauk, G., and Trievel, R. C. (2006) Nat. Struct. Mol. Biol. 13, 140–146 28. Trievel, R. C., Beach, B. M., Dirk, L. M., Houtz, R. L., and Hurley, J. H. (2002) Cell 111, 91–103 29. Kapust, R. B., To¨zse´r, J., Fox, J. D., Anderson, D. E., Cherry, S., Copeland, T. D., and Waugh, D. S. (2001) Protein Eng. 14, 993–1000 30. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307–326 31. Vagin, A., and Teplyakov, A. (2000) Acta Crystallogr. D Biol. Crystallogr. 56, 1622–1624 32. Emsley, P., and Cowtan, K. (2004) Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 33. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Acta Crystallogr. D Biol. Crystallogr. 53, 240–255 34. Davis, I. W., Leaver-Fay, A., Chen, V. B., Block, J. N., Kapral, G. J., Wang, X., Murray, L. W., Arendall, W. B., 3rd, Snoeyink, J., Richardson, J. S., and Richardson, D. C. (2007) Nucleic Acids Res. 35, W375–W383 35. Bru¨nger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. D Biol. Crystallogr. 54, 905–921 36. Collazo, E., Couture, J. F., Bulfer, S., and Trievel, R. C. (2005) Anal. Bio- chem. 342, 86–92 37. Chirpich, T. P., Zappia, V., Costilow, R. N., and Barker, H. A. (1970) J. Biol. Chem. 245, 1778–1789 38. Couture, J. F., Hauk, G., Thompson, M. J., Blackburn, G. M., and Trievel, R. C. (2006) J. Biol. Chem. 281, 19280–19287 39. Kwon, T., Chang, J. H., Kwak, E., Lee, C. W., Joachimiak, A., Kim, Y. C., Lee, J., and Cho, Y. (2003) EMBO J. 22, 292–303 40. Chandra, A., and Chowdhuri, S. (2002) J. Phys. Chem. B 106, 6779–6783 41. Guardia, E., Marti, J., Garcia-Tarres, L., and Laria, D. (2005) J. Mol. Liq. 117, 63–67 42. Dillon, S. C., Zhang, X., Trievel, R. C., and Cheng, X. (2005) Genome Biol. 6, 227 43. Zhang, X., Tamaru, H., Khan, S. I., Horton, J. R., Keefe, L. J., Selker, E. U., and Cheng, X. (2002) Cell 111, 117–127 Lysine Methylation by SET7/9 Mutants 31858 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010
3M56
SET7/9 Y305F in complex with TAF10-K189me2 peptide and AdoHcy
SET7/9 Catalytic Mutants Reveal the Role of Active Site Water Molecules in Lysine Multiple Methylation*□ S Received for publication,February 15, 2010, and in revised form, July 14, 2010 Published, JBC Papers in Press,August 1, 2010, DOI 10.1074/jbc.M110.114587 Paul A. Del Rizzo‡, Jean-Franc¸ois Couture‡§1, Lynnette M. A. Dirk¶, Bethany S. Strunk‡, Marijo S. Roiko‡, Joseph S. Brunzelle, Robert L. Houtz¶, and Raymond C. Trievel‡2 From the ‡Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109, the §Ottawa Institute of Systems Biology, Department of Biochemistry, Microbiology, and Immunology, University of Ottawa, Ottawa, Ontario K1H 8M5, Canada, the ¶Department of Horticulture, Plant Physiology/Biochemistry/Molecular Biology Program, University of Kentucky, Lexington, Kentucky 40546, and the Department of Molecular Pharmacology and Biological Chemistry, Northwestern University Center for Synchrotron Research, Life Sciences Collaborative Access Team, Argonne, Illinois 60439 SET domain lysine methyltransferases (KMTs) methylate specific lysine residues in histone and non-histone substrates. These enzymes also display product specificity by catalyzing dis- tinct degrees of methylation of the lysine -amino group. To elucidate the molecular mechanism underlying this specificity, we have characterized the Y245A and Y305F mutants of the human KMT SET7/9 (also known as KMT7) that alter its prod- uct specificity from a monomethyltransferase to a di- and a tri- methyltransferase, respectively. Crystal structures of these mutants in complex with peptides bearing unmodified, mono-, di-, and trimethylated lysines illustrate the roles of active site water molecules in aligning the lysine -amino group for methyl transfer with S-adenosylmethionine. Displacement or dissocia- tion of these solvent molecules enlarges the diameter of the active site, accommodating the increasing size of the methylated -amino group during successive methyl transfer reactions. Together, these results furnish new insights into the roles of active site water molecules in modulating lysine multiple meth- ylation by SET domain KMTs and provide the first molecular snapshots of the mono-, di-, and trimethyl transfer reactions catalyzed by these enzymes. SET domain enzymes represent a family of S-adenosylmethi- onine (AdoMet)3-dependent methyltransferases that catalyze the site-specific methylation of protein lysyl residues in a host of proteins, including histones, transcription factors, chroma- tin-modifying enzymes, ribosomal subunits, and other sub- strates (1–3). In many instances, these modifications serve to recruit effector proteins that recognize methyl-lysyl residues in a sequence-dependent fashion (4). In addition, SET domain KMTs exhibit product specificity, defined as their ability to cat- alyze mono-, di-, or trimethylation of the lysine -amino group. This specificity is biologically relevant because many methyl- lysine-binding proteins can discriminate among different degrees of lysine methylation (4). Thus, both the site and degree of lysine methylation are critical to recognition by effector proteins. Structural and functional studies have identified a Phe/Tyr switch in the active site of SET domain KMTs that governs their respective product specificities (5, 6). According to this model, KMTs that possess a tyrosine in the Phe/Tyr switch site are limited to catalyzing lysine monomethylation, whereas en- zymes that possess a phenylalanine or another hydrophobic residue in this position display di- or trimethyltransferase activ- ity. Mutational analysis of various SET domain KMTs, includ- ing DIM-5 (KMT1) (6), SET7/9 (6), G9A (KMT1C) (5), and SET8 (also known as PR-SET7 and KMT5A) (7, 8), has demon- strated that substitutions in the Phe/Tyr switch result in pre- dictable changes in product specificity. Several models have been proposed to explain the mechanism by which the Phe/Tyr switch site governs this specificity, including variations in the diameter of the active site due to the size of Phe/Tyr switch residue and steric hindrance by the tyrosine hydroxyl group (6, 9–11). However, our recent studies of the Phe/Tyr switch mutant Y334F in the human histone H4 Lys-20 (H4K20) meth- yltransferase SET8 indicate that the Phe/Tyr switch regulates product specificity via a more subtle mechanism (8). Specifi- cally, the switch modulates the binding of an active site water molecule that in turn regulates the transition from mono- methylation to multiple methylation. Among the KMTs that have been structurally characterized, SET7/9 has emerged as an archetypal model for studying the catalytic mechanism and product specificity of the SET domain family. Although initially isolated as a histone H3 Lys-4 (H3K4)-specific methyltransferase, this KMT has been shown to regulate the functions of numerous non-histone substrates through site-specific methylation (12–21). Early structural and functional studies of SET7/9 identified two active site mutants, * Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth Grant R01 GM073839 (to R. C. T.) and National Institutes of Health Admin- istrative Supplement GM073839-04S1 (to R. C. T.) funded through the American Recovery and Reinvestment Act. This work was also supported by Department of Energy Grant DE-FG02-92ER20075 (to R. L. H.). □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Table 1. The atomic coordinates and structure factors (codes 3M53, 3M54, 3M55, 3M56, 3M57,3M58,3M59,and3M5A)havebeendepositedintheProteinDataBank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). 1 Supported by a Canadian Institutes of Health Research postdoctoral fellowship. 2 To whom correspondence should be addressed: Dept. of Biological Chem- istry, University of Michigan Medical School, 1150 West Medical Center Dr., 5301 MSRB III, Ann Arbor, MI 48109. Tel.: 734-647-0889; Fax: 734-763-4581; E-mail: rtrievel@umich.edu. 3 The abbreviations used are: AdoMet, S-adenosylmethionine; AdoHcy, S-ad- enosylhomocysteine; KMT, lysine methyltransferase; ITC, isothermal titra- tion calorimetry; CH–O, carbon-oxygen hydrogen bond; BisTris, 2-[bis(2- hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 41, pp. 31849–31858, October 8, 2010 © 2010 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31849 Y245A and Y305F, which change its product specificity. The Phe/Tyr switch mutant Y305F alters SET7/9 product specificity from a mono- to dimethyltransferase (6), whereas the Y245A substitution converts the enzyme into a trimethyltransferase with weak monomethyltransferase activity (11). These mutants have been the subjects of numerous molecular modeling simu- lations that have led to various models to explain their distinct product specificities (22–26). However, the lack of structural data for the SET7/9 Y245A and Y305F mutants in complex with cognate methylated peptides has hindered our understanding of the mechanisms that define the respective product specific- ities of these mutants. Moreover, these structures would yield a framework for visualizing the mono-, di-, and trimethylation reactions catalyzed by SET domain KMTs. To gain insight into the molecular basis of their product specificities, we have determined high resolution crystal structures of the SET7/9 Y245A and Y305F mutants in com- plex with peptides of the TATA box-binding protein-associ- ated factor TAF10 bearing the Lys-189 methylation site in unmodified (K189), monomethylated (K189me1), dimethyl- ated (K189me2), and trimethylated (K189me3) states. The structures and accompanying biochemical data support a model whereby changes in the occupancy or position of water molecules in the active site are critical in establishing the prod- uct specificities of the SET7/9 Y245A and Y305F mutants. Together, our results provide new insights into the mechanisms that govern SET domain product specificity and provide step- wise snapshots of the lysine mono-, di-, and trimethyl transfer reactions catalyzed by KMTs. EXPERIMENTAL PROCEDURES Cloning, Expression, and Purification of the SET7/9 Mutants— The Y245A and Y305F mutants were introduced into the pHIS2 SET7/9 expression vector encoding residues 110–366 (27) via QuikChange site-directed mutagenesis (Stratagene) and were verified by dideoxy DNA sequencing. The plasmids encoding wild type (WT) SET7/9 and the Y245A and Y305F mutants were transformed into Rosetta2 DE3 cells (Novagen) and were expressed as described previously (27, 28). In the course of characterizing WT SET7/9, we observed that the enzyme co- purified with AdoMet or another contaminant that resulted in technical difficulties in the isothermal titration calorimetry (ITC) experiments and co-crystallization trials with the TAF10 peptides. To overcome this problem, a denaturation and refold- ing step was inserted in the purification scheme. The denatur- ation and refolding protocol involved adding 6 M guanidine HCl, 50 mM sodium phosphate, pH 7.5, and 500 mM NaCl to the protein while it was immobilized on a nickel-Sepharose column (GE Healthcare). The column was washed with this buffer, fol- lowed by a solution of 6 M urea, 50 mM sodium phosphate, pH 7.5, 150 mM NaCl, and 10 mM 2-mercaptoethanol to remove the cofactor from the denatured enzyme. A reverse gradient from 6 to 0 M urea was then performed in the same buffer to refold the protein, which was subsequently eluted from the column using a linear gradient of 0–500 mM imidazole in 50 mM sodium phosphate, pH 7.5, 500 NaCl, and 10 mM 2-mercaptoethanol. The refolded protein was digested with tobacco etch virus pro- tease (29) during dialysis against 20 mM Tris, pH 7.5, 150 mM NaCl, and 5 mM 2-mercaptoethanol and then purified using a Superdex 200 gel filtration column (GE Healthcare). Protein concentration was determined by its absorbance at 280 nm. Synthetic Peptides—The TAF10 peptides bearing K189, K189me1, K189me2, and K189me3 (sequence, acetyl- SKSK189DRKYTL-amide) and a biotinylated TAF10 peptide (sequence, acetyl-SKSK189DRKYTLT(K-EZLink-S-S-biotin)- amide) were synthesized and purified by New England Peptide, Inc. Peptide concentrations were measured using the absorb- ance of their tyrosine residue at 274 nm. Crystallization and Data Collection—Crystals were pro- duced by hanging drop vapor diffusion by mixing the crystalli- zation solution in a 1:1 ratio with 10–20 mg/ml SET7/9, 1 mM S-adenosylhomocysteine (AdoHcy), and 1.0–3.0 mM unmodi- fied or methylated TAF10-K189 peptide in 20 mM Tris, pH 8.0, 100 mM NaCl, and 2 mM tris(2-carboxyethyl)phosphine. Crys- tals were obtained at 20 °C in either 1.8–2.0 M (NH4)2SO4, with 0.1 M BisTris, pH 6.2–6.6, or in 0.95–1.1 M sodium citrate with 100 mM imidazole pH 8.0–8.4. In both crystallization condi- tions, the final pH values were between pH 8.0 and 9.0. Crystals in the (NH4)2SO4 condition were typically flash-frozen in the mother liquor containing 25–30% glycerol, and the crystals in the citrate condition were frozen in 1.6 M sodium citrate. Data were collected at the Advanced Photon Source beamlines 21-IDG (LS-CAT) and 23-IDD (GM/CA-CAT). Images were indexed, integrated, and scaled using HKL2000 (30). Structures of the mutants were solved by molecular replacement using MOLREP (31) with the coordinates of a previously reported SET7/9 ternary complex used as the search model (Protein Data Bank code 2F69). Successive rounds of model building and refinement were carried out using Coot (32) and REFMAC (33), respectively. The geometry of the models were verified by Mol- Probity (34). Simulated annealing omit maps were calculated using CNS (35) with the peptide and cofactor removed to elim- inate model bias in the active site. Structural figures were ren- dered using PyMOL (Schro¨dinger, LLC.). Fluorescent Methyltransferase Assay—A coupled fluorescent methyltransferase assay was used to measure the kinetic parameters of WT SET7/9 and the Y245A and Y305F mutants as reported previously, with the exception that 50–150 nM enzyme, 100 M AdoMet, and varying concentrations of TAF10 peptide substrate were used (27, 36). Assays were performed in triplicate, and a homocysteine calibration curve was used to calculate the initial velocities. Kinetic parameters were calcu- lated by plotting the velocities versus peptide concentration and by fitting the Michaelis-Menten equation to the data via non- linear regression using Prism 5.0 (GraphPad). In cases where the Km value was beyond the measurable range of the assay, the kcat/Km value was determined as described previously (7). Calorimetry Experiments—ITC was performed at 20 °C using a MicroCal VP-ITC calorimeter (GE Healthcare) with 0.12 mM protein and 4 mM AdoHcy in 20 mM sodium phosphate, pH 7, and 100 mM NaCl with 1.5 mM peptide as the injectant. Data were processed, and equilibrium dissociation constants (KD) and curve fitting errors were calculated from the binding iso- therms using Origin 7.0 (OriginLab Corp.). WT SET7/9 and the Y245A and Y305F mutants displayed ligand:protein binding stoichiometries (N values) between 0.8 and 1.0, demonstrating Lysine Methylation by SET7/9 Mutants 31850 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 that WT SET7/9 and its mutants were properly refolded due to their ability to bind peptides in an 1:1 molar ratio. TLC Product Analysis—Methyltransferase assays were per- formed in triplicate at 37 °C with the biotinylated TAF10 pep- tide (0.5 mM) for 30 min in the presence of WT SET7/9 (3 pmol), Y305F mutant (6 pmol), or the Y245A mutant (100 pmol). Assays contained 100 mM HEPES, pH 8.0, 50 mM NaCl, 1.0 mM [methyl-3H]AdoMet (diluted to a specific activity of 0.2 Ci/nmol with purified AdoMet (37)), 15 M Sulfolobus solfa- taricus AdoHcy hydrolase (36), and 2 units of adenosine deami- nase (Roche Applied Science) in a final volume of 20 l. The reactions were terminated by addition of an equal volume of 200 mM MES, pH 5.1, prior to addition of a 2-fold molar excess of immobilized avidin resin (UltraLink; Pierce). Biotinylated peptides were allowed to bind at room temperature for 30 min, and the resin was then collected by centrifugation (9000  g). The resin was washed three times with 300 mM NaCl, and the peptide was eluted overnight from the avidin resin by cleavage of the disulfide bond in the linker of the peptide using 10 mM tris(2-carboxyethyl)phosphine (Thermo Scientific). The resin was incubated with additional 10 mM tris(2-carboxyethyl)phos- phine the following day until the radiolabel was essentially removed from the resin. The recovered peptides were hydro- lyzed by incubation with 6 M HCl at 110 °C for 24 h. Subsequent steps in measuring the radiolabel incorporated into the mono-, di-, and trimethyl-lysine products were performed as reported previously (8). RESULTS Functional Analysis of the SET7/9 Y305F Mutant—Prior studies of SET7/9 by Zhang et al. (6) reported that mutation of the Phe/Tyr switch residue Tyr-305 to a phenylalanine alters its product specificity from a mono- to dimethyltransferase. We verified these findings by demon- strating that WT SET7/9 mono- methylated the TAF10-K189 pep- tide, whereas the Y305F mutant mono- and dimethylated this sub- strate, as demonstrated by mass spectrometry (data not shown). We next examined whether the Y305F substitution altered the affinity of SET7/9 for the TAF10-K189 peptides using ITC (Fig. 1). A comparison of the KD values revealed that SET7/9 Y305F bound the TAF10-K189 and TAF10-K189me1 peptides 4- and 6-fold more tightly, respec- tively, than the WT enzyme, whereas this mutant displayed a substantially diminished affinity for the TAF10-K189me2 pep- tide (Table 1). Although the WT enzyme and the Y305F mutant exhibited discernable differences in their affinities for the unmodified and monomethylated peptides, these variations are modest and cannot account for their distinct product specific- ities, suggesting that a kinetic effect during methylation may be responsible. To investigate this possibility, we characterized the kinetic parameters of WT SET7/9 and the Y305F mutant using the TAF10 peptides as substrates. Both enzymes methylated the unmodified peptide with comparable kcat and Km values (Table 2). In analyzing the kinetic parameters for the methylation of the monomethylated peptide by SET7/9 Y305F, we found that this substrate displayed an elevated Km value that was beyond the measurable range of the assay due to its limited solubility. In this case, we measured the catalytic efficiency (kcat/Km) for the methylation of this peptide and found that it was methylated 15-fold less efficiently than the unmodified peptide by SET7/9 Y305F. Given the fact that the Y305F mutant exhibited a higher binding affinity for the TAF10-K189me1 peptide than the WT enzyme (Table 1), the kinetic data suggest that a step in the reaction pathway following substrate binding limits the catalytic efficiency of this mutant. We next examined whether the Y305F mutant dimethylated the TAF10-K189 peptide via a processive or a distributive mechanism. In a processive mechanism, the methyl-lysine substrate would remain bound to the enzyme during successive methyl transfer reactions; thus, the concentration of an inter- mediate, such as monomethyl-lysine, cannot exceed the en- zyme concentration during the assay. In a distributive mecha- nism, the intermediates are released into solution where they accumulate prior to the next round of methylation, resulting in an intermediate concentration that is greater than that of the enzyme. Using a radiometric TLC assay and a biotinylated TAF10 peptide, we quantified the amounts of monomethylated products generated by the WT SET7/9 and the Y305F mutant FIGURE 1. ITC analysis of WT SET7/9 and the Y245A and Y305F mutants. Representative titrations and binding isotherms for the calorimetry experiments are shown for the unmodified TAF10-K189 peptide titrated into WT SET7/9 (left plot), SET7/9 Y305F (middle plot), and SET7/9 Y245A (right plot). In each panel, the ITC titration experiment (upper panel) is shown with the binding isotherms (lower panel) fit to a single binding site model. TABLE 1 Analysis of the binding affinity of WT SET7/9 and its catalytic mutants for unmodified and methylated TAF10 peptides SET7/9 TAF10 peptide KD a M WT K189 4.9  0.20 WT K189me1 4.0  0.36 Y305F K189 1.3  0.10 Y305F K189me1 0.62  0.065 Y305Fb K189me2 70 Y245A K189 4.0  0.25 Y245A K189me1 3.3  0.10 Y245A K189me2 5.8  0.22 Y245A K189me3 11  0.28 a Curve fitting errors were calculated from the binding isotherms. b An estimate of the affinity is reported due to weak peptide binding. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31851 (Table 3). The data demonstrate that comparable amounts of monomethyl-lysine were generated when the quantity of enzyme usedistakenintoaccount,inagreementwiththeirsimilar turnover numbers for the TAF10-K189 peptide (Table 2). The Y305F mutant also produced small but measurable quantities of radiolabeled dimethyl-lysine product that were substantially smaller than the amount of monomethyl-lysine generated. Therefore, the TLC data are consistent with a distributive mechanism for dimethylation by the Y305F mutant because the amount of monomethyl-lysine produced exceeded the quantity of enzyme used in the assay. Structures of WT SET7/9 and the Y305F Mutant in Complex with Unmodified and Methylated TAF10 Peptides—To deter- mine the mechanism by which the Y305F substitution alters the product specificity of SET7/9, we determined the crystal struc- tures of this mutant bound to AdoHcy and TAF10-K189, TAF10-K189me1, and TAF10-K189me2 peptides and com- pared these to the structures of the WT SET7/9AdoHcy TAF10-K189 complex (supplemental Table 1). The structures of these complexes were determined to 1.85 Å or higher reso- lution, permitting unambiguous modeling of the K189 side chains in the active site of the enzyme based on simulated annealing omit maps (Fig. 2). The ternary complexes of the WT and the Y305F mutant superimpose with overall root mean square differences of less than 0.3 Å for all aligned atoms, indi- cating that neither the Y305F mutation nor the binding of the various TAF10-K189 peptides results in substantial changes in its overall structure. An inspection of the active sites of the SET7/9 WT and Y305F complexes illustrates the binding modes of the unmod- ified and methylated forms of K189 in the TAF10 peptides (Fig. 2, A–D). The K189 side chain binds in an extended all trans conformation in a deep pocket, termed the lysine binding chan- nel, that is composed of Tyr-245, Thr-266, Leu-267, Ser-268, Tyr-305, Tyr-335, and Tyr-337 (not shown for clarity) (Fig. 2A). These residues interact with the aliphatic portion of the K189 side chain primarily through van der Waals contacts. The lysine binding channel connects to the AdoMet-binding site on the opposite face of the catalytic domain via an oxygen-lined methyl transfer pore (38). During catalysis, the methyl group of the cofactor is positioned within the methyl transfer pore for the SN2 reaction with the -amino group of the lysine or methyl-lysine substrate (see below). To lower the activation barrier for this reaction, the lysine -amine nucleophile is aligned for methyl transfer through a hydrogen bond network within the active site. In the WT enzyme, the K189 -amino group hydrogen bonds to the hydroxyl group of Tyr-245 as well as to two water molecules (Fig. 2A). One of the water molecules (termed water 1), is coor- dinated in a solvent pocket, through hydrogen bonds to the carbonyl oxygens of Gly-292 and Ala-295 and to the hydroxyl group of the Phe/Tyr switch residue Tyr-305. This solvent pocket is structurally conserved in SET domain KMTs and has an important role in defining product specificity through the adjacent Phe/Tyr switch residue, as shown in our prior studies of the human H4K20 methyltransferase SET8 (8). The other water molecule is bound within the methyl transfer pore between the lysine substrate and the thioether sulfur atom of AdoHcy through hydrogen bonds to Tyr-245, Asn-265, and His-293 in SET7/9 and the TAF10-K189 -amino group. This water is not observed in other structures of SET7/9 ternary complexes and may represent the approximate position that the AdoMet methyl group occupies in the methyl transfer pore in the Michaelis complex. In structures of the Y305F ternary complexes, the K189, K189me1, and K189me2 side chains also adopt extended trans side chain geometries within the lysine binding channel that are stabilized via hydrogen bonding to Tyr-245 in the enzyme (Fig. 2, B–D). The orientations of the K189me1 and K189me2 side chains are further maintained through carbon-oxygen (CH–O) hydrogen bonding between the methyl groups and oxygen atoms within the vicinity of the methyl transfer pore, as reported previously in other SET domain KMT structures (8, 10, 38). A superimposition of the SET7/9 WT and Y305F com- plexes underscores the similarity of the lysyl binding conforma- tions (Fig. 2E). However, there are notable differences in the hydrogen bond patterns and occupancy of water 1 within the solvent pocket in the Y305F mutant compared with the WT enzyme. Specifically, the Y305F substitution results in the loss of one hydrogen bond to water 1 in the structures of the TAF10- K189 and TAF10-K189me1 complexes (Fig. 2, B and C). In con- trast, water 1 is absent in TAF10-K189me2 complex, and the vacated solvent pocket is occupied by one of the methyl groups TABLE 2 Kinetic parameters of WT SET7/9 and the Y305F and Y245A mutants Enzyme TAF10 peptide substrate Km a kcat a kcat/Km a M min1 M1 min1  103 WT K189 160  17 17  0.62 110  17 Y305F K189 88  5.0 17  0.30 190  11 Y305Fb K189me1 11  0.50 Y245A K189 200  35 0.53  0.04 2.6  0.47 Y245A K189me1 210  23 5.9  0.23 28  3.3 Y245A K189me2 400  29 6.5  0.16 15  1.2 a Curve fitting errors were calculated from the hyperbolic fits for the Michaelis-Menten equation. b Km and kcat were not determined because the Km value was beyond the measurable range; therefore, the kcat/Km value is reported. TABLE 3 Product analysis of WT SET7/9 and the Y305F and Y245A mutants Enzyme Quantity of enzyme Measured product Amount of product formeda nmol nmol WT 0.003 Kme1 0.65  0.07 Y305F 0.006 Kme1 1.5  0.49 Kme2 0.033  0.009 Y245A 0.100 Kme1 0.80  0.22 Kme2 0.39  0.021 Kme3 0.076  0.019 a Standard deviation was calculated from triplicate measurements. Lysine Methylation by SET7/9 Mutants 31852 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 of the dimethyl -amine (Fig. 2D). This methyl group forms a 3.5-Å CH–O hydrogen bond to the carbonyl oxygen of Ala-295, further stabilizing the binding of the dimethyl-lysine side chain. A homologous dimethyl-lysine-binding mode and CH–O hydrogen bond was observed in our prior structural studies of the SET8 Y334F Phe/Tyr switch mutant that confers an analo- gous change in product specificity from a mono- to a dimeth- yltransferase (8). A structural alignment of the active sites of the SET7/9 Y305F and SET8 Y334F mutants bound to cognate dimethylated peptides illustrates that the coordinates of the dimethyl-lysyl side chains are virtually superimposable, with one methyl group oriented toward the methyltransfer pore and the second positioned within the vacant solvent pocket (Fig. 2F). Taken together, the structures of the SET7/9 Y305F com- plexes and the similarities in the dimethyl-lysine conforma- tions in the SET7/9 Y305F and SET8 Y334F mutants imply that the Phe/Tyr switch governs product specificity through a con- served mechanism whereby it indirectly influences the binding modes of the methyl-lysine side chain by modulating the affin- ity of the water molecule (water 1) bound in the solvent pocket. Biochemical Characterization of the SET7/9 Y245A Mutant—Previ- ous studies by Xiao et al. (11) reported that the Y245A mutation yields an unusual change in the product specificity of SET7/9, converting the enzyme to a trimeth- yltransferase with weak monometh- yltransferase activity. We deter- mined that the SET7/9 Y245A could mono-, di-, and trimethylate the TAF10-K189 peptide by mass spec- trometry (data not shown) and TLC (Table 3), confirming the earlier studies of Xiao et al. (11). ITC analysis revealed that the Y245A mutant displayed comparable KD values for the unmodified and methylated TAF10-K189 peptides (Fig. 1), although its affinity for the trimethylated peptide was modestly diminished in comparison with the other peptides (Table 1). The ITC data demonstrate that the Y245A mutant bound the unmodified, mono-, and dimethylated sub- strates with equivalent affinities, suggesting that a kinetic effect or a structural alteration in the active site may be responsible for its diminished activity toward un- modified substrates. To gain further insight into its peculiar product specificity, we characterized the kinetic properties of the SET7/9 Y245A mutant. Steady state analysis demonstrated that this mutant displayed similar Km values for the unmodified, mono- and dimethylated TAF10 peptides (Table 2). However, the turnover number for the TAF10-K189 peptide was diminished over 10-fold versus the methylated peptides and was reduced 30-fold versus the WT enzyme, in agreement with the weak monomethyltransferase activity reported by Xiao et al. (11). In addition, we investigated whether this mutant catalyzes lysine trimethylation via a pro- cessive or distributive mechanism as described for SET7/9 Y305F. The TLC data illustrate that the mono- and dimethyl- lysine intermediates accumulated at quantities greater than that of the enzyme used in the assay, indicating that SET7/9 Y245A obeys a distributive mechanism, analogous to the Y305F mutant (Table 3). Structures of SET7/9 Y245A Bound to Unmodified and Meth- ylated TAF10 Peptides—To elucidate the mechanism underly- ing its unusual product specificity, we determined the crystal structures of SET7/9 Y245A in complex with AdoHcy and unmodified, mono-, di-, and trimethylated TAF10 peptides (supplemental Table 1). These complexes superimpose with the structure of the WT SET7/9AdoHcyTAF10-K189 com- FIGURE 2. Crystal structures of WT SET7/9 and the Y305F mutant in complex with AdoHcy and unmodi- fied and methylated TAF10 peptides. A, active site of WT SET7/9 bound to AdoHcy and TAF10-K189 and the active site of SET7/9 Y305F bound to AdoHcy and TAF10-K189 (B), TAF10-K189me1 (C), and TAF10-K189me2 peptides (D). A–D, the K189 residue in the TAF10 peptide is depicted with yellow carbon atoms with the corresponding Fo – Fc simulated-annealing omit maps (contoured at 2.5 ) shown. AdoHcy is colored with green carbon atoms, and SET7/9 carbon atoms are colored gray, with the exception of Tyr-305/Y305F (magenta) and Tyr-245 (cyan). Gly-264 to Ser-268 are shown, and the main chain atoms of Gly-292 to Ala-295 are depicted for clarity. Red dashed lines indicate conventional hydrogen bonds, and blue dashed lines indicate CH–O hydrogen bonds. Cyan dashed lines indicate weak CH–O hydrogen bonds between 3.5 and 3.7 Å in length. The water molecule in the solvent binding pocket is numbered 1. E, overlay of the active sites of the WT enzyme (light blue carbon atoms) and the Y305F mutant (green, yellow, and pink carbon atoms correspond to K189, K189me1, and K189me2, respectively). Red dashed lines indicate coordination of the waters in the struc- turesoftheY305Fmutant,andthelightbluedashedlineindicateshydrogenbondingthatonlyoccursintheWT enzyme. F, superimposition of the active sites of the SET7/9 Y305FAdoHcyTAF10-K189me2 and SET8 Y334FAdoHcyH4K20me2 (Protein Data Bank code 3F9X) complexes shown with green and yellow carbon atoms, respectively. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31853 plex with root mean squared differences of less than 0.4 Å for all aligned atoms, indicating that the Y245A mutant does not per- turb the overall structure of the enzyme. Simulated annealing omit maps illustrate that K189 side chains are bound within the lysine binding channel through hydrogen bonds and van der Waals contacts (Fig. 3, A–D), although the interactions and binding modes are distinct from those in the complexes of WT SET7/9 and the Y305F mutant (Fig. 2, A–D). In the unmodified TAF10 peptide complex, the K189 -amino group forms a weak hydrogen bond with the carbonyl oxygen of Gly-264 (Fig. 3A), whereas the -amino groups of K189me1 and K189me2 hydro- gen bond to the hydroxyl group of Tyr-305 in the mono- and dimethylated peptide complexes (Fig. 3, B and C). The confor- mations of the K189me1 and K189me2 side chains are further stabilized by water-mediated hydrogen bonding and through CH–O hydrogen bonding to their methyl groups. In the TAF10-K189me3 peptide complex, the trimethyl-lysine side chain is coordinated exclusively through direct and water-me- diated CH–O hydrogen bonds to its methyl groups because the quaternary -ammonium cation cannot engage in hydrogen bonding (Fig. 3D). A structural alignment of the four SET7/9 Y245A complexes illustrates distinct binding modes for the unmodified versus the methylated K189 side chains, highlighting the selectivity of this mutant for methylated substrates. The side chains of K189me1, K189me2, and K189me3 roughly overlay with their respective -amino groups superimposed and adopt slightly kinked con- formations (Fig. 3E), as opposed to the extended trans geome- try of the unmodified and methylated lysines in the complexes of the WT enzyme and the Y305F mutant (Fig. 2, A–D). Con- versely, the unmodified K189 side chain does not superimpose with its methylated counterparts and is oriented in an alterna- tive configuration due to its hydrogen bonding to Gly-264 (Fig. 3, A and E). An overlay of the structures of the WT enzyme and Y245A mutant bound to the unmodified TAF10 peptide illus- trates that the side chains of K189 do not superimpose and that the K189 -amino group appears to be misaligned with AdoHcy in the Y245A complex (Fig. 3F). This suboptimal alignment may explain the diminished kcat value of SET7/9 Y245A mutant toward substrates with unmodified lysines (Table 2). A comparison of the structures of the SET7/9 Y245A and Y305F complexes yields a molecular explanation for the differ- ent product specificities of these two mutants. In the SET7/9 Y305F complexes, Tyr-245 aligns the K189 -amino group for methyl transfer through hydrogen bonding to its hydroxyl group (Fig. 2, B–D). Conversely, in the Y245A mutant, the K189me1 and K189me2 -amino groups are oriented through hydrogen bonding to Tyr-305 (Fig. 3, B and C). These distinct hydrogen bond patterns impart differences in the conforma- tions of the lysyl side chains due to the relative orientations of Tyr-245 and Tyr-305 in the lysine binding channel. Specifically, the kinked conformation adopted by the K189me1 and K189me2 side chains in the Y245A complexes (Fig. 3, B and C) may contribute to the differences in the turnover numbers of this mutant versus those of the WT enzyme and the Y305F mutant (Table 2). In addition, the dimethyl -amino group of the K189me2 side chain binds in distinct orientations in the Y245A and Y305F mutants due to their hydrogen bonding to FIGURE 3. Crystal structures of the SET7/9 Y245A mutant in complex with AdoHcy and unmodified and methylated TAF10 peptides. Active site of SET7/9 Y245A bound to AdoHcy and TAF10-K189 (A), TAF10-K189me1 (B), TAF10-K189me2 (C), and TAF10-K189me3 peptides (D). A–D, Fo – Fc simulated- annealing omit maps (contoured at 2.5 ) for the unmodified and methylated K189 side chains are illustrated. The residues and hydrogen bonds in each com- plexarecoloredasinFig.2.Thewatermoleculesinthelysinebindingchannelsof the Y245A complexes are numbered 1–4, as described in the text. E, superimpo- sition of the active sites of the Y245A complexes bound to the four methylated statesofTAF10-K189(unmodified,mono-,di-,andtrimethylatedshowningreen, yellow, pink, and blue, respectively). F, overlay of WT SET7/9 (blue carbons) and SET7/9 Y245A (green carbons) in complex with TAF10-K189. Waters correspond- ing to the WT and Y245A structures are colored cyan and green, respectively. G, alignment of the active sites of the Y245A (yellow carbons) and Y305F (green carbons) mutants in complex with the TAF10-K189me2 peptide. Hydrogen bonds from the Y305F structure are shown as green dashed lines, and waters and hydrogen bonds in the Y245A structure are shown in yellow and orange, respectively. Lysine Methylation by SET7/9 Mutants 31854 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 Tyr-305 and Tyr-245, respectively (Fig. 3G). In the Y305F mutant, hydrogen bonds to the dimethyl -amino group cou- pled with steric constraints in the lysine binding channel pre- vent the K189me2 side chain from undergoing a conforma- tional change that is conducive to trimethylation (Fig. 2D), consistent with its dimethyltransferase activity. However, in the Y245A mutant, the alanine substitution enlarges the diameter of the lysine binding channel, accommodating trimethyl-lysine (Fig. 3D). In addition, the larger diameter would permit the dimethyl-lysine substrate to undergo the conformational reor- ganization necessary to align the -amino group in a productive geometry for trimethylation. A major difference in the active site of the Y245A mutant versus the other SET7/9 structures is the presence of several water molecules bound in the cavity generated by the Y245A mutation. In the structure of the Y245A mutant bound to TAF10-K189, three water molecules (waters 2–4) occupy this cavity and are arranged in a triangular geometry (Fig. 3A). In addition, water 1 shifts 1.6 Å from its position in the solvent pocket toward water 2 to which it forms a hydrogen bond (Fig. 3, A and E). The shift in water 1 was unexpected given its con- served orientation in the solvent pocket of the SET7/9 WT and Y305F complexes (Fig. 2, A–C) as well as in the structures of other SET domain KMTs (8). This displacement is presumably related to the alternative conformation of the K189 side chain whose -amino group is too distant (4.3 Å) to form a productive hydrogen bond to water 1. Conversely, in the Y245A complexes bound to TAF10-K189me1 and TAF10-K189me2, water 1 remains tightly bound in the solvent pocket through hydrogen bonds to Tyr-305 hydroxyl group, the carbonyl oxygens of Gly- 292 and Ala-295, and the K189 -amino group (Fig. 3, B, C, and E), analogous to its binding in the WT enzyme (Fig. 2A). How- ever, in the TAF10-K189me3 complex, one of the methyl groups of the trimethyl -ammonium cation is oriented into the solvent pocket (Fig. 3D), similar to the dimethyl-lysine binding mode observed in the Y305F mutant (Fig. 2D). The binding of the methyl group in the solvent pocket displaces water 1 by 3.2 Å relative to its position in the TAF10-K189me1 complex (Fig. 3E), thereby avoiding a steric clash with the trimethylated -ammonium group. Variations in the occupancy of water 2 are also seen in the different Y245A structures. Water 2 is bound in similar orientations in the active site of the unmodified and monomethylated peptide complexes but is absent in the di- and trimethylated peptide complexes due to the binding of a methyl group in this position (Fig. 3, A–E). In summary, the changes in the positions or occupancies of waters 1 and 2 correlate with the binding modes of the unmodified and methylated K189 within the active site of the Y245A mutant. Catalytic Models of Lysine Multiple Methylation by SET7/9 Y245A, and Y305F—The structures of the SET7/9 complexes reported here offer a prime opportunity to generate stepwise models for lysine mono-, di-, and trimethylation by a SET domain KMT. We modeled the AdoMet-bound Michaelis complexes by superimposing the SET7/9 product complexes with the previously reported structure of the SET7/9-AdoMet binary complex (Fig. 4) (39). The conformations of the mono- and dimethyl -amino groups in the Michaelis complexes were inferred from the coordinates of the corresponding dimethyl- and trimethyl-lysine products, respectively. In addition, we modeled the -amino group in a deprotonated state with its hydrogen atoms oriented toward the hydrogen bond acceptors that align the lysyl side chain for methylation. As a basis for this comparison, we first modeled the monomethylation reaction catalyzed by WT SET7/9 (Fig. 4A). In the substrate ternary complex, the lysine -amine is aligned with the methyl group and sulfonium cation of AdoMet through a hydrogen bond to the Tyr-245 hydroxyl group and water 1 in the solvent pocket. The values of the reaction distance and angle are 2.8 Å and 153°, respectively, in approximate agreement with the linear geome- try of a SN2 methyl transfer reaction calculated in other mod- eled substrate complexes (8, 10). In the product complex, the monomethyl-lysine side chain is bound in an extended confor- mation with its methyl group oriented within the methyl trans- fer pore, thereby obstructing AdoMet binding. Furthermore, water 1 remains tightly coordinated in the solvent pocket through four hydrogen bonds to Gly-292, Ala-295, Tyr-305, and the monomethyl -amino group. These interactions hinder the dissociation of water 1 and the related rearrangement of the monomethyl-lysine side chain required for a second methyl transfer reaction, explaining why the WT enzyme cannot cata- lyze di- and trimethylation. These findings concur with the FIGURE 4. Catalytic models of the methyl transfer reactions catalyzed by WT SET7/9 and the Y305F mutant. A, monomethylation of TAF10-K189 by the WT enzyme. The reaction scheme depicts the modeled substrate ternary complex (left) and the product complex (right) for the transfer of the methyl group from AdoMet (green carbon atoms) to K189 in TAF10 (yellow carbons), yielding AdoHcy and K189me1. The red arrow indicates the direction of the nucleophilic attack of the deprotonated -amino group on the AdoMet methyl group. The transferred methyl group is colored green, and the white atoms represent the hydrogens of the -amino group. Hydrogen bonds and residues in the enzyme active site are illustrated as in Fig. 2. The reaction distance and angle are labeled in red. B and C, models of the Y305F mutant for the first methyl transfer reaction with TAF10-K189 (B) and second methyl transfer reaction with TAF10-K189me1 (C). Color schemes are the same as in A. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31855 model for SET7/9 product specificity reported in previous structural and functional studies (6, 11). Similar reaction geometry is observed in the model for the monomethyl transfer reaction catalyzed by SET7/9 Y305F. Hydrogen bonds from the Tyr-245 hydroxyl group and water 1 align the lysine -amino group with the AdoMet methyl group at a distance of 2.1 Å and an angle of 160°, equivalent to those measured in the Michaelis complex of the WT enzyme (Fig. 4, A and B). In the product complex, monomethyl-lysine adopts an extended trans configuration analogous to that in the WT enzyme. For dimethylation to occur, the monomethyl-lysine must undergo a conformational change in which its methyl group is rotated out of the methyl transfer path with AdoMet. The structure of the Y305F mutant bound to the dimethylated TAF10 peptide (Fig. 2D) implies that this rearrangement occurs through the dissociation of water 1 due to the loss of the Tyr- 305 hydrogen bond in the solvent pocket. The dissociation of water 1 would enable the monomethyl-lysine side chain to adopt an alternative conformation through a rotation about its C–N bond, projecting the methyl group into the solvent pocket (Fig. 4C). This rotation reorients the methyl group out of the methyl transfer path while realigning the monomethyl - amino group for a second methylation reaction through a direct hydrogen bond to the Tyr-245 hydroxyl group and a CH–O hydrogen bond between its methyl group and Ala-295. The modeled reaction geometry for monomethyl-lysine substrate complex of the Y305F mutant (2.7 Å, 160°) is equivalent to that of the first methyl transfer reaction in SET7/9 Y305F. These geometries concur with our previous models for mono- and dimethylation catalyzed by SET8 Y334F (8), illustrating that the orientation of a methyl group into the solvent pocket is a con- served feature of SET domain KMTs that catalyze multiple methylation. In addition, we modeled the methyl transfer reactions cata- lyzed by SET7/9 Y245A (Fig. 5). In the model of the lysine sub- strate complex, the -amino group is aligned for methyl transfer by a hydrogen bond to the carbonyl oxygen of Gly-264, result- ing in a short reaction distance (2.3 Å) and a suboptimal reac- tion angle (141°) with the methyl group of AdoMet (Fig. 5A). This misalignment appears to be a direct consequence of the Y245A mutation that abolishes hydrogen bonding to the - amino group, illustrating that the suboptimal orientation of the -amine likely contributes to the diminished activity of this mutant toward unmodified substrates (11). Conversely, in the modeled monomethyl-lysine substrate complex for SET7/9 Y245A, Tyr-305 hydrogen bonds to the -amino group, aligning it for methyl transfer (Fig. 5B). In addition, CH–O hydrogen bonds to the monomethyl-lysine methyl group and the dissoci- ation of water 2 from the active site also contribute to reposition- ing the -amino group for dimethylation. Collectively, these inter- actions orient the -amine in a reaction angle of 165° that is more conducive to methyl transfer. However, the reaction distance for dimethylation is 0.6 Å longer than that in the corresponding Y305F model because Tyr-305 is positioned further from AdoMet than Tyr-245 (Figs. 4C and 5B). In the third methyl transfer reaction catalyzed by SET7/9 Y245A, the lone pair of electrons of the dimethyl-lysine - amino group acts as the nucleophile and thus cannot engage in hydrogen bonding. The structure of the trimethyl-lysine prod- uct complex (Fig. 3D) implies that the dimethyl -amine is aligned via CH–O hydrogen bonds to its methyl groups, as shown in the model of the Michaelis complex for this reaction (Fig. 5C). These CH–O hydrogen bonds restrain the orienta- tion of the -amino group and position one of the methyl groups into the solvent pocket, displacing water 1 as discussed earlier (Fig. 3, D and E). These interactions cumulatively align the -amino group and AdoMet methyl group with a reaction distance of 3.0 Å and angle of 162° (Fig. 5C). Taken together, the models of the substrate complexes for SET7/9 Y245A suggest that CH–O hydrogen bonds play an increasingly important role in aligning the methylated -amino group in successive rounds of methyl transfer. DISCUSSION The structural and functional characterization of the SET7/9 Y245A and Y305F mutants presented here yields new insights into the mechanism underlying the product specificity of SET domain KMTs. Importantly, it resolves a general paradox concerning this specificity. How does the active site constrain the motion of the lysine -amino group to align it for methyl transfer with AdoMet, while providing adequate volume to accommodate the mono-, di-, and tri- methylated lysine side chain generated during multiple methyl transfer reactions? The structures of the Y305F and Y245A mutants resolve this paradox, illustrating that alter- ations in the positions or occupancies of water molecules within their active sites generate the space required to FIGURE 5. Models for the methyl transfer reactions catalyzed by the SET7/9 Y245A mutant. Models of the Y245A mutant for the first methyl transfer reaction with TAF10-K189 (A), the second methyl transfer reaction with TAF10-K189me1 (B), and the third methyl transfer reaction with TAF10- K189me2 (C). Residues and hydrogen bonds are depicted as in Fig. 3. Lysine Methylation by SET7/9 Mutants 31856 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 accommodate the multiply methylated -amine produced during successive catalytic cycles. Minor perturbations in the side chains of certain active site residues, such as Tyr- 305, are also observed in alignments of the WT enzyme and the Y245A and Y305F complexes, although these changes are modest compared with the displacement or dissociation of the water molecules in the active site. These findings suggest that the waters function as transient place holders that facil- itate the SN2 methyl transfer reaction. During monomethy- lation, they function to constrain the movement of the lysine -amino group by mediating hydrogen bonds between the substrate and enzyme, thereby promoting the linear align- ment with the methyl group and sulfonium cation of AdoMet (Fig. 4, A and B). During di- and trimethylation, the water molecules either relocate within the lysine binding channel or dissociate from the enzyme, yielding the space required to rotate the methyl group away from the methyl transfer pore and to realign the -amine in productive geom- etry for the next methyl transfer reaction (Figs. 4C and 5, B and C). These findings agree with our prior analysis of the SET8 Phe/Tyr switch mutant in which we demonstrated that the Y334F substitution attenuates hydrogen bonding to the water molecule bound in the solvent pocket, promoting its dissociation and the conformational changes necessary for lysine dimethylation (8). Indeed, there is a nearly identical alignment of the dimethyl-lysine side chains in the structures of SET7/9Y305FandSET8Y334Fcomplexes,despitethedifferences in the orientations of the Phe-305 and Phe-334 side chains in each structure (Fig. 2F). Finally, both the SET8 Y334F (8) and SET7/9 Y305F mutants (Table 2) displayed diminished catalytic efficien- cies for lysine dimethylation versus monomethylation. These dif- ferences may reflect the kinetics of the reorganization within the active site, including the dissociation of the water molecule from the solvent pocket and the concomitant realignment of the monomethyl-lysine into a productive geometry for dimethylation. In addition to their place-holding role, the active site waters may also facilitate the deprotonation of the lysine -amino group between methyl transfer reactions. For methylation to occur, the -amino group must be deprotonated to function as the nucleophile in the SN2 methyl transfer reaction with AdoMet (Figs. 4 and 5). Although the pKa value of the lysine -amine in solution is 10.5, molecular dynamics simulations by Zhang and Bruice (25, 26) indicate that this value diminishes to 8.2 upon formation of the SET7/9 Michaelis complex due to the proximity of the AdoMet sulfonium cation and the low dielectric constant of the active site. Furthermore, their simu- lations show that a chain of water molecules facilitates the dep- rotonation of the -amino group prior to methyl transfer, trans- ferring the proton to bulk solvent. Although these water molecule chains are not evident in our crystal structures, the Y305F and Y245A complexes suggest another potential mech- anism for deprotonation. In the dimethyl-lysine complexes of the Y305F and Y245A mutants, the dissociation of water 1 and 2, respectively, from the lysine binding channel requires that the solvent-mediated hydrogen bond to the -amino group is broken (Figs. 2D and 3C). It is conceivable that these waters dissociate from the active site as hydronium ions, promoting the realignment and deprotonation of the methyl -amino group for the next methyl transfer reaction. A comparison of the SET7/9 Y305F and SET8 Y334F com- plexes yields insights into the mechanism by which the Phe/Tyr switch influences water binding within the solvent pocket. The phenylalanine substitution in the Phe/Tyr switch results in the loss of a single hydrogen bond to the water molecule (water 1) in the solvent pocket compared with the four hydrogen bonds that coordinate the solvent molecule in WT SET7/9 (Fig. 2, A and B) and SET8 (7, 8). Although this attenuation in hydrogen bonding may appear insignificant, this difference is nonetheless impor- tant for at least two reasons. First, theoretical calculations indi- cate that, on average, water molecules form 3.5 hydrogen bonds in solutions (40, 41). This value is greater than the num- ber of hydrogen bonds coordinating water 1 in the solvent pocket in SET7/9 Y305F (Fig. 2, B and C) as well as in SET8 Y334F and other di- and trimethyltransferases that possess a hydrophobic residue in the Phe/Tyr switch site (8). From the perspective of the water molecule, the greater hydrogen bond- ing potential in solution would tend to thermodynamically favor its dissociation from the solvent pocket in SET domain KMTs that lack a tyrosine in the Phe/Tyr switch position. Sec- ond, the ordered binding of water molecules observed in the active sites of SET domain ternary complexes represents an unfavorable entropy compared with their diffusion in bulk sol- vent. In WT SET7/9 (Fig. 2A) and SET8 (7, 8), this entropic penalty can be partially offset through the favorable enthalpy of binding associated with the four hydrogen bonds that coordi- nate the water within the solvent pocket. It is conceivable that the loss of the tyrosine-mediated hydrogen bond in the Phe/Tyr switch shifts the equilibrium in favor of dissociation of the water molecule from the solvent pocket, thereby facilitating dimethylation in SET7/9 Y305F, SET8 Y334F, and other di- and trimethyltransferases. The structures of the SET7/9 Y245A and Y305F complexes illustrate the interactions that align the lysine -amino group during the methyl transfer reactions in each enzyme. In the WT enzyme and the Y305F mutant, hydrogen bonding to the hydroxyl group of Tyr-245 appears to be critical in properly aligning the -amine for methyl transfer (Fig. 4). Tyr-245 is conserved in the sequences of many SET domain KMTs (8, 42), and substitutions of this residue generally impair or abolish activity, indicating its importance in catalysis (8, 43). However, SET7/9 appears to be an exception to this rule, as the Y245A mutant is not only active but is capable of catalyzing lysine trimethylation. In this mutant, Tyr-305 appears to assume the role of Tyr-245 by hydrogen bonding to the monomethylated -amino group to align it for methyl transfer with AdoMet, as illustrated in the modeled substrate complex for the dimethy- lation reaction (Fig. 5B). Conversely, in the model for trimethy- lation, the Tyr-305 hydroxyl group does not hydrogen bond to the -amine but instead participates in a CH–O hydrogen bond with one of the methyl groups to assist in aligning the dimethy- lated -amine for the methyl transfer reaction (Fig. 5C). Addi- tional structural and functional studies of the SET domain tri- methyltransferases will aid in further illuminating the roles of CH–O hydrogen bonds in facilitating lysine multiple methylation. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31857 Acknowledgments—We acknowledge S. Schiebold for assistance in protein expression, purification, and crystallization and S. Anderson and R. Sanishvili for their assistance with x-ray data collection. We also thank S. Bulfer and S. Horowitz for reading the manuscript and providing useful comments. This work utilized the Protein Structure Facility of the Michigan Diabetes Research and Training Center, Uni- versity of Michigan, supported by National Institutes of Health Grant DK020572, NIDDK. Use of the Advanced Photon Source was sup- ported by the United States Department of Energy, Basic Energy Sci- ences, Office of Science, under Contract DE-AC02-06CH11357. GM/CA CAT has been funded in whole or in part by National Institutes of Health NCI Grant Y1-CO-1020 and NIGMS Grant Y1-GM-1104. Use of the LS-CAT Sector 21 was supported by Michi- gan Economic Development Corporation and the Michigan Technol- ogy Tri-Corridor Grant 085P1000817 for the support of this research program. REFERENCES 1. Huang, J., and Berger, S. L. (2008) Curr. Opin. Genet. Dev. 18, 152–158 2. Morgunkova, A., and Barlev, N. A. (2006) Cell Cycle 5, 1308–1312 3. Yang, X. D., Lamb, A., and Chen, L. F. (2009) Epigenetics 4, 429–433 4. Taverna, S. D., Li, H., Ruthenburg, A. J., Allis, C. D., and Patel, D. J. (2007) Nat. Struct. Mol. Biol. 14, 1025–1040 5. Collins, R. E., Tachibana, M., Tamaru, H., Smith, K. M., Jia, D., Zhang, X., Selker, E. U., Shinkai, Y., and Cheng, X. (2005) J. Biol. Chem. 280, 5563–5570 6. Zhang, X., Yang, Z., Khan, S. I., Horton, J. R., Tamaru, H., Selker, E. U., and Cheng, X. (2003) Mol. Cell 12, 177–185 7. Couture, J. F., Collazo, E., Brunzelle, J. S., and Trievel, R. C. (2005) Genes Dev. 19, 1455–1465 8. Couture, J. F., Dirk, L. M., Brunzelle, J. S., Houtz, R. L., and Trievel, R. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 20659–20664 9. Qian, C., Wang, X., Manzur, K., Sachchidanand, Farooq, A., Zeng, L., Wang, R., and Zhou, M. M. (2006) J. Mol. Biol. 359, 86–96 10. Trievel, R. C., Flynn, E. M., Houtz, R. L., and Hurley, J. H. (2003) Nat. Struct. Biol. 10, 545–552 11. Xiao, B., Jing, C., Wilson, J. R., Walker, P. A., Vasisht, N., Kelly, G., Howell, S., Taylor, I. A., Blackburn, G. M., and Gamblin, S. J. (2003) Nature 421, 652–656 12. Chuikov, S., Kurash, J. K., Wilson, J. R., Xiao, B., Justin, N., Ivanov, G. S., McKinney, K., Tempst, P., Prives, C., Gamblin, S. J., Barlev, N. A., and Reinberg, D. (2004) Nature 432, 353–360 13. Ea, C. K., and Baltimore, D. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 18972–18977 14. Este`ve, P. O., Chin, H. G., Benner, J., Feehery, G. R., Samaranayake, M., Horwitz, G. A., Jacobsen, S. E., and Pradhan, S. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 5076–5081 15. Kouskouti, A., Scheer, E., Staub, A., Tora, L., and Talianidis, I. (2004) Mol. Cell 14, 175–182 16. Masatsugu, T., and Yamamoto, K. (2009) Biochem. Biophys. Res. Commun. 381, 22–26 17. Munro, S., Khaire, N., Inche, A., Carr, S., and La Thangue, N. B. (2010) Oncogene 29, 2357–2367 18. Pagans, S., Kauder, S. E., Kaehlcke, K., Sakane, N., Schroeder, S., Dorm- eyer, W., Trievel, R. C., Verdin, E., Schnolzer, M., and Ott, M. (2010) Cell Host Microbe 7, 234–244 19. Subramanian, K., Jia, D., Kapoor-Vazirani, P., Powell, D. R., Collins, R. E., Sharma, D., Peng, J., Cheng, X., and Vertino, P. M. (2008) Mol. Cell 30, 336–347 20. Wang, J., Hevi, S., Kurash, J. K., Lei, H., Gay, F., Bajko, J., Su, H., Sun, W., Chang, H., Xu, G., Gaudet, F., Li, E., and Chen, T. (2009) Nat. Genet. 41, 125–129 21. Yang, X. D., Huang, B., Li, M., Lamb, A., Kelleher, N. L., and Chen, L. F. (2009) EMBO J. 28, 1055–1066 22. Guo, H. B., and Guo, H. (2007) Proc. Natl. Acad. Sci. U.S.A. 104, 8797–8802 23. Hu, P., Wang, S., and Zhang, Y. (2008) J. Am. Chem. Soc. 130, 3806–3813 24. Hu, P., and Zhang, Y. (2006) J. Am. Chem. Soc. 128, 1272–1278 25. Zhang, X., and Bruice, T. C. (2007) Biochemistry 46, 14838–14844 26. Zhang, X., and Bruice, T. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 5728–5732 27. Couture, J. F., Collazo, E., Hauk, G., and Trievel, R. C. (2006) Nat. Struct. Mol. Biol. 13, 140–146 28. Trievel, R. C., Beach, B. M., Dirk, L. M., Houtz, R. L., and Hurley, J. H. (2002) Cell 111, 91–103 29. Kapust, R. B., To¨zse´r, J., Fox, J. D., Anderson, D. E., Cherry, S., Copeland, T. D., and Waugh, D. S. (2001) Protein Eng. 14, 993–1000 30. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307–326 31. Vagin, A., and Teplyakov, A. (2000) Acta Crystallogr. D Biol. Crystallogr. 56, 1622–1624 32. Emsley, P., and Cowtan, K. (2004) Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 33. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Acta Crystallogr. D Biol. Crystallogr. 53, 240–255 34. Davis, I. W., Leaver-Fay, A., Chen, V. B., Block, J. N., Kapral, G. J., Wang, X., Murray, L. W., Arendall, W. B., 3rd, Snoeyink, J., Richardson, J. S., and Richardson, D. C. (2007) Nucleic Acids Res. 35, W375–W383 35. Bru¨nger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. D Biol. Crystallogr. 54, 905–921 36. Collazo, E., Couture, J. F., Bulfer, S., and Trievel, R. C. (2005) Anal. Bio- chem. 342, 86–92 37. Chirpich, T. P., Zappia, V., Costilow, R. N., and Barker, H. A. (1970) J. Biol. Chem. 245, 1778–1789 38. Couture, J. F., Hauk, G., Thompson, M. J., Blackburn, G. M., and Trievel, R. C. (2006) J. Biol. Chem. 281, 19280–19287 39. Kwon, T., Chang, J. H., Kwak, E., Lee, C. W., Joachimiak, A., Kim, Y. C., Lee, J., and Cho, Y. (2003) EMBO J. 22, 292–303 40. Chandra, A., and Chowdhuri, S. (2002) J. Phys. Chem. B 106, 6779–6783 41. Guardia, E., Marti, J., Garcia-Tarres, L., and Laria, D. (2005) J. Mol. Liq. 117, 63–67 42. Dillon, S. C., Zhang, X., Trievel, R. C., and Cheng, X. (2005) Genome Biol. 6, 227 43. Zhang, X., Tamaru, H., Khan, S. I., Horton, J. R., Keefe, L. J., Selker, E. U., and Cheng, X. (2002) Cell 111, 117–127 Lysine Methylation by SET7/9 Mutants 31858 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010
3M57
SET7/9 Y245A in complex with TAF10 peptide and AdoHcy
SET7/9 Catalytic Mutants Reveal the Role of Active Site Water Molecules in Lysine Multiple Methylation*□ S Received for publication,February 15, 2010, and in revised form, July 14, 2010 Published, JBC Papers in Press,August 1, 2010, DOI 10.1074/jbc.M110.114587 Paul A. Del Rizzo‡, Jean-Franc¸ois Couture‡§1, Lynnette M. A. Dirk¶, Bethany S. Strunk‡, Marijo S. Roiko‡, Joseph S. Brunzelle, Robert L. Houtz¶, and Raymond C. Trievel‡2 From the ‡Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109, the §Ottawa Institute of Systems Biology, Department of Biochemistry, Microbiology, and Immunology, University of Ottawa, Ottawa, Ontario K1H 8M5, Canada, the ¶Department of Horticulture, Plant Physiology/Biochemistry/Molecular Biology Program, University of Kentucky, Lexington, Kentucky 40546, and the Department of Molecular Pharmacology and Biological Chemistry, Northwestern University Center for Synchrotron Research, Life Sciences Collaborative Access Team, Argonne, Illinois 60439 SET domain lysine methyltransferases (KMTs) methylate specific lysine residues in histone and non-histone substrates. These enzymes also display product specificity by catalyzing dis- tinct degrees of methylation of the lysine -amino group. To elucidate the molecular mechanism underlying this specificity, we have characterized the Y245A and Y305F mutants of the human KMT SET7/9 (also known as KMT7) that alter its prod- uct specificity from a monomethyltransferase to a di- and a tri- methyltransferase, respectively. Crystal structures of these mutants in complex with peptides bearing unmodified, mono-, di-, and trimethylated lysines illustrate the roles of active site water molecules in aligning the lysine -amino group for methyl transfer with S-adenosylmethionine. Displacement or dissocia- tion of these solvent molecules enlarges the diameter of the active site, accommodating the increasing size of the methylated -amino group during successive methyl transfer reactions. Together, these results furnish new insights into the roles of active site water molecules in modulating lysine multiple meth- ylation by SET domain KMTs and provide the first molecular snapshots of the mono-, di-, and trimethyl transfer reactions catalyzed by these enzymes. SET domain enzymes represent a family of S-adenosylmethi- onine (AdoMet)3-dependent methyltransferases that catalyze the site-specific methylation of protein lysyl residues in a host of proteins, including histones, transcription factors, chroma- tin-modifying enzymes, ribosomal subunits, and other sub- strates (1–3). In many instances, these modifications serve to recruit effector proteins that recognize methyl-lysyl residues in a sequence-dependent fashion (4). In addition, SET domain KMTs exhibit product specificity, defined as their ability to cat- alyze mono-, di-, or trimethylation of the lysine -amino group. This specificity is biologically relevant because many methyl- lysine-binding proteins can discriminate among different degrees of lysine methylation (4). Thus, both the site and degree of lysine methylation are critical to recognition by effector proteins. Structural and functional studies have identified a Phe/Tyr switch in the active site of SET domain KMTs that governs their respective product specificities (5, 6). According to this model, KMTs that possess a tyrosine in the Phe/Tyr switch site are limited to catalyzing lysine monomethylation, whereas en- zymes that possess a phenylalanine or another hydrophobic residue in this position display di- or trimethyltransferase activ- ity. Mutational analysis of various SET domain KMTs, includ- ing DIM-5 (KMT1) (6), SET7/9 (6), G9A (KMT1C) (5), and SET8 (also known as PR-SET7 and KMT5A) (7, 8), has demon- strated that substitutions in the Phe/Tyr switch result in pre- dictable changes in product specificity. Several models have been proposed to explain the mechanism by which the Phe/Tyr switch site governs this specificity, including variations in the diameter of the active site due to the size of Phe/Tyr switch residue and steric hindrance by the tyrosine hydroxyl group (6, 9–11). However, our recent studies of the Phe/Tyr switch mutant Y334F in the human histone H4 Lys-20 (H4K20) meth- yltransferase SET8 indicate that the Phe/Tyr switch regulates product specificity via a more subtle mechanism (8). Specifi- cally, the switch modulates the binding of an active site water molecule that in turn regulates the transition from mono- methylation to multiple methylation. Among the KMTs that have been structurally characterized, SET7/9 has emerged as an archetypal model for studying the catalytic mechanism and product specificity of the SET domain family. Although initially isolated as a histone H3 Lys-4 (H3K4)-specific methyltransferase, this KMT has been shown to regulate the functions of numerous non-histone substrates through site-specific methylation (12–21). Early structural and functional studies of SET7/9 identified two active site mutants, * Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth Grant R01 GM073839 (to R. C. T.) and National Institutes of Health Admin- istrative Supplement GM073839-04S1 (to R. C. T.) funded through the American Recovery and Reinvestment Act. This work was also supported by Department of Energy Grant DE-FG02-92ER20075 (to R. L. H.). □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Table 1. The atomic coordinates and structure factors (codes 3M53, 3M54, 3M55, 3M56, 3M57,3M58,3M59,and3M5A)havebeendepositedintheProteinDataBank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). 1 Supported by a Canadian Institutes of Health Research postdoctoral fellowship. 2 To whom correspondence should be addressed: Dept. of Biological Chem- istry, University of Michigan Medical School, 1150 West Medical Center Dr., 5301 MSRB III, Ann Arbor, MI 48109. Tel.: 734-647-0889; Fax: 734-763-4581; E-mail: rtrievel@umich.edu. 3 The abbreviations used are: AdoMet, S-adenosylmethionine; AdoHcy, S-ad- enosylhomocysteine; KMT, lysine methyltransferase; ITC, isothermal titra- tion calorimetry; CH–O, carbon-oxygen hydrogen bond; BisTris, 2-[bis(2- hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 41, pp. 31849–31858, October 8, 2010 © 2010 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31849 Y245A and Y305F, which change its product specificity. The Phe/Tyr switch mutant Y305F alters SET7/9 product specificity from a mono- to dimethyltransferase (6), whereas the Y245A substitution converts the enzyme into a trimethyltransferase with weak monomethyltransferase activity (11). These mutants have been the subjects of numerous molecular modeling simu- lations that have led to various models to explain their distinct product specificities (22–26). However, the lack of structural data for the SET7/9 Y245A and Y305F mutants in complex with cognate methylated peptides has hindered our understanding of the mechanisms that define the respective product specific- ities of these mutants. Moreover, these structures would yield a framework for visualizing the mono-, di-, and trimethylation reactions catalyzed by SET domain KMTs. To gain insight into the molecular basis of their product specificities, we have determined high resolution crystal structures of the SET7/9 Y245A and Y305F mutants in com- plex with peptides of the TATA box-binding protein-associ- ated factor TAF10 bearing the Lys-189 methylation site in unmodified (K189), monomethylated (K189me1), dimethyl- ated (K189me2), and trimethylated (K189me3) states. The structures and accompanying biochemical data support a model whereby changes in the occupancy or position of water molecules in the active site are critical in establishing the prod- uct specificities of the SET7/9 Y245A and Y305F mutants. Together, our results provide new insights into the mechanisms that govern SET domain product specificity and provide step- wise snapshots of the lysine mono-, di-, and trimethyl transfer reactions catalyzed by KMTs. EXPERIMENTAL PROCEDURES Cloning, Expression, and Purification of the SET7/9 Mutants— The Y245A and Y305F mutants were introduced into the pHIS2 SET7/9 expression vector encoding residues 110–366 (27) via QuikChange site-directed mutagenesis (Stratagene) and were verified by dideoxy DNA sequencing. The plasmids encoding wild type (WT) SET7/9 and the Y245A and Y305F mutants were transformed into Rosetta2 DE3 cells (Novagen) and were expressed as described previously (27, 28). In the course of characterizing WT SET7/9, we observed that the enzyme co- purified with AdoMet or another contaminant that resulted in technical difficulties in the isothermal titration calorimetry (ITC) experiments and co-crystallization trials with the TAF10 peptides. To overcome this problem, a denaturation and refold- ing step was inserted in the purification scheme. The denatur- ation and refolding protocol involved adding 6 M guanidine HCl, 50 mM sodium phosphate, pH 7.5, and 500 mM NaCl to the protein while it was immobilized on a nickel-Sepharose column (GE Healthcare). The column was washed with this buffer, fol- lowed by a solution of 6 M urea, 50 mM sodium phosphate, pH 7.5, 150 mM NaCl, and 10 mM 2-mercaptoethanol to remove the cofactor from the denatured enzyme. A reverse gradient from 6 to 0 M urea was then performed in the same buffer to refold the protein, which was subsequently eluted from the column using a linear gradient of 0–500 mM imidazole in 50 mM sodium phosphate, pH 7.5, 500 NaCl, and 10 mM 2-mercaptoethanol. The refolded protein was digested with tobacco etch virus pro- tease (29) during dialysis against 20 mM Tris, pH 7.5, 150 mM NaCl, and 5 mM 2-mercaptoethanol and then purified using a Superdex 200 gel filtration column (GE Healthcare). Protein concentration was determined by its absorbance at 280 nm. Synthetic Peptides—The TAF10 peptides bearing K189, K189me1, K189me2, and K189me3 (sequence, acetyl- SKSK189DRKYTL-amide) and a biotinylated TAF10 peptide (sequence, acetyl-SKSK189DRKYTLT(K-EZLink-S-S-biotin)- amide) were synthesized and purified by New England Peptide, Inc. Peptide concentrations were measured using the absorb- ance of their tyrosine residue at 274 nm. Crystallization and Data Collection—Crystals were pro- duced by hanging drop vapor diffusion by mixing the crystalli- zation solution in a 1:1 ratio with 10–20 mg/ml SET7/9, 1 mM S-adenosylhomocysteine (AdoHcy), and 1.0–3.0 mM unmodi- fied or methylated TAF10-K189 peptide in 20 mM Tris, pH 8.0, 100 mM NaCl, and 2 mM tris(2-carboxyethyl)phosphine. Crys- tals were obtained at 20 °C in either 1.8–2.0 M (NH4)2SO4, with 0.1 M BisTris, pH 6.2–6.6, or in 0.95–1.1 M sodium citrate with 100 mM imidazole pH 8.0–8.4. In both crystallization condi- tions, the final pH values were between pH 8.0 and 9.0. Crystals in the (NH4)2SO4 condition were typically flash-frozen in the mother liquor containing 25–30% glycerol, and the crystals in the citrate condition were frozen in 1.6 M sodium citrate. Data were collected at the Advanced Photon Source beamlines 21-IDG (LS-CAT) and 23-IDD (GM/CA-CAT). Images were indexed, integrated, and scaled using HKL2000 (30). Structures of the mutants were solved by molecular replacement using MOLREP (31) with the coordinates of a previously reported SET7/9 ternary complex used as the search model (Protein Data Bank code 2F69). Successive rounds of model building and refinement were carried out using Coot (32) and REFMAC (33), respectively. The geometry of the models were verified by Mol- Probity (34). Simulated annealing omit maps were calculated using CNS (35) with the peptide and cofactor removed to elim- inate model bias in the active site. Structural figures were ren- dered using PyMOL (Schro¨dinger, LLC.). Fluorescent Methyltransferase Assay—A coupled fluorescent methyltransferase assay was used to measure the kinetic parameters of WT SET7/9 and the Y245A and Y305F mutants as reported previously, with the exception that 50–150 nM enzyme, 100 M AdoMet, and varying concentrations of TAF10 peptide substrate were used (27, 36). Assays were performed in triplicate, and a homocysteine calibration curve was used to calculate the initial velocities. Kinetic parameters were calcu- lated by plotting the velocities versus peptide concentration and by fitting the Michaelis-Menten equation to the data via non- linear regression using Prism 5.0 (GraphPad). In cases where the Km value was beyond the measurable range of the assay, the kcat/Km value was determined as described previously (7). Calorimetry Experiments—ITC was performed at 20 °C using a MicroCal VP-ITC calorimeter (GE Healthcare) with 0.12 mM protein and 4 mM AdoHcy in 20 mM sodium phosphate, pH 7, and 100 mM NaCl with 1.5 mM peptide as the injectant. Data were processed, and equilibrium dissociation constants (KD) and curve fitting errors were calculated from the binding iso- therms using Origin 7.0 (OriginLab Corp.). WT SET7/9 and the Y245A and Y305F mutants displayed ligand:protein binding stoichiometries (N values) between 0.8 and 1.0, demonstrating Lysine Methylation by SET7/9 Mutants 31850 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 that WT SET7/9 and its mutants were properly refolded due to their ability to bind peptides in an 1:1 molar ratio. TLC Product Analysis—Methyltransferase assays were per- formed in triplicate at 37 °C with the biotinylated TAF10 pep- tide (0.5 mM) for 30 min in the presence of WT SET7/9 (3 pmol), Y305F mutant (6 pmol), or the Y245A mutant (100 pmol). Assays contained 100 mM HEPES, pH 8.0, 50 mM NaCl, 1.0 mM [methyl-3H]AdoMet (diluted to a specific activity of 0.2 Ci/nmol with purified AdoMet (37)), 15 M Sulfolobus solfa- taricus AdoHcy hydrolase (36), and 2 units of adenosine deami- nase (Roche Applied Science) in a final volume of 20 l. The reactions were terminated by addition of an equal volume of 200 mM MES, pH 5.1, prior to addition of a 2-fold molar excess of immobilized avidin resin (UltraLink; Pierce). Biotinylated peptides were allowed to bind at room temperature for 30 min, and the resin was then collected by centrifugation (9000  g). The resin was washed three times with 300 mM NaCl, and the peptide was eluted overnight from the avidin resin by cleavage of the disulfide bond in the linker of the peptide using 10 mM tris(2-carboxyethyl)phosphine (Thermo Scientific). The resin was incubated with additional 10 mM tris(2-carboxyethyl)phos- phine the following day until the radiolabel was essentially removed from the resin. The recovered peptides were hydro- lyzed by incubation with 6 M HCl at 110 °C for 24 h. Subsequent steps in measuring the radiolabel incorporated into the mono-, di-, and trimethyl-lysine products were performed as reported previously (8). RESULTS Functional Analysis of the SET7/9 Y305F Mutant—Prior studies of SET7/9 by Zhang et al. (6) reported that mutation of the Phe/Tyr switch residue Tyr-305 to a phenylalanine alters its product specificity from a mono- to dimethyltransferase. We verified these findings by demon- strating that WT SET7/9 mono- methylated the TAF10-K189 pep- tide, whereas the Y305F mutant mono- and dimethylated this sub- strate, as demonstrated by mass spectrometry (data not shown). We next examined whether the Y305F substitution altered the affinity of SET7/9 for the TAF10-K189 peptides using ITC (Fig. 1). A comparison of the KD values revealed that SET7/9 Y305F bound the TAF10-K189 and TAF10-K189me1 peptides 4- and 6-fold more tightly, respec- tively, than the WT enzyme, whereas this mutant displayed a substantially diminished affinity for the TAF10-K189me2 pep- tide (Table 1). Although the WT enzyme and the Y305F mutant exhibited discernable differences in their affinities for the unmodified and monomethylated peptides, these variations are modest and cannot account for their distinct product specific- ities, suggesting that a kinetic effect during methylation may be responsible. To investigate this possibility, we characterized the kinetic parameters of WT SET7/9 and the Y305F mutant using the TAF10 peptides as substrates. Both enzymes methylated the unmodified peptide with comparable kcat and Km values (Table 2). In analyzing the kinetic parameters for the methylation of the monomethylated peptide by SET7/9 Y305F, we found that this substrate displayed an elevated Km value that was beyond the measurable range of the assay due to its limited solubility. In this case, we measured the catalytic efficiency (kcat/Km) for the methylation of this peptide and found that it was methylated 15-fold less efficiently than the unmodified peptide by SET7/9 Y305F. Given the fact that the Y305F mutant exhibited a higher binding affinity for the TAF10-K189me1 peptide than the WT enzyme (Table 1), the kinetic data suggest that a step in the reaction pathway following substrate binding limits the catalytic efficiency of this mutant. We next examined whether the Y305F mutant dimethylated the TAF10-K189 peptide via a processive or a distributive mechanism. In a processive mechanism, the methyl-lysine substrate would remain bound to the enzyme during successive methyl transfer reactions; thus, the concentration of an inter- mediate, such as monomethyl-lysine, cannot exceed the en- zyme concentration during the assay. In a distributive mecha- nism, the intermediates are released into solution where they accumulate prior to the next round of methylation, resulting in an intermediate concentration that is greater than that of the enzyme. Using a radiometric TLC assay and a biotinylated TAF10 peptide, we quantified the amounts of monomethylated products generated by the WT SET7/9 and the Y305F mutant FIGURE 1. ITC analysis of WT SET7/9 and the Y245A and Y305F mutants. Representative titrations and binding isotherms for the calorimetry experiments are shown for the unmodified TAF10-K189 peptide titrated into WT SET7/9 (left plot), SET7/9 Y305F (middle plot), and SET7/9 Y245A (right plot). In each panel, the ITC titration experiment (upper panel) is shown with the binding isotherms (lower panel) fit to a single binding site model. TABLE 1 Analysis of the binding affinity of WT SET7/9 and its catalytic mutants for unmodified and methylated TAF10 peptides SET7/9 TAF10 peptide KD a M WT K189 4.9  0.20 WT K189me1 4.0  0.36 Y305F K189 1.3  0.10 Y305F K189me1 0.62  0.065 Y305Fb K189me2 70 Y245A K189 4.0  0.25 Y245A K189me1 3.3  0.10 Y245A K189me2 5.8  0.22 Y245A K189me3 11  0.28 a Curve fitting errors were calculated from the binding isotherms. b An estimate of the affinity is reported due to weak peptide binding. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31851 (Table 3). The data demonstrate that comparable amounts of monomethyl-lysine were generated when the quantity of enzyme usedistakenintoaccount,inagreementwiththeirsimilar turnover numbers for the TAF10-K189 peptide (Table 2). The Y305F mutant also produced small but measurable quantities of radiolabeled dimethyl-lysine product that were substantially smaller than the amount of monomethyl-lysine generated. Therefore, the TLC data are consistent with a distributive mechanism for dimethylation by the Y305F mutant because the amount of monomethyl-lysine produced exceeded the quantity of enzyme used in the assay. Structures of WT SET7/9 and the Y305F Mutant in Complex with Unmodified and Methylated TAF10 Peptides—To deter- mine the mechanism by which the Y305F substitution alters the product specificity of SET7/9, we determined the crystal struc- tures of this mutant bound to AdoHcy and TAF10-K189, TAF10-K189me1, and TAF10-K189me2 peptides and com- pared these to the structures of the WT SET7/9AdoHcy TAF10-K189 complex (supplemental Table 1). The structures of these complexes were determined to 1.85 Å or higher reso- lution, permitting unambiguous modeling of the K189 side chains in the active site of the enzyme based on simulated annealing omit maps (Fig. 2). The ternary complexes of the WT and the Y305F mutant superimpose with overall root mean square differences of less than 0.3 Å for all aligned atoms, indi- cating that neither the Y305F mutation nor the binding of the various TAF10-K189 peptides results in substantial changes in its overall structure. An inspection of the active sites of the SET7/9 WT and Y305F complexes illustrates the binding modes of the unmod- ified and methylated forms of K189 in the TAF10 peptides (Fig. 2, A–D). The K189 side chain binds in an extended all trans conformation in a deep pocket, termed the lysine binding chan- nel, that is composed of Tyr-245, Thr-266, Leu-267, Ser-268, Tyr-305, Tyr-335, and Tyr-337 (not shown for clarity) (Fig. 2A). These residues interact with the aliphatic portion of the K189 side chain primarily through van der Waals contacts. The lysine binding channel connects to the AdoMet-binding site on the opposite face of the catalytic domain via an oxygen-lined methyl transfer pore (38). During catalysis, the methyl group of the cofactor is positioned within the methyl transfer pore for the SN2 reaction with the -amino group of the lysine or methyl-lysine substrate (see below). To lower the activation barrier for this reaction, the lysine -amine nucleophile is aligned for methyl transfer through a hydrogen bond network within the active site. In the WT enzyme, the K189 -amino group hydrogen bonds to the hydroxyl group of Tyr-245 as well as to two water molecules (Fig. 2A). One of the water molecules (termed water 1), is coor- dinated in a solvent pocket, through hydrogen bonds to the carbonyl oxygens of Gly-292 and Ala-295 and to the hydroxyl group of the Phe/Tyr switch residue Tyr-305. This solvent pocket is structurally conserved in SET domain KMTs and has an important role in defining product specificity through the adjacent Phe/Tyr switch residue, as shown in our prior studies of the human H4K20 methyltransferase SET8 (8). The other water molecule is bound within the methyl transfer pore between the lysine substrate and the thioether sulfur atom of AdoHcy through hydrogen bonds to Tyr-245, Asn-265, and His-293 in SET7/9 and the TAF10-K189 -amino group. This water is not observed in other structures of SET7/9 ternary complexes and may represent the approximate position that the AdoMet methyl group occupies in the methyl transfer pore in the Michaelis complex. In structures of the Y305F ternary complexes, the K189, K189me1, and K189me2 side chains also adopt extended trans side chain geometries within the lysine binding channel that are stabilized via hydrogen bonding to Tyr-245 in the enzyme (Fig. 2, B–D). The orientations of the K189me1 and K189me2 side chains are further maintained through carbon-oxygen (CH–O) hydrogen bonding between the methyl groups and oxygen atoms within the vicinity of the methyl transfer pore, as reported previously in other SET domain KMT structures (8, 10, 38). A superimposition of the SET7/9 WT and Y305F com- plexes underscores the similarity of the lysyl binding conforma- tions (Fig. 2E). However, there are notable differences in the hydrogen bond patterns and occupancy of water 1 within the solvent pocket in the Y305F mutant compared with the WT enzyme. Specifically, the Y305F substitution results in the loss of one hydrogen bond to water 1 in the structures of the TAF10- K189 and TAF10-K189me1 complexes (Fig. 2, B and C). In con- trast, water 1 is absent in TAF10-K189me2 complex, and the vacated solvent pocket is occupied by one of the methyl groups TABLE 2 Kinetic parameters of WT SET7/9 and the Y305F and Y245A mutants Enzyme TAF10 peptide substrate Km a kcat a kcat/Km a M min1 M1 min1  103 WT K189 160  17 17  0.62 110  17 Y305F K189 88  5.0 17  0.30 190  11 Y305Fb K189me1 11  0.50 Y245A K189 200  35 0.53  0.04 2.6  0.47 Y245A K189me1 210  23 5.9  0.23 28  3.3 Y245A K189me2 400  29 6.5  0.16 15  1.2 a Curve fitting errors were calculated from the hyperbolic fits for the Michaelis-Menten equation. b Km and kcat were not determined because the Km value was beyond the measurable range; therefore, the kcat/Km value is reported. TABLE 3 Product analysis of WT SET7/9 and the Y305F and Y245A mutants Enzyme Quantity of enzyme Measured product Amount of product formeda nmol nmol WT 0.003 Kme1 0.65  0.07 Y305F 0.006 Kme1 1.5  0.49 Kme2 0.033  0.009 Y245A 0.100 Kme1 0.80  0.22 Kme2 0.39  0.021 Kme3 0.076  0.019 a Standard deviation was calculated from triplicate measurements. Lysine Methylation by SET7/9 Mutants 31852 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 of the dimethyl -amine (Fig. 2D). This methyl group forms a 3.5-Å CH–O hydrogen bond to the carbonyl oxygen of Ala-295, further stabilizing the binding of the dimethyl-lysine side chain. A homologous dimethyl-lysine-binding mode and CH–O hydrogen bond was observed in our prior structural studies of the SET8 Y334F Phe/Tyr switch mutant that confers an analo- gous change in product specificity from a mono- to a dimeth- yltransferase (8). A structural alignment of the active sites of the SET7/9 Y305F and SET8 Y334F mutants bound to cognate dimethylated peptides illustrates that the coordinates of the dimethyl-lysyl side chains are virtually superimposable, with one methyl group oriented toward the methyltransfer pore and the second positioned within the vacant solvent pocket (Fig. 2F). Taken together, the structures of the SET7/9 Y305F com- plexes and the similarities in the dimethyl-lysine conforma- tions in the SET7/9 Y305F and SET8 Y334F mutants imply that the Phe/Tyr switch governs product specificity through a con- served mechanism whereby it indirectly influences the binding modes of the methyl-lysine side chain by modulating the affin- ity of the water molecule (water 1) bound in the solvent pocket. Biochemical Characterization of the SET7/9 Y245A Mutant—Previ- ous studies by Xiao et al. (11) reported that the Y245A mutation yields an unusual change in the product specificity of SET7/9, converting the enzyme to a trimeth- yltransferase with weak monometh- yltransferase activity. We deter- mined that the SET7/9 Y245A could mono-, di-, and trimethylate the TAF10-K189 peptide by mass spec- trometry (data not shown) and TLC (Table 3), confirming the earlier studies of Xiao et al. (11). ITC analysis revealed that the Y245A mutant displayed comparable KD values for the unmodified and methylated TAF10-K189 peptides (Fig. 1), although its affinity for the trimethylated peptide was modestly diminished in comparison with the other peptides (Table 1). The ITC data demonstrate that the Y245A mutant bound the unmodified, mono-, and dimethylated sub- strates with equivalent affinities, suggesting that a kinetic effect or a structural alteration in the active site may be responsible for its diminished activity toward un- modified substrates. To gain further insight into its peculiar product specificity, we characterized the kinetic properties of the SET7/9 Y245A mutant. Steady state analysis demonstrated that this mutant displayed similar Km values for the unmodified, mono- and dimethylated TAF10 peptides (Table 2). However, the turnover number for the TAF10-K189 peptide was diminished over 10-fold versus the methylated peptides and was reduced 30-fold versus the WT enzyme, in agreement with the weak monomethyltransferase activity reported by Xiao et al. (11). In addition, we investigated whether this mutant catalyzes lysine trimethylation via a pro- cessive or distributive mechanism as described for SET7/9 Y305F. The TLC data illustrate that the mono- and dimethyl- lysine intermediates accumulated at quantities greater than that of the enzyme used in the assay, indicating that SET7/9 Y245A obeys a distributive mechanism, analogous to the Y305F mutant (Table 3). Structures of SET7/9 Y245A Bound to Unmodified and Meth- ylated TAF10 Peptides—To elucidate the mechanism underly- ing its unusual product specificity, we determined the crystal structures of SET7/9 Y245A in complex with AdoHcy and unmodified, mono-, di-, and trimethylated TAF10 peptides (supplemental Table 1). These complexes superimpose with the structure of the WT SET7/9AdoHcyTAF10-K189 com- FIGURE 2. Crystal structures of WT SET7/9 and the Y305F mutant in complex with AdoHcy and unmodi- fied and methylated TAF10 peptides. A, active site of WT SET7/9 bound to AdoHcy and TAF10-K189 and the active site of SET7/9 Y305F bound to AdoHcy and TAF10-K189 (B), TAF10-K189me1 (C), and TAF10-K189me2 peptides (D). A–D, the K189 residue in the TAF10 peptide is depicted with yellow carbon atoms with the corresponding Fo – Fc simulated-annealing omit maps (contoured at 2.5 ) shown. AdoHcy is colored with green carbon atoms, and SET7/9 carbon atoms are colored gray, with the exception of Tyr-305/Y305F (magenta) and Tyr-245 (cyan). Gly-264 to Ser-268 are shown, and the main chain atoms of Gly-292 to Ala-295 are depicted for clarity. Red dashed lines indicate conventional hydrogen bonds, and blue dashed lines indicate CH–O hydrogen bonds. Cyan dashed lines indicate weak CH–O hydrogen bonds between 3.5 and 3.7 Å in length. The water molecule in the solvent binding pocket is numbered 1. E, overlay of the active sites of the WT enzyme (light blue carbon atoms) and the Y305F mutant (green, yellow, and pink carbon atoms correspond to K189, K189me1, and K189me2, respectively). Red dashed lines indicate coordination of the waters in the struc- turesoftheY305Fmutant,andthelightbluedashedlineindicateshydrogenbondingthatonlyoccursintheWT enzyme. F, superimposition of the active sites of the SET7/9 Y305FAdoHcyTAF10-K189me2 and SET8 Y334FAdoHcyH4K20me2 (Protein Data Bank code 3F9X) complexes shown with green and yellow carbon atoms, respectively. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31853 plex with root mean squared differences of less than 0.4 Å for all aligned atoms, indicating that the Y245A mutant does not per- turb the overall structure of the enzyme. Simulated annealing omit maps illustrate that K189 side chains are bound within the lysine binding channel through hydrogen bonds and van der Waals contacts (Fig. 3, A–D), although the interactions and binding modes are distinct from those in the complexes of WT SET7/9 and the Y305F mutant (Fig. 2, A–D). In the unmodified TAF10 peptide complex, the K189 -amino group forms a weak hydrogen bond with the carbonyl oxygen of Gly-264 (Fig. 3A), whereas the -amino groups of K189me1 and K189me2 hydro- gen bond to the hydroxyl group of Tyr-305 in the mono- and dimethylated peptide complexes (Fig. 3, B and C). The confor- mations of the K189me1 and K189me2 side chains are further stabilized by water-mediated hydrogen bonding and through CH–O hydrogen bonding to their methyl groups. In the TAF10-K189me3 peptide complex, the trimethyl-lysine side chain is coordinated exclusively through direct and water-me- diated CH–O hydrogen bonds to its methyl groups because the quaternary -ammonium cation cannot engage in hydrogen bonding (Fig. 3D). A structural alignment of the four SET7/9 Y245A complexes illustrates distinct binding modes for the unmodified versus the methylated K189 side chains, highlighting the selectivity of this mutant for methylated substrates. The side chains of K189me1, K189me2, and K189me3 roughly overlay with their respective -amino groups superimposed and adopt slightly kinked con- formations (Fig. 3E), as opposed to the extended trans geome- try of the unmodified and methylated lysines in the complexes of the WT enzyme and the Y305F mutant (Fig. 2, A–D). Con- versely, the unmodified K189 side chain does not superimpose with its methylated counterparts and is oriented in an alterna- tive configuration due to its hydrogen bonding to Gly-264 (Fig. 3, A and E). An overlay of the structures of the WT enzyme and Y245A mutant bound to the unmodified TAF10 peptide illus- trates that the side chains of K189 do not superimpose and that the K189 -amino group appears to be misaligned with AdoHcy in the Y245A complex (Fig. 3F). This suboptimal alignment may explain the diminished kcat value of SET7/9 Y245A mutant toward substrates with unmodified lysines (Table 2). A comparison of the structures of the SET7/9 Y245A and Y305F complexes yields a molecular explanation for the differ- ent product specificities of these two mutants. In the SET7/9 Y305F complexes, Tyr-245 aligns the K189 -amino group for methyl transfer through hydrogen bonding to its hydroxyl group (Fig. 2, B–D). Conversely, in the Y245A mutant, the K189me1 and K189me2 -amino groups are oriented through hydrogen bonding to Tyr-305 (Fig. 3, B and C). These distinct hydrogen bond patterns impart differences in the conforma- tions of the lysyl side chains due to the relative orientations of Tyr-245 and Tyr-305 in the lysine binding channel. Specifically, the kinked conformation adopted by the K189me1 and K189me2 side chains in the Y245A complexes (Fig. 3, B and C) may contribute to the differences in the turnover numbers of this mutant versus those of the WT enzyme and the Y305F mutant (Table 2). In addition, the dimethyl -amino group of the K189me2 side chain binds in distinct orientations in the Y245A and Y305F mutants due to their hydrogen bonding to FIGURE 3. Crystal structures of the SET7/9 Y245A mutant in complex with AdoHcy and unmodified and methylated TAF10 peptides. Active site of SET7/9 Y245A bound to AdoHcy and TAF10-K189 (A), TAF10-K189me1 (B), TAF10-K189me2 (C), and TAF10-K189me3 peptides (D). A–D, Fo – Fc simulated- annealing omit maps (contoured at 2.5 ) for the unmodified and methylated K189 side chains are illustrated. The residues and hydrogen bonds in each com- plexarecoloredasinFig.2.Thewatermoleculesinthelysinebindingchannelsof the Y245A complexes are numbered 1–4, as described in the text. E, superimpo- sition of the active sites of the Y245A complexes bound to the four methylated statesofTAF10-K189(unmodified,mono-,di-,andtrimethylatedshowningreen, yellow, pink, and blue, respectively). F, overlay of WT SET7/9 (blue carbons) and SET7/9 Y245A (green carbons) in complex with TAF10-K189. Waters correspond- ing to the WT and Y245A structures are colored cyan and green, respectively. G, alignment of the active sites of the Y245A (yellow carbons) and Y305F (green carbons) mutants in complex with the TAF10-K189me2 peptide. Hydrogen bonds from the Y305F structure are shown as green dashed lines, and waters and hydrogen bonds in the Y245A structure are shown in yellow and orange, respectively. Lysine Methylation by SET7/9 Mutants 31854 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 Tyr-305 and Tyr-245, respectively (Fig. 3G). In the Y305F mutant, hydrogen bonds to the dimethyl -amino group cou- pled with steric constraints in the lysine binding channel pre- vent the K189me2 side chain from undergoing a conforma- tional change that is conducive to trimethylation (Fig. 2D), consistent with its dimethyltransferase activity. However, in the Y245A mutant, the alanine substitution enlarges the diameter of the lysine binding channel, accommodating trimethyl-lysine (Fig. 3D). In addition, the larger diameter would permit the dimethyl-lysine substrate to undergo the conformational reor- ganization necessary to align the -amino group in a productive geometry for trimethylation. A major difference in the active site of the Y245A mutant versus the other SET7/9 structures is the presence of several water molecules bound in the cavity generated by the Y245A mutation. In the structure of the Y245A mutant bound to TAF10-K189, three water molecules (waters 2–4) occupy this cavity and are arranged in a triangular geometry (Fig. 3A). In addition, water 1 shifts 1.6 Å from its position in the solvent pocket toward water 2 to which it forms a hydrogen bond (Fig. 3, A and E). The shift in water 1 was unexpected given its con- served orientation in the solvent pocket of the SET7/9 WT and Y305F complexes (Fig. 2, A–C) as well as in the structures of other SET domain KMTs (8). This displacement is presumably related to the alternative conformation of the K189 side chain whose -amino group is too distant (4.3 Å) to form a productive hydrogen bond to water 1. Conversely, in the Y245A complexes bound to TAF10-K189me1 and TAF10-K189me2, water 1 remains tightly bound in the solvent pocket through hydrogen bonds to Tyr-305 hydroxyl group, the carbonyl oxygens of Gly- 292 and Ala-295, and the K189 -amino group (Fig. 3, B, C, and E), analogous to its binding in the WT enzyme (Fig. 2A). How- ever, in the TAF10-K189me3 complex, one of the methyl groups of the trimethyl -ammonium cation is oriented into the solvent pocket (Fig. 3D), similar to the dimethyl-lysine binding mode observed in the Y305F mutant (Fig. 2D). The binding of the methyl group in the solvent pocket displaces water 1 by 3.2 Å relative to its position in the TAF10-K189me1 complex (Fig. 3E), thereby avoiding a steric clash with the trimethylated -ammonium group. Variations in the occupancy of water 2 are also seen in the different Y245A structures. Water 2 is bound in similar orientations in the active site of the unmodified and monomethylated peptide complexes but is absent in the di- and trimethylated peptide complexes due to the binding of a methyl group in this position (Fig. 3, A–E). In summary, the changes in the positions or occupancies of waters 1 and 2 correlate with the binding modes of the unmodified and methylated K189 within the active site of the Y245A mutant. Catalytic Models of Lysine Multiple Methylation by SET7/9 Y245A, and Y305F—The structures of the SET7/9 complexes reported here offer a prime opportunity to generate stepwise models for lysine mono-, di-, and trimethylation by a SET domain KMT. We modeled the AdoMet-bound Michaelis complexes by superimposing the SET7/9 product complexes with the previously reported structure of the SET7/9-AdoMet binary complex (Fig. 4) (39). The conformations of the mono- and dimethyl -amino groups in the Michaelis complexes were inferred from the coordinates of the corresponding dimethyl- and trimethyl-lysine products, respectively. In addition, we modeled the -amino group in a deprotonated state with its hydrogen atoms oriented toward the hydrogen bond acceptors that align the lysyl side chain for methylation. As a basis for this comparison, we first modeled the monomethylation reaction catalyzed by WT SET7/9 (Fig. 4A). In the substrate ternary complex, the lysine -amine is aligned with the methyl group and sulfonium cation of AdoMet through a hydrogen bond to the Tyr-245 hydroxyl group and water 1 in the solvent pocket. The values of the reaction distance and angle are 2.8 Å and 153°, respectively, in approximate agreement with the linear geome- try of a SN2 methyl transfer reaction calculated in other mod- eled substrate complexes (8, 10). In the product complex, the monomethyl-lysine side chain is bound in an extended confor- mation with its methyl group oriented within the methyl trans- fer pore, thereby obstructing AdoMet binding. Furthermore, water 1 remains tightly coordinated in the solvent pocket through four hydrogen bonds to Gly-292, Ala-295, Tyr-305, and the monomethyl -amino group. These interactions hinder the dissociation of water 1 and the related rearrangement of the monomethyl-lysine side chain required for a second methyl transfer reaction, explaining why the WT enzyme cannot cata- lyze di- and trimethylation. These findings concur with the FIGURE 4. Catalytic models of the methyl transfer reactions catalyzed by WT SET7/9 and the Y305F mutant. A, monomethylation of TAF10-K189 by the WT enzyme. The reaction scheme depicts the modeled substrate ternary complex (left) and the product complex (right) for the transfer of the methyl group from AdoMet (green carbon atoms) to K189 in TAF10 (yellow carbons), yielding AdoHcy and K189me1. The red arrow indicates the direction of the nucleophilic attack of the deprotonated -amino group on the AdoMet methyl group. The transferred methyl group is colored green, and the white atoms represent the hydrogens of the -amino group. Hydrogen bonds and residues in the enzyme active site are illustrated as in Fig. 2. The reaction distance and angle are labeled in red. B and C, models of the Y305F mutant for the first methyl transfer reaction with TAF10-K189 (B) and second methyl transfer reaction with TAF10-K189me1 (C). Color schemes are the same as in A. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31855 model for SET7/9 product specificity reported in previous structural and functional studies (6, 11). Similar reaction geometry is observed in the model for the monomethyl transfer reaction catalyzed by SET7/9 Y305F. Hydrogen bonds from the Tyr-245 hydroxyl group and water 1 align the lysine -amino group with the AdoMet methyl group at a distance of 2.1 Å and an angle of 160°, equivalent to those measured in the Michaelis complex of the WT enzyme (Fig. 4, A and B). In the product complex, monomethyl-lysine adopts an extended trans configuration analogous to that in the WT enzyme. For dimethylation to occur, the monomethyl-lysine must undergo a conformational change in which its methyl group is rotated out of the methyl transfer path with AdoMet. The structure of the Y305F mutant bound to the dimethylated TAF10 peptide (Fig. 2D) implies that this rearrangement occurs through the dissociation of water 1 due to the loss of the Tyr- 305 hydrogen bond in the solvent pocket. The dissociation of water 1 would enable the monomethyl-lysine side chain to adopt an alternative conformation through a rotation about its C–N bond, projecting the methyl group into the solvent pocket (Fig. 4C). This rotation reorients the methyl group out of the methyl transfer path while realigning the monomethyl - amino group for a second methylation reaction through a direct hydrogen bond to the Tyr-245 hydroxyl group and a CH–O hydrogen bond between its methyl group and Ala-295. The modeled reaction geometry for monomethyl-lysine substrate complex of the Y305F mutant (2.7 Å, 160°) is equivalent to that of the first methyl transfer reaction in SET7/9 Y305F. These geometries concur with our previous models for mono- and dimethylation catalyzed by SET8 Y334F (8), illustrating that the orientation of a methyl group into the solvent pocket is a con- served feature of SET domain KMTs that catalyze multiple methylation. In addition, we modeled the methyl transfer reactions cata- lyzed by SET7/9 Y245A (Fig. 5). In the model of the lysine sub- strate complex, the -amino group is aligned for methyl transfer by a hydrogen bond to the carbonyl oxygen of Gly-264, result- ing in a short reaction distance (2.3 Å) and a suboptimal reac- tion angle (141°) with the methyl group of AdoMet (Fig. 5A). This misalignment appears to be a direct consequence of the Y245A mutation that abolishes hydrogen bonding to the - amino group, illustrating that the suboptimal orientation of the -amine likely contributes to the diminished activity of this mutant toward unmodified substrates (11). Conversely, in the modeled monomethyl-lysine substrate complex for SET7/9 Y245A, Tyr-305 hydrogen bonds to the -amino group, aligning it for methyl transfer (Fig. 5B). In addition, CH–O hydrogen bonds to the monomethyl-lysine methyl group and the dissoci- ation of water 2 from the active site also contribute to reposition- ing the -amino group for dimethylation. Collectively, these inter- actions orient the -amine in a reaction angle of 165° that is more conducive to methyl transfer. However, the reaction distance for dimethylation is 0.6 Å longer than that in the corresponding Y305F model because Tyr-305 is positioned further from AdoMet than Tyr-245 (Figs. 4C and 5B). In the third methyl transfer reaction catalyzed by SET7/9 Y245A, the lone pair of electrons of the dimethyl-lysine - amino group acts as the nucleophile and thus cannot engage in hydrogen bonding. The structure of the trimethyl-lysine prod- uct complex (Fig. 3D) implies that the dimethyl -amine is aligned via CH–O hydrogen bonds to its methyl groups, as shown in the model of the Michaelis complex for this reaction (Fig. 5C). These CH–O hydrogen bonds restrain the orienta- tion of the -amino group and position one of the methyl groups into the solvent pocket, displacing water 1 as discussed earlier (Fig. 3, D and E). These interactions cumulatively align the -amino group and AdoMet methyl group with a reaction distance of 3.0 Å and angle of 162° (Fig. 5C). Taken together, the models of the substrate complexes for SET7/9 Y245A suggest that CH–O hydrogen bonds play an increasingly important role in aligning the methylated -amino group in successive rounds of methyl transfer. DISCUSSION The structural and functional characterization of the SET7/9 Y245A and Y305F mutants presented here yields new insights into the mechanism underlying the product specificity of SET domain KMTs. Importantly, it resolves a general paradox concerning this specificity. How does the active site constrain the motion of the lysine -amino group to align it for methyl transfer with AdoMet, while providing adequate volume to accommodate the mono-, di-, and tri- methylated lysine side chain generated during multiple methyl transfer reactions? The structures of the Y305F and Y245A mutants resolve this paradox, illustrating that alter- ations in the positions or occupancies of water molecules within their active sites generate the space required to FIGURE 5. Models for the methyl transfer reactions catalyzed by the SET7/9 Y245A mutant. Models of the Y245A mutant for the first methyl transfer reaction with TAF10-K189 (A), the second methyl transfer reaction with TAF10-K189me1 (B), and the third methyl transfer reaction with TAF10- K189me2 (C). Residues and hydrogen bonds are depicted as in Fig. 3. Lysine Methylation by SET7/9 Mutants 31856 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 accommodate the multiply methylated -amine produced during successive catalytic cycles. Minor perturbations in the side chains of certain active site residues, such as Tyr- 305, are also observed in alignments of the WT enzyme and the Y245A and Y305F complexes, although these changes are modest compared with the displacement or dissociation of the water molecules in the active site. These findings suggest that the waters function as transient place holders that facil- itate the SN2 methyl transfer reaction. During monomethy- lation, they function to constrain the movement of the lysine -amino group by mediating hydrogen bonds between the substrate and enzyme, thereby promoting the linear align- ment with the methyl group and sulfonium cation of AdoMet (Fig. 4, A and B). During di- and trimethylation, the water molecules either relocate within the lysine binding channel or dissociate from the enzyme, yielding the space required to rotate the methyl group away from the methyl transfer pore and to realign the -amine in productive geom- etry for the next methyl transfer reaction (Figs. 4C and 5, B and C). These findings agree with our prior analysis of the SET8 Phe/Tyr switch mutant in which we demonstrated that the Y334F substitution attenuates hydrogen bonding to the water molecule bound in the solvent pocket, promoting its dissociation and the conformational changes necessary for lysine dimethylation (8). Indeed, there is a nearly identical alignment of the dimethyl-lysine side chains in the structures of SET7/9Y305FandSET8Y334Fcomplexes,despitethedifferences in the orientations of the Phe-305 and Phe-334 side chains in each structure (Fig. 2F). Finally, both the SET8 Y334F (8) and SET7/9 Y305F mutants (Table 2) displayed diminished catalytic efficien- cies for lysine dimethylation versus monomethylation. These dif- ferences may reflect the kinetics of the reorganization within the active site, including the dissociation of the water molecule from the solvent pocket and the concomitant realignment of the monomethyl-lysine into a productive geometry for dimethylation. In addition to their place-holding role, the active site waters may also facilitate the deprotonation of the lysine -amino group between methyl transfer reactions. For methylation to occur, the -amino group must be deprotonated to function as the nucleophile in the SN2 methyl transfer reaction with AdoMet (Figs. 4 and 5). Although the pKa value of the lysine -amine in solution is 10.5, molecular dynamics simulations by Zhang and Bruice (25, 26) indicate that this value diminishes to 8.2 upon formation of the SET7/9 Michaelis complex due to the proximity of the AdoMet sulfonium cation and the low dielectric constant of the active site. Furthermore, their simu- lations show that a chain of water molecules facilitates the dep- rotonation of the -amino group prior to methyl transfer, trans- ferring the proton to bulk solvent. Although these water molecule chains are not evident in our crystal structures, the Y305F and Y245A complexes suggest another potential mech- anism for deprotonation. In the dimethyl-lysine complexes of the Y305F and Y245A mutants, the dissociation of water 1 and 2, respectively, from the lysine binding channel requires that the solvent-mediated hydrogen bond to the -amino group is broken (Figs. 2D and 3C). It is conceivable that these waters dissociate from the active site as hydronium ions, promoting the realignment and deprotonation of the methyl -amino group for the next methyl transfer reaction. A comparison of the SET7/9 Y305F and SET8 Y334F com- plexes yields insights into the mechanism by which the Phe/Tyr switch influences water binding within the solvent pocket. The phenylalanine substitution in the Phe/Tyr switch results in the loss of a single hydrogen bond to the water molecule (water 1) in the solvent pocket compared with the four hydrogen bonds that coordinate the solvent molecule in WT SET7/9 (Fig. 2, A and B) and SET8 (7, 8). Although this attenuation in hydrogen bonding may appear insignificant, this difference is nonetheless impor- tant for at least two reasons. First, theoretical calculations indi- cate that, on average, water molecules form 3.5 hydrogen bonds in solutions (40, 41). This value is greater than the num- ber of hydrogen bonds coordinating water 1 in the solvent pocket in SET7/9 Y305F (Fig. 2, B and C) as well as in SET8 Y334F and other di- and trimethyltransferases that possess a hydrophobic residue in the Phe/Tyr switch site (8). From the perspective of the water molecule, the greater hydrogen bond- ing potential in solution would tend to thermodynamically favor its dissociation from the solvent pocket in SET domain KMTs that lack a tyrosine in the Phe/Tyr switch position. Sec- ond, the ordered binding of water molecules observed in the active sites of SET domain ternary complexes represents an unfavorable entropy compared with their diffusion in bulk sol- vent. In WT SET7/9 (Fig. 2A) and SET8 (7, 8), this entropic penalty can be partially offset through the favorable enthalpy of binding associated with the four hydrogen bonds that coordi- nate the water within the solvent pocket. It is conceivable that the loss of the tyrosine-mediated hydrogen bond in the Phe/Tyr switch shifts the equilibrium in favor of dissociation of the water molecule from the solvent pocket, thereby facilitating dimethylation in SET7/9 Y305F, SET8 Y334F, and other di- and trimethyltransferases. The structures of the SET7/9 Y245A and Y305F complexes illustrate the interactions that align the lysine -amino group during the methyl transfer reactions in each enzyme. In the WT enzyme and the Y305F mutant, hydrogen bonding to the hydroxyl group of Tyr-245 appears to be critical in properly aligning the -amine for methyl transfer (Fig. 4). Tyr-245 is conserved in the sequences of many SET domain KMTs (8, 42), and substitutions of this residue generally impair or abolish activity, indicating its importance in catalysis (8, 43). However, SET7/9 appears to be an exception to this rule, as the Y245A mutant is not only active but is capable of catalyzing lysine trimethylation. In this mutant, Tyr-305 appears to assume the role of Tyr-245 by hydrogen bonding to the monomethylated -amino group to align it for methyl transfer with AdoMet, as illustrated in the modeled substrate complex for the dimethy- lation reaction (Fig. 5B). Conversely, in the model for trimethy- lation, the Tyr-305 hydroxyl group does not hydrogen bond to the -amine but instead participates in a CH–O hydrogen bond with one of the methyl groups to assist in aligning the dimethy- lated -amine for the methyl transfer reaction (Fig. 5C). Addi- tional structural and functional studies of the SET domain tri- methyltransferases will aid in further illuminating the roles of CH–O hydrogen bonds in facilitating lysine multiple methylation. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31857 Acknowledgments—We acknowledge S. Schiebold for assistance in protein expression, purification, and crystallization and S. Anderson and R. Sanishvili for their assistance with x-ray data collection. We also thank S. Bulfer and S. Horowitz for reading the manuscript and providing useful comments. This work utilized the Protein Structure Facility of the Michigan Diabetes Research and Training Center, Uni- versity of Michigan, supported by National Institutes of Health Grant DK020572, NIDDK. Use of the Advanced Photon Source was sup- ported by the United States Department of Energy, Basic Energy Sci- ences, Office of Science, under Contract DE-AC02-06CH11357. GM/CA CAT has been funded in whole or in part by National Institutes of Health NCI Grant Y1-CO-1020 and NIGMS Grant Y1-GM-1104. Use of the LS-CAT Sector 21 was supported by Michi- gan Economic Development Corporation and the Michigan Technol- ogy Tri-Corridor Grant 085P1000817 for the support of this research program. REFERENCES 1. Huang, J., and Berger, S. L. (2008) Curr. Opin. Genet. Dev. 18, 152–158 2. Morgunkova, A., and Barlev, N. A. (2006) Cell Cycle 5, 1308–1312 3. Yang, X. D., Lamb, A., and Chen, L. F. (2009) Epigenetics 4, 429–433 4. Taverna, S. D., Li, H., Ruthenburg, A. J., Allis, C. D., and Patel, D. J. (2007) Nat. Struct. Mol. Biol. 14, 1025–1040 5. Collins, R. E., Tachibana, M., Tamaru, H., Smith, K. M., Jia, D., Zhang, X., Selker, E. U., Shinkai, Y., and Cheng, X. (2005) J. Biol. Chem. 280, 5563–5570 6. Zhang, X., Yang, Z., Khan, S. I., Horton, J. R., Tamaru, H., Selker, E. U., and Cheng, X. (2003) Mol. Cell 12, 177–185 7. Couture, J. F., Collazo, E., Brunzelle, J. S., and Trievel, R. C. (2005) Genes Dev. 19, 1455–1465 8. Couture, J. F., Dirk, L. M., Brunzelle, J. S., Houtz, R. L., and Trievel, R. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 20659–20664 9. Qian, C., Wang, X., Manzur, K., Sachchidanand, Farooq, A., Zeng, L., Wang, R., and Zhou, M. M. (2006) J. Mol. Biol. 359, 86–96 10. Trievel, R. C., Flynn, E. M., Houtz, R. L., and Hurley, J. H. (2003) Nat. Struct. Biol. 10, 545–552 11. Xiao, B., Jing, C., Wilson, J. R., Walker, P. A., Vasisht, N., Kelly, G., Howell, S., Taylor, I. A., Blackburn, G. M., and Gamblin, S. J. (2003) Nature 421, 652–656 12. Chuikov, S., Kurash, J. K., Wilson, J. R., Xiao, B., Justin, N., Ivanov, G. S., McKinney, K., Tempst, P., Prives, C., Gamblin, S. J., Barlev, N. A., and Reinberg, D. (2004) Nature 432, 353–360 13. Ea, C. K., and Baltimore, D. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 18972–18977 14. Este`ve, P. O., Chin, H. G., Benner, J., Feehery, G. R., Samaranayake, M., Horwitz, G. A., Jacobsen, S. E., and Pradhan, S. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 5076–5081 15. Kouskouti, A., Scheer, E., Staub, A., Tora, L., and Talianidis, I. (2004) Mol. Cell 14, 175–182 16. Masatsugu, T., and Yamamoto, K. (2009) Biochem. Biophys. Res. Commun. 381, 22–26 17. Munro, S., Khaire, N., Inche, A., Carr, S., and La Thangue, N. B. (2010) Oncogene 29, 2357–2367 18. Pagans, S., Kauder, S. E., Kaehlcke, K., Sakane, N., Schroeder, S., Dorm- eyer, W., Trievel, R. C., Verdin, E., Schnolzer, M., and Ott, M. (2010) Cell Host Microbe 7, 234–244 19. Subramanian, K., Jia, D., Kapoor-Vazirani, P., Powell, D. R., Collins, R. E., Sharma, D., Peng, J., Cheng, X., and Vertino, P. M. (2008) Mol. Cell 30, 336–347 20. Wang, J., Hevi, S., Kurash, J. K., Lei, H., Gay, F., Bajko, J., Su, H., Sun, W., Chang, H., Xu, G., Gaudet, F., Li, E., and Chen, T. (2009) Nat. Genet. 41, 125–129 21. Yang, X. D., Huang, B., Li, M., Lamb, A., Kelleher, N. L., and Chen, L. F. (2009) EMBO J. 28, 1055–1066 22. Guo, H. B., and Guo, H. (2007) Proc. Natl. Acad. Sci. U.S.A. 104, 8797–8802 23. Hu, P., Wang, S., and Zhang, Y. (2008) J. Am. Chem. Soc. 130, 3806–3813 24. Hu, P., and Zhang, Y. (2006) J. Am. Chem. Soc. 128, 1272–1278 25. Zhang, X., and Bruice, T. C. (2007) Biochemistry 46, 14838–14844 26. Zhang, X., and Bruice, T. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 5728–5732 27. Couture, J. F., Collazo, E., Hauk, G., and Trievel, R. C. (2006) Nat. Struct. Mol. Biol. 13, 140–146 28. Trievel, R. C., Beach, B. M., Dirk, L. M., Houtz, R. L., and Hurley, J. H. (2002) Cell 111, 91–103 29. Kapust, R. B., To¨zse´r, J., Fox, J. D., Anderson, D. E., Cherry, S., Copeland, T. D., and Waugh, D. S. (2001) Protein Eng. 14, 993–1000 30. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307–326 31. Vagin, A., and Teplyakov, A. (2000) Acta Crystallogr. D Biol. Crystallogr. 56, 1622–1624 32. Emsley, P., and Cowtan, K. (2004) Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 33. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Acta Crystallogr. D Biol. Crystallogr. 53, 240–255 34. Davis, I. W., Leaver-Fay, A., Chen, V. B., Block, J. N., Kapral, G. J., Wang, X., Murray, L. W., Arendall, W. B., 3rd, Snoeyink, J., Richardson, J. S., and Richardson, D. C. (2007) Nucleic Acids Res. 35, W375–W383 35. Bru¨nger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. D Biol. Crystallogr. 54, 905–921 36. Collazo, E., Couture, J. F., Bulfer, S., and Trievel, R. C. (2005) Anal. Bio- chem. 342, 86–92 37. Chirpich, T. P., Zappia, V., Costilow, R. N., and Barker, H. A. (1970) J. Biol. Chem. 245, 1778–1789 38. Couture, J. F., Hauk, G., Thompson, M. J., Blackburn, G. M., and Trievel, R. C. (2006) J. Biol. Chem. 281, 19280–19287 39. Kwon, T., Chang, J. H., Kwak, E., Lee, C. W., Joachimiak, A., Kim, Y. C., Lee, J., and Cho, Y. (2003) EMBO J. 22, 292–303 40. Chandra, A., and Chowdhuri, S. (2002) J. Phys. Chem. B 106, 6779–6783 41. Guardia, E., Marti, J., Garcia-Tarres, L., and Laria, D. (2005) J. Mol. Liq. 117, 63–67 42. Dillon, S. C., Zhang, X., Trievel, R. C., and Cheng, X. (2005) Genome Biol. 6, 227 43. Zhang, X., Tamaru, H., Khan, S. I., Horton, J. R., Keefe, L. J., Selker, E. U., and Cheng, X. (2002) Cell 111, 117–127 Lysine Methylation by SET7/9 Mutants 31858 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010
3M58
SET7/9 Y245A in complex with TAF10-K189me1 peptide and AdoHcy
SET7/9 Catalytic Mutants Reveal the Role of Active Site Water Molecules in Lysine Multiple Methylation*□ S Received for publication,February 15, 2010, and in revised form, July 14, 2010 Published, JBC Papers in Press,August 1, 2010, DOI 10.1074/jbc.M110.114587 Paul A. Del Rizzo‡, Jean-Franc¸ois Couture‡§1, Lynnette M. A. Dirk¶, Bethany S. Strunk‡, Marijo S. Roiko‡, Joseph S. Brunzelle, Robert L. Houtz¶, and Raymond C. Trievel‡2 From the ‡Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109, the §Ottawa Institute of Systems Biology, Department of Biochemistry, Microbiology, and Immunology, University of Ottawa, Ottawa, Ontario K1H 8M5, Canada, the ¶Department of Horticulture, Plant Physiology/Biochemistry/Molecular Biology Program, University of Kentucky, Lexington, Kentucky 40546, and the Department of Molecular Pharmacology and Biological Chemistry, Northwestern University Center for Synchrotron Research, Life Sciences Collaborative Access Team, Argonne, Illinois 60439 SET domain lysine methyltransferases (KMTs) methylate specific lysine residues in histone and non-histone substrates. These enzymes also display product specificity by catalyzing dis- tinct degrees of methylation of the lysine -amino group. To elucidate the molecular mechanism underlying this specificity, we have characterized the Y245A and Y305F mutants of the human KMT SET7/9 (also known as KMT7) that alter its prod- uct specificity from a monomethyltransferase to a di- and a tri- methyltransferase, respectively. Crystal structures of these mutants in complex with peptides bearing unmodified, mono-, di-, and trimethylated lysines illustrate the roles of active site water molecules in aligning the lysine -amino group for methyl transfer with S-adenosylmethionine. Displacement or dissocia- tion of these solvent molecules enlarges the diameter of the active site, accommodating the increasing size of the methylated -amino group during successive methyl transfer reactions. Together, these results furnish new insights into the roles of active site water molecules in modulating lysine multiple meth- ylation by SET domain KMTs and provide the first molecular snapshots of the mono-, di-, and trimethyl transfer reactions catalyzed by these enzymes. SET domain enzymes represent a family of S-adenosylmethi- onine (AdoMet)3-dependent methyltransferases that catalyze the site-specific methylation of protein lysyl residues in a host of proteins, including histones, transcription factors, chroma- tin-modifying enzymes, ribosomal subunits, and other sub- strates (1–3). In many instances, these modifications serve to recruit effector proteins that recognize methyl-lysyl residues in a sequence-dependent fashion (4). In addition, SET domain KMTs exhibit product specificity, defined as their ability to cat- alyze mono-, di-, or trimethylation of the lysine -amino group. This specificity is biologically relevant because many methyl- lysine-binding proteins can discriminate among different degrees of lysine methylation (4). Thus, both the site and degree of lysine methylation are critical to recognition by effector proteins. Structural and functional studies have identified a Phe/Tyr switch in the active site of SET domain KMTs that governs their respective product specificities (5, 6). According to this model, KMTs that possess a tyrosine in the Phe/Tyr switch site are limited to catalyzing lysine monomethylation, whereas en- zymes that possess a phenylalanine or another hydrophobic residue in this position display di- or trimethyltransferase activ- ity. Mutational analysis of various SET domain KMTs, includ- ing DIM-5 (KMT1) (6), SET7/9 (6), G9A (KMT1C) (5), and SET8 (also known as PR-SET7 and KMT5A) (7, 8), has demon- strated that substitutions in the Phe/Tyr switch result in pre- dictable changes in product specificity. Several models have been proposed to explain the mechanism by which the Phe/Tyr switch site governs this specificity, including variations in the diameter of the active site due to the size of Phe/Tyr switch residue and steric hindrance by the tyrosine hydroxyl group (6, 9–11). However, our recent studies of the Phe/Tyr switch mutant Y334F in the human histone H4 Lys-20 (H4K20) meth- yltransferase SET8 indicate that the Phe/Tyr switch regulates product specificity via a more subtle mechanism (8). Specifi- cally, the switch modulates the binding of an active site water molecule that in turn regulates the transition from mono- methylation to multiple methylation. Among the KMTs that have been structurally characterized, SET7/9 has emerged as an archetypal model for studying the catalytic mechanism and product specificity of the SET domain family. Although initially isolated as a histone H3 Lys-4 (H3K4)-specific methyltransferase, this KMT has been shown to regulate the functions of numerous non-histone substrates through site-specific methylation (12–21). Early structural and functional studies of SET7/9 identified two active site mutants, * Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth Grant R01 GM073839 (to R. C. T.) and National Institutes of Health Admin- istrative Supplement GM073839-04S1 (to R. C. T.) funded through the American Recovery and Reinvestment Act. This work was also supported by Department of Energy Grant DE-FG02-92ER20075 (to R. L. H.). □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Table 1. The atomic coordinates and structure factors (codes 3M53, 3M54, 3M55, 3M56, 3M57,3M58,3M59,and3M5A)havebeendepositedintheProteinDataBank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). 1 Supported by a Canadian Institutes of Health Research postdoctoral fellowship. 2 To whom correspondence should be addressed: Dept. of Biological Chem- istry, University of Michigan Medical School, 1150 West Medical Center Dr., 5301 MSRB III, Ann Arbor, MI 48109. Tel.: 734-647-0889; Fax: 734-763-4581; E-mail: rtrievel@umich.edu. 3 The abbreviations used are: AdoMet, S-adenosylmethionine; AdoHcy, S-ad- enosylhomocysteine; KMT, lysine methyltransferase; ITC, isothermal titra- tion calorimetry; CH–O, carbon-oxygen hydrogen bond; BisTris, 2-[bis(2- hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 41, pp. 31849–31858, October 8, 2010 © 2010 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31849 Y245A and Y305F, which change its product specificity. The Phe/Tyr switch mutant Y305F alters SET7/9 product specificity from a mono- to dimethyltransferase (6), whereas the Y245A substitution converts the enzyme into a trimethyltransferase with weak monomethyltransferase activity (11). These mutants have been the subjects of numerous molecular modeling simu- lations that have led to various models to explain their distinct product specificities (22–26). However, the lack of structural data for the SET7/9 Y245A and Y305F mutants in complex with cognate methylated peptides has hindered our understanding of the mechanisms that define the respective product specific- ities of these mutants. Moreover, these structures would yield a framework for visualizing the mono-, di-, and trimethylation reactions catalyzed by SET domain KMTs. To gain insight into the molecular basis of their product specificities, we have determined high resolution crystal structures of the SET7/9 Y245A and Y305F mutants in com- plex with peptides of the TATA box-binding protein-associ- ated factor TAF10 bearing the Lys-189 methylation site in unmodified (K189), monomethylated (K189me1), dimethyl- ated (K189me2), and trimethylated (K189me3) states. The structures and accompanying biochemical data support a model whereby changes in the occupancy or position of water molecules in the active site are critical in establishing the prod- uct specificities of the SET7/9 Y245A and Y305F mutants. Together, our results provide new insights into the mechanisms that govern SET domain product specificity and provide step- wise snapshots of the lysine mono-, di-, and trimethyl transfer reactions catalyzed by KMTs. EXPERIMENTAL PROCEDURES Cloning, Expression, and Purification of the SET7/9 Mutants— The Y245A and Y305F mutants were introduced into the pHIS2 SET7/9 expression vector encoding residues 110–366 (27) via QuikChange site-directed mutagenesis (Stratagene) and were verified by dideoxy DNA sequencing. The plasmids encoding wild type (WT) SET7/9 and the Y245A and Y305F mutants were transformed into Rosetta2 DE3 cells (Novagen) and were expressed as described previously (27, 28). In the course of characterizing WT SET7/9, we observed that the enzyme co- purified with AdoMet or another contaminant that resulted in technical difficulties in the isothermal titration calorimetry (ITC) experiments and co-crystallization trials with the TAF10 peptides. To overcome this problem, a denaturation and refold- ing step was inserted in the purification scheme. The denatur- ation and refolding protocol involved adding 6 M guanidine HCl, 50 mM sodium phosphate, pH 7.5, and 500 mM NaCl to the protein while it was immobilized on a nickel-Sepharose column (GE Healthcare). The column was washed with this buffer, fol- lowed by a solution of 6 M urea, 50 mM sodium phosphate, pH 7.5, 150 mM NaCl, and 10 mM 2-mercaptoethanol to remove the cofactor from the denatured enzyme. A reverse gradient from 6 to 0 M urea was then performed in the same buffer to refold the protein, which was subsequently eluted from the column using a linear gradient of 0–500 mM imidazole in 50 mM sodium phosphate, pH 7.5, 500 NaCl, and 10 mM 2-mercaptoethanol. The refolded protein was digested with tobacco etch virus pro- tease (29) during dialysis against 20 mM Tris, pH 7.5, 150 mM NaCl, and 5 mM 2-mercaptoethanol and then purified using a Superdex 200 gel filtration column (GE Healthcare). Protein concentration was determined by its absorbance at 280 nm. Synthetic Peptides—The TAF10 peptides bearing K189, K189me1, K189me2, and K189me3 (sequence, acetyl- SKSK189DRKYTL-amide) and a biotinylated TAF10 peptide (sequence, acetyl-SKSK189DRKYTLT(K-EZLink-S-S-biotin)- amide) were synthesized and purified by New England Peptide, Inc. Peptide concentrations were measured using the absorb- ance of their tyrosine residue at 274 nm. Crystallization and Data Collection—Crystals were pro- duced by hanging drop vapor diffusion by mixing the crystalli- zation solution in a 1:1 ratio with 10–20 mg/ml SET7/9, 1 mM S-adenosylhomocysteine (AdoHcy), and 1.0–3.0 mM unmodi- fied or methylated TAF10-K189 peptide in 20 mM Tris, pH 8.0, 100 mM NaCl, and 2 mM tris(2-carboxyethyl)phosphine. Crys- tals were obtained at 20 °C in either 1.8–2.0 M (NH4)2SO4, with 0.1 M BisTris, pH 6.2–6.6, or in 0.95–1.1 M sodium citrate with 100 mM imidazole pH 8.0–8.4. In both crystallization condi- tions, the final pH values were between pH 8.0 and 9.0. Crystals in the (NH4)2SO4 condition were typically flash-frozen in the mother liquor containing 25–30% glycerol, and the crystals in the citrate condition were frozen in 1.6 M sodium citrate. Data were collected at the Advanced Photon Source beamlines 21-IDG (LS-CAT) and 23-IDD (GM/CA-CAT). Images were indexed, integrated, and scaled using HKL2000 (30). Structures of the mutants were solved by molecular replacement using MOLREP (31) with the coordinates of a previously reported SET7/9 ternary complex used as the search model (Protein Data Bank code 2F69). Successive rounds of model building and refinement were carried out using Coot (32) and REFMAC (33), respectively. The geometry of the models were verified by Mol- Probity (34). Simulated annealing omit maps were calculated using CNS (35) with the peptide and cofactor removed to elim- inate model bias in the active site. Structural figures were ren- dered using PyMOL (Schro¨dinger, LLC.). Fluorescent Methyltransferase Assay—A coupled fluorescent methyltransferase assay was used to measure the kinetic parameters of WT SET7/9 and the Y245A and Y305F mutants as reported previously, with the exception that 50–150 nM enzyme, 100 M AdoMet, and varying concentrations of TAF10 peptide substrate were used (27, 36). Assays were performed in triplicate, and a homocysteine calibration curve was used to calculate the initial velocities. Kinetic parameters were calcu- lated by plotting the velocities versus peptide concentration and by fitting the Michaelis-Menten equation to the data via non- linear regression using Prism 5.0 (GraphPad). In cases where the Km value was beyond the measurable range of the assay, the kcat/Km value was determined as described previously (7). Calorimetry Experiments—ITC was performed at 20 °C using a MicroCal VP-ITC calorimeter (GE Healthcare) with 0.12 mM protein and 4 mM AdoHcy in 20 mM sodium phosphate, pH 7, and 100 mM NaCl with 1.5 mM peptide as the injectant. Data were processed, and equilibrium dissociation constants (KD) and curve fitting errors were calculated from the binding iso- therms using Origin 7.0 (OriginLab Corp.). WT SET7/9 and the Y245A and Y305F mutants displayed ligand:protein binding stoichiometries (N values) between 0.8 and 1.0, demonstrating Lysine Methylation by SET7/9 Mutants 31850 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 that WT SET7/9 and its mutants were properly refolded due to their ability to bind peptides in an 1:1 molar ratio. TLC Product Analysis—Methyltransferase assays were per- formed in triplicate at 37 °C with the biotinylated TAF10 pep- tide (0.5 mM) for 30 min in the presence of WT SET7/9 (3 pmol), Y305F mutant (6 pmol), or the Y245A mutant (100 pmol). Assays contained 100 mM HEPES, pH 8.0, 50 mM NaCl, 1.0 mM [methyl-3H]AdoMet (diluted to a specific activity of 0.2 Ci/nmol with purified AdoMet (37)), 15 M Sulfolobus solfa- taricus AdoHcy hydrolase (36), and 2 units of adenosine deami- nase (Roche Applied Science) in a final volume of 20 l. The reactions were terminated by addition of an equal volume of 200 mM MES, pH 5.1, prior to addition of a 2-fold molar excess of immobilized avidin resin (UltraLink; Pierce). Biotinylated peptides were allowed to bind at room temperature for 30 min, and the resin was then collected by centrifugation (9000  g). The resin was washed three times with 300 mM NaCl, and the peptide was eluted overnight from the avidin resin by cleavage of the disulfide bond in the linker of the peptide using 10 mM tris(2-carboxyethyl)phosphine (Thermo Scientific). The resin was incubated with additional 10 mM tris(2-carboxyethyl)phos- phine the following day until the radiolabel was essentially removed from the resin. The recovered peptides were hydro- lyzed by incubation with 6 M HCl at 110 °C for 24 h. Subsequent steps in measuring the radiolabel incorporated into the mono-, di-, and trimethyl-lysine products were performed as reported previously (8). RESULTS Functional Analysis of the SET7/9 Y305F Mutant—Prior studies of SET7/9 by Zhang et al. (6) reported that mutation of the Phe/Tyr switch residue Tyr-305 to a phenylalanine alters its product specificity from a mono- to dimethyltransferase. We verified these findings by demon- strating that WT SET7/9 mono- methylated the TAF10-K189 pep- tide, whereas the Y305F mutant mono- and dimethylated this sub- strate, as demonstrated by mass spectrometry (data not shown). We next examined whether the Y305F substitution altered the affinity of SET7/9 for the TAF10-K189 peptides using ITC (Fig. 1). A comparison of the KD values revealed that SET7/9 Y305F bound the TAF10-K189 and TAF10-K189me1 peptides 4- and 6-fold more tightly, respec- tively, than the WT enzyme, whereas this mutant displayed a substantially diminished affinity for the TAF10-K189me2 pep- tide (Table 1). Although the WT enzyme and the Y305F mutant exhibited discernable differences in their affinities for the unmodified and monomethylated peptides, these variations are modest and cannot account for their distinct product specific- ities, suggesting that a kinetic effect during methylation may be responsible. To investigate this possibility, we characterized the kinetic parameters of WT SET7/9 and the Y305F mutant using the TAF10 peptides as substrates. Both enzymes methylated the unmodified peptide with comparable kcat and Km values (Table 2). In analyzing the kinetic parameters for the methylation of the monomethylated peptide by SET7/9 Y305F, we found that this substrate displayed an elevated Km value that was beyond the measurable range of the assay due to its limited solubility. In this case, we measured the catalytic efficiency (kcat/Km) for the methylation of this peptide and found that it was methylated 15-fold less efficiently than the unmodified peptide by SET7/9 Y305F. Given the fact that the Y305F mutant exhibited a higher binding affinity for the TAF10-K189me1 peptide than the WT enzyme (Table 1), the kinetic data suggest that a step in the reaction pathway following substrate binding limits the catalytic efficiency of this mutant. We next examined whether the Y305F mutant dimethylated the TAF10-K189 peptide via a processive or a distributive mechanism. In a processive mechanism, the methyl-lysine substrate would remain bound to the enzyme during successive methyl transfer reactions; thus, the concentration of an inter- mediate, such as monomethyl-lysine, cannot exceed the en- zyme concentration during the assay. In a distributive mecha- nism, the intermediates are released into solution where they accumulate prior to the next round of methylation, resulting in an intermediate concentration that is greater than that of the enzyme. Using a radiometric TLC assay and a biotinylated TAF10 peptide, we quantified the amounts of monomethylated products generated by the WT SET7/9 and the Y305F mutant FIGURE 1. ITC analysis of WT SET7/9 and the Y245A and Y305F mutants. Representative titrations and binding isotherms for the calorimetry experiments are shown for the unmodified TAF10-K189 peptide titrated into WT SET7/9 (left plot), SET7/9 Y305F (middle plot), and SET7/9 Y245A (right plot). In each panel, the ITC titration experiment (upper panel) is shown with the binding isotherms (lower panel) fit to a single binding site model. TABLE 1 Analysis of the binding affinity of WT SET7/9 and its catalytic mutants for unmodified and methylated TAF10 peptides SET7/9 TAF10 peptide KD a M WT K189 4.9  0.20 WT K189me1 4.0  0.36 Y305F K189 1.3  0.10 Y305F K189me1 0.62  0.065 Y305Fb K189me2 70 Y245A K189 4.0  0.25 Y245A K189me1 3.3  0.10 Y245A K189me2 5.8  0.22 Y245A K189me3 11  0.28 a Curve fitting errors were calculated from the binding isotherms. b An estimate of the affinity is reported due to weak peptide binding. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31851 (Table 3). The data demonstrate that comparable amounts of monomethyl-lysine were generated when the quantity of enzyme usedistakenintoaccount,inagreementwiththeirsimilar turnover numbers for the TAF10-K189 peptide (Table 2). The Y305F mutant also produced small but measurable quantities of radiolabeled dimethyl-lysine product that were substantially smaller than the amount of monomethyl-lysine generated. Therefore, the TLC data are consistent with a distributive mechanism for dimethylation by the Y305F mutant because the amount of monomethyl-lysine produced exceeded the quantity of enzyme used in the assay. Structures of WT SET7/9 and the Y305F Mutant in Complex with Unmodified and Methylated TAF10 Peptides—To deter- mine the mechanism by which the Y305F substitution alters the product specificity of SET7/9, we determined the crystal struc- tures of this mutant bound to AdoHcy and TAF10-K189, TAF10-K189me1, and TAF10-K189me2 peptides and com- pared these to the structures of the WT SET7/9AdoHcy TAF10-K189 complex (supplemental Table 1). The structures of these complexes were determined to 1.85 Å or higher reso- lution, permitting unambiguous modeling of the K189 side chains in the active site of the enzyme based on simulated annealing omit maps (Fig. 2). The ternary complexes of the WT and the Y305F mutant superimpose with overall root mean square differences of less than 0.3 Å for all aligned atoms, indi- cating that neither the Y305F mutation nor the binding of the various TAF10-K189 peptides results in substantial changes in its overall structure. An inspection of the active sites of the SET7/9 WT and Y305F complexes illustrates the binding modes of the unmod- ified and methylated forms of K189 in the TAF10 peptides (Fig. 2, A–D). The K189 side chain binds in an extended all trans conformation in a deep pocket, termed the lysine binding chan- nel, that is composed of Tyr-245, Thr-266, Leu-267, Ser-268, Tyr-305, Tyr-335, and Tyr-337 (not shown for clarity) (Fig. 2A). These residues interact with the aliphatic portion of the K189 side chain primarily through van der Waals contacts. The lysine binding channel connects to the AdoMet-binding site on the opposite face of the catalytic domain via an oxygen-lined methyl transfer pore (38). During catalysis, the methyl group of the cofactor is positioned within the methyl transfer pore for the SN2 reaction with the -amino group of the lysine or methyl-lysine substrate (see below). To lower the activation barrier for this reaction, the lysine -amine nucleophile is aligned for methyl transfer through a hydrogen bond network within the active site. In the WT enzyme, the K189 -amino group hydrogen bonds to the hydroxyl group of Tyr-245 as well as to two water molecules (Fig. 2A). One of the water molecules (termed water 1), is coor- dinated in a solvent pocket, through hydrogen bonds to the carbonyl oxygens of Gly-292 and Ala-295 and to the hydroxyl group of the Phe/Tyr switch residue Tyr-305. This solvent pocket is structurally conserved in SET domain KMTs and has an important role in defining product specificity through the adjacent Phe/Tyr switch residue, as shown in our prior studies of the human H4K20 methyltransferase SET8 (8). The other water molecule is bound within the methyl transfer pore between the lysine substrate and the thioether sulfur atom of AdoHcy through hydrogen bonds to Tyr-245, Asn-265, and His-293 in SET7/9 and the TAF10-K189 -amino group. This water is not observed in other structures of SET7/9 ternary complexes and may represent the approximate position that the AdoMet methyl group occupies in the methyl transfer pore in the Michaelis complex. In structures of the Y305F ternary complexes, the K189, K189me1, and K189me2 side chains also adopt extended trans side chain geometries within the lysine binding channel that are stabilized via hydrogen bonding to Tyr-245 in the enzyme (Fig. 2, B–D). The orientations of the K189me1 and K189me2 side chains are further maintained through carbon-oxygen (CH–O) hydrogen bonding between the methyl groups and oxygen atoms within the vicinity of the methyl transfer pore, as reported previously in other SET domain KMT structures (8, 10, 38). A superimposition of the SET7/9 WT and Y305F com- plexes underscores the similarity of the lysyl binding conforma- tions (Fig. 2E). However, there are notable differences in the hydrogen bond patterns and occupancy of water 1 within the solvent pocket in the Y305F mutant compared with the WT enzyme. Specifically, the Y305F substitution results in the loss of one hydrogen bond to water 1 in the structures of the TAF10- K189 and TAF10-K189me1 complexes (Fig. 2, B and C). In con- trast, water 1 is absent in TAF10-K189me2 complex, and the vacated solvent pocket is occupied by one of the methyl groups TABLE 2 Kinetic parameters of WT SET7/9 and the Y305F and Y245A mutants Enzyme TAF10 peptide substrate Km a kcat a kcat/Km a M min1 M1 min1  103 WT K189 160  17 17  0.62 110  17 Y305F K189 88  5.0 17  0.30 190  11 Y305Fb K189me1 11  0.50 Y245A K189 200  35 0.53  0.04 2.6  0.47 Y245A K189me1 210  23 5.9  0.23 28  3.3 Y245A K189me2 400  29 6.5  0.16 15  1.2 a Curve fitting errors were calculated from the hyperbolic fits for the Michaelis-Menten equation. b Km and kcat were not determined because the Km value was beyond the measurable range; therefore, the kcat/Km value is reported. TABLE 3 Product analysis of WT SET7/9 and the Y305F and Y245A mutants Enzyme Quantity of enzyme Measured product Amount of product formeda nmol nmol WT 0.003 Kme1 0.65  0.07 Y305F 0.006 Kme1 1.5  0.49 Kme2 0.033  0.009 Y245A 0.100 Kme1 0.80  0.22 Kme2 0.39  0.021 Kme3 0.076  0.019 a Standard deviation was calculated from triplicate measurements. Lysine Methylation by SET7/9 Mutants 31852 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 of the dimethyl -amine (Fig. 2D). This methyl group forms a 3.5-Å CH–O hydrogen bond to the carbonyl oxygen of Ala-295, further stabilizing the binding of the dimethyl-lysine side chain. A homologous dimethyl-lysine-binding mode and CH–O hydrogen bond was observed in our prior structural studies of the SET8 Y334F Phe/Tyr switch mutant that confers an analo- gous change in product specificity from a mono- to a dimeth- yltransferase (8). A structural alignment of the active sites of the SET7/9 Y305F and SET8 Y334F mutants bound to cognate dimethylated peptides illustrates that the coordinates of the dimethyl-lysyl side chains are virtually superimposable, with one methyl group oriented toward the methyltransfer pore and the second positioned within the vacant solvent pocket (Fig. 2F). Taken together, the structures of the SET7/9 Y305F com- plexes and the similarities in the dimethyl-lysine conforma- tions in the SET7/9 Y305F and SET8 Y334F mutants imply that the Phe/Tyr switch governs product specificity through a con- served mechanism whereby it indirectly influences the binding modes of the methyl-lysine side chain by modulating the affin- ity of the water molecule (water 1) bound in the solvent pocket. Biochemical Characterization of the SET7/9 Y245A Mutant—Previ- ous studies by Xiao et al. (11) reported that the Y245A mutation yields an unusual change in the product specificity of SET7/9, converting the enzyme to a trimeth- yltransferase with weak monometh- yltransferase activity. We deter- mined that the SET7/9 Y245A could mono-, di-, and trimethylate the TAF10-K189 peptide by mass spec- trometry (data not shown) and TLC (Table 3), confirming the earlier studies of Xiao et al. (11). ITC analysis revealed that the Y245A mutant displayed comparable KD values for the unmodified and methylated TAF10-K189 peptides (Fig. 1), although its affinity for the trimethylated peptide was modestly diminished in comparison with the other peptides (Table 1). The ITC data demonstrate that the Y245A mutant bound the unmodified, mono-, and dimethylated sub- strates with equivalent affinities, suggesting that a kinetic effect or a structural alteration in the active site may be responsible for its diminished activity toward un- modified substrates. To gain further insight into its peculiar product specificity, we characterized the kinetic properties of the SET7/9 Y245A mutant. Steady state analysis demonstrated that this mutant displayed similar Km values for the unmodified, mono- and dimethylated TAF10 peptides (Table 2). However, the turnover number for the TAF10-K189 peptide was diminished over 10-fold versus the methylated peptides and was reduced 30-fold versus the WT enzyme, in agreement with the weak monomethyltransferase activity reported by Xiao et al. (11). In addition, we investigated whether this mutant catalyzes lysine trimethylation via a pro- cessive or distributive mechanism as described for SET7/9 Y305F. The TLC data illustrate that the mono- and dimethyl- lysine intermediates accumulated at quantities greater than that of the enzyme used in the assay, indicating that SET7/9 Y245A obeys a distributive mechanism, analogous to the Y305F mutant (Table 3). Structures of SET7/9 Y245A Bound to Unmodified and Meth- ylated TAF10 Peptides—To elucidate the mechanism underly- ing its unusual product specificity, we determined the crystal structures of SET7/9 Y245A in complex with AdoHcy and unmodified, mono-, di-, and trimethylated TAF10 peptides (supplemental Table 1). These complexes superimpose with the structure of the WT SET7/9AdoHcyTAF10-K189 com- FIGURE 2. Crystal structures of WT SET7/9 and the Y305F mutant in complex with AdoHcy and unmodi- fied and methylated TAF10 peptides. A, active site of WT SET7/9 bound to AdoHcy and TAF10-K189 and the active site of SET7/9 Y305F bound to AdoHcy and TAF10-K189 (B), TAF10-K189me1 (C), and TAF10-K189me2 peptides (D). A–D, the K189 residue in the TAF10 peptide is depicted with yellow carbon atoms with the corresponding Fo – Fc simulated-annealing omit maps (contoured at 2.5 ) shown. AdoHcy is colored with green carbon atoms, and SET7/9 carbon atoms are colored gray, with the exception of Tyr-305/Y305F (magenta) and Tyr-245 (cyan). Gly-264 to Ser-268 are shown, and the main chain atoms of Gly-292 to Ala-295 are depicted for clarity. Red dashed lines indicate conventional hydrogen bonds, and blue dashed lines indicate CH–O hydrogen bonds. Cyan dashed lines indicate weak CH–O hydrogen bonds between 3.5 and 3.7 Å in length. The water molecule in the solvent binding pocket is numbered 1. E, overlay of the active sites of the WT enzyme (light blue carbon atoms) and the Y305F mutant (green, yellow, and pink carbon atoms correspond to K189, K189me1, and K189me2, respectively). Red dashed lines indicate coordination of the waters in the struc- turesoftheY305Fmutant,andthelightbluedashedlineindicateshydrogenbondingthatonlyoccursintheWT enzyme. F, superimposition of the active sites of the SET7/9 Y305FAdoHcyTAF10-K189me2 and SET8 Y334FAdoHcyH4K20me2 (Protein Data Bank code 3F9X) complexes shown with green and yellow carbon atoms, respectively. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31853 plex with root mean squared differences of less than 0.4 Å for all aligned atoms, indicating that the Y245A mutant does not per- turb the overall structure of the enzyme. Simulated annealing omit maps illustrate that K189 side chains are bound within the lysine binding channel through hydrogen bonds and van der Waals contacts (Fig. 3, A–D), although the interactions and binding modes are distinct from those in the complexes of WT SET7/9 and the Y305F mutant (Fig. 2, A–D). In the unmodified TAF10 peptide complex, the K189 -amino group forms a weak hydrogen bond with the carbonyl oxygen of Gly-264 (Fig. 3A), whereas the -amino groups of K189me1 and K189me2 hydro- gen bond to the hydroxyl group of Tyr-305 in the mono- and dimethylated peptide complexes (Fig. 3, B and C). The confor- mations of the K189me1 and K189me2 side chains are further stabilized by water-mediated hydrogen bonding and through CH–O hydrogen bonding to their methyl groups. In the TAF10-K189me3 peptide complex, the trimethyl-lysine side chain is coordinated exclusively through direct and water-me- diated CH–O hydrogen bonds to its methyl groups because the quaternary -ammonium cation cannot engage in hydrogen bonding (Fig. 3D). A structural alignment of the four SET7/9 Y245A complexes illustrates distinct binding modes for the unmodified versus the methylated K189 side chains, highlighting the selectivity of this mutant for methylated substrates. The side chains of K189me1, K189me2, and K189me3 roughly overlay with their respective -amino groups superimposed and adopt slightly kinked con- formations (Fig. 3E), as opposed to the extended trans geome- try of the unmodified and methylated lysines in the complexes of the WT enzyme and the Y305F mutant (Fig. 2, A–D). Con- versely, the unmodified K189 side chain does not superimpose with its methylated counterparts and is oriented in an alterna- tive configuration due to its hydrogen bonding to Gly-264 (Fig. 3, A and E). An overlay of the structures of the WT enzyme and Y245A mutant bound to the unmodified TAF10 peptide illus- trates that the side chains of K189 do not superimpose and that the K189 -amino group appears to be misaligned with AdoHcy in the Y245A complex (Fig. 3F). This suboptimal alignment may explain the diminished kcat value of SET7/9 Y245A mutant toward substrates with unmodified lysines (Table 2). A comparison of the structures of the SET7/9 Y245A and Y305F complexes yields a molecular explanation for the differ- ent product specificities of these two mutants. In the SET7/9 Y305F complexes, Tyr-245 aligns the K189 -amino group for methyl transfer through hydrogen bonding to its hydroxyl group (Fig. 2, B–D). Conversely, in the Y245A mutant, the K189me1 and K189me2 -amino groups are oriented through hydrogen bonding to Tyr-305 (Fig. 3, B and C). These distinct hydrogen bond patterns impart differences in the conforma- tions of the lysyl side chains due to the relative orientations of Tyr-245 and Tyr-305 in the lysine binding channel. Specifically, the kinked conformation adopted by the K189me1 and K189me2 side chains in the Y245A complexes (Fig. 3, B and C) may contribute to the differences in the turnover numbers of this mutant versus those of the WT enzyme and the Y305F mutant (Table 2). In addition, the dimethyl -amino group of the K189me2 side chain binds in distinct orientations in the Y245A and Y305F mutants due to their hydrogen bonding to FIGURE 3. Crystal structures of the SET7/9 Y245A mutant in complex with AdoHcy and unmodified and methylated TAF10 peptides. Active site of SET7/9 Y245A bound to AdoHcy and TAF10-K189 (A), TAF10-K189me1 (B), TAF10-K189me2 (C), and TAF10-K189me3 peptides (D). A–D, Fo – Fc simulated- annealing omit maps (contoured at 2.5 ) for the unmodified and methylated K189 side chains are illustrated. The residues and hydrogen bonds in each com- plexarecoloredasinFig.2.Thewatermoleculesinthelysinebindingchannelsof the Y245A complexes are numbered 1–4, as described in the text. E, superimpo- sition of the active sites of the Y245A complexes bound to the four methylated statesofTAF10-K189(unmodified,mono-,di-,andtrimethylatedshowningreen, yellow, pink, and blue, respectively). F, overlay of WT SET7/9 (blue carbons) and SET7/9 Y245A (green carbons) in complex with TAF10-K189. Waters correspond- ing to the WT and Y245A structures are colored cyan and green, respectively. G, alignment of the active sites of the Y245A (yellow carbons) and Y305F (green carbons) mutants in complex with the TAF10-K189me2 peptide. Hydrogen bonds from the Y305F structure are shown as green dashed lines, and waters and hydrogen bonds in the Y245A structure are shown in yellow and orange, respectively. Lysine Methylation by SET7/9 Mutants 31854 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 Tyr-305 and Tyr-245, respectively (Fig. 3G). In the Y305F mutant, hydrogen bonds to the dimethyl -amino group cou- pled with steric constraints in the lysine binding channel pre- vent the K189me2 side chain from undergoing a conforma- tional change that is conducive to trimethylation (Fig. 2D), consistent with its dimethyltransferase activity. However, in the Y245A mutant, the alanine substitution enlarges the diameter of the lysine binding channel, accommodating trimethyl-lysine (Fig. 3D). In addition, the larger diameter would permit the dimethyl-lysine substrate to undergo the conformational reor- ganization necessary to align the -amino group in a productive geometry for trimethylation. A major difference in the active site of the Y245A mutant versus the other SET7/9 structures is the presence of several water molecules bound in the cavity generated by the Y245A mutation. In the structure of the Y245A mutant bound to TAF10-K189, three water molecules (waters 2–4) occupy this cavity and are arranged in a triangular geometry (Fig. 3A). In addition, water 1 shifts 1.6 Å from its position in the solvent pocket toward water 2 to which it forms a hydrogen bond (Fig. 3, A and E). The shift in water 1 was unexpected given its con- served orientation in the solvent pocket of the SET7/9 WT and Y305F complexes (Fig. 2, A–C) as well as in the structures of other SET domain KMTs (8). This displacement is presumably related to the alternative conformation of the K189 side chain whose -amino group is too distant (4.3 Å) to form a productive hydrogen bond to water 1. Conversely, in the Y245A complexes bound to TAF10-K189me1 and TAF10-K189me2, water 1 remains tightly bound in the solvent pocket through hydrogen bonds to Tyr-305 hydroxyl group, the carbonyl oxygens of Gly- 292 and Ala-295, and the K189 -amino group (Fig. 3, B, C, and E), analogous to its binding in the WT enzyme (Fig. 2A). How- ever, in the TAF10-K189me3 complex, one of the methyl groups of the trimethyl -ammonium cation is oriented into the solvent pocket (Fig. 3D), similar to the dimethyl-lysine binding mode observed in the Y305F mutant (Fig. 2D). The binding of the methyl group in the solvent pocket displaces water 1 by 3.2 Å relative to its position in the TAF10-K189me1 complex (Fig. 3E), thereby avoiding a steric clash with the trimethylated -ammonium group. Variations in the occupancy of water 2 are also seen in the different Y245A structures. Water 2 is bound in similar orientations in the active site of the unmodified and monomethylated peptide complexes but is absent in the di- and trimethylated peptide complexes due to the binding of a methyl group in this position (Fig. 3, A–E). In summary, the changes in the positions or occupancies of waters 1 and 2 correlate with the binding modes of the unmodified and methylated K189 within the active site of the Y245A mutant. Catalytic Models of Lysine Multiple Methylation by SET7/9 Y245A, and Y305F—The structures of the SET7/9 complexes reported here offer a prime opportunity to generate stepwise models for lysine mono-, di-, and trimethylation by a SET domain KMT. We modeled the AdoMet-bound Michaelis complexes by superimposing the SET7/9 product complexes with the previously reported structure of the SET7/9-AdoMet binary complex (Fig. 4) (39). The conformations of the mono- and dimethyl -amino groups in the Michaelis complexes were inferred from the coordinates of the corresponding dimethyl- and trimethyl-lysine products, respectively. In addition, we modeled the -amino group in a deprotonated state with its hydrogen atoms oriented toward the hydrogen bond acceptors that align the lysyl side chain for methylation. As a basis for this comparison, we first modeled the monomethylation reaction catalyzed by WT SET7/9 (Fig. 4A). In the substrate ternary complex, the lysine -amine is aligned with the methyl group and sulfonium cation of AdoMet through a hydrogen bond to the Tyr-245 hydroxyl group and water 1 in the solvent pocket. The values of the reaction distance and angle are 2.8 Å and 153°, respectively, in approximate agreement with the linear geome- try of a SN2 methyl transfer reaction calculated in other mod- eled substrate complexes (8, 10). In the product complex, the monomethyl-lysine side chain is bound in an extended confor- mation with its methyl group oriented within the methyl trans- fer pore, thereby obstructing AdoMet binding. Furthermore, water 1 remains tightly coordinated in the solvent pocket through four hydrogen bonds to Gly-292, Ala-295, Tyr-305, and the monomethyl -amino group. These interactions hinder the dissociation of water 1 and the related rearrangement of the monomethyl-lysine side chain required for a second methyl transfer reaction, explaining why the WT enzyme cannot cata- lyze di- and trimethylation. These findings concur with the FIGURE 4. Catalytic models of the methyl transfer reactions catalyzed by WT SET7/9 and the Y305F mutant. A, monomethylation of TAF10-K189 by the WT enzyme. The reaction scheme depicts the modeled substrate ternary complex (left) and the product complex (right) for the transfer of the methyl group from AdoMet (green carbon atoms) to K189 in TAF10 (yellow carbons), yielding AdoHcy and K189me1. The red arrow indicates the direction of the nucleophilic attack of the deprotonated -amino group on the AdoMet methyl group. The transferred methyl group is colored green, and the white atoms represent the hydrogens of the -amino group. Hydrogen bonds and residues in the enzyme active site are illustrated as in Fig. 2. The reaction distance and angle are labeled in red. B and C, models of the Y305F mutant for the first methyl transfer reaction with TAF10-K189 (B) and second methyl transfer reaction with TAF10-K189me1 (C). Color schemes are the same as in A. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31855 model for SET7/9 product specificity reported in previous structural and functional studies (6, 11). Similar reaction geometry is observed in the model for the monomethyl transfer reaction catalyzed by SET7/9 Y305F. Hydrogen bonds from the Tyr-245 hydroxyl group and water 1 align the lysine -amino group with the AdoMet methyl group at a distance of 2.1 Å and an angle of 160°, equivalent to those measured in the Michaelis complex of the WT enzyme (Fig. 4, A and B). In the product complex, monomethyl-lysine adopts an extended trans configuration analogous to that in the WT enzyme. For dimethylation to occur, the monomethyl-lysine must undergo a conformational change in which its methyl group is rotated out of the methyl transfer path with AdoMet. The structure of the Y305F mutant bound to the dimethylated TAF10 peptide (Fig. 2D) implies that this rearrangement occurs through the dissociation of water 1 due to the loss of the Tyr- 305 hydrogen bond in the solvent pocket. The dissociation of water 1 would enable the monomethyl-lysine side chain to adopt an alternative conformation through a rotation about its C–N bond, projecting the methyl group into the solvent pocket (Fig. 4C). This rotation reorients the methyl group out of the methyl transfer path while realigning the monomethyl - amino group for a second methylation reaction through a direct hydrogen bond to the Tyr-245 hydroxyl group and a CH–O hydrogen bond between its methyl group and Ala-295. The modeled reaction geometry for monomethyl-lysine substrate complex of the Y305F mutant (2.7 Å, 160°) is equivalent to that of the first methyl transfer reaction in SET7/9 Y305F. These geometries concur with our previous models for mono- and dimethylation catalyzed by SET8 Y334F (8), illustrating that the orientation of a methyl group into the solvent pocket is a con- served feature of SET domain KMTs that catalyze multiple methylation. In addition, we modeled the methyl transfer reactions cata- lyzed by SET7/9 Y245A (Fig. 5). In the model of the lysine sub- strate complex, the -amino group is aligned for methyl transfer by a hydrogen bond to the carbonyl oxygen of Gly-264, result- ing in a short reaction distance (2.3 Å) and a suboptimal reac- tion angle (141°) with the methyl group of AdoMet (Fig. 5A). This misalignment appears to be a direct consequence of the Y245A mutation that abolishes hydrogen bonding to the - amino group, illustrating that the suboptimal orientation of the -amine likely contributes to the diminished activity of this mutant toward unmodified substrates (11). Conversely, in the modeled monomethyl-lysine substrate complex for SET7/9 Y245A, Tyr-305 hydrogen bonds to the -amino group, aligning it for methyl transfer (Fig. 5B). In addition, CH–O hydrogen bonds to the monomethyl-lysine methyl group and the dissoci- ation of water 2 from the active site also contribute to reposition- ing the -amino group for dimethylation. Collectively, these inter- actions orient the -amine in a reaction angle of 165° that is more conducive to methyl transfer. However, the reaction distance for dimethylation is 0.6 Å longer than that in the corresponding Y305F model because Tyr-305 is positioned further from AdoMet than Tyr-245 (Figs. 4C and 5B). In the third methyl transfer reaction catalyzed by SET7/9 Y245A, the lone pair of electrons of the dimethyl-lysine - amino group acts as the nucleophile and thus cannot engage in hydrogen bonding. The structure of the trimethyl-lysine prod- uct complex (Fig. 3D) implies that the dimethyl -amine is aligned via CH–O hydrogen bonds to its methyl groups, as shown in the model of the Michaelis complex for this reaction (Fig. 5C). These CH–O hydrogen bonds restrain the orienta- tion of the -amino group and position one of the methyl groups into the solvent pocket, displacing water 1 as discussed earlier (Fig. 3, D and E). These interactions cumulatively align the -amino group and AdoMet methyl group with a reaction distance of 3.0 Å and angle of 162° (Fig. 5C). Taken together, the models of the substrate complexes for SET7/9 Y245A suggest that CH–O hydrogen bonds play an increasingly important role in aligning the methylated -amino group in successive rounds of methyl transfer. DISCUSSION The structural and functional characterization of the SET7/9 Y245A and Y305F mutants presented here yields new insights into the mechanism underlying the product specificity of SET domain KMTs. Importantly, it resolves a general paradox concerning this specificity. How does the active site constrain the motion of the lysine -amino group to align it for methyl transfer with AdoMet, while providing adequate volume to accommodate the mono-, di-, and tri- methylated lysine side chain generated during multiple methyl transfer reactions? The structures of the Y305F and Y245A mutants resolve this paradox, illustrating that alter- ations in the positions or occupancies of water molecules within their active sites generate the space required to FIGURE 5. Models for the methyl transfer reactions catalyzed by the SET7/9 Y245A mutant. Models of the Y245A mutant for the first methyl transfer reaction with TAF10-K189 (A), the second methyl transfer reaction with TAF10-K189me1 (B), and the third methyl transfer reaction with TAF10- K189me2 (C). Residues and hydrogen bonds are depicted as in Fig. 3. Lysine Methylation by SET7/9 Mutants 31856 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 accommodate the multiply methylated -amine produced during successive catalytic cycles. Minor perturbations in the side chains of certain active site residues, such as Tyr- 305, are also observed in alignments of the WT enzyme and the Y245A and Y305F complexes, although these changes are modest compared with the displacement or dissociation of the water molecules in the active site. These findings suggest that the waters function as transient place holders that facil- itate the SN2 methyl transfer reaction. During monomethy- lation, they function to constrain the movement of the lysine -amino group by mediating hydrogen bonds between the substrate and enzyme, thereby promoting the linear align- ment with the methyl group and sulfonium cation of AdoMet (Fig. 4, A and B). During di- and trimethylation, the water molecules either relocate within the lysine binding channel or dissociate from the enzyme, yielding the space required to rotate the methyl group away from the methyl transfer pore and to realign the -amine in productive geom- etry for the next methyl transfer reaction (Figs. 4C and 5, B and C). These findings agree with our prior analysis of the SET8 Phe/Tyr switch mutant in which we demonstrated that the Y334F substitution attenuates hydrogen bonding to the water molecule bound in the solvent pocket, promoting its dissociation and the conformational changes necessary for lysine dimethylation (8). Indeed, there is a nearly identical alignment of the dimethyl-lysine side chains in the structures of SET7/9Y305FandSET8Y334Fcomplexes,despitethedifferences in the orientations of the Phe-305 and Phe-334 side chains in each structure (Fig. 2F). Finally, both the SET8 Y334F (8) and SET7/9 Y305F mutants (Table 2) displayed diminished catalytic efficien- cies for lysine dimethylation versus monomethylation. These dif- ferences may reflect the kinetics of the reorganization within the active site, including the dissociation of the water molecule from the solvent pocket and the concomitant realignment of the monomethyl-lysine into a productive geometry for dimethylation. In addition to their place-holding role, the active site waters may also facilitate the deprotonation of the lysine -amino group between methyl transfer reactions. For methylation to occur, the -amino group must be deprotonated to function as the nucleophile in the SN2 methyl transfer reaction with AdoMet (Figs. 4 and 5). Although the pKa value of the lysine -amine in solution is 10.5, molecular dynamics simulations by Zhang and Bruice (25, 26) indicate that this value diminishes to 8.2 upon formation of the SET7/9 Michaelis complex due to the proximity of the AdoMet sulfonium cation and the low dielectric constant of the active site. Furthermore, their simu- lations show that a chain of water molecules facilitates the dep- rotonation of the -amino group prior to methyl transfer, trans- ferring the proton to bulk solvent. Although these water molecule chains are not evident in our crystal structures, the Y305F and Y245A complexes suggest another potential mech- anism for deprotonation. In the dimethyl-lysine complexes of the Y305F and Y245A mutants, the dissociation of water 1 and 2, respectively, from the lysine binding channel requires that the solvent-mediated hydrogen bond to the -amino group is broken (Figs. 2D and 3C). It is conceivable that these waters dissociate from the active site as hydronium ions, promoting the realignment and deprotonation of the methyl -amino group for the next methyl transfer reaction. A comparison of the SET7/9 Y305F and SET8 Y334F com- plexes yields insights into the mechanism by which the Phe/Tyr switch influences water binding within the solvent pocket. The phenylalanine substitution in the Phe/Tyr switch results in the loss of a single hydrogen bond to the water molecule (water 1) in the solvent pocket compared with the four hydrogen bonds that coordinate the solvent molecule in WT SET7/9 (Fig. 2, A and B) and SET8 (7, 8). Although this attenuation in hydrogen bonding may appear insignificant, this difference is nonetheless impor- tant for at least two reasons. First, theoretical calculations indi- cate that, on average, water molecules form 3.5 hydrogen bonds in solutions (40, 41). This value is greater than the num- ber of hydrogen bonds coordinating water 1 in the solvent pocket in SET7/9 Y305F (Fig. 2, B and C) as well as in SET8 Y334F and other di- and trimethyltransferases that possess a hydrophobic residue in the Phe/Tyr switch site (8). From the perspective of the water molecule, the greater hydrogen bond- ing potential in solution would tend to thermodynamically favor its dissociation from the solvent pocket in SET domain KMTs that lack a tyrosine in the Phe/Tyr switch position. Sec- ond, the ordered binding of water molecules observed in the active sites of SET domain ternary complexes represents an unfavorable entropy compared with their diffusion in bulk sol- vent. In WT SET7/9 (Fig. 2A) and SET8 (7, 8), this entropic penalty can be partially offset through the favorable enthalpy of binding associated with the four hydrogen bonds that coordi- nate the water within the solvent pocket. It is conceivable that the loss of the tyrosine-mediated hydrogen bond in the Phe/Tyr switch shifts the equilibrium in favor of dissociation of the water molecule from the solvent pocket, thereby facilitating dimethylation in SET7/9 Y305F, SET8 Y334F, and other di- and trimethyltransferases. The structures of the SET7/9 Y245A and Y305F complexes illustrate the interactions that align the lysine -amino group during the methyl transfer reactions in each enzyme. In the WT enzyme and the Y305F mutant, hydrogen bonding to the hydroxyl group of Tyr-245 appears to be critical in properly aligning the -amine for methyl transfer (Fig. 4). Tyr-245 is conserved in the sequences of many SET domain KMTs (8, 42), and substitutions of this residue generally impair or abolish activity, indicating its importance in catalysis (8, 43). However, SET7/9 appears to be an exception to this rule, as the Y245A mutant is not only active but is capable of catalyzing lysine trimethylation. In this mutant, Tyr-305 appears to assume the role of Tyr-245 by hydrogen bonding to the monomethylated -amino group to align it for methyl transfer with AdoMet, as illustrated in the modeled substrate complex for the dimethy- lation reaction (Fig. 5B). Conversely, in the model for trimethy- lation, the Tyr-305 hydroxyl group does not hydrogen bond to the -amine but instead participates in a CH–O hydrogen bond with one of the methyl groups to assist in aligning the dimethy- lated -amine for the methyl transfer reaction (Fig. 5C). Addi- tional structural and functional studies of the SET domain tri- methyltransferases will aid in further illuminating the roles of CH–O hydrogen bonds in facilitating lysine multiple methylation. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31857 Acknowledgments—We acknowledge S. Schiebold for assistance in protein expression, purification, and crystallization and S. Anderson and R. Sanishvili for their assistance with x-ray data collection. We also thank S. Bulfer and S. Horowitz for reading the manuscript and providing useful comments. This work utilized the Protein Structure Facility of the Michigan Diabetes Research and Training Center, Uni- versity of Michigan, supported by National Institutes of Health Grant DK020572, NIDDK. Use of the Advanced Photon Source was sup- ported by the United States Department of Energy, Basic Energy Sci- ences, Office of Science, under Contract DE-AC02-06CH11357. GM/CA CAT has been funded in whole or in part by National Institutes of Health NCI Grant Y1-CO-1020 and NIGMS Grant Y1-GM-1104. Use of the LS-CAT Sector 21 was supported by Michi- gan Economic Development Corporation and the Michigan Technol- ogy Tri-Corridor Grant 085P1000817 for the support of this research program. REFERENCES 1. Huang, J., and Berger, S. L. (2008) Curr. Opin. Genet. Dev. 18, 152–158 2. Morgunkova, A., and Barlev, N. A. (2006) Cell Cycle 5, 1308–1312 3. Yang, X. D., Lamb, A., and Chen, L. F. (2009) Epigenetics 4, 429–433 4. Taverna, S. D., Li, H., Ruthenburg, A. J., Allis, C. D., and Patel, D. J. (2007) Nat. Struct. Mol. Biol. 14, 1025–1040 5. Collins, R. E., Tachibana, M., Tamaru, H., Smith, K. M., Jia, D., Zhang, X., Selker, E. U., Shinkai, Y., and Cheng, X. (2005) J. Biol. Chem. 280, 5563–5570 6. Zhang, X., Yang, Z., Khan, S. I., Horton, J. R., Tamaru, H., Selker, E. U., and Cheng, X. (2003) Mol. Cell 12, 177–185 7. Couture, J. F., Collazo, E., Brunzelle, J. S., and Trievel, R. C. (2005) Genes Dev. 19, 1455–1465 8. Couture, J. F., Dirk, L. M., Brunzelle, J. S., Houtz, R. L., and Trievel, R. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 20659–20664 9. Qian, C., Wang, X., Manzur, K., Sachchidanand, Farooq, A., Zeng, L., Wang, R., and Zhou, M. M. (2006) J. Mol. Biol. 359, 86–96 10. Trievel, R. C., Flynn, E. M., Houtz, R. L., and Hurley, J. H. (2003) Nat. Struct. Biol. 10, 545–552 11. Xiao, B., Jing, C., Wilson, J. R., Walker, P. A., Vasisht, N., Kelly, G., Howell, S., Taylor, I. A., Blackburn, G. M., and Gamblin, S. J. (2003) Nature 421, 652–656 12. Chuikov, S., Kurash, J. K., Wilson, J. R., Xiao, B., Justin, N., Ivanov, G. S., McKinney, K., Tempst, P., Prives, C., Gamblin, S. J., Barlev, N. A., and Reinberg, D. (2004) Nature 432, 353–360 13. Ea, C. K., and Baltimore, D. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 18972–18977 14. Este`ve, P. O., Chin, H. G., Benner, J., Feehery, G. R., Samaranayake, M., Horwitz, G. A., Jacobsen, S. E., and Pradhan, S. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 5076–5081 15. Kouskouti, A., Scheer, E., Staub, A., Tora, L., and Talianidis, I. (2004) Mol. Cell 14, 175–182 16. Masatsugu, T., and Yamamoto, K. (2009) Biochem. Biophys. Res. Commun. 381, 22–26 17. Munro, S., Khaire, N., Inche, A., Carr, S., and La Thangue, N. B. (2010) Oncogene 29, 2357–2367 18. Pagans, S., Kauder, S. E., Kaehlcke, K., Sakane, N., Schroeder, S., Dorm- eyer, W., Trievel, R. C., Verdin, E., Schnolzer, M., and Ott, M. (2010) Cell Host Microbe 7, 234–244 19. Subramanian, K., Jia, D., Kapoor-Vazirani, P., Powell, D. R., Collins, R. E., Sharma, D., Peng, J., Cheng, X., and Vertino, P. M. (2008) Mol. Cell 30, 336–347 20. Wang, J., Hevi, S., Kurash, J. K., Lei, H., Gay, F., Bajko, J., Su, H., Sun, W., Chang, H., Xu, G., Gaudet, F., Li, E., and Chen, T. (2009) Nat. Genet. 41, 125–129 21. Yang, X. D., Huang, B., Li, M., Lamb, A., Kelleher, N. L., and Chen, L. F. (2009) EMBO J. 28, 1055–1066 22. Guo, H. B., and Guo, H. (2007) Proc. Natl. Acad. Sci. U.S.A. 104, 8797–8802 23. Hu, P., Wang, S., and Zhang, Y. (2008) J. Am. Chem. Soc. 130, 3806–3813 24. Hu, P., and Zhang, Y. (2006) J. Am. Chem. Soc. 128, 1272–1278 25. Zhang, X., and Bruice, T. C. (2007) Biochemistry 46, 14838–14844 26. Zhang, X., and Bruice, T. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 5728–5732 27. Couture, J. F., Collazo, E., Hauk, G., and Trievel, R. C. (2006) Nat. Struct. Mol. Biol. 13, 140–146 28. Trievel, R. C., Beach, B. M., Dirk, L. M., Houtz, R. L., and Hurley, J. H. (2002) Cell 111, 91–103 29. Kapust, R. B., To¨zse´r, J., Fox, J. D., Anderson, D. E., Cherry, S., Copeland, T. D., and Waugh, D. S. (2001) Protein Eng. 14, 993–1000 30. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307–326 31. Vagin, A., and Teplyakov, A. (2000) Acta Crystallogr. D Biol. Crystallogr. 56, 1622–1624 32. Emsley, P., and Cowtan, K. (2004) Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 33. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Acta Crystallogr. D Biol. Crystallogr. 53, 240–255 34. Davis, I. W., Leaver-Fay, A., Chen, V. B., Block, J. N., Kapral, G. J., Wang, X., Murray, L. W., Arendall, W. B., 3rd, Snoeyink, J., Richardson, J. S., and Richardson, D. C. (2007) Nucleic Acids Res. 35, W375–W383 35. Bru¨nger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. D Biol. Crystallogr. 54, 905–921 36. Collazo, E., Couture, J. F., Bulfer, S., and Trievel, R. C. (2005) Anal. Bio- chem. 342, 86–92 37. Chirpich, T. P., Zappia, V., Costilow, R. N., and Barker, H. A. (1970) J. Biol. Chem. 245, 1778–1789 38. Couture, J. F., Hauk, G., Thompson, M. J., Blackburn, G. M., and Trievel, R. C. (2006) J. Biol. Chem. 281, 19280–19287 39. Kwon, T., Chang, J. H., Kwak, E., Lee, C. W., Joachimiak, A., Kim, Y. C., Lee, J., and Cho, Y. (2003) EMBO J. 22, 292–303 40. Chandra, A., and Chowdhuri, S. (2002) J. Phys. Chem. B 106, 6779–6783 41. Guardia, E., Marti, J., Garcia-Tarres, L., and Laria, D. (2005) J. Mol. Liq. 117, 63–67 42. Dillon, S. C., Zhang, X., Trievel, R. C., and Cheng, X. (2005) Genome Biol. 6, 227 43. Zhang, X., Tamaru, H., Khan, S. I., Horton, J. R., Keefe, L. J., Selker, E. U., and Cheng, X. (2002) Cell 111, 117–127 Lysine Methylation by SET7/9 Mutants 31858 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010
3M59
SET7/9 Y245A in complex with TAF10-K189me2 peptide and AdoHcy
SET7/9 Catalytic Mutants Reveal the Role of Active Site Water Molecules in Lysine Multiple Methylation*□ S Received for publication,February 15, 2010, and in revised form, July 14, 2010 Published, JBC Papers in Press,August 1, 2010, DOI 10.1074/jbc.M110.114587 Paul A. Del Rizzo‡, Jean-Franc¸ois Couture‡§1, Lynnette M. A. Dirk¶, Bethany S. Strunk‡, Marijo S. Roiko‡, Joseph S. Brunzelle, Robert L. Houtz¶, and Raymond C. Trievel‡2 From the ‡Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109, the §Ottawa Institute of Systems Biology, Department of Biochemistry, Microbiology, and Immunology, University of Ottawa, Ottawa, Ontario K1H 8M5, Canada, the ¶Department of Horticulture, Plant Physiology/Biochemistry/Molecular Biology Program, University of Kentucky, Lexington, Kentucky 40546, and the Department of Molecular Pharmacology and Biological Chemistry, Northwestern University Center for Synchrotron Research, Life Sciences Collaborative Access Team, Argonne, Illinois 60439 SET domain lysine methyltransferases (KMTs) methylate specific lysine residues in histone and non-histone substrates. These enzymes also display product specificity by catalyzing dis- tinct degrees of methylation of the lysine -amino group. To elucidate the molecular mechanism underlying this specificity, we have characterized the Y245A and Y305F mutants of the human KMT SET7/9 (also known as KMT7) that alter its prod- uct specificity from a monomethyltransferase to a di- and a tri- methyltransferase, respectively. Crystal structures of these mutants in complex with peptides bearing unmodified, mono-, di-, and trimethylated lysines illustrate the roles of active site water molecules in aligning the lysine -amino group for methyl transfer with S-adenosylmethionine. Displacement or dissocia- tion of these solvent molecules enlarges the diameter of the active site, accommodating the increasing size of the methylated -amino group during successive methyl transfer reactions. Together, these results furnish new insights into the roles of active site water molecules in modulating lysine multiple meth- ylation by SET domain KMTs and provide the first molecular snapshots of the mono-, di-, and trimethyl transfer reactions catalyzed by these enzymes. SET domain enzymes represent a family of S-adenosylmethi- onine (AdoMet)3-dependent methyltransferases that catalyze the site-specific methylation of protein lysyl residues in a host of proteins, including histones, transcription factors, chroma- tin-modifying enzymes, ribosomal subunits, and other sub- strates (1–3). In many instances, these modifications serve to recruit effector proteins that recognize methyl-lysyl residues in a sequence-dependent fashion (4). In addition, SET domain KMTs exhibit product specificity, defined as their ability to cat- alyze mono-, di-, or trimethylation of the lysine -amino group. This specificity is biologically relevant because many methyl- lysine-binding proteins can discriminate among different degrees of lysine methylation (4). Thus, both the site and degree of lysine methylation are critical to recognition by effector proteins. Structural and functional studies have identified a Phe/Tyr switch in the active site of SET domain KMTs that governs their respective product specificities (5, 6). According to this model, KMTs that possess a tyrosine in the Phe/Tyr switch site are limited to catalyzing lysine monomethylation, whereas en- zymes that possess a phenylalanine or another hydrophobic residue in this position display di- or trimethyltransferase activ- ity. Mutational analysis of various SET domain KMTs, includ- ing DIM-5 (KMT1) (6), SET7/9 (6), G9A (KMT1C) (5), and SET8 (also known as PR-SET7 and KMT5A) (7, 8), has demon- strated that substitutions in the Phe/Tyr switch result in pre- dictable changes in product specificity. Several models have been proposed to explain the mechanism by which the Phe/Tyr switch site governs this specificity, including variations in the diameter of the active site due to the size of Phe/Tyr switch residue and steric hindrance by the tyrosine hydroxyl group (6, 9–11). However, our recent studies of the Phe/Tyr switch mutant Y334F in the human histone H4 Lys-20 (H4K20) meth- yltransferase SET8 indicate that the Phe/Tyr switch regulates product specificity via a more subtle mechanism (8). Specifi- cally, the switch modulates the binding of an active site water molecule that in turn regulates the transition from mono- methylation to multiple methylation. Among the KMTs that have been structurally characterized, SET7/9 has emerged as an archetypal model for studying the catalytic mechanism and product specificity of the SET domain family. Although initially isolated as a histone H3 Lys-4 (H3K4)-specific methyltransferase, this KMT has been shown to regulate the functions of numerous non-histone substrates through site-specific methylation (12–21). Early structural and functional studies of SET7/9 identified two active site mutants, * Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth Grant R01 GM073839 (to R. C. T.) and National Institutes of Health Admin- istrative Supplement GM073839-04S1 (to R. C. T.) funded through the American Recovery and Reinvestment Act. This work was also supported by Department of Energy Grant DE-FG02-92ER20075 (to R. L. H.). □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Table 1. The atomic coordinates and structure factors (codes 3M53, 3M54, 3M55, 3M56, 3M57,3M58,3M59,and3M5A)havebeendepositedintheProteinDataBank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). 1 Supported by a Canadian Institutes of Health Research postdoctoral fellowship. 2 To whom correspondence should be addressed: Dept. of Biological Chem- istry, University of Michigan Medical School, 1150 West Medical Center Dr., 5301 MSRB III, Ann Arbor, MI 48109. Tel.: 734-647-0889; Fax: 734-763-4581; E-mail: rtrievel@umich.edu. 3 The abbreviations used are: AdoMet, S-adenosylmethionine; AdoHcy, S-ad- enosylhomocysteine; KMT, lysine methyltransferase; ITC, isothermal titra- tion calorimetry; CH–O, carbon-oxygen hydrogen bond; BisTris, 2-[bis(2- hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 41, pp. 31849–31858, October 8, 2010 © 2010 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31849 Y245A and Y305F, which change its product specificity. The Phe/Tyr switch mutant Y305F alters SET7/9 product specificity from a mono- to dimethyltransferase (6), whereas the Y245A substitution converts the enzyme into a trimethyltransferase with weak monomethyltransferase activity (11). These mutants have been the subjects of numerous molecular modeling simu- lations that have led to various models to explain their distinct product specificities (22–26). However, the lack of structural data for the SET7/9 Y245A and Y305F mutants in complex with cognate methylated peptides has hindered our understanding of the mechanisms that define the respective product specific- ities of these mutants. Moreover, these structures would yield a framework for visualizing the mono-, di-, and trimethylation reactions catalyzed by SET domain KMTs. To gain insight into the molecular basis of their product specificities, we have determined high resolution crystal structures of the SET7/9 Y245A and Y305F mutants in com- plex with peptides of the TATA box-binding protein-associ- ated factor TAF10 bearing the Lys-189 methylation site in unmodified (K189), monomethylated (K189me1), dimethyl- ated (K189me2), and trimethylated (K189me3) states. The structures and accompanying biochemical data support a model whereby changes in the occupancy or position of water molecules in the active site are critical in establishing the prod- uct specificities of the SET7/9 Y245A and Y305F mutants. Together, our results provide new insights into the mechanisms that govern SET domain product specificity and provide step- wise snapshots of the lysine mono-, di-, and trimethyl transfer reactions catalyzed by KMTs. EXPERIMENTAL PROCEDURES Cloning, Expression, and Purification of the SET7/9 Mutants— The Y245A and Y305F mutants were introduced into the pHIS2 SET7/9 expression vector encoding residues 110–366 (27) via QuikChange site-directed mutagenesis (Stratagene) and were verified by dideoxy DNA sequencing. The plasmids encoding wild type (WT) SET7/9 and the Y245A and Y305F mutants were transformed into Rosetta2 DE3 cells (Novagen) and were expressed as described previously (27, 28). In the course of characterizing WT SET7/9, we observed that the enzyme co- purified with AdoMet or another contaminant that resulted in technical difficulties in the isothermal titration calorimetry (ITC) experiments and co-crystallization trials with the TAF10 peptides. To overcome this problem, a denaturation and refold- ing step was inserted in the purification scheme. The denatur- ation and refolding protocol involved adding 6 M guanidine HCl, 50 mM sodium phosphate, pH 7.5, and 500 mM NaCl to the protein while it was immobilized on a nickel-Sepharose column (GE Healthcare). The column was washed with this buffer, fol- lowed by a solution of 6 M urea, 50 mM sodium phosphate, pH 7.5, 150 mM NaCl, and 10 mM 2-mercaptoethanol to remove the cofactor from the denatured enzyme. A reverse gradient from 6 to 0 M urea was then performed in the same buffer to refold the protein, which was subsequently eluted from the column using a linear gradient of 0–500 mM imidazole in 50 mM sodium phosphate, pH 7.5, 500 NaCl, and 10 mM 2-mercaptoethanol. The refolded protein was digested with tobacco etch virus pro- tease (29) during dialysis against 20 mM Tris, pH 7.5, 150 mM NaCl, and 5 mM 2-mercaptoethanol and then purified using a Superdex 200 gel filtration column (GE Healthcare). Protein concentration was determined by its absorbance at 280 nm. Synthetic Peptides—The TAF10 peptides bearing K189, K189me1, K189me2, and K189me3 (sequence, acetyl- SKSK189DRKYTL-amide) and a biotinylated TAF10 peptide (sequence, acetyl-SKSK189DRKYTLT(K-EZLink-S-S-biotin)- amide) were synthesized and purified by New England Peptide, Inc. Peptide concentrations were measured using the absorb- ance of their tyrosine residue at 274 nm. Crystallization and Data Collection—Crystals were pro- duced by hanging drop vapor diffusion by mixing the crystalli- zation solution in a 1:1 ratio with 10–20 mg/ml SET7/9, 1 mM S-adenosylhomocysteine (AdoHcy), and 1.0–3.0 mM unmodi- fied or methylated TAF10-K189 peptide in 20 mM Tris, pH 8.0, 100 mM NaCl, and 2 mM tris(2-carboxyethyl)phosphine. Crys- tals were obtained at 20 °C in either 1.8–2.0 M (NH4)2SO4, with 0.1 M BisTris, pH 6.2–6.6, or in 0.95–1.1 M sodium citrate with 100 mM imidazole pH 8.0–8.4. In both crystallization condi- tions, the final pH values were between pH 8.0 and 9.0. Crystals in the (NH4)2SO4 condition were typically flash-frozen in the mother liquor containing 25–30% glycerol, and the crystals in the citrate condition were frozen in 1.6 M sodium citrate. Data were collected at the Advanced Photon Source beamlines 21-IDG (LS-CAT) and 23-IDD (GM/CA-CAT). Images were indexed, integrated, and scaled using HKL2000 (30). Structures of the mutants were solved by molecular replacement using MOLREP (31) with the coordinates of a previously reported SET7/9 ternary complex used as the search model (Protein Data Bank code 2F69). Successive rounds of model building and refinement were carried out using Coot (32) and REFMAC (33), respectively. The geometry of the models were verified by Mol- Probity (34). Simulated annealing omit maps were calculated using CNS (35) with the peptide and cofactor removed to elim- inate model bias in the active site. Structural figures were ren- dered using PyMOL (Schro¨dinger, LLC.). Fluorescent Methyltransferase Assay—A coupled fluorescent methyltransferase assay was used to measure the kinetic parameters of WT SET7/9 and the Y245A and Y305F mutants as reported previously, with the exception that 50–150 nM enzyme, 100 M AdoMet, and varying concentrations of TAF10 peptide substrate were used (27, 36). Assays were performed in triplicate, and a homocysteine calibration curve was used to calculate the initial velocities. Kinetic parameters were calcu- lated by plotting the velocities versus peptide concentration and by fitting the Michaelis-Menten equation to the data via non- linear regression using Prism 5.0 (GraphPad). In cases where the Km value was beyond the measurable range of the assay, the kcat/Km value was determined as described previously (7). Calorimetry Experiments—ITC was performed at 20 °C using a MicroCal VP-ITC calorimeter (GE Healthcare) with 0.12 mM protein and 4 mM AdoHcy in 20 mM sodium phosphate, pH 7, and 100 mM NaCl with 1.5 mM peptide as the injectant. Data were processed, and equilibrium dissociation constants (KD) and curve fitting errors were calculated from the binding iso- therms using Origin 7.0 (OriginLab Corp.). WT SET7/9 and the Y245A and Y305F mutants displayed ligand:protein binding stoichiometries (N values) between 0.8 and 1.0, demonstrating Lysine Methylation by SET7/9 Mutants 31850 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 that WT SET7/9 and its mutants were properly refolded due to their ability to bind peptides in an 1:1 molar ratio. TLC Product Analysis—Methyltransferase assays were per- formed in triplicate at 37 °C with the biotinylated TAF10 pep- tide (0.5 mM) for 30 min in the presence of WT SET7/9 (3 pmol), Y305F mutant (6 pmol), or the Y245A mutant (100 pmol). Assays contained 100 mM HEPES, pH 8.0, 50 mM NaCl, 1.0 mM [methyl-3H]AdoMet (diluted to a specific activity of 0.2 Ci/nmol with purified AdoMet (37)), 15 M Sulfolobus solfa- taricus AdoHcy hydrolase (36), and 2 units of adenosine deami- nase (Roche Applied Science) in a final volume of 20 l. The reactions were terminated by addition of an equal volume of 200 mM MES, pH 5.1, prior to addition of a 2-fold molar excess of immobilized avidin resin (UltraLink; Pierce). Biotinylated peptides were allowed to bind at room temperature for 30 min, and the resin was then collected by centrifugation (9000  g). The resin was washed three times with 300 mM NaCl, and the peptide was eluted overnight from the avidin resin by cleavage of the disulfide bond in the linker of the peptide using 10 mM tris(2-carboxyethyl)phosphine (Thermo Scientific). The resin was incubated with additional 10 mM tris(2-carboxyethyl)phos- phine the following day until the radiolabel was essentially removed from the resin. The recovered peptides were hydro- lyzed by incubation with 6 M HCl at 110 °C for 24 h. Subsequent steps in measuring the radiolabel incorporated into the mono-, di-, and trimethyl-lysine products were performed as reported previously (8). RESULTS Functional Analysis of the SET7/9 Y305F Mutant—Prior studies of SET7/9 by Zhang et al. (6) reported that mutation of the Phe/Tyr switch residue Tyr-305 to a phenylalanine alters its product specificity from a mono- to dimethyltransferase. We verified these findings by demon- strating that WT SET7/9 mono- methylated the TAF10-K189 pep- tide, whereas the Y305F mutant mono- and dimethylated this sub- strate, as demonstrated by mass spectrometry (data not shown). We next examined whether the Y305F substitution altered the affinity of SET7/9 for the TAF10-K189 peptides using ITC (Fig. 1). A comparison of the KD values revealed that SET7/9 Y305F bound the TAF10-K189 and TAF10-K189me1 peptides 4- and 6-fold more tightly, respec- tively, than the WT enzyme, whereas this mutant displayed a substantially diminished affinity for the TAF10-K189me2 pep- tide (Table 1). Although the WT enzyme and the Y305F mutant exhibited discernable differences in their affinities for the unmodified and monomethylated peptides, these variations are modest and cannot account for their distinct product specific- ities, suggesting that a kinetic effect during methylation may be responsible. To investigate this possibility, we characterized the kinetic parameters of WT SET7/9 and the Y305F mutant using the TAF10 peptides as substrates. Both enzymes methylated the unmodified peptide with comparable kcat and Km values (Table 2). In analyzing the kinetic parameters for the methylation of the monomethylated peptide by SET7/9 Y305F, we found that this substrate displayed an elevated Km value that was beyond the measurable range of the assay due to its limited solubility. In this case, we measured the catalytic efficiency (kcat/Km) for the methylation of this peptide and found that it was methylated 15-fold less efficiently than the unmodified peptide by SET7/9 Y305F. Given the fact that the Y305F mutant exhibited a higher binding affinity for the TAF10-K189me1 peptide than the WT enzyme (Table 1), the kinetic data suggest that a step in the reaction pathway following substrate binding limits the catalytic efficiency of this mutant. We next examined whether the Y305F mutant dimethylated the TAF10-K189 peptide via a processive or a distributive mechanism. In a processive mechanism, the methyl-lysine substrate would remain bound to the enzyme during successive methyl transfer reactions; thus, the concentration of an inter- mediate, such as monomethyl-lysine, cannot exceed the en- zyme concentration during the assay. In a distributive mecha- nism, the intermediates are released into solution where they accumulate prior to the next round of methylation, resulting in an intermediate concentration that is greater than that of the enzyme. Using a radiometric TLC assay and a biotinylated TAF10 peptide, we quantified the amounts of monomethylated products generated by the WT SET7/9 and the Y305F mutant FIGURE 1. ITC analysis of WT SET7/9 and the Y245A and Y305F mutants. Representative titrations and binding isotherms for the calorimetry experiments are shown for the unmodified TAF10-K189 peptide titrated into WT SET7/9 (left plot), SET7/9 Y305F (middle plot), and SET7/9 Y245A (right plot). In each panel, the ITC titration experiment (upper panel) is shown with the binding isotherms (lower panel) fit to a single binding site model. TABLE 1 Analysis of the binding affinity of WT SET7/9 and its catalytic mutants for unmodified and methylated TAF10 peptides SET7/9 TAF10 peptide KD a M WT K189 4.9  0.20 WT K189me1 4.0  0.36 Y305F K189 1.3  0.10 Y305F K189me1 0.62  0.065 Y305Fb K189me2 70 Y245A K189 4.0  0.25 Y245A K189me1 3.3  0.10 Y245A K189me2 5.8  0.22 Y245A K189me3 11  0.28 a Curve fitting errors were calculated from the binding isotherms. b An estimate of the affinity is reported due to weak peptide binding. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31851 (Table 3). The data demonstrate that comparable amounts of monomethyl-lysine were generated when the quantity of enzyme usedistakenintoaccount,inagreementwiththeirsimilar turnover numbers for the TAF10-K189 peptide (Table 2). The Y305F mutant also produced small but measurable quantities of radiolabeled dimethyl-lysine product that were substantially smaller than the amount of monomethyl-lysine generated. Therefore, the TLC data are consistent with a distributive mechanism for dimethylation by the Y305F mutant because the amount of monomethyl-lysine produced exceeded the quantity of enzyme used in the assay. Structures of WT SET7/9 and the Y305F Mutant in Complex with Unmodified and Methylated TAF10 Peptides—To deter- mine the mechanism by which the Y305F substitution alters the product specificity of SET7/9, we determined the crystal struc- tures of this mutant bound to AdoHcy and TAF10-K189, TAF10-K189me1, and TAF10-K189me2 peptides and com- pared these to the structures of the WT SET7/9AdoHcy TAF10-K189 complex (supplemental Table 1). The structures of these complexes were determined to 1.85 Å or higher reso- lution, permitting unambiguous modeling of the K189 side chains in the active site of the enzyme based on simulated annealing omit maps (Fig. 2). The ternary complexes of the WT and the Y305F mutant superimpose with overall root mean square differences of less than 0.3 Å for all aligned atoms, indi- cating that neither the Y305F mutation nor the binding of the various TAF10-K189 peptides results in substantial changes in its overall structure. An inspection of the active sites of the SET7/9 WT and Y305F complexes illustrates the binding modes of the unmod- ified and methylated forms of K189 in the TAF10 peptides (Fig. 2, A–D). The K189 side chain binds in an extended all trans conformation in a deep pocket, termed the lysine binding chan- nel, that is composed of Tyr-245, Thr-266, Leu-267, Ser-268, Tyr-305, Tyr-335, and Tyr-337 (not shown for clarity) (Fig. 2A). These residues interact with the aliphatic portion of the K189 side chain primarily through van der Waals contacts. The lysine binding channel connects to the AdoMet-binding site on the opposite face of the catalytic domain via an oxygen-lined methyl transfer pore (38). During catalysis, the methyl group of the cofactor is positioned within the methyl transfer pore for the SN2 reaction with the -amino group of the lysine or methyl-lysine substrate (see below). To lower the activation barrier for this reaction, the lysine -amine nucleophile is aligned for methyl transfer through a hydrogen bond network within the active site. In the WT enzyme, the K189 -amino group hydrogen bonds to the hydroxyl group of Tyr-245 as well as to two water molecules (Fig. 2A). One of the water molecules (termed water 1), is coor- dinated in a solvent pocket, through hydrogen bonds to the carbonyl oxygens of Gly-292 and Ala-295 and to the hydroxyl group of the Phe/Tyr switch residue Tyr-305. This solvent pocket is structurally conserved in SET domain KMTs and has an important role in defining product specificity through the adjacent Phe/Tyr switch residue, as shown in our prior studies of the human H4K20 methyltransferase SET8 (8). The other water molecule is bound within the methyl transfer pore between the lysine substrate and the thioether sulfur atom of AdoHcy through hydrogen bonds to Tyr-245, Asn-265, and His-293 in SET7/9 and the TAF10-K189 -amino group. This water is not observed in other structures of SET7/9 ternary complexes and may represent the approximate position that the AdoMet methyl group occupies in the methyl transfer pore in the Michaelis complex. In structures of the Y305F ternary complexes, the K189, K189me1, and K189me2 side chains also adopt extended trans side chain geometries within the lysine binding channel that are stabilized via hydrogen bonding to Tyr-245 in the enzyme (Fig. 2, B–D). The orientations of the K189me1 and K189me2 side chains are further maintained through carbon-oxygen (CH–O) hydrogen bonding between the methyl groups and oxygen atoms within the vicinity of the methyl transfer pore, as reported previously in other SET domain KMT structures (8, 10, 38). A superimposition of the SET7/9 WT and Y305F com- plexes underscores the similarity of the lysyl binding conforma- tions (Fig. 2E). However, there are notable differences in the hydrogen bond patterns and occupancy of water 1 within the solvent pocket in the Y305F mutant compared with the WT enzyme. Specifically, the Y305F substitution results in the loss of one hydrogen bond to water 1 in the structures of the TAF10- K189 and TAF10-K189me1 complexes (Fig. 2, B and C). In con- trast, water 1 is absent in TAF10-K189me2 complex, and the vacated solvent pocket is occupied by one of the methyl groups TABLE 2 Kinetic parameters of WT SET7/9 and the Y305F and Y245A mutants Enzyme TAF10 peptide substrate Km a kcat a kcat/Km a M min1 M1 min1  103 WT K189 160  17 17  0.62 110  17 Y305F K189 88  5.0 17  0.30 190  11 Y305Fb K189me1 11  0.50 Y245A K189 200  35 0.53  0.04 2.6  0.47 Y245A K189me1 210  23 5.9  0.23 28  3.3 Y245A K189me2 400  29 6.5  0.16 15  1.2 a Curve fitting errors were calculated from the hyperbolic fits for the Michaelis-Menten equation. b Km and kcat were not determined because the Km value was beyond the measurable range; therefore, the kcat/Km value is reported. TABLE 3 Product analysis of WT SET7/9 and the Y305F and Y245A mutants Enzyme Quantity of enzyme Measured product Amount of product formeda nmol nmol WT 0.003 Kme1 0.65  0.07 Y305F 0.006 Kme1 1.5  0.49 Kme2 0.033  0.009 Y245A 0.100 Kme1 0.80  0.22 Kme2 0.39  0.021 Kme3 0.076  0.019 a Standard deviation was calculated from triplicate measurements. Lysine Methylation by SET7/9 Mutants 31852 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 of the dimethyl -amine (Fig. 2D). This methyl group forms a 3.5-Å CH–O hydrogen bond to the carbonyl oxygen of Ala-295, further stabilizing the binding of the dimethyl-lysine side chain. A homologous dimethyl-lysine-binding mode and CH–O hydrogen bond was observed in our prior structural studies of the SET8 Y334F Phe/Tyr switch mutant that confers an analo- gous change in product specificity from a mono- to a dimeth- yltransferase (8). A structural alignment of the active sites of the SET7/9 Y305F and SET8 Y334F mutants bound to cognate dimethylated peptides illustrates that the coordinates of the dimethyl-lysyl side chains are virtually superimposable, with one methyl group oriented toward the methyltransfer pore and the second positioned within the vacant solvent pocket (Fig. 2F). Taken together, the structures of the SET7/9 Y305F com- plexes and the similarities in the dimethyl-lysine conforma- tions in the SET7/9 Y305F and SET8 Y334F mutants imply that the Phe/Tyr switch governs product specificity through a con- served mechanism whereby it indirectly influences the binding modes of the methyl-lysine side chain by modulating the affin- ity of the water molecule (water 1) bound in the solvent pocket. Biochemical Characterization of the SET7/9 Y245A Mutant—Previ- ous studies by Xiao et al. (11) reported that the Y245A mutation yields an unusual change in the product specificity of SET7/9, converting the enzyme to a trimeth- yltransferase with weak monometh- yltransferase activity. We deter- mined that the SET7/9 Y245A could mono-, di-, and trimethylate the TAF10-K189 peptide by mass spec- trometry (data not shown) and TLC (Table 3), confirming the earlier studies of Xiao et al. (11). ITC analysis revealed that the Y245A mutant displayed comparable KD values for the unmodified and methylated TAF10-K189 peptides (Fig. 1), although its affinity for the trimethylated peptide was modestly diminished in comparison with the other peptides (Table 1). The ITC data demonstrate that the Y245A mutant bound the unmodified, mono-, and dimethylated sub- strates with equivalent affinities, suggesting that a kinetic effect or a structural alteration in the active site may be responsible for its diminished activity toward un- modified substrates. To gain further insight into its peculiar product specificity, we characterized the kinetic properties of the SET7/9 Y245A mutant. Steady state analysis demonstrated that this mutant displayed similar Km values for the unmodified, mono- and dimethylated TAF10 peptides (Table 2). However, the turnover number for the TAF10-K189 peptide was diminished over 10-fold versus the methylated peptides and was reduced 30-fold versus the WT enzyme, in agreement with the weak monomethyltransferase activity reported by Xiao et al. (11). In addition, we investigated whether this mutant catalyzes lysine trimethylation via a pro- cessive or distributive mechanism as described for SET7/9 Y305F. The TLC data illustrate that the mono- and dimethyl- lysine intermediates accumulated at quantities greater than that of the enzyme used in the assay, indicating that SET7/9 Y245A obeys a distributive mechanism, analogous to the Y305F mutant (Table 3). Structures of SET7/9 Y245A Bound to Unmodified and Meth- ylated TAF10 Peptides—To elucidate the mechanism underly- ing its unusual product specificity, we determined the crystal structures of SET7/9 Y245A in complex with AdoHcy and unmodified, mono-, di-, and trimethylated TAF10 peptides (supplemental Table 1). These complexes superimpose with the structure of the WT SET7/9AdoHcyTAF10-K189 com- FIGURE 2. Crystal structures of WT SET7/9 and the Y305F mutant in complex with AdoHcy and unmodi- fied and methylated TAF10 peptides. A, active site of WT SET7/9 bound to AdoHcy and TAF10-K189 and the active site of SET7/9 Y305F bound to AdoHcy and TAF10-K189 (B), TAF10-K189me1 (C), and TAF10-K189me2 peptides (D). A–D, the K189 residue in the TAF10 peptide is depicted with yellow carbon atoms with the corresponding Fo – Fc simulated-annealing omit maps (contoured at 2.5 ) shown. AdoHcy is colored with green carbon atoms, and SET7/9 carbon atoms are colored gray, with the exception of Tyr-305/Y305F (magenta) and Tyr-245 (cyan). Gly-264 to Ser-268 are shown, and the main chain atoms of Gly-292 to Ala-295 are depicted for clarity. Red dashed lines indicate conventional hydrogen bonds, and blue dashed lines indicate CH–O hydrogen bonds. Cyan dashed lines indicate weak CH–O hydrogen bonds between 3.5 and 3.7 Å in length. The water molecule in the solvent binding pocket is numbered 1. E, overlay of the active sites of the WT enzyme (light blue carbon atoms) and the Y305F mutant (green, yellow, and pink carbon atoms correspond to K189, K189me1, and K189me2, respectively). Red dashed lines indicate coordination of the waters in the struc- turesoftheY305Fmutant,andthelightbluedashedlineindicateshydrogenbondingthatonlyoccursintheWT enzyme. F, superimposition of the active sites of the SET7/9 Y305FAdoHcyTAF10-K189me2 and SET8 Y334FAdoHcyH4K20me2 (Protein Data Bank code 3F9X) complexes shown with green and yellow carbon atoms, respectively. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31853 plex with root mean squared differences of less than 0.4 Å for all aligned atoms, indicating that the Y245A mutant does not per- turb the overall structure of the enzyme. Simulated annealing omit maps illustrate that K189 side chains are bound within the lysine binding channel through hydrogen bonds and van der Waals contacts (Fig. 3, A–D), although the interactions and binding modes are distinct from those in the complexes of WT SET7/9 and the Y305F mutant (Fig. 2, A–D). In the unmodified TAF10 peptide complex, the K189 -amino group forms a weak hydrogen bond with the carbonyl oxygen of Gly-264 (Fig. 3A), whereas the -amino groups of K189me1 and K189me2 hydro- gen bond to the hydroxyl group of Tyr-305 in the mono- and dimethylated peptide complexes (Fig. 3, B and C). The confor- mations of the K189me1 and K189me2 side chains are further stabilized by water-mediated hydrogen bonding and through CH–O hydrogen bonding to their methyl groups. In the TAF10-K189me3 peptide complex, the trimethyl-lysine side chain is coordinated exclusively through direct and water-me- diated CH–O hydrogen bonds to its methyl groups because the quaternary -ammonium cation cannot engage in hydrogen bonding (Fig. 3D). A structural alignment of the four SET7/9 Y245A complexes illustrates distinct binding modes for the unmodified versus the methylated K189 side chains, highlighting the selectivity of this mutant for methylated substrates. The side chains of K189me1, K189me2, and K189me3 roughly overlay with their respective -amino groups superimposed and adopt slightly kinked con- formations (Fig. 3E), as opposed to the extended trans geome- try of the unmodified and methylated lysines in the complexes of the WT enzyme and the Y305F mutant (Fig. 2, A–D). Con- versely, the unmodified K189 side chain does not superimpose with its methylated counterparts and is oriented in an alterna- tive configuration due to its hydrogen bonding to Gly-264 (Fig. 3, A and E). An overlay of the structures of the WT enzyme and Y245A mutant bound to the unmodified TAF10 peptide illus- trates that the side chains of K189 do not superimpose and that the K189 -amino group appears to be misaligned with AdoHcy in the Y245A complex (Fig. 3F). This suboptimal alignment may explain the diminished kcat value of SET7/9 Y245A mutant toward substrates with unmodified lysines (Table 2). A comparison of the structures of the SET7/9 Y245A and Y305F complexes yields a molecular explanation for the differ- ent product specificities of these two mutants. In the SET7/9 Y305F complexes, Tyr-245 aligns the K189 -amino group for methyl transfer through hydrogen bonding to its hydroxyl group (Fig. 2, B–D). Conversely, in the Y245A mutant, the K189me1 and K189me2 -amino groups are oriented through hydrogen bonding to Tyr-305 (Fig. 3, B and C). These distinct hydrogen bond patterns impart differences in the conforma- tions of the lysyl side chains due to the relative orientations of Tyr-245 and Tyr-305 in the lysine binding channel. Specifically, the kinked conformation adopted by the K189me1 and K189me2 side chains in the Y245A complexes (Fig. 3, B and C) may contribute to the differences in the turnover numbers of this mutant versus those of the WT enzyme and the Y305F mutant (Table 2). In addition, the dimethyl -amino group of the K189me2 side chain binds in distinct orientations in the Y245A and Y305F mutants due to their hydrogen bonding to FIGURE 3. Crystal structures of the SET7/9 Y245A mutant in complex with AdoHcy and unmodified and methylated TAF10 peptides. Active site of SET7/9 Y245A bound to AdoHcy and TAF10-K189 (A), TAF10-K189me1 (B), TAF10-K189me2 (C), and TAF10-K189me3 peptides (D). A–D, Fo – Fc simulated- annealing omit maps (contoured at 2.5 ) for the unmodified and methylated K189 side chains are illustrated. The residues and hydrogen bonds in each com- plexarecoloredasinFig.2.Thewatermoleculesinthelysinebindingchannelsof the Y245A complexes are numbered 1–4, as described in the text. E, superimpo- sition of the active sites of the Y245A complexes bound to the four methylated statesofTAF10-K189(unmodified,mono-,di-,andtrimethylatedshowningreen, yellow, pink, and blue, respectively). F, overlay of WT SET7/9 (blue carbons) and SET7/9 Y245A (green carbons) in complex with TAF10-K189. Waters correspond- ing to the WT and Y245A structures are colored cyan and green, respectively. G, alignment of the active sites of the Y245A (yellow carbons) and Y305F (green carbons) mutants in complex with the TAF10-K189me2 peptide. Hydrogen bonds from the Y305F structure are shown as green dashed lines, and waters and hydrogen bonds in the Y245A structure are shown in yellow and orange, respectively. Lysine Methylation by SET7/9 Mutants 31854 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 Tyr-305 and Tyr-245, respectively (Fig. 3G). In the Y305F mutant, hydrogen bonds to the dimethyl -amino group cou- pled with steric constraints in the lysine binding channel pre- vent the K189me2 side chain from undergoing a conforma- tional change that is conducive to trimethylation (Fig. 2D), consistent with its dimethyltransferase activity. However, in the Y245A mutant, the alanine substitution enlarges the diameter of the lysine binding channel, accommodating trimethyl-lysine (Fig. 3D). In addition, the larger diameter would permit the dimethyl-lysine substrate to undergo the conformational reor- ganization necessary to align the -amino group in a productive geometry for trimethylation. A major difference in the active site of the Y245A mutant versus the other SET7/9 structures is the presence of several water molecules bound in the cavity generated by the Y245A mutation. In the structure of the Y245A mutant bound to TAF10-K189, three water molecules (waters 2–4) occupy this cavity and are arranged in a triangular geometry (Fig. 3A). In addition, water 1 shifts 1.6 Å from its position in the solvent pocket toward water 2 to which it forms a hydrogen bond (Fig. 3, A and E). The shift in water 1 was unexpected given its con- served orientation in the solvent pocket of the SET7/9 WT and Y305F complexes (Fig. 2, A–C) as well as in the structures of other SET domain KMTs (8). This displacement is presumably related to the alternative conformation of the K189 side chain whose -amino group is too distant (4.3 Å) to form a productive hydrogen bond to water 1. Conversely, in the Y245A complexes bound to TAF10-K189me1 and TAF10-K189me2, water 1 remains tightly bound in the solvent pocket through hydrogen bonds to Tyr-305 hydroxyl group, the carbonyl oxygens of Gly- 292 and Ala-295, and the K189 -amino group (Fig. 3, B, C, and E), analogous to its binding in the WT enzyme (Fig. 2A). How- ever, in the TAF10-K189me3 complex, one of the methyl groups of the trimethyl -ammonium cation is oriented into the solvent pocket (Fig. 3D), similar to the dimethyl-lysine binding mode observed in the Y305F mutant (Fig. 2D). The binding of the methyl group in the solvent pocket displaces water 1 by 3.2 Å relative to its position in the TAF10-K189me1 complex (Fig. 3E), thereby avoiding a steric clash with the trimethylated -ammonium group. Variations in the occupancy of water 2 are also seen in the different Y245A structures. Water 2 is bound in similar orientations in the active site of the unmodified and monomethylated peptide complexes but is absent in the di- and trimethylated peptide complexes due to the binding of a methyl group in this position (Fig. 3, A–E). In summary, the changes in the positions or occupancies of waters 1 and 2 correlate with the binding modes of the unmodified and methylated K189 within the active site of the Y245A mutant. Catalytic Models of Lysine Multiple Methylation by SET7/9 Y245A, and Y305F—The structures of the SET7/9 complexes reported here offer a prime opportunity to generate stepwise models for lysine mono-, di-, and trimethylation by a SET domain KMT. We modeled the AdoMet-bound Michaelis complexes by superimposing the SET7/9 product complexes with the previously reported structure of the SET7/9-AdoMet binary complex (Fig. 4) (39). The conformations of the mono- and dimethyl -amino groups in the Michaelis complexes were inferred from the coordinates of the corresponding dimethyl- and trimethyl-lysine products, respectively. In addition, we modeled the -amino group in a deprotonated state with its hydrogen atoms oriented toward the hydrogen bond acceptors that align the lysyl side chain for methylation. As a basis for this comparison, we first modeled the monomethylation reaction catalyzed by WT SET7/9 (Fig. 4A). In the substrate ternary complex, the lysine -amine is aligned with the methyl group and sulfonium cation of AdoMet through a hydrogen bond to the Tyr-245 hydroxyl group and water 1 in the solvent pocket. The values of the reaction distance and angle are 2.8 Å and 153°, respectively, in approximate agreement with the linear geome- try of a SN2 methyl transfer reaction calculated in other mod- eled substrate complexes (8, 10). In the product complex, the monomethyl-lysine side chain is bound in an extended confor- mation with its methyl group oriented within the methyl trans- fer pore, thereby obstructing AdoMet binding. Furthermore, water 1 remains tightly coordinated in the solvent pocket through four hydrogen bonds to Gly-292, Ala-295, Tyr-305, and the monomethyl -amino group. These interactions hinder the dissociation of water 1 and the related rearrangement of the monomethyl-lysine side chain required for a second methyl transfer reaction, explaining why the WT enzyme cannot cata- lyze di- and trimethylation. These findings concur with the FIGURE 4. Catalytic models of the methyl transfer reactions catalyzed by WT SET7/9 and the Y305F mutant. A, monomethylation of TAF10-K189 by the WT enzyme. The reaction scheme depicts the modeled substrate ternary complex (left) and the product complex (right) for the transfer of the methyl group from AdoMet (green carbon atoms) to K189 in TAF10 (yellow carbons), yielding AdoHcy and K189me1. The red arrow indicates the direction of the nucleophilic attack of the deprotonated -amino group on the AdoMet methyl group. The transferred methyl group is colored green, and the white atoms represent the hydrogens of the -amino group. Hydrogen bonds and residues in the enzyme active site are illustrated as in Fig. 2. The reaction distance and angle are labeled in red. B and C, models of the Y305F mutant for the first methyl transfer reaction with TAF10-K189 (B) and second methyl transfer reaction with TAF10-K189me1 (C). Color schemes are the same as in A. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31855 model for SET7/9 product specificity reported in previous structural and functional studies (6, 11). Similar reaction geometry is observed in the model for the monomethyl transfer reaction catalyzed by SET7/9 Y305F. Hydrogen bonds from the Tyr-245 hydroxyl group and water 1 align the lysine -amino group with the AdoMet methyl group at a distance of 2.1 Å and an angle of 160°, equivalent to those measured in the Michaelis complex of the WT enzyme (Fig. 4, A and B). In the product complex, monomethyl-lysine adopts an extended trans configuration analogous to that in the WT enzyme. For dimethylation to occur, the monomethyl-lysine must undergo a conformational change in which its methyl group is rotated out of the methyl transfer path with AdoMet. The structure of the Y305F mutant bound to the dimethylated TAF10 peptide (Fig. 2D) implies that this rearrangement occurs through the dissociation of water 1 due to the loss of the Tyr- 305 hydrogen bond in the solvent pocket. The dissociation of water 1 would enable the monomethyl-lysine side chain to adopt an alternative conformation through a rotation about its C–N bond, projecting the methyl group into the solvent pocket (Fig. 4C). This rotation reorients the methyl group out of the methyl transfer path while realigning the monomethyl - amino group for a second methylation reaction through a direct hydrogen bond to the Tyr-245 hydroxyl group and a CH–O hydrogen bond between its methyl group and Ala-295. The modeled reaction geometry for monomethyl-lysine substrate complex of the Y305F mutant (2.7 Å, 160°) is equivalent to that of the first methyl transfer reaction in SET7/9 Y305F. These geometries concur with our previous models for mono- and dimethylation catalyzed by SET8 Y334F (8), illustrating that the orientation of a methyl group into the solvent pocket is a con- served feature of SET domain KMTs that catalyze multiple methylation. In addition, we modeled the methyl transfer reactions cata- lyzed by SET7/9 Y245A (Fig. 5). In the model of the lysine sub- strate complex, the -amino group is aligned for methyl transfer by a hydrogen bond to the carbonyl oxygen of Gly-264, result- ing in a short reaction distance (2.3 Å) and a suboptimal reac- tion angle (141°) with the methyl group of AdoMet (Fig. 5A). This misalignment appears to be a direct consequence of the Y245A mutation that abolishes hydrogen bonding to the - amino group, illustrating that the suboptimal orientation of the -amine likely contributes to the diminished activity of this mutant toward unmodified substrates (11). Conversely, in the modeled monomethyl-lysine substrate complex for SET7/9 Y245A, Tyr-305 hydrogen bonds to the -amino group, aligning it for methyl transfer (Fig. 5B). In addition, CH–O hydrogen bonds to the monomethyl-lysine methyl group and the dissoci- ation of water 2 from the active site also contribute to reposition- ing the -amino group for dimethylation. Collectively, these inter- actions orient the -amine in a reaction angle of 165° that is more conducive to methyl transfer. However, the reaction distance for dimethylation is 0.6 Å longer than that in the corresponding Y305F model because Tyr-305 is positioned further from AdoMet than Tyr-245 (Figs. 4C and 5B). In the third methyl transfer reaction catalyzed by SET7/9 Y245A, the lone pair of electrons of the dimethyl-lysine - amino group acts as the nucleophile and thus cannot engage in hydrogen bonding. The structure of the trimethyl-lysine prod- uct complex (Fig. 3D) implies that the dimethyl -amine is aligned via CH–O hydrogen bonds to its methyl groups, as shown in the model of the Michaelis complex for this reaction (Fig. 5C). These CH–O hydrogen bonds restrain the orienta- tion of the -amino group and position one of the methyl groups into the solvent pocket, displacing water 1 as discussed earlier (Fig. 3, D and E). These interactions cumulatively align the -amino group and AdoMet methyl group with a reaction distance of 3.0 Å and angle of 162° (Fig. 5C). Taken together, the models of the substrate complexes for SET7/9 Y245A suggest that CH–O hydrogen bonds play an increasingly important role in aligning the methylated -amino group in successive rounds of methyl transfer. DISCUSSION The structural and functional characterization of the SET7/9 Y245A and Y305F mutants presented here yields new insights into the mechanism underlying the product specificity of SET domain KMTs. Importantly, it resolves a general paradox concerning this specificity. How does the active site constrain the motion of the lysine -amino group to align it for methyl transfer with AdoMet, while providing adequate volume to accommodate the mono-, di-, and tri- methylated lysine side chain generated during multiple methyl transfer reactions? The structures of the Y305F and Y245A mutants resolve this paradox, illustrating that alter- ations in the positions or occupancies of water molecules within their active sites generate the space required to FIGURE 5. Models for the methyl transfer reactions catalyzed by the SET7/9 Y245A mutant. Models of the Y245A mutant for the first methyl transfer reaction with TAF10-K189 (A), the second methyl transfer reaction with TAF10-K189me1 (B), and the third methyl transfer reaction with TAF10- K189me2 (C). Residues and hydrogen bonds are depicted as in Fig. 3. Lysine Methylation by SET7/9 Mutants 31856 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 accommodate the multiply methylated -amine produced during successive catalytic cycles. Minor perturbations in the side chains of certain active site residues, such as Tyr- 305, are also observed in alignments of the WT enzyme and the Y245A and Y305F complexes, although these changes are modest compared with the displacement or dissociation of the water molecules in the active site. These findings suggest that the waters function as transient place holders that facil- itate the SN2 methyl transfer reaction. During monomethy- lation, they function to constrain the movement of the lysine -amino group by mediating hydrogen bonds between the substrate and enzyme, thereby promoting the linear align- ment with the methyl group and sulfonium cation of AdoMet (Fig. 4, A and B). During di- and trimethylation, the water molecules either relocate within the lysine binding channel or dissociate from the enzyme, yielding the space required to rotate the methyl group away from the methyl transfer pore and to realign the -amine in productive geom- etry for the next methyl transfer reaction (Figs. 4C and 5, B and C). These findings agree with our prior analysis of the SET8 Phe/Tyr switch mutant in which we demonstrated that the Y334F substitution attenuates hydrogen bonding to the water molecule bound in the solvent pocket, promoting its dissociation and the conformational changes necessary for lysine dimethylation (8). Indeed, there is a nearly identical alignment of the dimethyl-lysine side chains in the structures of SET7/9Y305FandSET8Y334Fcomplexes,despitethedifferences in the orientations of the Phe-305 and Phe-334 side chains in each structure (Fig. 2F). Finally, both the SET8 Y334F (8) and SET7/9 Y305F mutants (Table 2) displayed diminished catalytic efficien- cies for lysine dimethylation versus monomethylation. These dif- ferences may reflect the kinetics of the reorganization within the active site, including the dissociation of the water molecule from the solvent pocket and the concomitant realignment of the monomethyl-lysine into a productive geometry for dimethylation. In addition to their place-holding role, the active site waters may also facilitate the deprotonation of the lysine -amino group between methyl transfer reactions. For methylation to occur, the -amino group must be deprotonated to function as the nucleophile in the SN2 methyl transfer reaction with AdoMet (Figs. 4 and 5). Although the pKa value of the lysine -amine in solution is 10.5, molecular dynamics simulations by Zhang and Bruice (25, 26) indicate that this value diminishes to 8.2 upon formation of the SET7/9 Michaelis complex due to the proximity of the AdoMet sulfonium cation and the low dielectric constant of the active site. Furthermore, their simu- lations show that a chain of water molecules facilitates the dep- rotonation of the -amino group prior to methyl transfer, trans- ferring the proton to bulk solvent. Although these water molecule chains are not evident in our crystal structures, the Y305F and Y245A complexes suggest another potential mech- anism for deprotonation. In the dimethyl-lysine complexes of the Y305F and Y245A mutants, the dissociation of water 1 and 2, respectively, from the lysine binding channel requires that the solvent-mediated hydrogen bond to the -amino group is broken (Figs. 2D and 3C). It is conceivable that these waters dissociate from the active site as hydronium ions, promoting the realignment and deprotonation of the methyl -amino group for the next methyl transfer reaction. A comparison of the SET7/9 Y305F and SET8 Y334F com- plexes yields insights into the mechanism by which the Phe/Tyr switch influences water binding within the solvent pocket. The phenylalanine substitution in the Phe/Tyr switch results in the loss of a single hydrogen bond to the water molecule (water 1) in the solvent pocket compared with the four hydrogen bonds that coordinate the solvent molecule in WT SET7/9 (Fig. 2, A and B) and SET8 (7, 8). Although this attenuation in hydrogen bonding may appear insignificant, this difference is nonetheless impor- tant for at least two reasons. First, theoretical calculations indi- cate that, on average, water molecules form 3.5 hydrogen bonds in solutions (40, 41). This value is greater than the num- ber of hydrogen bonds coordinating water 1 in the solvent pocket in SET7/9 Y305F (Fig. 2, B and C) as well as in SET8 Y334F and other di- and trimethyltransferases that possess a hydrophobic residue in the Phe/Tyr switch site (8). From the perspective of the water molecule, the greater hydrogen bond- ing potential in solution would tend to thermodynamically favor its dissociation from the solvent pocket in SET domain KMTs that lack a tyrosine in the Phe/Tyr switch position. Sec- ond, the ordered binding of water molecules observed in the active sites of SET domain ternary complexes represents an unfavorable entropy compared with their diffusion in bulk sol- vent. In WT SET7/9 (Fig. 2A) and SET8 (7, 8), this entropic penalty can be partially offset through the favorable enthalpy of binding associated with the four hydrogen bonds that coordi- nate the water within the solvent pocket. It is conceivable that the loss of the tyrosine-mediated hydrogen bond in the Phe/Tyr switch shifts the equilibrium in favor of dissociation of the water molecule from the solvent pocket, thereby facilitating dimethylation in SET7/9 Y305F, SET8 Y334F, and other di- and trimethyltransferases. The structures of the SET7/9 Y245A and Y305F complexes illustrate the interactions that align the lysine -amino group during the methyl transfer reactions in each enzyme. In the WT enzyme and the Y305F mutant, hydrogen bonding to the hydroxyl group of Tyr-245 appears to be critical in properly aligning the -amine for methyl transfer (Fig. 4). Tyr-245 is conserved in the sequences of many SET domain KMTs (8, 42), and substitutions of this residue generally impair or abolish activity, indicating its importance in catalysis (8, 43). However, SET7/9 appears to be an exception to this rule, as the Y245A mutant is not only active but is capable of catalyzing lysine trimethylation. In this mutant, Tyr-305 appears to assume the role of Tyr-245 by hydrogen bonding to the monomethylated -amino group to align it for methyl transfer with AdoMet, as illustrated in the modeled substrate complex for the dimethy- lation reaction (Fig. 5B). Conversely, in the model for trimethy- lation, the Tyr-305 hydroxyl group does not hydrogen bond to the -amine but instead participates in a CH–O hydrogen bond with one of the methyl groups to assist in aligning the dimethy- lated -amine for the methyl transfer reaction (Fig. 5C). Addi- tional structural and functional studies of the SET domain tri- methyltransferases will aid in further illuminating the roles of CH–O hydrogen bonds in facilitating lysine multiple methylation. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31857 Acknowledgments—We acknowledge S. Schiebold for assistance in protein expression, purification, and crystallization and S. Anderson and R. Sanishvili for their assistance with x-ray data collection. We also thank S. Bulfer and S. Horowitz for reading the manuscript and providing useful comments. This work utilized the Protein Structure Facility of the Michigan Diabetes Research and Training Center, Uni- versity of Michigan, supported by National Institutes of Health Grant DK020572, NIDDK. Use of the Advanced Photon Source was sup- ported by the United States Department of Energy, Basic Energy Sci- ences, Office of Science, under Contract DE-AC02-06CH11357. GM/CA CAT has been funded in whole or in part by National Institutes of Health NCI Grant Y1-CO-1020 and NIGMS Grant Y1-GM-1104. Use of the LS-CAT Sector 21 was supported by Michi- gan Economic Development Corporation and the Michigan Technol- ogy Tri-Corridor Grant 085P1000817 for the support of this research program. REFERENCES 1. Huang, J., and Berger, S. L. (2008) Curr. Opin. Genet. Dev. 18, 152–158 2. Morgunkova, A., and Barlev, N. A. (2006) Cell Cycle 5, 1308–1312 3. Yang, X. D., Lamb, A., and Chen, L. F. (2009) Epigenetics 4, 429–433 4. Taverna, S. D., Li, H., Ruthenburg, A. J., Allis, C. D., and Patel, D. J. (2007) Nat. Struct. Mol. Biol. 14, 1025–1040 5. Collins, R. E., Tachibana, M., Tamaru, H., Smith, K. M., Jia, D., Zhang, X., Selker, E. U., Shinkai, Y., and Cheng, X. (2005) J. Biol. Chem. 280, 5563–5570 6. Zhang, X., Yang, Z., Khan, S. I., Horton, J. R., Tamaru, H., Selker, E. U., and Cheng, X. (2003) Mol. Cell 12, 177–185 7. Couture, J. F., Collazo, E., Brunzelle, J. S., and Trievel, R. C. (2005) Genes Dev. 19, 1455–1465 8. Couture, J. F., Dirk, L. M., Brunzelle, J. S., Houtz, R. L., and Trievel, R. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 20659–20664 9. Qian, C., Wang, X., Manzur, K., Sachchidanand, Farooq, A., Zeng, L., Wang, R., and Zhou, M. M. (2006) J. Mol. Biol. 359, 86–96 10. Trievel, R. C., Flynn, E. M., Houtz, R. L., and Hurley, J. H. (2003) Nat. Struct. Biol. 10, 545–552 11. Xiao, B., Jing, C., Wilson, J. R., Walker, P. A., Vasisht, N., Kelly, G., Howell, S., Taylor, I. A., Blackburn, G. M., and Gamblin, S. J. (2003) Nature 421, 652–656 12. Chuikov, S., Kurash, J. K., Wilson, J. R., Xiao, B., Justin, N., Ivanov, G. S., McKinney, K., Tempst, P., Prives, C., Gamblin, S. J., Barlev, N. A., and Reinberg, D. (2004) Nature 432, 353–360 13. Ea, C. K., and Baltimore, D. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 18972–18977 14. Este`ve, P. O., Chin, H. G., Benner, J., Feehery, G. R., Samaranayake, M., Horwitz, G. A., Jacobsen, S. E., and Pradhan, S. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 5076–5081 15. Kouskouti, A., Scheer, E., Staub, A., Tora, L., and Talianidis, I. (2004) Mol. Cell 14, 175–182 16. Masatsugu, T., and Yamamoto, K. (2009) Biochem. Biophys. Res. Commun. 381, 22–26 17. Munro, S., Khaire, N., Inche, A., Carr, S., and La Thangue, N. B. (2010) Oncogene 29, 2357–2367 18. Pagans, S., Kauder, S. E., Kaehlcke, K., Sakane, N., Schroeder, S., Dorm- eyer, W., Trievel, R. C., Verdin, E., Schnolzer, M., and Ott, M. (2010) Cell Host Microbe 7, 234–244 19. Subramanian, K., Jia, D., Kapoor-Vazirani, P., Powell, D. R., Collins, R. E., Sharma, D., Peng, J., Cheng, X., and Vertino, P. M. (2008) Mol. Cell 30, 336–347 20. Wang, J., Hevi, S., Kurash, J. K., Lei, H., Gay, F., Bajko, J., Su, H., Sun, W., Chang, H., Xu, G., Gaudet, F., Li, E., and Chen, T. (2009) Nat. Genet. 41, 125–129 21. Yang, X. D., Huang, B., Li, M., Lamb, A., Kelleher, N. L., and Chen, L. F. (2009) EMBO J. 28, 1055–1066 22. Guo, H. B., and Guo, H. (2007) Proc. Natl. Acad. Sci. U.S.A. 104, 8797–8802 23. Hu, P., Wang, S., and Zhang, Y. (2008) J. Am. Chem. Soc. 130, 3806–3813 24. Hu, P., and Zhang, Y. (2006) J. Am. Chem. Soc. 128, 1272–1278 25. Zhang, X., and Bruice, T. C. (2007) Biochemistry 46, 14838–14844 26. Zhang, X., and Bruice, T. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 5728–5732 27. Couture, J. F., Collazo, E., Hauk, G., and Trievel, R. C. (2006) Nat. Struct. Mol. Biol. 13, 140–146 28. Trievel, R. C., Beach, B. M., Dirk, L. M., Houtz, R. L., and Hurley, J. H. (2002) Cell 111, 91–103 29. Kapust, R. B., To¨zse´r, J., Fox, J. D., Anderson, D. E., Cherry, S., Copeland, T. D., and Waugh, D. S. (2001) Protein Eng. 14, 993–1000 30. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307–326 31. Vagin, A., and Teplyakov, A. (2000) Acta Crystallogr. D Biol. Crystallogr. 56, 1622–1624 32. Emsley, P., and Cowtan, K. (2004) Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 33. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Acta Crystallogr. D Biol. Crystallogr. 53, 240–255 34. Davis, I. W., Leaver-Fay, A., Chen, V. B., Block, J. N., Kapral, G. J., Wang, X., Murray, L. W., Arendall, W. B., 3rd, Snoeyink, J., Richardson, J. S., and Richardson, D. C. (2007) Nucleic Acids Res. 35, W375–W383 35. Bru¨nger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. D Biol. Crystallogr. 54, 905–921 36. Collazo, E., Couture, J. F., Bulfer, S., and Trievel, R. C. (2005) Anal. Bio- chem. 342, 86–92 37. Chirpich, T. P., Zappia, V., Costilow, R. N., and Barker, H. A. (1970) J. Biol. Chem. 245, 1778–1789 38. Couture, J. F., Hauk, G., Thompson, M. J., Blackburn, G. M., and Trievel, R. C. (2006) J. Biol. Chem. 281, 19280–19287 39. Kwon, T., Chang, J. H., Kwak, E., Lee, C. W., Joachimiak, A., Kim, Y. C., Lee, J., and Cho, Y. (2003) EMBO J. 22, 292–303 40. Chandra, A., and Chowdhuri, S. (2002) J. Phys. Chem. B 106, 6779–6783 41. Guardia, E., Marti, J., Garcia-Tarres, L., and Laria, D. (2005) J. Mol. Liq. 117, 63–67 42. Dillon, S. C., Zhang, X., Trievel, R. C., and Cheng, X. (2005) Genome Biol. 6, 227 43. Zhang, X., Tamaru, H., Khan, S. I., Horton, J. R., Keefe, L. J., Selker, E. U., and Cheng, X. (2002) Cell 111, 117–127 Lysine Methylation by SET7/9 Mutants 31858 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010
3M5A
SET7/9 Y245A in complex with TAF10-K189me3 peptide and AdoHcy
SET7/9 Catalytic Mutants Reveal the Role of Active Site Water Molecules in Lysine Multiple Methylation*□ S Received for publication,February 15, 2010, and in revised form, July 14, 2010 Published, JBC Papers in Press,August 1, 2010, DOI 10.1074/jbc.M110.114587 Paul A. Del Rizzo‡, Jean-Franc¸ois Couture‡§1, Lynnette M. A. Dirk¶, Bethany S. Strunk‡, Marijo S. Roiko‡, Joseph S. Brunzelle, Robert L. Houtz¶, and Raymond C. Trievel‡2 From the ‡Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109, the §Ottawa Institute of Systems Biology, Department of Biochemistry, Microbiology, and Immunology, University of Ottawa, Ottawa, Ontario K1H 8M5, Canada, the ¶Department of Horticulture, Plant Physiology/Biochemistry/Molecular Biology Program, University of Kentucky, Lexington, Kentucky 40546, and the Department of Molecular Pharmacology and Biological Chemistry, Northwestern University Center for Synchrotron Research, Life Sciences Collaborative Access Team, Argonne, Illinois 60439 SET domain lysine methyltransferases (KMTs) methylate specific lysine residues in histone and non-histone substrates. These enzymes also display product specificity by catalyzing dis- tinct degrees of methylation of the lysine -amino group. To elucidate the molecular mechanism underlying this specificity, we have characterized the Y245A and Y305F mutants of the human KMT SET7/9 (also known as KMT7) that alter its prod- uct specificity from a monomethyltransferase to a di- and a tri- methyltransferase, respectively. Crystal structures of these mutants in complex with peptides bearing unmodified, mono-, di-, and trimethylated lysines illustrate the roles of active site water molecules in aligning the lysine -amino group for methyl transfer with S-adenosylmethionine. Displacement or dissocia- tion of these solvent molecules enlarges the diameter of the active site, accommodating the increasing size of the methylated -amino group during successive methyl transfer reactions. Together, these results furnish new insights into the roles of active site water molecules in modulating lysine multiple meth- ylation by SET domain KMTs and provide the first molecular snapshots of the mono-, di-, and trimethyl transfer reactions catalyzed by these enzymes. SET domain enzymes represent a family of S-adenosylmethi- onine (AdoMet)3-dependent methyltransferases that catalyze the site-specific methylation of protein lysyl residues in a host of proteins, including histones, transcription factors, chroma- tin-modifying enzymes, ribosomal subunits, and other sub- strates (1–3). In many instances, these modifications serve to recruit effector proteins that recognize methyl-lysyl residues in a sequence-dependent fashion (4). In addition, SET domain KMTs exhibit product specificity, defined as their ability to cat- alyze mono-, di-, or trimethylation of the lysine -amino group. This specificity is biologically relevant because many methyl- lysine-binding proteins can discriminate among different degrees of lysine methylation (4). Thus, both the site and degree of lysine methylation are critical to recognition by effector proteins. Structural and functional studies have identified a Phe/Tyr switch in the active site of SET domain KMTs that governs their respective product specificities (5, 6). According to this model, KMTs that possess a tyrosine in the Phe/Tyr switch site are limited to catalyzing lysine monomethylation, whereas en- zymes that possess a phenylalanine or another hydrophobic residue in this position display di- or trimethyltransferase activ- ity. Mutational analysis of various SET domain KMTs, includ- ing DIM-5 (KMT1) (6), SET7/9 (6), G9A (KMT1C) (5), and SET8 (also known as PR-SET7 and KMT5A) (7, 8), has demon- strated that substitutions in the Phe/Tyr switch result in pre- dictable changes in product specificity. Several models have been proposed to explain the mechanism by which the Phe/Tyr switch site governs this specificity, including variations in the diameter of the active site due to the size of Phe/Tyr switch residue and steric hindrance by the tyrosine hydroxyl group (6, 9–11). However, our recent studies of the Phe/Tyr switch mutant Y334F in the human histone H4 Lys-20 (H4K20) meth- yltransferase SET8 indicate that the Phe/Tyr switch regulates product specificity via a more subtle mechanism (8). Specifi- cally, the switch modulates the binding of an active site water molecule that in turn regulates the transition from mono- methylation to multiple methylation. Among the KMTs that have been structurally characterized, SET7/9 has emerged as an archetypal model for studying the catalytic mechanism and product specificity of the SET domain family. Although initially isolated as a histone H3 Lys-4 (H3K4)-specific methyltransferase, this KMT has been shown to regulate the functions of numerous non-histone substrates through site-specific methylation (12–21). Early structural and functional studies of SET7/9 identified two active site mutants, * Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth Grant R01 GM073839 (to R. C. T.) and National Institutes of Health Admin- istrative Supplement GM073839-04S1 (to R. C. T.) funded through the American Recovery and Reinvestment Act. This work was also supported by Department of Energy Grant DE-FG02-92ER20075 (to R. L. H.). □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Table 1. The atomic coordinates and structure factors (codes 3M53, 3M54, 3M55, 3M56, 3M57,3M58,3M59,and3M5A)havebeendepositedintheProteinDataBank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). 1 Supported by a Canadian Institutes of Health Research postdoctoral fellowship. 2 To whom correspondence should be addressed: Dept. of Biological Chem- istry, University of Michigan Medical School, 1150 West Medical Center Dr., 5301 MSRB III, Ann Arbor, MI 48109. Tel.: 734-647-0889; Fax: 734-763-4581; E-mail: rtrievel@umich.edu. 3 The abbreviations used are: AdoMet, S-adenosylmethionine; AdoHcy, S-ad- enosylhomocysteine; KMT, lysine methyltransferase; ITC, isothermal titra- tion calorimetry; CH–O, carbon-oxygen hydrogen bond; BisTris, 2-[bis(2- hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 41, pp. 31849–31858, October 8, 2010 © 2010 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31849 Y245A and Y305F, which change its product specificity. The Phe/Tyr switch mutant Y305F alters SET7/9 product specificity from a mono- to dimethyltransferase (6), whereas the Y245A substitution converts the enzyme into a trimethyltransferase with weak monomethyltransferase activity (11). These mutants have been the subjects of numerous molecular modeling simu- lations that have led to various models to explain their distinct product specificities (22–26). However, the lack of structural data for the SET7/9 Y245A and Y305F mutants in complex with cognate methylated peptides has hindered our understanding of the mechanisms that define the respective product specific- ities of these mutants. Moreover, these structures would yield a framework for visualizing the mono-, di-, and trimethylation reactions catalyzed by SET domain KMTs. To gain insight into the molecular basis of their product specificities, we have determined high resolution crystal structures of the SET7/9 Y245A and Y305F mutants in com- plex with peptides of the TATA box-binding protein-associ- ated factor TAF10 bearing the Lys-189 methylation site in unmodified (K189), monomethylated (K189me1), dimethyl- ated (K189me2), and trimethylated (K189me3) states. The structures and accompanying biochemical data support a model whereby changes in the occupancy or position of water molecules in the active site are critical in establishing the prod- uct specificities of the SET7/9 Y245A and Y305F mutants. Together, our results provide new insights into the mechanisms that govern SET domain product specificity and provide step- wise snapshots of the lysine mono-, di-, and trimethyl transfer reactions catalyzed by KMTs. EXPERIMENTAL PROCEDURES Cloning, Expression, and Purification of the SET7/9 Mutants— The Y245A and Y305F mutants were introduced into the pHIS2 SET7/9 expression vector encoding residues 110–366 (27) via QuikChange site-directed mutagenesis (Stratagene) and were verified by dideoxy DNA sequencing. The plasmids encoding wild type (WT) SET7/9 and the Y245A and Y305F mutants were transformed into Rosetta2 DE3 cells (Novagen) and were expressed as described previously (27, 28). In the course of characterizing WT SET7/9, we observed that the enzyme co- purified with AdoMet or another contaminant that resulted in technical difficulties in the isothermal titration calorimetry (ITC) experiments and co-crystallization trials with the TAF10 peptides. To overcome this problem, a denaturation and refold- ing step was inserted in the purification scheme. The denatur- ation and refolding protocol involved adding 6 M guanidine HCl, 50 mM sodium phosphate, pH 7.5, and 500 mM NaCl to the protein while it was immobilized on a nickel-Sepharose column (GE Healthcare). The column was washed with this buffer, fol- lowed by a solution of 6 M urea, 50 mM sodium phosphate, pH 7.5, 150 mM NaCl, and 10 mM 2-mercaptoethanol to remove the cofactor from the denatured enzyme. A reverse gradient from 6 to 0 M urea was then performed in the same buffer to refold the protein, which was subsequently eluted from the column using a linear gradient of 0–500 mM imidazole in 50 mM sodium phosphate, pH 7.5, 500 NaCl, and 10 mM 2-mercaptoethanol. The refolded protein was digested with tobacco etch virus pro- tease (29) during dialysis against 20 mM Tris, pH 7.5, 150 mM NaCl, and 5 mM 2-mercaptoethanol and then purified using a Superdex 200 gel filtration column (GE Healthcare). Protein concentration was determined by its absorbance at 280 nm. Synthetic Peptides—The TAF10 peptides bearing K189, K189me1, K189me2, and K189me3 (sequence, acetyl- SKSK189DRKYTL-amide) and a biotinylated TAF10 peptide (sequence, acetyl-SKSK189DRKYTLT(K-EZLink-S-S-biotin)- amide) were synthesized and purified by New England Peptide, Inc. Peptide concentrations were measured using the absorb- ance of their tyrosine residue at 274 nm. Crystallization and Data Collection—Crystals were pro- duced by hanging drop vapor diffusion by mixing the crystalli- zation solution in a 1:1 ratio with 10–20 mg/ml SET7/9, 1 mM S-adenosylhomocysteine (AdoHcy), and 1.0–3.0 mM unmodi- fied or methylated TAF10-K189 peptide in 20 mM Tris, pH 8.0, 100 mM NaCl, and 2 mM tris(2-carboxyethyl)phosphine. Crys- tals were obtained at 20 °C in either 1.8–2.0 M (NH4)2SO4, with 0.1 M BisTris, pH 6.2–6.6, or in 0.95–1.1 M sodium citrate with 100 mM imidazole pH 8.0–8.4. In both crystallization condi- tions, the final pH values were between pH 8.0 and 9.0. Crystals in the (NH4)2SO4 condition were typically flash-frozen in the mother liquor containing 25–30% glycerol, and the crystals in the citrate condition were frozen in 1.6 M sodium citrate. Data were collected at the Advanced Photon Source beamlines 21-IDG (LS-CAT) and 23-IDD (GM/CA-CAT). Images were indexed, integrated, and scaled using HKL2000 (30). Structures of the mutants were solved by molecular replacement using MOLREP (31) with the coordinates of a previously reported SET7/9 ternary complex used as the search model (Protein Data Bank code 2F69). Successive rounds of model building and refinement were carried out using Coot (32) and REFMAC (33), respectively. The geometry of the models were verified by Mol- Probity (34). Simulated annealing omit maps were calculated using CNS (35) with the peptide and cofactor removed to elim- inate model bias in the active site. Structural figures were ren- dered using PyMOL (Schro¨dinger, LLC.). Fluorescent Methyltransferase Assay—A coupled fluorescent methyltransferase assay was used to measure the kinetic parameters of WT SET7/9 and the Y245A and Y305F mutants as reported previously, with the exception that 50–150 nM enzyme, 100 M AdoMet, and varying concentrations of TAF10 peptide substrate were used (27, 36). Assays were performed in triplicate, and a homocysteine calibration curve was used to calculate the initial velocities. Kinetic parameters were calcu- lated by plotting the velocities versus peptide concentration and by fitting the Michaelis-Menten equation to the data via non- linear regression using Prism 5.0 (GraphPad). In cases where the Km value was beyond the measurable range of the assay, the kcat/Km value was determined as described previously (7). Calorimetry Experiments—ITC was performed at 20 °C using a MicroCal VP-ITC calorimeter (GE Healthcare) with 0.12 mM protein and 4 mM AdoHcy in 20 mM sodium phosphate, pH 7, and 100 mM NaCl with 1.5 mM peptide as the injectant. Data were processed, and equilibrium dissociation constants (KD) and curve fitting errors were calculated from the binding iso- therms using Origin 7.0 (OriginLab Corp.). WT SET7/9 and the Y245A and Y305F mutants displayed ligand:protein binding stoichiometries (N values) between 0.8 and 1.0, demonstrating Lysine Methylation by SET7/9 Mutants 31850 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 that WT SET7/9 and its mutants were properly refolded due to their ability to bind peptides in an 1:1 molar ratio. TLC Product Analysis—Methyltransferase assays were per- formed in triplicate at 37 °C with the biotinylated TAF10 pep- tide (0.5 mM) for 30 min in the presence of WT SET7/9 (3 pmol), Y305F mutant (6 pmol), or the Y245A mutant (100 pmol). Assays contained 100 mM HEPES, pH 8.0, 50 mM NaCl, 1.0 mM [methyl-3H]AdoMet (diluted to a specific activity of 0.2 Ci/nmol with purified AdoMet (37)), 15 M Sulfolobus solfa- taricus AdoHcy hydrolase (36), and 2 units of adenosine deami- nase (Roche Applied Science) in a final volume of 20 l. The reactions were terminated by addition of an equal volume of 200 mM MES, pH 5.1, prior to addition of a 2-fold molar excess of immobilized avidin resin (UltraLink; Pierce). Biotinylated peptides were allowed to bind at room temperature for 30 min, and the resin was then collected by centrifugation (9000  g). The resin was washed three times with 300 mM NaCl, and the peptide was eluted overnight from the avidin resin by cleavage of the disulfide bond in the linker of the peptide using 10 mM tris(2-carboxyethyl)phosphine (Thermo Scientific). The resin was incubated with additional 10 mM tris(2-carboxyethyl)phos- phine the following day until the radiolabel was essentially removed from the resin. The recovered peptides were hydro- lyzed by incubation with 6 M HCl at 110 °C for 24 h. Subsequent steps in measuring the radiolabel incorporated into the mono-, di-, and trimethyl-lysine products were performed as reported previously (8). RESULTS Functional Analysis of the SET7/9 Y305F Mutant—Prior studies of SET7/9 by Zhang et al. (6) reported that mutation of the Phe/Tyr switch residue Tyr-305 to a phenylalanine alters its product specificity from a mono- to dimethyltransferase. We verified these findings by demon- strating that WT SET7/9 mono- methylated the TAF10-K189 pep- tide, whereas the Y305F mutant mono- and dimethylated this sub- strate, as demonstrated by mass spectrometry (data not shown). We next examined whether the Y305F substitution altered the affinity of SET7/9 for the TAF10-K189 peptides using ITC (Fig. 1). A comparison of the KD values revealed that SET7/9 Y305F bound the TAF10-K189 and TAF10-K189me1 peptides 4- and 6-fold more tightly, respec- tively, than the WT enzyme, whereas this mutant displayed a substantially diminished affinity for the TAF10-K189me2 pep- tide (Table 1). Although the WT enzyme and the Y305F mutant exhibited discernable differences in their affinities for the unmodified and monomethylated peptides, these variations are modest and cannot account for their distinct product specific- ities, suggesting that a kinetic effect during methylation may be responsible. To investigate this possibility, we characterized the kinetic parameters of WT SET7/9 and the Y305F mutant using the TAF10 peptides as substrates. Both enzymes methylated the unmodified peptide with comparable kcat and Km values (Table 2). In analyzing the kinetic parameters for the methylation of the monomethylated peptide by SET7/9 Y305F, we found that this substrate displayed an elevated Km value that was beyond the measurable range of the assay due to its limited solubility. In this case, we measured the catalytic efficiency (kcat/Km) for the methylation of this peptide and found that it was methylated 15-fold less efficiently than the unmodified peptide by SET7/9 Y305F. Given the fact that the Y305F mutant exhibited a higher binding affinity for the TAF10-K189me1 peptide than the WT enzyme (Table 1), the kinetic data suggest that a step in the reaction pathway following substrate binding limits the catalytic efficiency of this mutant. We next examined whether the Y305F mutant dimethylated the TAF10-K189 peptide via a processive or a distributive mechanism. In a processive mechanism, the methyl-lysine substrate would remain bound to the enzyme during successive methyl transfer reactions; thus, the concentration of an inter- mediate, such as monomethyl-lysine, cannot exceed the en- zyme concentration during the assay. In a distributive mecha- nism, the intermediates are released into solution where they accumulate prior to the next round of methylation, resulting in an intermediate concentration that is greater than that of the enzyme. Using a radiometric TLC assay and a biotinylated TAF10 peptide, we quantified the amounts of monomethylated products generated by the WT SET7/9 and the Y305F mutant FIGURE 1. ITC analysis of WT SET7/9 and the Y245A and Y305F mutants. Representative titrations and binding isotherms for the calorimetry experiments are shown for the unmodified TAF10-K189 peptide titrated into WT SET7/9 (left plot), SET7/9 Y305F (middle plot), and SET7/9 Y245A (right plot). In each panel, the ITC titration experiment (upper panel) is shown with the binding isotherms (lower panel) fit to a single binding site model. TABLE 1 Analysis of the binding affinity of WT SET7/9 and its catalytic mutants for unmodified and methylated TAF10 peptides SET7/9 TAF10 peptide KD a M WT K189 4.9  0.20 WT K189me1 4.0  0.36 Y305F K189 1.3  0.10 Y305F K189me1 0.62  0.065 Y305Fb K189me2 70 Y245A K189 4.0  0.25 Y245A K189me1 3.3  0.10 Y245A K189me2 5.8  0.22 Y245A K189me3 11  0.28 a Curve fitting errors were calculated from the binding isotherms. b An estimate of the affinity is reported due to weak peptide binding. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31851 (Table 3). The data demonstrate that comparable amounts of monomethyl-lysine were generated when the quantity of enzyme usedistakenintoaccount,inagreementwiththeirsimilar turnover numbers for the TAF10-K189 peptide (Table 2). The Y305F mutant also produced small but measurable quantities of radiolabeled dimethyl-lysine product that were substantially smaller than the amount of monomethyl-lysine generated. Therefore, the TLC data are consistent with a distributive mechanism for dimethylation by the Y305F mutant because the amount of monomethyl-lysine produced exceeded the quantity of enzyme used in the assay. Structures of WT SET7/9 and the Y305F Mutant in Complex with Unmodified and Methylated TAF10 Peptides—To deter- mine the mechanism by which the Y305F substitution alters the product specificity of SET7/9, we determined the crystal struc- tures of this mutant bound to AdoHcy and TAF10-K189, TAF10-K189me1, and TAF10-K189me2 peptides and com- pared these to the structures of the WT SET7/9AdoHcy TAF10-K189 complex (supplemental Table 1). The structures of these complexes were determined to 1.85 Å or higher reso- lution, permitting unambiguous modeling of the K189 side chains in the active site of the enzyme based on simulated annealing omit maps (Fig. 2). The ternary complexes of the WT and the Y305F mutant superimpose with overall root mean square differences of less than 0.3 Å for all aligned atoms, indi- cating that neither the Y305F mutation nor the binding of the various TAF10-K189 peptides results in substantial changes in its overall structure. An inspection of the active sites of the SET7/9 WT and Y305F complexes illustrates the binding modes of the unmod- ified and methylated forms of K189 in the TAF10 peptides (Fig. 2, A–D). The K189 side chain binds in an extended all trans conformation in a deep pocket, termed the lysine binding chan- nel, that is composed of Tyr-245, Thr-266, Leu-267, Ser-268, Tyr-305, Tyr-335, and Tyr-337 (not shown for clarity) (Fig. 2A). These residues interact with the aliphatic portion of the K189 side chain primarily through van der Waals contacts. The lysine binding channel connects to the AdoMet-binding site on the opposite face of the catalytic domain via an oxygen-lined methyl transfer pore (38). During catalysis, the methyl group of the cofactor is positioned within the methyl transfer pore for the SN2 reaction with the -amino group of the lysine or methyl-lysine substrate (see below). To lower the activation barrier for this reaction, the lysine -amine nucleophile is aligned for methyl transfer through a hydrogen bond network within the active site. In the WT enzyme, the K189 -amino group hydrogen bonds to the hydroxyl group of Tyr-245 as well as to two water molecules (Fig. 2A). One of the water molecules (termed water 1), is coor- dinated in a solvent pocket, through hydrogen bonds to the carbonyl oxygens of Gly-292 and Ala-295 and to the hydroxyl group of the Phe/Tyr switch residue Tyr-305. This solvent pocket is structurally conserved in SET domain KMTs and has an important role in defining product specificity through the adjacent Phe/Tyr switch residue, as shown in our prior studies of the human H4K20 methyltransferase SET8 (8). The other water molecule is bound within the methyl transfer pore between the lysine substrate and the thioether sulfur atom of AdoHcy through hydrogen bonds to Tyr-245, Asn-265, and His-293 in SET7/9 and the TAF10-K189 -amino group. This water is not observed in other structures of SET7/9 ternary complexes and may represent the approximate position that the AdoMet methyl group occupies in the methyl transfer pore in the Michaelis complex. In structures of the Y305F ternary complexes, the K189, K189me1, and K189me2 side chains also adopt extended trans side chain geometries within the lysine binding channel that are stabilized via hydrogen bonding to Tyr-245 in the enzyme (Fig. 2, B–D). The orientations of the K189me1 and K189me2 side chains are further maintained through carbon-oxygen (CH–O) hydrogen bonding between the methyl groups and oxygen atoms within the vicinity of the methyl transfer pore, as reported previously in other SET domain KMT structures (8, 10, 38). A superimposition of the SET7/9 WT and Y305F com- plexes underscores the similarity of the lysyl binding conforma- tions (Fig. 2E). However, there are notable differences in the hydrogen bond patterns and occupancy of water 1 within the solvent pocket in the Y305F mutant compared with the WT enzyme. Specifically, the Y305F substitution results in the loss of one hydrogen bond to water 1 in the structures of the TAF10- K189 and TAF10-K189me1 complexes (Fig. 2, B and C). In con- trast, water 1 is absent in TAF10-K189me2 complex, and the vacated solvent pocket is occupied by one of the methyl groups TABLE 2 Kinetic parameters of WT SET7/9 and the Y305F and Y245A mutants Enzyme TAF10 peptide substrate Km a kcat a kcat/Km a M min1 M1 min1  103 WT K189 160  17 17  0.62 110  17 Y305F K189 88  5.0 17  0.30 190  11 Y305Fb K189me1 11  0.50 Y245A K189 200  35 0.53  0.04 2.6  0.47 Y245A K189me1 210  23 5.9  0.23 28  3.3 Y245A K189me2 400  29 6.5  0.16 15  1.2 a Curve fitting errors were calculated from the hyperbolic fits for the Michaelis-Menten equation. b Km and kcat were not determined because the Km value was beyond the measurable range; therefore, the kcat/Km value is reported. TABLE 3 Product analysis of WT SET7/9 and the Y305F and Y245A mutants Enzyme Quantity of enzyme Measured product Amount of product formeda nmol nmol WT 0.003 Kme1 0.65  0.07 Y305F 0.006 Kme1 1.5  0.49 Kme2 0.033  0.009 Y245A 0.100 Kme1 0.80  0.22 Kme2 0.39  0.021 Kme3 0.076  0.019 a Standard deviation was calculated from triplicate measurements. Lysine Methylation by SET7/9 Mutants 31852 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 of the dimethyl -amine (Fig. 2D). This methyl group forms a 3.5-Å CH–O hydrogen bond to the carbonyl oxygen of Ala-295, further stabilizing the binding of the dimethyl-lysine side chain. A homologous dimethyl-lysine-binding mode and CH–O hydrogen bond was observed in our prior structural studies of the SET8 Y334F Phe/Tyr switch mutant that confers an analo- gous change in product specificity from a mono- to a dimeth- yltransferase (8). A structural alignment of the active sites of the SET7/9 Y305F and SET8 Y334F mutants bound to cognate dimethylated peptides illustrates that the coordinates of the dimethyl-lysyl side chains are virtually superimposable, with one methyl group oriented toward the methyltransfer pore and the second positioned within the vacant solvent pocket (Fig. 2F). Taken together, the structures of the SET7/9 Y305F com- plexes and the similarities in the dimethyl-lysine conforma- tions in the SET7/9 Y305F and SET8 Y334F mutants imply that the Phe/Tyr switch governs product specificity through a con- served mechanism whereby it indirectly influences the binding modes of the methyl-lysine side chain by modulating the affin- ity of the water molecule (water 1) bound in the solvent pocket. Biochemical Characterization of the SET7/9 Y245A Mutant—Previ- ous studies by Xiao et al. (11) reported that the Y245A mutation yields an unusual change in the product specificity of SET7/9, converting the enzyme to a trimeth- yltransferase with weak monometh- yltransferase activity. We deter- mined that the SET7/9 Y245A could mono-, di-, and trimethylate the TAF10-K189 peptide by mass spec- trometry (data not shown) and TLC (Table 3), confirming the earlier studies of Xiao et al. (11). ITC analysis revealed that the Y245A mutant displayed comparable KD values for the unmodified and methylated TAF10-K189 peptides (Fig. 1), although its affinity for the trimethylated peptide was modestly diminished in comparison with the other peptides (Table 1). The ITC data demonstrate that the Y245A mutant bound the unmodified, mono-, and dimethylated sub- strates with equivalent affinities, suggesting that a kinetic effect or a structural alteration in the active site may be responsible for its diminished activity toward un- modified substrates. To gain further insight into its peculiar product specificity, we characterized the kinetic properties of the SET7/9 Y245A mutant. Steady state analysis demonstrated that this mutant displayed similar Km values for the unmodified, mono- and dimethylated TAF10 peptides (Table 2). However, the turnover number for the TAF10-K189 peptide was diminished over 10-fold versus the methylated peptides and was reduced 30-fold versus the WT enzyme, in agreement with the weak monomethyltransferase activity reported by Xiao et al. (11). In addition, we investigated whether this mutant catalyzes lysine trimethylation via a pro- cessive or distributive mechanism as described for SET7/9 Y305F. The TLC data illustrate that the mono- and dimethyl- lysine intermediates accumulated at quantities greater than that of the enzyme used in the assay, indicating that SET7/9 Y245A obeys a distributive mechanism, analogous to the Y305F mutant (Table 3). Structures of SET7/9 Y245A Bound to Unmodified and Meth- ylated TAF10 Peptides—To elucidate the mechanism underly- ing its unusual product specificity, we determined the crystal structures of SET7/9 Y245A in complex with AdoHcy and unmodified, mono-, di-, and trimethylated TAF10 peptides (supplemental Table 1). These complexes superimpose with the structure of the WT SET7/9AdoHcyTAF10-K189 com- FIGURE 2. Crystal structures of WT SET7/9 and the Y305F mutant in complex with AdoHcy and unmodi- fied and methylated TAF10 peptides. A, active site of WT SET7/9 bound to AdoHcy and TAF10-K189 and the active site of SET7/9 Y305F bound to AdoHcy and TAF10-K189 (B), TAF10-K189me1 (C), and TAF10-K189me2 peptides (D). A–D, the K189 residue in the TAF10 peptide is depicted with yellow carbon atoms with the corresponding Fo – Fc simulated-annealing omit maps (contoured at 2.5 ) shown. AdoHcy is colored with green carbon atoms, and SET7/9 carbon atoms are colored gray, with the exception of Tyr-305/Y305F (magenta) and Tyr-245 (cyan). Gly-264 to Ser-268 are shown, and the main chain atoms of Gly-292 to Ala-295 are depicted for clarity. Red dashed lines indicate conventional hydrogen bonds, and blue dashed lines indicate CH–O hydrogen bonds. Cyan dashed lines indicate weak CH–O hydrogen bonds between 3.5 and 3.7 Å in length. The water molecule in the solvent binding pocket is numbered 1. E, overlay of the active sites of the WT enzyme (light blue carbon atoms) and the Y305F mutant (green, yellow, and pink carbon atoms correspond to K189, K189me1, and K189me2, respectively). Red dashed lines indicate coordination of the waters in the struc- turesoftheY305Fmutant,andthelightbluedashedlineindicateshydrogenbondingthatonlyoccursintheWT enzyme. F, superimposition of the active sites of the SET7/9 Y305FAdoHcyTAF10-K189me2 and SET8 Y334FAdoHcyH4K20me2 (Protein Data Bank code 3F9X) complexes shown with green and yellow carbon atoms, respectively. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31853 plex with root mean squared differences of less than 0.4 Å for all aligned atoms, indicating that the Y245A mutant does not per- turb the overall structure of the enzyme. Simulated annealing omit maps illustrate that K189 side chains are bound within the lysine binding channel through hydrogen bonds and van der Waals contacts (Fig. 3, A–D), although the interactions and binding modes are distinct from those in the complexes of WT SET7/9 and the Y305F mutant (Fig. 2, A–D). In the unmodified TAF10 peptide complex, the K189 -amino group forms a weak hydrogen bond with the carbonyl oxygen of Gly-264 (Fig. 3A), whereas the -amino groups of K189me1 and K189me2 hydro- gen bond to the hydroxyl group of Tyr-305 in the mono- and dimethylated peptide complexes (Fig. 3, B and C). The confor- mations of the K189me1 and K189me2 side chains are further stabilized by water-mediated hydrogen bonding and through CH–O hydrogen bonding to their methyl groups. In the TAF10-K189me3 peptide complex, the trimethyl-lysine side chain is coordinated exclusively through direct and water-me- diated CH–O hydrogen bonds to its methyl groups because the quaternary -ammonium cation cannot engage in hydrogen bonding (Fig. 3D). A structural alignment of the four SET7/9 Y245A complexes illustrates distinct binding modes for the unmodified versus the methylated K189 side chains, highlighting the selectivity of this mutant for methylated substrates. The side chains of K189me1, K189me2, and K189me3 roughly overlay with their respective -amino groups superimposed and adopt slightly kinked con- formations (Fig. 3E), as opposed to the extended trans geome- try of the unmodified and methylated lysines in the complexes of the WT enzyme and the Y305F mutant (Fig. 2, A–D). Con- versely, the unmodified K189 side chain does not superimpose with its methylated counterparts and is oriented in an alterna- tive configuration due to its hydrogen bonding to Gly-264 (Fig. 3, A and E). An overlay of the structures of the WT enzyme and Y245A mutant bound to the unmodified TAF10 peptide illus- trates that the side chains of K189 do not superimpose and that the K189 -amino group appears to be misaligned with AdoHcy in the Y245A complex (Fig. 3F). This suboptimal alignment may explain the diminished kcat value of SET7/9 Y245A mutant toward substrates with unmodified lysines (Table 2). A comparison of the structures of the SET7/9 Y245A and Y305F complexes yields a molecular explanation for the differ- ent product specificities of these two mutants. In the SET7/9 Y305F complexes, Tyr-245 aligns the K189 -amino group for methyl transfer through hydrogen bonding to its hydroxyl group (Fig. 2, B–D). Conversely, in the Y245A mutant, the K189me1 and K189me2 -amino groups are oriented through hydrogen bonding to Tyr-305 (Fig. 3, B and C). These distinct hydrogen bond patterns impart differences in the conforma- tions of the lysyl side chains due to the relative orientations of Tyr-245 and Tyr-305 in the lysine binding channel. Specifically, the kinked conformation adopted by the K189me1 and K189me2 side chains in the Y245A complexes (Fig. 3, B and C) may contribute to the differences in the turnover numbers of this mutant versus those of the WT enzyme and the Y305F mutant (Table 2). In addition, the dimethyl -amino group of the K189me2 side chain binds in distinct orientations in the Y245A and Y305F mutants due to their hydrogen bonding to FIGURE 3. Crystal structures of the SET7/9 Y245A mutant in complex with AdoHcy and unmodified and methylated TAF10 peptides. Active site of SET7/9 Y245A bound to AdoHcy and TAF10-K189 (A), TAF10-K189me1 (B), TAF10-K189me2 (C), and TAF10-K189me3 peptides (D). A–D, Fo – Fc simulated- annealing omit maps (contoured at 2.5 ) for the unmodified and methylated K189 side chains are illustrated. The residues and hydrogen bonds in each com- plexarecoloredasinFig.2.Thewatermoleculesinthelysinebindingchannelsof the Y245A complexes are numbered 1–4, as described in the text. E, superimpo- sition of the active sites of the Y245A complexes bound to the four methylated statesofTAF10-K189(unmodified,mono-,di-,andtrimethylatedshowningreen, yellow, pink, and blue, respectively). F, overlay of WT SET7/9 (blue carbons) and SET7/9 Y245A (green carbons) in complex with TAF10-K189. Waters correspond- ing to the WT and Y245A structures are colored cyan and green, respectively. G, alignment of the active sites of the Y245A (yellow carbons) and Y305F (green carbons) mutants in complex with the TAF10-K189me2 peptide. Hydrogen bonds from the Y305F structure are shown as green dashed lines, and waters and hydrogen bonds in the Y245A structure are shown in yellow and orange, respectively. Lysine Methylation by SET7/9 Mutants 31854 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 Tyr-305 and Tyr-245, respectively (Fig. 3G). In the Y305F mutant, hydrogen bonds to the dimethyl -amino group cou- pled with steric constraints in the lysine binding channel pre- vent the K189me2 side chain from undergoing a conforma- tional change that is conducive to trimethylation (Fig. 2D), consistent with its dimethyltransferase activity. However, in the Y245A mutant, the alanine substitution enlarges the diameter of the lysine binding channel, accommodating trimethyl-lysine (Fig. 3D). In addition, the larger diameter would permit the dimethyl-lysine substrate to undergo the conformational reor- ganization necessary to align the -amino group in a productive geometry for trimethylation. A major difference in the active site of the Y245A mutant versus the other SET7/9 structures is the presence of several water molecules bound in the cavity generated by the Y245A mutation. In the structure of the Y245A mutant bound to TAF10-K189, three water molecules (waters 2–4) occupy this cavity and are arranged in a triangular geometry (Fig. 3A). In addition, water 1 shifts 1.6 Å from its position in the solvent pocket toward water 2 to which it forms a hydrogen bond (Fig. 3, A and E). The shift in water 1 was unexpected given its con- served orientation in the solvent pocket of the SET7/9 WT and Y305F complexes (Fig. 2, A–C) as well as in the structures of other SET domain KMTs (8). This displacement is presumably related to the alternative conformation of the K189 side chain whose -amino group is too distant (4.3 Å) to form a productive hydrogen bond to water 1. Conversely, in the Y245A complexes bound to TAF10-K189me1 and TAF10-K189me2, water 1 remains tightly bound in the solvent pocket through hydrogen bonds to Tyr-305 hydroxyl group, the carbonyl oxygens of Gly- 292 and Ala-295, and the K189 -amino group (Fig. 3, B, C, and E), analogous to its binding in the WT enzyme (Fig. 2A). How- ever, in the TAF10-K189me3 complex, one of the methyl groups of the trimethyl -ammonium cation is oriented into the solvent pocket (Fig. 3D), similar to the dimethyl-lysine binding mode observed in the Y305F mutant (Fig. 2D). The binding of the methyl group in the solvent pocket displaces water 1 by 3.2 Å relative to its position in the TAF10-K189me1 complex (Fig. 3E), thereby avoiding a steric clash with the trimethylated -ammonium group. Variations in the occupancy of water 2 are also seen in the different Y245A structures. Water 2 is bound in similar orientations in the active site of the unmodified and monomethylated peptide complexes but is absent in the di- and trimethylated peptide complexes due to the binding of a methyl group in this position (Fig. 3, A–E). In summary, the changes in the positions or occupancies of waters 1 and 2 correlate with the binding modes of the unmodified and methylated K189 within the active site of the Y245A mutant. Catalytic Models of Lysine Multiple Methylation by SET7/9 Y245A, and Y305F—The structures of the SET7/9 complexes reported here offer a prime opportunity to generate stepwise models for lysine mono-, di-, and trimethylation by a SET domain KMT. We modeled the AdoMet-bound Michaelis complexes by superimposing the SET7/9 product complexes with the previously reported structure of the SET7/9-AdoMet binary complex (Fig. 4) (39). The conformations of the mono- and dimethyl -amino groups in the Michaelis complexes were inferred from the coordinates of the corresponding dimethyl- and trimethyl-lysine products, respectively. In addition, we modeled the -amino group in a deprotonated state with its hydrogen atoms oriented toward the hydrogen bond acceptors that align the lysyl side chain for methylation. As a basis for this comparison, we first modeled the monomethylation reaction catalyzed by WT SET7/9 (Fig. 4A). In the substrate ternary complex, the lysine -amine is aligned with the methyl group and sulfonium cation of AdoMet through a hydrogen bond to the Tyr-245 hydroxyl group and water 1 in the solvent pocket. The values of the reaction distance and angle are 2.8 Å and 153°, respectively, in approximate agreement with the linear geome- try of a SN2 methyl transfer reaction calculated in other mod- eled substrate complexes (8, 10). In the product complex, the monomethyl-lysine side chain is bound in an extended confor- mation with its methyl group oriented within the methyl trans- fer pore, thereby obstructing AdoMet binding. Furthermore, water 1 remains tightly coordinated in the solvent pocket through four hydrogen bonds to Gly-292, Ala-295, Tyr-305, and the monomethyl -amino group. These interactions hinder the dissociation of water 1 and the related rearrangement of the monomethyl-lysine side chain required for a second methyl transfer reaction, explaining why the WT enzyme cannot cata- lyze di- and trimethylation. These findings concur with the FIGURE 4. Catalytic models of the methyl transfer reactions catalyzed by WT SET7/9 and the Y305F mutant. A, monomethylation of TAF10-K189 by the WT enzyme. The reaction scheme depicts the modeled substrate ternary complex (left) and the product complex (right) for the transfer of the methyl group from AdoMet (green carbon atoms) to K189 in TAF10 (yellow carbons), yielding AdoHcy and K189me1. The red arrow indicates the direction of the nucleophilic attack of the deprotonated -amino group on the AdoMet methyl group. The transferred methyl group is colored green, and the white atoms represent the hydrogens of the -amino group. Hydrogen bonds and residues in the enzyme active site are illustrated as in Fig. 2. The reaction distance and angle are labeled in red. B and C, models of the Y305F mutant for the first methyl transfer reaction with TAF10-K189 (B) and second methyl transfer reaction with TAF10-K189me1 (C). Color schemes are the same as in A. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31855 model for SET7/9 product specificity reported in previous structural and functional studies (6, 11). Similar reaction geometry is observed in the model for the monomethyl transfer reaction catalyzed by SET7/9 Y305F. Hydrogen bonds from the Tyr-245 hydroxyl group and water 1 align the lysine -amino group with the AdoMet methyl group at a distance of 2.1 Å and an angle of 160°, equivalent to those measured in the Michaelis complex of the WT enzyme (Fig. 4, A and B). In the product complex, monomethyl-lysine adopts an extended trans configuration analogous to that in the WT enzyme. For dimethylation to occur, the monomethyl-lysine must undergo a conformational change in which its methyl group is rotated out of the methyl transfer path with AdoMet. The structure of the Y305F mutant bound to the dimethylated TAF10 peptide (Fig. 2D) implies that this rearrangement occurs through the dissociation of water 1 due to the loss of the Tyr- 305 hydrogen bond in the solvent pocket. The dissociation of water 1 would enable the monomethyl-lysine side chain to adopt an alternative conformation through a rotation about its C–N bond, projecting the methyl group into the solvent pocket (Fig. 4C). This rotation reorients the methyl group out of the methyl transfer path while realigning the monomethyl - amino group for a second methylation reaction through a direct hydrogen bond to the Tyr-245 hydroxyl group and a CH–O hydrogen bond between its methyl group and Ala-295. The modeled reaction geometry for monomethyl-lysine substrate complex of the Y305F mutant (2.7 Å, 160°) is equivalent to that of the first methyl transfer reaction in SET7/9 Y305F. These geometries concur with our previous models for mono- and dimethylation catalyzed by SET8 Y334F (8), illustrating that the orientation of a methyl group into the solvent pocket is a con- served feature of SET domain KMTs that catalyze multiple methylation. In addition, we modeled the methyl transfer reactions cata- lyzed by SET7/9 Y245A (Fig. 5). In the model of the lysine sub- strate complex, the -amino group is aligned for methyl transfer by a hydrogen bond to the carbonyl oxygen of Gly-264, result- ing in a short reaction distance (2.3 Å) and a suboptimal reac- tion angle (141°) with the methyl group of AdoMet (Fig. 5A). This misalignment appears to be a direct consequence of the Y245A mutation that abolishes hydrogen bonding to the - amino group, illustrating that the suboptimal orientation of the -amine likely contributes to the diminished activity of this mutant toward unmodified substrates (11). Conversely, in the modeled monomethyl-lysine substrate complex for SET7/9 Y245A, Tyr-305 hydrogen bonds to the -amino group, aligning it for methyl transfer (Fig. 5B). In addition, CH–O hydrogen bonds to the monomethyl-lysine methyl group and the dissoci- ation of water 2 from the active site also contribute to reposition- ing the -amino group for dimethylation. Collectively, these inter- actions orient the -amine in a reaction angle of 165° that is more conducive to methyl transfer. However, the reaction distance for dimethylation is 0.6 Å longer than that in the corresponding Y305F model because Tyr-305 is positioned further from AdoMet than Tyr-245 (Figs. 4C and 5B). In the third methyl transfer reaction catalyzed by SET7/9 Y245A, the lone pair of electrons of the dimethyl-lysine - amino group acts as the nucleophile and thus cannot engage in hydrogen bonding. The structure of the trimethyl-lysine prod- uct complex (Fig. 3D) implies that the dimethyl -amine is aligned via CH–O hydrogen bonds to its methyl groups, as shown in the model of the Michaelis complex for this reaction (Fig. 5C). These CH–O hydrogen bonds restrain the orienta- tion of the -amino group and position one of the methyl groups into the solvent pocket, displacing water 1 as discussed earlier (Fig. 3, D and E). These interactions cumulatively align the -amino group and AdoMet methyl group with a reaction distance of 3.0 Å and angle of 162° (Fig. 5C). Taken together, the models of the substrate complexes for SET7/9 Y245A suggest that CH–O hydrogen bonds play an increasingly important role in aligning the methylated -amino group in successive rounds of methyl transfer. DISCUSSION The structural and functional characterization of the SET7/9 Y245A and Y305F mutants presented here yields new insights into the mechanism underlying the product specificity of SET domain KMTs. Importantly, it resolves a general paradox concerning this specificity. How does the active site constrain the motion of the lysine -amino group to align it for methyl transfer with AdoMet, while providing adequate volume to accommodate the mono-, di-, and tri- methylated lysine side chain generated during multiple methyl transfer reactions? The structures of the Y305F and Y245A mutants resolve this paradox, illustrating that alter- ations in the positions or occupancies of water molecules within their active sites generate the space required to FIGURE 5. Models for the methyl transfer reactions catalyzed by the SET7/9 Y245A mutant. Models of the Y245A mutant for the first methyl transfer reaction with TAF10-K189 (A), the second methyl transfer reaction with TAF10-K189me1 (B), and the third methyl transfer reaction with TAF10- K189me2 (C). Residues and hydrogen bonds are depicted as in Fig. 3. Lysine Methylation by SET7/9 Mutants 31856 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010 accommodate the multiply methylated -amine produced during successive catalytic cycles. Minor perturbations in the side chains of certain active site residues, such as Tyr- 305, are also observed in alignments of the WT enzyme and the Y245A and Y305F complexes, although these changes are modest compared with the displacement or dissociation of the water molecules in the active site. These findings suggest that the waters function as transient place holders that facil- itate the SN2 methyl transfer reaction. During monomethy- lation, they function to constrain the movement of the lysine -amino group by mediating hydrogen bonds between the substrate and enzyme, thereby promoting the linear align- ment with the methyl group and sulfonium cation of AdoMet (Fig. 4, A and B). During di- and trimethylation, the water molecules either relocate within the lysine binding channel or dissociate from the enzyme, yielding the space required to rotate the methyl group away from the methyl transfer pore and to realign the -amine in productive geom- etry for the next methyl transfer reaction (Figs. 4C and 5, B and C). These findings agree with our prior analysis of the SET8 Phe/Tyr switch mutant in which we demonstrated that the Y334F substitution attenuates hydrogen bonding to the water molecule bound in the solvent pocket, promoting its dissociation and the conformational changes necessary for lysine dimethylation (8). Indeed, there is a nearly identical alignment of the dimethyl-lysine side chains in the structures of SET7/9Y305FandSET8Y334Fcomplexes,despitethedifferences in the orientations of the Phe-305 and Phe-334 side chains in each structure (Fig. 2F). Finally, both the SET8 Y334F (8) and SET7/9 Y305F mutants (Table 2) displayed diminished catalytic efficien- cies for lysine dimethylation versus monomethylation. These dif- ferences may reflect the kinetics of the reorganization within the active site, including the dissociation of the water molecule from the solvent pocket and the concomitant realignment of the monomethyl-lysine into a productive geometry for dimethylation. In addition to their place-holding role, the active site waters may also facilitate the deprotonation of the lysine -amino group between methyl transfer reactions. For methylation to occur, the -amino group must be deprotonated to function as the nucleophile in the SN2 methyl transfer reaction with AdoMet (Figs. 4 and 5). Although the pKa value of the lysine -amine in solution is 10.5, molecular dynamics simulations by Zhang and Bruice (25, 26) indicate that this value diminishes to 8.2 upon formation of the SET7/9 Michaelis complex due to the proximity of the AdoMet sulfonium cation and the low dielectric constant of the active site. Furthermore, their simu- lations show that a chain of water molecules facilitates the dep- rotonation of the -amino group prior to methyl transfer, trans- ferring the proton to bulk solvent. Although these water molecule chains are not evident in our crystal structures, the Y305F and Y245A complexes suggest another potential mech- anism for deprotonation. In the dimethyl-lysine complexes of the Y305F and Y245A mutants, the dissociation of water 1 and 2, respectively, from the lysine binding channel requires that the solvent-mediated hydrogen bond to the -amino group is broken (Figs. 2D and 3C). It is conceivable that these waters dissociate from the active site as hydronium ions, promoting the realignment and deprotonation of the methyl -amino group for the next methyl transfer reaction. A comparison of the SET7/9 Y305F and SET8 Y334F com- plexes yields insights into the mechanism by which the Phe/Tyr switch influences water binding within the solvent pocket. The phenylalanine substitution in the Phe/Tyr switch results in the loss of a single hydrogen bond to the water molecule (water 1) in the solvent pocket compared with the four hydrogen bonds that coordinate the solvent molecule in WT SET7/9 (Fig. 2, A and B) and SET8 (7, 8). Although this attenuation in hydrogen bonding may appear insignificant, this difference is nonetheless impor- tant for at least two reasons. First, theoretical calculations indi- cate that, on average, water molecules form 3.5 hydrogen bonds in solutions (40, 41). This value is greater than the num- ber of hydrogen bonds coordinating water 1 in the solvent pocket in SET7/9 Y305F (Fig. 2, B and C) as well as in SET8 Y334F and other di- and trimethyltransferases that possess a hydrophobic residue in the Phe/Tyr switch site (8). From the perspective of the water molecule, the greater hydrogen bond- ing potential in solution would tend to thermodynamically favor its dissociation from the solvent pocket in SET domain KMTs that lack a tyrosine in the Phe/Tyr switch position. Sec- ond, the ordered binding of water molecules observed in the active sites of SET domain ternary complexes represents an unfavorable entropy compared with their diffusion in bulk sol- vent. In WT SET7/9 (Fig. 2A) and SET8 (7, 8), this entropic penalty can be partially offset through the favorable enthalpy of binding associated with the four hydrogen bonds that coordi- nate the water within the solvent pocket. It is conceivable that the loss of the tyrosine-mediated hydrogen bond in the Phe/Tyr switch shifts the equilibrium in favor of dissociation of the water molecule from the solvent pocket, thereby facilitating dimethylation in SET7/9 Y305F, SET8 Y334F, and other di- and trimethyltransferases. The structures of the SET7/9 Y245A and Y305F complexes illustrate the interactions that align the lysine -amino group during the methyl transfer reactions in each enzyme. In the WT enzyme and the Y305F mutant, hydrogen bonding to the hydroxyl group of Tyr-245 appears to be critical in properly aligning the -amine for methyl transfer (Fig. 4). Tyr-245 is conserved in the sequences of many SET domain KMTs (8, 42), and substitutions of this residue generally impair or abolish activity, indicating its importance in catalysis (8, 43). However, SET7/9 appears to be an exception to this rule, as the Y245A mutant is not only active but is capable of catalyzing lysine trimethylation. In this mutant, Tyr-305 appears to assume the role of Tyr-245 by hydrogen bonding to the monomethylated -amino group to align it for methyl transfer with AdoMet, as illustrated in the modeled substrate complex for the dimethy- lation reaction (Fig. 5B). Conversely, in the model for trimethy- lation, the Tyr-305 hydroxyl group does not hydrogen bond to the -amine but instead participates in a CH–O hydrogen bond with one of the methyl groups to assist in aligning the dimethy- lated -amine for the methyl transfer reaction (Fig. 5C). Addi- tional structural and functional studies of the SET domain tri- methyltransferases will aid in further illuminating the roles of CH–O hydrogen bonds in facilitating lysine multiple methylation. Lysine Methylation by SET7/9 Mutants OCTOBER 8, 2010•VOLUME 285•NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 31857 Acknowledgments—We acknowledge S. Schiebold for assistance in protein expression, purification, and crystallization and S. Anderson and R. Sanishvili for their assistance with x-ray data collection. We also thank S. Bulfer and S. Horowitz for reading the manuscript and providing useful comments. This work utilized the Protein Structure Facility of the Michigan Diabetes Research and Training Center, Uni- versity of Michigan, supported by National Institutes of Health Grant DK020572, NIDDK. Use of the Advanced Photon Source was sup- ported by the United States Department of Energy, Basic Energy Sci- ences, Office of Science, under Contract DE-AC02-06CH11357. GM/CA CAT has been funded in whole or in part by National Institutes of Health NCI Grant Y1-CO-1020 and NIGMS Grant Y1-GM-1104. Use of the LS-CAT Sector 21 was supported by Michi- gan Economic Development Corporation and the Michigan Technol- ogy Tri-Corridor Grant 085P1000817 for the support of this research program. REFERENCES 1. Huang, J., and Berger, S. L. (2008) Curr. Opin. Genet. Dev. 18, 152–158 2. Morgunkova, A., and Barlev, N. A. (2006) Cell Cycle 5, 1308–1312 3. Yang, X. D., Lamb, A., and Chen, L. F. (2009) Epigenetics 4, 429–433 4. Taverna, S. D., Li, H., Ruthenburg, A. J., Allis, C. D., and Patel, D. J. (2007) Nat. Struct. Mol. Biol. 14, 1025–1040 5. Collins, R. E., Tachibana, M., Tamaru, H., Smith, K. M., Jia, D., Zhang, X., Selker, E. U., Shinkai, Y., and Cheng, X. (2005) J. Biol. Chem. 280, 5563–5570 6. Zhang, X., Yang, Z., Khan, S. I., Horton, J. R., Tamaru, H., Selker, E. U., and Cheng, X. (2003) Mol. Cell 12, 177–185 7. Couture, J. F., Collazo, E., Brunzelle, J. S., and Trievel, R. C. (2005) Genes Dev. 19, 1455–1465 8. Couture, J. F., Dirk, L. M., Brunzelle, J. S., Houtz, R. L., and Trievel, R. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 20659–20664 9. Qian, C., Wang, X., Manzur, K., Sachchidanand, Farooq, A., Zeng, L., Wang, R., and Zhou, M. M. (2006) J. Mol. Biol. 359, 86–96 10. Trievel, R. C., Flynn, E. M., Houtz, R. L., and Hurley, J. H. (2003) Nat. Struct. Biol. 10, 545–552 11. Xiao, B., Jing, C., Wilson, J. R., Walker, P. A., Vasisht, N., Kelly, G., Howell, S., Taylor, I. A., Blackburn, G. M., and Gamblin, S. J. (2003) Nature 421, 652–656 12. Chuikov, S., Kurash, J. K., Wilson, J. R., Xiao, B., Justin, N., Ivanov, G. S., McKinney, K., Tempst, P., Prives, C., Gamblin, S. J., Barlev, N. A., and Reinberg, D. (2004) Nature 432, 353–360 13. Ea, C. K., and Baltimore, D. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 18972–18977 14. Este`ve, P. O., Chin, H. G., Benner, J., Feehery, G. R., Samaranayake, M., Horwitz, G. A., Jacobsen, S. E., and Pradhan, S. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 5076–5081 15. Kouskouti, A., Scheer, E., Staub, A., Tora, L., and Talianidis, I. (2004) Mol. Cell 14, 175–182 16. Masatsugu, T., and Yamamoto, K. (2009) Biochem. Biophys. Res. Commun. 381, 22–26 17. Munro, S., Khaire, N., Inche, A., Carr, S., and La Thangue, N. B. (2010) Oncogene 29, 2357–2367 18. Pagans, S., Kauder, S. E., Kaehlcke, K., Sakane, N., Schroeder, S., Dorm- eyer, W., Trievel, R. C., Verdin, E., Schnolzer, M., and Ott, M. (2010) Cell Host Microbe 7, 234–244 19. Subramanian, K., Jia, D., Kapoor-Vazirani, P., Powell, D. R., Collins, R. E., Sharma, D., Peng, J., Cheng, X., and Vertino, P. M. (2008) Mol. Cell 30, 336–347 20. Wang, J., Hevi, S., Kurash, J. K., Lei, H., Gay, F., Bajko, J., Su, H., Sun, W., Chang, H., Xu, G., Gaudet, F., Li, E., and Chen, T. (2009) Nat. Genet. 41, 125–129 21. Yang, X. D., Huang, B., Li, M., Lamb, A., Kelleher, N. L., and Chen, L. F. (2009) EMBO J. 28, 1055–1066 22. Guo, H. B., and Guo, H. (2007) Proc. Natl. Acad. Sci. U.S.A. 104, 8797–8802 23. Hu, P., Wang, S., and Zhang, Y. (2008) J. Am. Chem. Soc. 130, 3806–3813 24. Hu, P., and Zhang, Y. (2006) J. Am. Chem. Soc. 128, 1272–1278 25. Zhang, X., and Bruice, T. C. (2007) Biochemistry 46, 14838–14844 26. Zhang, X., and Bruice, T. C. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 5728–5732 27. Couture, J. F., Collazo, E., Hauk, G., and Trievel, R. C. (2006) Nat. Struct. Mol. Biol. 13, 140–146 28. Trievel, R. C., Beach, B. M., Dirk, L. M., Houtz, R. L., and Hurley, J. H. (2002) Cell 111, 91–103 29. Kapust, R. B., To¨zse´r, J., Fox, J. D., Anderson, D. E., Cherry, S., Copeland, T. D., and Waugh, D. S. (2001) Protein Eng. 14, 993–1000 30. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307–326 31. Vagin, A., and Teplyakov, A. (2000) Acta Crystallogr. D Biol. Crystallogr. 56, 1622–1624 32. Emsley, P., and Cowtan, K. (2004) Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 33. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Acta Crystallogr. D Biol. Crystallogr. 53, 240–255 34. Davis, I. W., Leaver-Fay, A., Chen, V. B., Block, J. N., Kapral, G. J., Wang, X., Murray, L. W., Arendall, W. B., 3rd, Snoeyink, J., Richardson, J. S., and Richardson, D. C. (2007) Nucleic Acids Res. 35, W375–W383 35. Bru¨nger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. D Biol. Crystallogr. 54, 905–921 36. Collazo, E., Couture, J. F., Bulfer, S., and Trievel, R. C. (2005) Anal. Bio- chem. 342, 86–92 37. Chirpich, T. P., Zappia, V., Costilow, R. N., and Barker, H. A. (1970) J. Biol. Chem. 245, 1778–1789 38. Couture, J. F., Hauk, G., Thompson, M. J., Blackburn, G. M., and Trievel, R. C. (2006) J. Biol. Chem. 281, 19280–19287 39. Kwon, T., Chang, J. H., Kwak, E., Lee, C. W., Joachimiak, A., Kim, Y. C., Lee, J., and Cho, Y. (2003) EMBO J. 22, 292–303 40. Chandra, A., and Chowdhuri, S. (2002) J. Phys. Chem. B 106, 6779–6783 41. Guardia, E., Marti, J., Garcia-Tarres, L., and Laria, D. (2005) J. Mol. Liq. 117, 63–67 42. Dillon, S. C., Zhang, X., Trievel, R. C., and Cheng, X. (2005) Genome Biol. 6, 227 43. Zhang, X., Tamaru, H., Khan, S. I., Horton, J. R., Keefe, L. J., Selker, E. U., and Cheng, X. (2002) Cell 111, 117–127 Lysine Methylation by SET7/9 Mutants 31858 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 41•OCTOBER 8, 2010
3M5C
Crystal structure of N-acetyl-L-ornithine transcarbamylase K302E mutant complexed with PALAO
Reversible Post-Translational Carboxylation Modulates The Enzymatic Activity Of N-Acetyl-L-Ornithine Transcarbamylase† Yongdong Li1,2, Xiaolin Yu1, Jeremy Ho1, David Fushman3, Norma M. Allewell3, Mendel Tuchman1, and Dashuang Shi1,‡ 1Research Center for Genetic Medicine and Department of Integrative Systems Biology, Children’s National Medical Center, The George Washington University, Washington, DC 20010, USA. 2Key Laboratory of Organo-Pharmaceutical Chemistry, Jiangxi Province, Gannan Normal University, Ganzhou 341000, China. 3Department of Chemistry and Biochemistry, College of Chemical and Life Sciences, University of Maryland, College Park, MD 20742, USA. Abstract N-acetyl-L-ornithine transcarbamylase (AOTCase), rather than ornithine transcarbamylase (OTCase), is the essential carbamylase enzyme in the arginine biosynthesis of several plant and human pathogens. The specificity of this unique enzyme provides a potential target for controlling the spread of these pathogens. Recently, several crystal structures of AOTCase from Xanthomonas campestris (xc) have been determined. In these structures, an unexplained electron density at the tip of Lys302 side-chain was observed. Using 13C NMR spectroscopy, we show herein that Lys302 is post-translationally carboxylated. The structure of wild-type AOTCase complexed with the bisubstrate analogue, Nδ-(phosphonoacetyl)-Nα-acetyl-L-ornithine (PALAO), indicates that the carboxyl group on Lys302 forms a strong hydrogen bonding network with surrounding active site residues, Lys252, Ser253, His293, and Glu92 from the adjacent subunit either directly or via a water molecule. Furthermore, the carboxyl group is involved in binding N-acetyl-L-ornithine via a water molecule. Activity assays with the wild-type enzyme and several mutants demonstrate that the post translational modification of lysine 302 has an important role in catalysis. Post-translational modification of the ε-amino group of lysine residues in proteins is a common mechanism used by organisms to regulate protein functions including DNA-protein interactions, subcellular localization, transcriptional activity, and protein stability and activity (1). Lysine residues can be modified by the addition of functional groups to become acetylated, methylated, carbamylated or carboxylated. The role of histone lysine acetylation and methylation in affecting chromatin structure and gene expression has been well established for more than a decade (2). However, the biological roles for lysine carbamylation and carboxylation have rarely been investigated. †This work was supported by Public Health Service grants DK-47870 (MT) and DK-067935 (DS) from the National Institute of Diabetes, Digestive and Kidney Diseases. JH was supported by a Scholarship from the Doug and Lynn Parsons Family Foundation. The Cornell High Energy Synchrotron Source (CHESS) is supported by the National Science Foundation under award DMR 0225180 and the Macromolecular Diffraction Facility at CHESS (MacCHESS) is supported by award RR-01646 from the National Institutes of Health, through its National Center for Research Resources. ‡Corresponding author. dshi@cnmcresearch.org. Phone: 202-476-5817. Fax: 202-476-6014. SUPPORTING INFORMATION AVAILABLE Figure S1. Structure and hydrogen bonding network around residue 302 for previously determined AOTCase structures. This material is available free of charge via the Internet at http://pubs.acs.org. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 August 17. Published in final edited form as: Biochemistry. 2010 August 17; 49(32): 6887–6895. doi:10.1021/bi1007386. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript In vivo, lysine acetylation and methylation are usually carried out by acetyltransferase and methyltransferase enzymes, respectively (3). In addition, some proteins such as hemoglobin and human serum albumin can be acetylated non-enzymatically by chemicals such as aspirin, methyl acetyl phosphate, and other acetylating agents such as acetyl-CoA (4–8). Lysine can also be methylated by small chemicals in vitro, and this has routinely been used as a rescue method for protein crystallization (9). Lysine carbamylation and lysine carboxylation have only been achieved by using chemicals and no enzyme has yet been found to catalyze these modifications. Lysine carbamylation was one of the earliest post- translational modification of proteins to be elucidated when it was identified as a product of reversible denaturation-renaturation studies of proteins with urea (10,11). This carbamylation, which produces homocitrulline, has also been detected in uremic patients (12) and in patients with elevated plasma and/or urinary lysine levels (13). In contrast, lysine carboxylation is not as commonly reported, but has been identified in a number of proteins via crystal structure determinations. In most of these proteins, the carboxyl groups of modified lysines are involved in bridging two metal ions that play a structural role in the active site. In several other proteins, however, a direct role for a carboxylated lysine in the catalytic mechanism has been reported (14–16). N-acetyl-L-ornithine transcarbamylase (AOTCase, EC 2.1.3.9) was recently discovered to be part of a novel arginine biosynthesis pathway in plant pathogens of the Xanthomonadaceae family such as Xylella and Xanthomonas (17–19). These pathogens attack a variety of economically important crops including citrus fruits, cotton, tomatoes, and rice (20,21). Genome sequence analyses showed that an AOTCase-like gene is also present in some human pathogens such as Stenotrophomonas maltophilia and members of the genus Bacteroides (22). In the case of Bacteroides fragilis, this gene was later confirmed to encode another novel transcarbamylase, N-succinyl-L-ornithine transcarbamylase (SOTCase, EC 2.1.3.11) (23). Crystal structures of both AOTCase and SOTCase bound with substrate or substrate analogues have recently been determined (17,18,23). An extended density at the side-chain tip of Lys302 in AOTCase was observed suggesting a post- translational modification. Since Lys302 is located within the active site of AOTCase and is not present in SOTCase, it was proposed as one of three key signature residues to distinguish the two carbamylases (22). Here, we demonstrate that Lys302 is post- translationally modified by carboxylation and that this change affects the catalytic function of the enzyme. MATERIALS AND METHODS Materials All chemicals were purchased from Sigma Chemical Company unless otherwise specified. ANOR was purchased from Indofine Chemical Co., Inc. N-acetyl-L-citrulline was custom synthesized and purified by Chiral Quest Company. PALAO (>95% purity) was synthesized by IMI TAMI Institute of Research and Development Ltd. (19). xcAOTCase was prepared and purified as previously described (18). Mutants K302A (primer: 5’- CTGCGTCGCAACGTCGCGGCTACTGATGCGGTG-3’), K302E (primer:5’- CTGCGTCGCAACGTCGAGGCTACTGATGCGGTG-3’) and K302R (primer: 5’- CTGCGTCGCAACGTCAGGGCTACTGATGCGGTG-3’) were generated by site-directed mutagenesis using the “Quik Change” mutagenesis kit (Stratagene) according to the manufacturer’s protocol. The correct mutants were confirmed by DNA sequencing. Recombinant mutant proteins were expressed and purified in the same manner as the wild- type enzyme. Li et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Activity assay The modified colorimetric assay method, which detects the formation of the ureido group during the transcarbamylation reaction (24), was used to measure enzyme activity. CP and AORN concentration were kept constant as 4.0 mM. After an incubation of 5 minutes, the reaction was stopped by the addition of 1 ml of color reagent, as described previously (19). A set of tubes containing known amounts of N-acetyl-L-citrulline was included with each rack of enzyme assays to produce a standard curve for calculation of the enzyme specific activity. Mass spectrometric analysis In order to identify the post-translational modification, mass spectrometric analysis was carried out on a 4700 ABI TOF/TOF mass spectrometer (Applier Biosystems) based on the method described previously (25). In brief, 10 µg of native protein were digested overnight at 310 K using trypsin in 50 mM ammonium bicarbonate pH 7.4. After desalting using a C18 ZipTip micropipette tip, the resulting peptides were eluted in 10 µl of acetonitrile/0.1% TFA [70:30(v:v)]. The sample was mixed with matrix solution and spotted on a MALDI plate to be submitted to the mass spectrometric analysis. Chemical rescue experiments The assay in the presence of various selected chemical was conducted as described above. The stock solutions of small chemicals were titrated to the pH of the assay with KOH or HCl. 13C NMR experiments The wild-type and K302A mutant protein of AOTCase (~10 mg) was precipitated by degassed buffer (pH 4.5) containing 25 mM sodium acetate. After centrifugation, the precipitate was re-dissolved by adding a buffer containing 20 mM NaH13CO3, 100 mM Tris HCl (pH 8.0) and 50 mM NaCl. Before NMR experiments, 40 µl D2O was added to 500 µl protein sample. The 13C NMR spectra were collected on a Bruker Avance 600 spectrometer (operating at 14.1 T) equipped with a direct 13C-detection probe at 298 K. The experimental settings and processing parameters for the wild-type protein and K302A variant were identical. 512 transients were collected with 4K time domain points and a spectral width of 3019 Hz centered at 160 ppm. The spectra were processed using exponential multiplication with the line broadening factor set to 3Hz. The similarity of protein concentration in both samples was verified by 1H NMR (not shown). Crystallization, data collection and processing PALAO-bound wild-type and mutant AOTCase crystals were grown using the hanging-drop vapor diffusion method, with conditions similar to those used to produce native and ligand- complexed AOTCase crystals (18,23). 2.0 µl of ~10 mg/ml solution of AOTCase were mixed with 1.6 µl of reservoir solution and 0.4 µl PALAO solution (~0.01 M). The reservoir solution contained 20% (w/v) PEG 3350, 0.2 M lithium sulfate and 0.1 M bis-Tris, pH 6.0. Diffraction data for the PALAO-bound crystal were collected at 100 K at the F1 beam line of the Cornell High Energy Synchrotron Source. Data sets for the PALAO-bound mutant AOTCase crystals were obtained using a Rigaku anode x-ray generator in the Molecular Structure Section of the National Institute of Health. All data were processed using HKL2000 package (26) and reduced using the program TRUNCATE in the CCP4 suite (27). Data collection parameters are listed in Table 1. Li et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Structure solution and refinement Molecular replacement was used for phase determination of the PALAO-bound wild-type and mutant AOTCase structures. The coordinates of AOTCase (PDB code: 3KZO) after removal of ligands or water molecules were used for phase determination. Upon rigid-body refinement, electron density corresponding to the ligands could be clearly visualized. The ligands were built into the map using the graphics program O (28). Refinements were carried out using molecular annealing, energy minimization and restrained B factor refinement with the program CNS1.1 (29). During refinements, 5% of the reflections at various resolutions were randomly selected to set aside to calculate Rfree to monitor the progress of refinement (30). After every cycle of refinement, the model was manually adjusted using the program O (28). Water molecules were automatically assigned using the program WATERPICK of CNS. Model quality was checked using the program PROCHECK (31) to ensure good stereochemistry for all three models. The final refinement statistics are listed in Table 1. Figures 1 was drawn using the programs Pymol (http://www.pymol.org). Figure 3 was drawn using ChemDraw 8.0. The coordinates have been deposited with the RCSB PDB as entries 3M4J, 3M4N, 3M5C and 3M5D. RESULTS Lys302 in AOTCase is carboxylated To investigate the nature of the modification of Lys302 and how it affects catalytic activity, we revisited all AOTCase structures. In the PALAO-bound AOTCase structure, the electron density map clearly indicates that Lys302 is post-translationally modified (Figure 1A). The type of modification can include methylation, acetylation, carbamylation, and carboxylation. The shape of the electron density can been used to distinguish methyl groups from larger functional groups, but it is difficult to distinguish between acetyl, carbamyl, and carboxyl groups, all of which have three non-hydrogen atoms in a plane. Given the hydrogen bonding network with surrounding residues (Lys252, Ser253 and His293, Table 2), a carboxylated modification is the most likely choice for the modification of Lys302 in AOTCase. To exclude that the modification’s identity represents chemically stable moieties (methyl, acetyl, carbamyl), we analyzed trypsin digested fragments of purified AOTCase by TOF- TOF mass spectroscopy. As expected, only a peptide fragment with an unmodified Lys302 was observed, consistent with the lability of the carboxylic group in acidic solutions. At low pH, the carboxyl group is spontaneously released as carbon dioxide (14, 32), in contrast to other modified groups that are stably bound and can be observed by mass spectrometry analysis after proteolysis (33). The putative carboxyl group on the modified Lys302 forms direct hydrogen bonds with main-chain or side-chain nitrogen atoms of Lys252, Ser253 and His293 (Figure 1A and Table 2). Among these, Lys252 is involved in direct hydrogen bonding to the carboxyl group of the AORN moiety of PALAO, and His293 forms a strong hydrogen bond with the main-chain nitrogen atom of Leu295 in the conserved His293-Cys294-Leu295-Pro206 (HCLP) motif. The hydrogen bonding network between the carboxyl group of modified Lys302, His293 and the main-chain nitrogen atom of Leu295 is reminiscent of the similar hydrogen bonding network, Glu310-His302-Leu304 and Glu299-Leu272-Leu274, found in human and E. coli OTCase, respectively (34, 35). These three residues are conserved in all OTCase sequences, and the interactions between them are important for maintaining the HCLP motif in a specific conformation to orientate their main-chain oxygen atoms towards the active site. In all known transcarbamylase structures, a leucine residue corresponding to Leu295 is in an energetically unfavorable conformation and the peptide bond between this Li et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript leucine and Pro296 is in the cis conformation. In addition to the direct hydrogen bonding interaction above, the carboxyl group on the modified Lys302 interacts with the α-amino nitrogen atom of the AORN moiety of PALAO and Glu92 from the adjacent subunit via water molecules. When we revisited all previously determined AOTCase structures (see supplementary Figure S1) we found: (1) Lys302 was carboxylated in the absence of substrate binding, but substrate binding immobilizes the side-chain of Lys302 further by hydrogen bonding interaction via water molecules. (2) Water-mediated hydrogen bonding promotes interactions of carboxylated Lys302 with Glu92 from the adjacent subunit and the α-amino nitrogen atom of AORN. (3) Similarly to AOTCase, in the structure of SOTCase E92Z (Z = Ala, Ser, Pro, Val), mutant with N-succinyl-L-norvaline bound (22) the carboxylated Lys302 hydrogen bonds to the α-amino nitrogen atom and the succinyl carboxyl group of N- succinyl-L-norvaline via water molecules (Figure S1). To obtain direct, independent evidence for the carboxylation of Lys302, 13C NMR experiments were carried out with both wild-type protein and the K302A mutant. As observed for other proteins with carboxylated lysine (36,37), the strong 13C NMR signal at 164 ppm characteristic of a carboxyl group was clearly detectable in AOTCase wild-type protein labeled by 13C-bicarbonate, in contrast to the K302A mutant where the signal was weak (Figure 2). Since there are 17 other lysine residues in the protein, the weak signal seen for the K302A mutant might be caused by the adventitious carboxylation of another lysine with reduced pKa, as has been observed for the K392A mutant of the sensor domain of the BlaR protein (38). Functional and structural studies of Lys302 mutants To investigate the effect of lysine carboxylation on enzyme activity, Lys302 was mutated to alanine, glutamate or arginine. Each of these variants was expressed in E. coli and gave similar yields. Enzymatic assays demonstrated a significant decrease in enzymatic activity in all three mutants, reflecting the functional importance of Lys302 (Table 3). The level of enzymatic activity for the wild-type (WT) and three mutants was WT > K302A > K302E ≫ K302R. To determine the structural basis of these results, the WT and mutant enzymes bound with PALAO were crystallized and their structures were determined at 1.8–2.2 Å resolution. Only the K302R mutation had and appreciable effect on the structure of the protein. Since K302 is located near the AORN binding site, the mutations would weaken AORN binding to the active site. In the structure of the K302A mutant, three additional water molecules (labeled as w3, w4 and w5 in Figure 1B) replace the carboxylated lysine. The two water molecules (labeled w1 and w2 in Figure 1A–1D) that mediate the hydrogen bonding interaction of carboxylated Lys302 with PALAO and Glu92 from the adjacent subunit are also found in the K302A mutant structure. Furthermore, these water molecules maintain a similar hydrogen-bonding network to the wild-type enzyme. These results might explain why the K302A mutant retains significant catalytic activity (Table 3). To investigate whether adding short-chain carboxylic acids to the K302A mutant increases its activity as other enzymes (14, 15, 39, 40), the activity of the K302A mutant was measured in the presence of high formate and acetate concentration (0.5 M). Surprisingly, the activity of the K302A mutant was not significantly improved. The crystal structure of the K302A mutant soaking with the crystallization buffer in the presence of 0.5 M acetate was also determined (not shown) and it was observed that the same five water molecules were present in the cavity that replaced the side chain of the carboxylated lysine. This, the acetate’s inability to replace the water molecules in the crystal structure, is consistent with the unchanged activity assay results. Li et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The side-chain of Glu302 in the K302E mutant structure is well defined and anchored by hydrogen bonding interaction with the main-chain nitrogen atom of Arg298 and weakly hydrogen bonded to the main-chain nitrogen atom of Ser253 (Figure 1C). Two of three additional water molecules (w4 and w5) observed in the K302A mutant structure occupied the same position as the carboxyl oxygen atoms of Glu302 and form a similar hydrogen- bonding network. Relative to the PALAO-bound wild-type structure, there is only one more water molecule (w3) at the position of the carboxyl group of the carboxylated Lys302. This water molecule mediates a hydrogen bonding interaction between Glu302 and Lys252. Two common water molecules (w1 and w2) that interact with PALAO and Glu92 from the adjacent subunit, respectively, were also identified in the K302E structure. Our observation that the K302E mutant had lower enzymatic activity than that of the K302A mutant (Table 3) was surprising since the carboxyl group of the glutamate could conceivably function similarly to a carboxylated lysine. The explanation may be that, in the K302A mutant, the hydrogen bonding network is well maintained by water molecules in the cavity that replaces the carboxylated lysine. In particular, w3 is optimally located for strong hydrogen bonding to w1 (2.7 Å), which in turn binds AORN. The distances between w1 and the carboxyl oxygen of carboxylated Lys302 in all wild-type crystal structures are within 2.4–2.7 Å, but the distance between w1 and w2 in the K302E structure is significantly greater (3.2 Å). The weaker hydrogen bonding interaction may be a reason for lower enzymatic activity of the K302E mutant. In contrast to the K302A and K302E structures, the K302R structure shows a much larger reduction in enzyme activity relative to the wild-type enzyme. The electron density for the side-chain of Arg302 is weak and the temperature factor of its side-chain is 54.4 Å2, significantly higher than those of carboxylated Lys302 (44.7 Å2) and Glu302 (33.4 Å2), implying greater flexibility. Furthermore, the side-chain of Arg302 is oriented differently from the carboxyl group of carboxylated Lys302 and pushes the nearby residues His180, Pro181 and Lys182 outwards about 1.0 Å (Figure 1D). However, the water molecules involved in hydrogen bonding to the α-amino nitrogen atom of PALAO (w1) and the side- chains of Lys252 (w3) and Glu92 (w2) from the adjacent subunit are conserved. Consistent with the K302E structure, the distance between w1 and w2 is even greater (3.4 Å) than in the WT structure and the hydrogen bonding interaction between w2 and w3 is no longer observed. Thus, the almost undetectable enzymatic activity of the K302R mutant probably results from the changes at its active site, including the weakened hydrogen bonding network involved in substrate binding. DISCUSSION Several lines of evidence clearly indicate that Lys302 in AOTCase is carboxylated. First, the extra electron density indicates that the side-chain of Lys302 is modified. Second, the hydrogen bonding environment of Lys302 for hydrogen bonding interactions is compatible with a carboxyl group, but not for a positively charged lysine side-chain. Third, the modification is labile at low pH, since mass spectroscopy of samples prepared at low pH indicated that Lys302 was no longer modified. Fourth, the clear presence of the indicative 13C NMR signal at 164 ppm for wild-type protein and its absence in the K302A mutant confirms carboxylation of Lys302. It is well known that lysine carboxylation is non-enzymatic and reversible, while other post- translational modifications such as methylation, acetylation, and carbamylation are irreversible and detectable by mass spectroscopy. Furthermore, lysine methylation and acetylation usually require an enzyme-catalyzed reaction in vivo (41). Therefore, it is unlikely that such lysine modifications will be observed in recombinant proteins Li et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript overexpressed in a foreign host (e.g. E. coli). Lysine methylation can be achieved by using special chemicals in vitro, but these chemicals are not present in vivo. Lysine carbamylation and carboxylation use completely different mechanisms to form functionally different groups (Figure 3). Carbamylation can be achieved by cyanate produced from myeloperoxidase-catalyzed oxidation of thiocyanate, an anion abundant in plasma and increased in smokers, or from urea in the plasma. Lysine carboxylation, on the other hand, occurs readily in aqueous solution in the presence of carbon dioxide at a basic pH (32,42). Even though carbamylation and carboxylation use very different mechanisms, the two are confused in the literatures. Lysine carbamylation (or carbamoylation) is referred to in several publications (15,32,42–44), when the actual reaction is in fact carboxylation. The activity of the K302A mutant is almost half of that of the wild-type enzyme raising the question of why AOTCase retains a lysine in this position. Perhaps this lysine was maintained through evolution to distinguish AOTCase from SOTCase which uses N- succinyl-L-ornithine (SORN) rather than AORN (22), and OTCase which uses L-ornithine. An alternative explanation may be found in the very low activity of the K302R mutant. The side-chain of arginine has a positive charge while carboxylated lysine has a negative charge. The side chain of unmodified lysine is usually located in a similar position as that of arginine, as observed in the structure of UV damage endonuclease (14). It would be expected that the activity of AOTCase with an uncarboxylated lysine would be as low as the K302R mutant’s. It could further be surmised that, the respective organisms need to use carboxylation as a switch to turn “on” or “off” the arginine biosynthetic pathway. It has been well known that rubulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) in plant cells uses the carboxylation on Lys201 as a switch to turn the enzyme “on” during the day and “off” at night by removing the carboxyl group (45,46). Carbon dioxide and bicarbonate have been found to play an important biological role in modulating several biological processes including photosynthetic carbon fixation (47), pH homeostasis (48), carbon metabolism (49), activation of virulence in pathogenic organisms (50), sperm maturation (51), stimulation of mammalian G-protein-responsive adenylyl cyclase (52), and as an alarmone in Drosophila (53,54). Whether or not carboxylation of a key lysine in their related proteins is used as an underlying regulatory mechanism should be investigated further. There are 197 structures with carboxylated lysine residue (modified residue indicated as Kcx) in the Protein Data Bank (PDB). If structures with 90% identity are counted only once, there are still 52 unique structures remaining in this pool (Table 4). These proteins include hydantoinase (40,55), folylpolyglutamate synthase (43), UV damage endonuclease (14), OXA10, OXA-1 class D β-lactamase (38,56,57), urease (42), phosphotriesterase (58), dihydroorotase (59), dihydropyrimidinase (60), organophosphate hydrolase (61) and MurE and MurD ligases (44,62). In most of these proteins, the carboxylated lysine bridges two metal ions, similar to the role of glutamate or aspartate in proteins with two metal-binding sites (26 structures among 52). However, the urease apoenzyme can be activated in vitro only in the presence of carbon dioxide prior to nickel binding (63), suggesting that the carboxylated lysine may have other structural roles beyond binding metals. In some proteins such as β-lactamase, UV damage endonuclease, Rubisco, MurD and MurE ligase and BlaR signal transducer protein, a carboxylated lysine plays an essential catalytic role. More interestingly, in three structures (PDB ID: 1HL9, 1PU6 and 2UYN for fucosiadase, 3- methyladenine DNA glycosylase and TdcF protein of unknown function, respectively), the carboxylated lysines are located near the surface of proteins, presumably playing primarily a structure stabilizing role (64–66). Since the carboxyl group is labile at acidic pH, but easily formed in the presence of carbon dioxide at basic pH, the number of proteins with lysine carboxylation must be underestimated. Furthermore, the carboxylated lysine must be fixed in place by metal ions (either one or two) or hydrogen bonding with other protein residues (at least one). Therefore, any detection method involving denaturing the proteins will result Li et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript in release of the carboxyl group. With current technology, 13C NMR (38) and crystallography are the only methods that can detect this modification. However, these methods are not amenable to high-throughput investigations. The majority (49 out of 52 structures in the PDB) of known lysine carboxylation modifications were found to be located at or near the active site, probably because these sites receive the most attention. Revisiting the structures in PDB with more attention to surface lysines might reveal more structures with carboxylated lysines. In conclusion, we have shown that Lys302 in AOTCase is post-translationally modified by carboxylation and that this modification may be functionally important for enzymatic activity. Lysine carboxylation is likely to be a more common event than currently appreciated and may play a critical role in enzymatic activity and protein stability. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Abbreviations ACIT N-acteyl-L-citrulline ANOR N-acetyl-L-norvaline AORN N-acetyl-L-Ornithine AOTCase N-acetyl-L-ornithine transcarbamylase ATCase aspartate transcarbamlyase OTCase ornithine transcarbamylase CP carbamyl phosphate ORN L-ornithine PALAO Nδ-(phosphonacetyl)-Nα-acetyl-L-ornithine SORN N-succinyl-L-ornithine WT wild-type xc Xanthomonas campestris Acknowledgments We thank Dr. David Davies for facilitating our use of the diffraction equipment in the Molecular Structure Section of the National Institute of Health and Dr. Fred Dyda for help in data collection and processing, and Dr. Yui-Fai Lam in the University of Maryland for help in setting up NMR measurements. REFERENCES 1. Close P, Creppe C, Gillard M, Ladang A, Chapelle JP, Nguyen L, Chariot A. The emerging role of lysine acetylation of non-nuclear proteins. Cell Mol Life Sci. 2010; 67:1255–1264. [PubMed: 20082207] 2. Geiman TM, Robertson KD. Chromatin remodeling, histone modifications, and DNA methylation- how does it all fit together? J Cell Biochem. 2002; 87:117–125. [PubMed: 12244565] 3. An W. Histone acetylation and methylation: combinatorial players for transcriptional regulation. Subcell Biochem. 2007; 41:351–369. [PubMed: 17484136] Li et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 4. Yang F, Bian C, Zhu L, Zhao G, Huang Z, Huang M. Effect of human serum albumin on drug metabolism: structural evidence of esterase activity of human serum albumin. J Struct Biol. 2007; 157:348–355. [PubMed: 17067818] 5. Xu AS, Labotka RJ, London RE. Acetylation of human hemoglobin by methyl acetylphosphate. Evidence of broad regio-selectivity revealed by NMR studies. J Biol Chem. 1999; 274:26629– 26632. [PubMed: 10480863] 6. Ueno H, Pospischil MA, Manning JM. Methyl acetyl phosphate as a covalent probe for anion- binding sites in human and bovine hemoglobins. J Biol Chem. 1989; 264:12344–12351. [PubMed: 2745446] 7. Stavropoulos P, Nagy V, Blobel G, Hoelz A. Molecular basis for the autoregulation of the protein acetyl transferase Rtt109. Proc Natl Acad Sci U S A. 2008; 105:12236–12241. [PubMed: 18719104] 8. Bewley MC, Graziano V, Jiang J, Matz E, Studier FW, Pegg AE, Coleman CS, Flanagan JM. Structures of wild-type and mutant human spermidine/spermine N1-acetyltransferase, a potential therapeutic drug target. Proc Natl Acad Sci U S A. 2006; 103:2063–2068. [PubMed: 16455797] 9. Walter TS, Meier C, Assenberg R, Au KF, Ren J, Verma A, Nettleship JE, Owens RJ, Stuart DI, Grimes JM. Lysine methylation as a routine rescue strategy for protein crystallization. Structure. 2006; 14:1617–1622. [PubMed: 17098187] 10. Stark GRSW, Moore S. Reactions of the cyanante present in aqueous urea with amino acids and proteins. J Biol Chem. 1960; 235:3177–3181. 11. Bobb D, Hofstee BH. Gel isoelectric focusing for following the successive carbamylations of amino groups in chymotrypsinogen A. Anal Biochem. 1971; 40:209–217. [PubMed: 5550146] 12. Kraus LM, Kraus AP Jr. Carbamoylation of amino acids and proteins in uremia. Kidney Int Suppl. 2001; 78:S102–S107. [PubMed: 11168993] 13. Al-Dirbashi OY, Al-Hassnan ZN, Rashed MS. Determination of homocitrulline in urine of patients with HHH syndrome by liquid chromatography tandem mass spectrometry. Anal Bioanal Chem. 2006; 386:2013–2017. [PubMed: 17053917] 14. Meulenbroek EM, Paspaleva K, Thomassen EA, Abrahams JP, Goosen N, Pannu NS. Involvement of a carboxylated lysine in UV damage endonuclease. Protein Sci. 2009; 18:549–558. [PubMed: 19241382] 15. Dementin S, Bouhss A, Auger G, Parquet C, Mengin-Lecreulx D, Dideberg O, van Heijenoort J, Blanot D. Evidence of a functional requirement for a carbamoylated lysine residue in MurD, MurE and MurF synthetases as established by chemical rescue experiments. Eur J Biochem. 2001; 268:5800–5807. [PubMed: 11722566] 16. Cha J, Mobashery S. Lysine N(zeta)-decarboxylation in the BlaR1 protein from Staphylococcus aureus at the root of its function as an antibiotic sensor. J Am Chem Soc. 2007; 129:3834–3835. [PubMed: 17343387] 17. Shi D, Yu X, Roth L, Morizono H, Tuchman M, Allewell NM. Structures of N-acetylornithine transcarbamoylase from Xanthomonas campestris complexed with substrates and substrate analogs imply mechanisms for substrate binding and catalysis. Proteins. 2006; 64:532–542. [PubMed: 16741992] 18. Shi D, Morizono H, Yu X, Roth L, Caldovic L, Allewell NM, Malamy MH, Tuchman M. Crystal structure of N-acetylornithine transcarbamylase from Xanthomonas campestris: a novel enzyme in a new arginine biosynthetic pathway found in several eubacteria. J Biol Chem. 2005; 280:14366– 14369. [PubMed: 15731101] 19. Morizono H, Cabrera-Luque J, Shi D, Gallegos R, Yamaguchi S, Yu X, Allewell NM, Malamy MH, Tuchman M. Acetylornithine transcarbamylase: a novel enzyme in arginine biosynthesis. J Bacteriol. 2006; 188:2974–2982. [PubMed: 16585758] 20. da Silva FR, Vettore AL, Kemper EL, Leite A, Arruda P. Fastidian gum: the Xylella fastidiosa exopolysaccharide possibly involved in bacterial pathogenicity. FEMS Microbiol Lett. 2001; 203:165–171. [PubMed: 11583843] 21. da Silva AC, Ferro JA, Reinach FC, Farah CS, Furlan LR, Quaggio RB, Monteiro-Vitorello CB, Van Sluys MA, Almeida NF, Alves LM, do Amaral AM, Bertolini MC, Camargo LE, Camarotte G, Cannavan F, Cardozo J, Chambergo F, Ciapina LP, Cicarelli RM, Coutinho LL, Cursino-Santos Li et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript JR, El-Dorry H, Faria JB, Ferreira AJ, Ferreira RC, Ferro MI, Formighieri EF, Franco MC, Greggio CC, Gruber A, Katsuyama AM, Kishi LT, Leite RP, Lemos EG, Lemos MV, Locali EC, Machado MA, Madeira AM, Martinez-Rossi NM, Martins EC, Meidanis J, Menck CF, Miyaki CY, Moon DH, Moreira LM, Novo MT, Okura VK, Oliveira MC, Oliveira VR, Pereira HA, Rossi A, Sena JA, Silva C, de Souza RF, Spinola LA, Takita MA, Tamura RE, Teixeira EC, Tezza RI, Trindade dos Santos M, Truffi D, Tsai SM, White FF, Setubal JC, Kitajima JP. Comparison of the genomes of two Xanthomonas pathogens with differing host specificities. Nature. 2002; 417:459– 463. [PubMed: 12024217] 22. Shi D, Yu X, Cabrera-Luque J, Chen TY, Roth L, Morizono H, Allewell NM, Tuchman M. A single mutation in the active site swaps the substrate specificity of N-acetyl-L-ornithine transcarbamylase and N-succinyl-L-ornithine transcarbamylase. Protein Sci. 2007; 16:1689–1699. [PubMed: 17600144] 23. Shi D, Morizono H, Cabrera-Luque J, Yu X, Roth L, Malamy MH, Allewell NM, Tuchman M. Structure and catalytic mechanism of a novel N-succinyl-L-ornithine transcarbamylase in arginine biosynthesis of Bacteroides fragilis. J Biol Chem. 2006; 281:20623–20631. [PubMed: 16704984] 24. Pastra-Landis SC, Foote J, Kantrowitz ER. An improved colorimetric assay for aspartate and ornithine transcarbamylases. Anal Biochem. 1981; 118:358–363. [PubMed: 7337232] 25. Shi D, Yu X, Roth L, Morizono H, Hathout Y, Allewell NM, Tuchman M. Expression, purification, crystallization and preliminary X-ray crystallographic studies of a novel acetylcitrulline deacetylase from Xanthomonas campestris. Acta Crystallogr Sect F Struct Biol Cryst Commun. 2005; 61:676–679. 26. Otwinowski Z, Minor W. Processing of X-ray Diffraction Data Collected in Oscillation Mode. Methods in Enzymology. 1997; 276:307–326. 27. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr. 1994; 50:760–763. [PubMed: 15299374] 28. Jones TA, Zou JY, Cowan SW, Kjeldgaard M. Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr A. 1991; 47(Pt 2):110–119. [PubMed: 2025413] 29. Brunger AT, Adams PD, Clore GM, DeLano WL, Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges M, Pannu NS, Read RJ, Rice LM, Simonson T, Warren GL. Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr. 1998; 54:905–921. [PubMed: 9757107] 30. Brunger AT. Free R value: a novel statistical quantity for assessing the accuracy of crystal structures. Nature. 1992; 355:472–475. [PubMed: 18481394] 31. Laskowski RA, MacArthur MW, Moss DS, Thornton JM. PROCHECK: a program to check the stereochemical quality of protein structures. J Appl Crystallogr. 1993; 26:283–291. 32. Golemi D, Maveyraud L, Vakulenko S, Samama JP, Mobashery S. Critical involvement of a carbamylated lysine in catalytic function of class D beta-lactamases. Proc Natl Acad Sci U S A. 2001; 98:14280–14285. [PubMed: 11724923] 33. Lapko VN, Smith DL, Smith JB. In vivo carbamylation and acetylation of water-soluble human lens alphaB-crystallin lysine 92. Protein Sci. 2001; 10:1130–1136. [PubMed: 11369851] 34. Shi D, Morizono H, Ha Y, Aoyagi M, Tuchman M, Allewell NM. 1.85-A resolution crystal structure of human ornithine transcarbamoylase complexed with N-phosphonacetyl-L-ornithine. Catalytic mechanism and correlation with inherited deficiency. J Biol Chem. 1998; 273:34247– 34254. [PubMed: 9852088] 35. Langley DB, Templeton MD, Fields BA, Mitchell RE, Collyer CA. Mechanism of inactivation of ornithine transcarbamoylase by Ndelta -(N'-Sulfodiaminophosphinyl)-L-ornithine, a true transition state analogue? Crystal structure and implications for catalytic mechanism. J Biol Chem. 2000; 275:20012–20019. [PubMed: 10747936] 36. Cha J, Vakulenko SB, Mobashery S. Characterization of the beta-lactam antibiotic sensor domain of the MecR1 signal sensor/transducer protein from methicillin-resistant Staphylococcus aureus. Biochemistry. 2007; 46:7822–7831. [PubMed: 17550272] Li et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 37. O'Leary MH, Jaworski RJ, Hartman FC. C nuclear magnetic resonance study of the CO(2) activation of ribulosebisphosphate carboxylase from Rhodospirillum rubrum. Proc Natl Acad Sci U S A. 1979; 76:673–675. [PubMed: 16592618] 38. Golemi-Kotra D, Cha JY, Meroueh SO, Vakulenko SB, Mobashery S. Resistance to beta-lactam antibiotics and its mediation by the sensor domain of the transmembrane BlaR signaling pathway in Staphylococcus aureus. J Biol Chem. 2003; 278:18419–18425. [PubMed: 12591921] 39. Schneider KD, Bethel CR, Distler AM, Hujer AM, Bonomo RA, Leonard DA. Mutation of the active site carboxy-lysine (K70) of OXA-1 beta-lactamase results in a deacylation-deficient enzyme. Biochemistry. 2009; 48:6136–6145. [PubMed: 19485421] 40. Huang CY, Hsu CC, Chen MC, Yang YS. Effect of metal binding and posttranslational lysine carboxylation on the activity of recombinant hydantoinase. J Biol Inorg Chem. 2009; 14:111–121. [PubMed: 18781344] 41. Zhang Q, Wang Y. High mobility group proteins and their post-translational modifications. Biochim Biophys Acta. 2008; 1784:1159–1166. [PubMed: 18513496] 42. Jabri E, Carr MB, Hausinger RP, Karplus PA. The crystal structure of urease from Klebsiella aerogenes. Science. 1995; 268:998–1004. [PubMed: 7754395] 43. Young PG, Smith CA, Metcalf P, Baker EN. Structures of Mycobacterium tuberculosis folylpolyglutamate synthase complexed with ADP and AMPPCP. Acta Crystallogr D Biol Crystallogr D. 2008; 64:745–753. 44. Gordon E, Flouret B, Chantalat L, van Heijenoort J, Mengin-Lecreulx D, Dideberg O. Crystal structure of UDP-N-acetylmuramoyl-L-alanyl-D-glutamate: meso-diaminopimelate ligase from Escherichia coli. J Biol Chem. 2001; 276:10999–11006. [PubMed: 11124264] 45. Taylor TC, Andersson I. Structure of a product complex of spinach ribulose-1,5-bisphosphate carboxylase/oxygenase. Biochemistry. 1997; 36:4041–4046. [PubMed: 9092835] 46. Jensen R. Activation of Rubisco controls CO(2) assimilation in light: a perspective on its discovery. Photosynth Res. 2004; 82:187–193. [PubMed: 16151874] 47. Falkowski PG. Photosynthesis: the paradox of carbon dioxide efflux. Curr Biol. 1997; 7:R637– R639. [PubMed: 9368746] 48. Roos A, Boron WF. Intracellular pH. Physiol Rev. 1981; 61:296–434. [PubMed: 7012859] 49. Smith KS, Ferry JG. Prokaryotic carbonic anhydrases. FEMS Microbiol Rev. 2000; 24:335–366. [PubMed: 10978542] 50. Bahn YS, Muhlschlegel FA. CO2 sensing in fungi and beyond. Curr Opin Microbiol. 2006; 9:572– 578. [PubMed: 17045514] 51. Esposito G, Jaiswal BS, Xie F, Krajnc-Franken MA, Robben TJ, Strik AM, Kuil C, Philipsen RL, van Duin M, Conti M, Gossen JA. Mice deficient for soluble adenylyl cyclase are infertile because of a severe sperm-motility defect. Proc Natl Acad Sci U S A. 2004; 101:2993–2998. [PubMed: 14976244] 52. Townsend PD, Holliday PM, Fenyk S, Hess KC, Gray MA, Hodgson DR, Cann MJ. Stimulation of mammalian G-protein-responsive adenylyl cyclases by carbon dioxide. J Biol Chem. 2009; 284:784–791. [PubMed: 19008230] 53. Kwon JY, Dahanukar A, Weiss LA, Carlson JR. The molecular basis of CO2 reception in Drosophila. Proc Natl Acad Sci U S A. 2007; 104:3574–3578. [PubMed: 17360684] 54. Jones WD, Cayirlioglu P, Kadow IG, Vosshall LB. Two chemosensory receptors together mediate carbon dioxide detection in Drosophila. Nature. 2007; 445:86–90. [PubMed: 17167414] 55. Xu Z, Liu Y, Yang Y, Jiang W, Arnold E, Ding J. Crystal structure of D-Hydantoinase from Burkholderia pickettii at a resolution of 2.7 Angstroms: insights into the molecular basis of enzyme thermostability. J Bacteriol. 2003; 185:4038–4049. [PubMed: 12837777] 56. Sun T, Nukaga M, Mayama K, Braswell EH, Knox JR. Comparison of beta-lactamases of classes A and D: 1.5-A crystallographic structure of the class D OXA-1 oxacillinase. Protein Sci. 2003; 12:82–91. [PubMed: 12493831] 57. Maveyraud L, Golemi D, Kotra LP, Tranier S, Vakulenko S, Mobashery S, Samama JP. Insights into class D beta-lactamases are revealed by the crystal structure of the OXA10 enzyme from Pseudomonas aeruginosa. Structure. 2000; 8:1289–1298. [PubMed: 11188693] Li et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 58. Benning MM, Kuo JM, Raushel FM, Holden HM. Three-dimensional structure of the binuclear metal center of phosphotriesterase. Biochemistry. 1995; 34:7973–7978. [PubMed: 7794910] 59. Thoden JB, Phillips GN Jr, Neal TM, Raushel FM, Holden HM. Molecular structure of dihydroorotase: a paradigm for catalysis through the use of a binuclear metal center. Biochemistry. 2001; 40:6989–6997. [PubMed: 11401542] 60. Abendroth J, Niefind K, Schomburg D. X-ray structure of a dihydropyrimidinase from Thermus sp. at 1.3 A resolution. J Mol Biol. 2002; 320:143–156. [PubMed: 12079340] 61. Yang H, Carr PD, McLoughlin SY, Liu JW, Horne I, Qiu X, Jeffries CM, Russell RJ, Oakeshott JG, Ollis DL. Evolution of an organophosphate-degrading enzyme: a comparison of natural and directed evolution. Protein Eng. 2003; 16:135–145. [PubMed: 12676982] 62. Bertrand JA, Auger G, Martin L, Fanchon E, Blanot D, Le Beller D, van Heijenoort J, Dideberg O. Determination of the MurD mechanism through crystallographic analysis of enzyme complexes. J Mol Biol. 1999; 289:579–590. [PubMed: 10356330] 63. Park IS, Hausinger RP. Requirement of carbon dioxide for in vitro assembly of the urease nickel metallocenter. Science. 1995; 267:1156–1158. [PubMed: 7855593] 64. Sulzenbacher G, Bignon C, Nishimura T, Tarling CA, Withers SG, Henrissat B, Bourne Y. Crystal structure of Thermotoga maritima alpha-L-fucosidase. Insights into the catalytic mechanism and the molecular basis for fucosidosis. J Biol Chem. 2004; 279:13119–13128. [PubMed: 14715651] 65. Eichman BF, O'Rourke EJ, Radicella JP, Ellenberger T. Crystal structures of 3-methyladenine DNA glycosylase MagIII and the recognition of alkylated bases. Embo J. 2003; 22:4898–4909. [PubMed: 14517230] 66. Burman JD, Stevenson CE, Sawers RG, Lawson DM. The crystal structure of Escherichia coli TdcF, a member of the highly conserved YjgF/YER057c/UK114 family. BMC Struct Biol. 2007; 7:30. [PubMed: 17506874] Li et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Stereo view of the structure and hydrogen bonding network surrounding residue 302. A, PALAO bound wild-type AOTCase, B, PALAO bound K302A AOTCase C, PALAO bound K302E AOTCase, D, PALAO bound K302R AOTCase. Contours of the electron density maps (2Fo-Fc) around PALAO, residue 302 and water molecules are shown as a brown cage at 1.0σ. The final refined positions of the ligands and surrounding protein residues are represented as colored sticks. The predicted hydrogen bonding interactions are in pink dashed lines. The water molecules are represented as pink balls. The carbon of PALAO, residue 302 and other protein residues are shown in pink, light blue and green sticks, respectively. Li et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. 13C NMR spectra of wild-type (upper panel) and K302A mutant (lower panel) AOTCase (1 mM). Experiments were performed in 100 mM Tris HCl, 50 mM NaCl, 7% D2O, pH 8.0, supplemented with 20 mM NaH13CO3. The position of the resonance attributed to carboxylated lysine in the enzyme is around 164 ppm. Li et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Chemical structure of carbamylated vs. carboxylated lysine. Li et al. Page 16 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 17 Table 1 Data collection and refinement statistics Dataset PALAO K302A K302E K302R Space group I213 I213 I213 I213 Resolution (Å) 2.2 1.9 1.85 2.2 Unit-cell parameters (Å) a = b = c =128.88 a = b = c =128.92 a = b = c =129.29 a = b = c =127.39 Measurements 219,475 305,757 390,128 246,817 Unique reflections 18,269 (1,832) a 28,236 (1,365) 30,622 (1,456) 17,635 (879) Redundancy 12.0 (11.8) 10.8(5.4) 12.8 (5.4) 14.0 (13.1) Completeness (%) 99.8 (100.0) 100.0 (100.0) 99.7 (95.1) 100.0 (100.0) <I/σ (I)> 15.0 (4.9) 16.4 (2.3) 19.8 (2.8) 8.7 (3.7) Rmerg b 7.4 (48.4) 6.5(64.9) 5.2 (55.3) 9.8 (79.1) Wilson B (Å2) 30.4 27.6 28.6 21.9 Refinement Resolution range (Å) 50.0-2.2 50-1.9 50-1.85 50-2.2 No. of protein atoms 2620 2613 2617 2619 No. of water atoms 90 219 193 146 No. of hetero atoms 24 24 24 24 Rmsd of bond lengths (Å) 0.006 0.005 0.005 0.005 Rmsd of bond angle (°) 1.1 1.2 1.2 1.2 Rwork (%)c 20.0 19.8 20.0 18.9 Rfree (%)d 24.3 23.2 23.2 22.2 Average B factor (Å2) 41.7 32.2 32.3 35.3 aFigures in brackets apply to the highest-resolution shell. bRmerg = ΣhΣi|I(h,i)-<I(h)>|/∑hΣiI(h,i), where I(h,i) is the intensity of the ith observation of reflection h, and < I(h)> is the average intensity of redundant measurements of reflection h. cRwork= Σh‖Fobs| – |Fcalc‖/Σh|Fobs|. dRfree = Σh‖Fobs| – |Fcalc‖/Σh|Fobs| for 5% of the reserved reflections. Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 18 Table 2 Interactions between carboxylated lysine and other residues at the active site of AOTCase Kcx302 Other residues Bound ligands PALAO CPa AORNb CP +ANORc SO4+ACITd OQ1 K252 NZ 2.6 2.6 2.7 2.6 2.6 OQ1 W1e 2.6 2.6 2.6 2.7 OQ2 S253 N 3.0 3.1 2.8 2.9 2.9 OQ2 H293 NE2 3.0 3.2 3.0 2.9 2.9 NZ W2f 3.1 2.9 3.0 3.0 aThe values were calculated based on PDB ID 3KZM. bThe values were calculated based on PDB ID 3KZN. cThe values were calculated based on PDB ID 3KZO. dThe values were calculated based on PDB ID 3KZK. eThis water molecule hydrogen bonds to N1 atom of PALAO, AORN or ANOR, and backbone O atom of Pro296 as well. fThis water molecule hydrogen bonds to OE1 atom of Glu92 from adjacent subunit as well. Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 19 Table 3 Specific activity of wild-type and mutant AOTCase in the presence of acids (0.5M). Compounds added Specific activity(µmol/min/mg) Wild-type K302A K302E K302R None 43.4 ± 0.4a 23.0 ± 0.5 7.1 ± 0.1 0.059±0.01 Formate 44.1 ± 1.2 26.4 ± 0.6 6.7 ± 0.2 0.093±0.01 Acetate 48.5 ± 1.1 21.2 ± 0.8 6.6 ± 0.5 0.104±0.03 aThe Mean ± S.D. are shown (n = 3). Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 20 Table 4 Protein structures with lysine carboxylation modification PDB ID Enzyme name Residue Organism source Funciton 2OGJ Dihydroorotase 175 A.tumefaciens Bridging two Zn(II) 2Z26 Dihydroorotase 102 E.coli Bridging two Zn(II) 3JZE Dihydroorotase 103 S.enterica Bridging two Zn(II) 2GWN Dihydroorotase 149 P. gingivalis Bridging two Zn(II) 3F4C Organophosphorus hydrolase 243 G. stearothermophilus Bridging two Co(II) 3ICJ Metal-dependent hydrolase 294 P. furiosus Bridging two Zn(II) 3GTX Organophosphorus hydrolase 243 D. radiodurans Bridging two Co(II) 2QPX Metal-dependent hydrolase 166 L. casei Bridging two Zn(II) 2FTW Dihydropyrimidinase 158 D. discoideum Bridging two Zn(II) 2FVK Dihydropyrimidinase 167 S. kluyveri Bridging two Zn(II) 3DC8 Dihydropyrimidinase 147 S. meliloti Bridging two Zn(II) 3GNH L-Lys/Arg carboxypeptidase 211 C. crescentus cb15 Bridging two Zn(II) 3DUG Arginine carboxypeptidase 182 Unidentified Bridging two Zn(II) 2VC7 Phosphotriesterase 137 S. solfataricus Bridging two Co(II) 2R1N Metallophosphotriesterases 169 A. tumefaciens Bridging two Co(II) 2OB3 Phosphotriesterase 169 B. diminuta Bridging two Zn(II) 3E74 Allantoinase 146 E. coli Bridging two Fe(III) 1EJX Urease 217 K. aerogenes Bridging two Ni(II) 1E9Z Urease 219 H. pylori Bridging two Ni(II) 4UBP Urease 220 B. pasteurii Bridging two Ni(II) 1ONW Isoaspartyl dipeptidase 162 E. coli Bridging two Zn(II) 1K1D D-hydanroinase 150 G. stearothermophilus Bridging two Zn(II) 1GKR L-hydanroinase 147 A. aurescens Bridging two Zn(II) 1GKP D-hydanroinase 150 Thermus sp. Bridging two Zn(II) 1NFG D-hydantoinase 148 R. pickettii Bridging two Zn(II) 2ICS Adenine deaminase 154 E. faecalis Bridging two Zn(II) 1RQB Transcarboxylase 184 P. freudenreichii Binding one Co(II) 2QF7 Pyruvate carboxylase 718 R. etli Binding one Zn(II) 3BG3 Pyruvate carboxylase 741 H. sapiens Binding one Mn(II) 2OEM Rubisco-like protein 173 G. kaustophilus Binding one Mg(II) 1WDD Rubisco 201 O. sativa Binding one Mg(II) 1GK8 Rubisco 201 C. reinhardtii Binding one Mg(II) 1BWV Rubisco 201 G. partita Binding one Mg(II) 2WTZ ATP-dependent MurE ligase 262 M. tuberculosis Binding one Mg(II) 2JFG MurD ligase 198 E. coli Catalytic role? 1E8C MurE ligase 224 E. coli Catalytic role? 1JBW Folypolyglutamate synthetase 185 L. casei Catalytic role? 1W78 FolC bifunctional protein 188 E. coli Binding one Mg(II) 3HBR OXA-48 β-lactamase 73 K. pneumoniae Catalytic role Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 21 PDB ID Enzyme name Residue Organism source Funciton 3ISG Class D β-lactamase 70 E. coli Catalytic role 2P9V AmpC beta-lactamase 315 E. coli Catalytic role 1K55 OXA-10 β-lactamase 70 P. aeruginosa Catalytic role 1K38 β-lactamase OXA-2 70 S. typhimurium Catalytic role 1XQL Alanine racemase 129 G. stearothermophilus Binding substrate? 1VFS Alanine racemase 129 S. lavendulae Binding substrate? 1RCQ Alanine racemase 122 P. aeruginosa Binding substrate? 2J6V UV damage endonuclease 229 T. thermophilus Catalytic role 1H01 Cell division protein kinase 2 33 H. sapiens Catalytic role? 2UYN Protein TdcF A58 E. coli Structural role? 1HL9 Fucosidase 338 T. maritime Structural role? 1PU6 3-methyladenine DNA glycosylase 205 H. pylori Structural role? Biochemistry. Author manuscript; available in PMC 2011 August 17.
3M5D
Crystal structure of N-acetyl-L-ornithine transcarbamylase K302R mutant complexed with PALAO
Reversible Post-Translational Carboxylation Modulates The Enzymatic Activity Of N-Acetyl-L-Ornithine Transcarbamylase† Yongdong Li1,2, Xiaolin Yu1, Jeremy Ho1, David Fushman3, Norma M. Allewell3, Mendel Tuchman1, and Dashuang Shi1,‡ 1Research Center for Genetic Medicine and Department of Integrative Systems Biology, Children’s National Medical Center, The George Washington University, Washington, DC 20010, USA. 2Key Laboratory of Organo-Pharmaceutical Chemistry, Jiangxi Province, Gannan Normal University, Ganzhou 341000, China. 3Department of Chemistry and Biochemistry, College of Chemical and Life Sciences, University of Maryland, College Park, MD 20742, USA. Abstract N-acetyl-L-ornithine transcarbamylase (AOTCase), rather than ornithine transcarbamylase (OTCase), is the essential carbamylase enzyme in the arginine biosynthesis of several plant and human pathogens. The specificity of this unique enzyme provides a potential target for controlling the spread of these pathogens. Recently, several crystal structures of AOTCase from Xanthomonas campestris (xc) have been determined. In these structures, an unexplained electron density at the tip of Lys302 side-chain was observed. Using 13C NMR spectroscopy, we show herein that Lys302 is post-translationally carboxylated. The structure of wild-type AOTCase complexed with the bisubstrate analogue, Nδ-(phosphonoacetyl)-Nα-acetyl-L-ornithine (PALAO), indicates that the carboxyl group on Lys302 forms a strong hydrogen bonding network with surrounding active site residues, Lys252, Ser253, His293, and Glu92 from the adjacent subunit either directly or via a water molecule. Furthermore, the carboxyl group is involved in binding N-acetyl-L-ornithine via a water molecule. Activity assays with the wild-type enzyme and several mutants demonstrate that the post translational modification of lysine 302 has an important role in catalysis. Post-translational modification of the ε-amino group of lysine residues in proteins is a common mechanism used by organisms to regulate protein functions including DNA-protein interactions, subcellular localization, transcriptional activity, and protein stability and activity (1). Lysine residues can be modified by the addition of functional groups to become acetylated, methylated, carbamylated or carboxylated. The role of histone lysine acetylation and methylation in affecting chromatin structure and gene expression has been well established for more than a decade (2). However, the biological roles for lysine carbamylation and carboxylation have rarely been investigated. †This work was supported by Public Health Service grants DK-47870 (MT) and DK-067935 (DS) from the National Institute of Diabetes, Digestive and Kidney Diseases. JH was supported by a Scholarship from the Doug and Lynn Parsons Family Foundation. The Cornell High Energy Synchrotron Source (CHESS) is supported by the National Science Foundation under award DMR 0225180 and the Macromolecular Diffraction Facility at CHESS (MacCHESS) is supported by award RR-01646 from the National Institutes of Health, through its National Center for Research Resources. ‡Corresponding author. dshi@cnmcresearch.org. Phone: 202-476-5817. Fax: 202-476-6014. SUPPORTING INFORMATION AVAILABLE Figure S1. Structure and hydrogen bonding network around residue 302 for previously determined AOTCase structures. This material is available free of charge via the Internet at http://pubs.acs.org. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 August 17. Published in final edited form as: Biochemistry. 2010 August 17; 49(32): 6887–6895. doi:10.1021/bi1007386. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript In vivo, lysine acetylation and methylation are usually carried out by acetyltransferase and methyltransferase enzymes, respectively (3). In addition, some proteins such as hemoglobin and human serum albumin can be acetylated non-enzymatically by chemicals such as aspirin, methyl acetyl phosphate, and other acetylating agents such as acetyl-CoA (4–8). Lysine can also be methylated by small chemicals in vitro, and this has routinely been used as a rescue method for protein crystallization (9). Lysine carbamylation and lysine carboxylation have only been achieved by using chemicals and no enzyme has yet been found to catalyze these modifications. Lysine carbamylation was one of the earliest post- translational modification of proteins to be elucidated when it was identified as a product of reversible denaturation-renaturation studies of proteins with urea (10,11). This carbamylation, which produces homocitrulline, has also been detected in uremic patients (12) and in patients with elevated plasma and/or urinary lysine levels (13). In contrast, lysine carboxylation is not as commonly reported, but has been identified in a number of proteins via crystal structure determinations. In most of these proteins, the carboxyl groups of modified lysines are involved in bridging two metal ions that play a structural role in the active site. In several other proteins, however, a direct role for a carboxylated lysine in the catalytic mechanism has been reported (14–16). N-acetyl-L-ornithine transcarbamylase (AOTCase, EC 2.1.3.9) was recently discovered to be part of a novel arginine biosynthesis pathway in plant pathogens of the Xanthomonadaceae family such as Xylella and Xanthomonas (17–19). These pathogens attack a variety of economically important crops including citrus fruits, cotton, tomatoes, and rice (20,21). Genome sequence analyses showed that an AOTCase-like gene is also present in some human pathogens such as Stenotrophomonas maltophilia and members of the genus Bacteroides (22). In the case of Bacteroides fragilis, this gene was later confirmed to encode another novel transcarbamylase, N-succinyl-L-ornithine transcarbamylase (SOTCase, EC 2.1.3.11) (23). Crystal structures of both AOTCase and SOTCase bound with substrate or substrate analogues have recently been determined (17,18,23). An extended density at the side-chain tip of Lys302 in AOTCase was observed suggesting a post- translational modification. Since Lys302 is located within the active site of AOTCase and is not present in SOTCase, it was proposed as one of three key signature residues to distinguish the two carbamylases (22). Here, we demonstrate that Lys302 is post- translationally modified by carboxylation and that this change affects the catalytic function of the enzyme. MATERIALS AND METHODS Materials All chemicals were purchased from Sigma Chemical Company unless otherwise specified. ANOR was purchased from Indofine Chemical Co., Inc. N-acetyl-L-citrulline was custom synthesized and purified by Chiral Quest Company. PALAO (>95% purity) was synthesized by IMI TAMI Institute of Research and Development Ltd. (19). xcAOTCase was prepared and purified as previously described (18). Mutants K302A (primer: 5’- CTGCGTCGCAACGTCGCGGCTACTGATGCGGTG-3’), K302E (primer:5’- CTGCGTCGCAACGTCGAGGCTACTGATGCGGTG-3’) and K302R (primer: 5’- CTGCGTCGCAACGTCAGGGCTACTGATGCGGTG-3’) were generated by site-directed mutagenesis using the “Quik Change” mutagenesis kit (Stratagene) according to the manufacturer’s protocol. The correct mutants were confirmed by DNA sequencing. Recombinant mutant proteins were expressed and purified in the same manner as the wild- type enzyme. Li et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Activity assay The modified colorimetric assay method, which detects the formation of the ureido group during the transcarbamylation reaction (24), was used to measure enzyme activity. CP and AORN concentration were kept constant as 4.0 mM. After an incubation of 5 minutes, the reaction was stopped by the addition of 1 ml of color reagent, as described previously (19). A set of tubes containing known amounts of N-acetyl-L-citrulline was included with each rack of enzyme assays to produce a standard curve for calculation of the enzyme specific activity. Mass spectrometric analysis In order to identify the post-translational modification, mass spectrometric analysis was carried out on a 4700 ABI TOF/TOF mass spectrometer (Applier Biosystems) based on the method described previously (25). In brief, 10 µg of native protein were digested overnight at 310 K using trypsin in 50 mM ammonium bicarbonate pH 7.4. After desalting using a C18 ZipTip micropipette tip, the resulting peptides were eluted in 10 µl of acetonitrile/0.1% TFA [70:30(v:v)]. The sample was mixed with matrix solution and spotted on a MALDI plate to be submitted to the mass spectrometric analysis. Chemical rescue experiments The assay in the presence of various selected chemical was conducted as described above. The stock solutions of small chemicals were titrated to the pH of the assay with KOH or HCl. 13C NMR experiments The wild-type and K302A mutant protein of AOTCase (~10 mg) was precipitated by degassed buffer (pH 4.5) containing 25 mM sodium acetate. After centrifugation, the precipitate was re-dissolved by adding a buffer containing 20 mM NaH13CO3, 100 mM Tris HCl (pH 8.0) and 50 mM NaCl. Before NMR experiments, 40 µl D2O was added to 500 µl protein sample. The 13C NMR spectra were collected on a Bruker Avance 600 spectrometer (operating at 14.1 T) equipped with a direct 13C-detection probe at 298 K. The experimental settings and processing parameters for the wild-type protein and K302A variant were identical. 512 transients were collected with 4K time domain points and a spectral width of 3019 Hz centered at 160 ppm. The spectra were processed using exponential multiplication with the line broadening factor set to 3Hz. The similarity of protein concentration in both samples was verified by 1H NMR (not shown). Crystallization, data collection and processing PALAO-bound wild-type and mutant AOTCase crystals were grown using the hanging-drop vapor diffusion method, with conditions similar to those used to produce native and ligand- complexed AOTCase crystals (18,23). 2.0 µl of ~10 mg/ml solution of AOTCase were mixed with 1.6 µl of reservoir solution and 0.4 µl PALAO solution (~0.01 M). The reservoir solution contained 20% (w/v) PEG 3350, 0.2 M lithium sulfate and 0.1 M bis-Tris, pH 6.0. Diffraction data for the PALAO-bound crystal were collected at 100 K at the F1 beam line of the Cornell High Energy Synchrotron Source. Data sets for the PALAO-bound mutant AOTCase crystals were obtained using a Rigaku anode x-ray generator in the Molecular Structure Section of the National Institute of Health. All data were processed using HKL2000 package (26) and reduced using the program TRUNCATE in the CCP4 suite (27). Data collection parameters are listed in Table 1. Li et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Structure solution and refinement Molecular replacement was used for phase determination of the PALAO-bound wild-type and mutant AOTCase structures. The coordinates of AOTCase (PDB code: 3KZO) after removal of ligands or water molecules were used for phase determination. Upon rigid-body refinement, electron density corresponding to the ligands could be clearly visualized. The ligands were built into the map using the graphics program O (28). Refinements were carried out using molecular annealing, energy minimization and restrained B factor refinement with the program CNS1.1 (29). During refinements, 5% of the reflections at various resolutions were randomly selected to set aside to calculate Rfree to monitor the progress of refinement (30). After every cycle of refinement, the model was manually adjusted using the program O (28). Water molecules were automatically assigned using the program WATERPICK of CNS. Model quality was checked using the program PROCHECK (31) to ensure good stereochemistry for all three models. The final refinement statistics are listed in Table 1. Figures 1 was drawn using the programs Pymol (http://www.pymol.org). Figure 3 was drawn using ChemDraw 8.0. The coordinates have been deposited with the RCSB PDB as entries 3M4J, 3M4N, 3M5C and 3M5D. RESULTS Lys302 in AOTCase is carboxylated To investigate the nature of the modification of Lys302 and how it affects catalytic activity, we revisited all AOTCase structures. In the PALAO-bound AOTCase structure, the electron density map clearly indicates that Lys302 is post-translationally modified (Figure 1A). The type of modification can include methylation, acetylation, carbamylation, and carboxylation. The shape of the electron density can been used to distinguish methyl groups from larger functional groups, but it is difficult to distinguish between acetyl, carbamyl, and carboxyl groups, all of which have three non-hydrogen atoms in a plane. Given the hydrogen bonding network with surrounding residues (Lys252, Ser253 and His293, Table 2), a carboxylated modification is the most likely choice for the modification of Lys302 in AOTCase. To exclude that the modification’s identity represents chemically stable moieties (methyl, acetyl, carbamyl), we analyzed trypsin digested fragments of purified AOTCase by TOF- TOF mass spectroscopy. As expected, only a peptide fragment with an unmodified Lys302 was observed, consistent with the lability of the carboxylic group in acidic solutions. At low pH, the carboxyl group is spontaneously released as carbon dioxide (14, 32), in contrast to other modified groups that are stably bound and can be observed by mass spectrometry analysis after proteolysis (33). The putative carboxyl group on the modified Lys302 forms direct hydrogen bonds with main-chain or side-chain nitrogen atoms of Lys252, Ser253 and His293 (Figure 1A and Table 2). Among these, Lys252 is involved in direct hydrogen bonding to the carboxyl group of the AORN moiety of PALAO, and His293 forms a strong hydrogen bond with the main-chain nitrogen atom of Leu295 in the conserved His293-Cys294-Leu295-Pro206 (HCLP) motif. The hydrogen bonding network between the carboxyl group of modified Lys302, His293 and the main-chain nitrogen atom of Leu295 is reminiscent of the similar hydrogen bonding network, Glu310-His302-Leu304 and Glu299-Leu272-Leu274, found in human and E. coli OTCase, respectively (34, 35). These three residues are conserved in all OTCase sequences, and the interactions between them are important for maintaining the HCLP motif in a specific conformation to orientate their main-chain oxygen atoms towards the active site. In all known transcarbamylase structures, a leucine residue corresponding to Leu295 is in an energetically unfavorable conformation and the peptide bond between this Li et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript leucine and Pro296 is in the cis conformation. In addition to the direct hydrogen bonding interaction above, the carboxyl group on the modified Lys302 interacts with the α-amino nitrogen atom of the AORN moiety of PALAO and Glu92 from the adjacent subunit via water molecules. When we revisited all previously determined AOTCase structures (see supplementary Figure S1) we found: (1) Lys302 was carboxylated in the absence of substrate binding, but substrate binding immobilizes the side-chain of Lys302 further by hydrogen bonding interaction via water molecules. (2) Water-mediated hydrogen bonding promotes interactions of carboxylated Lys302 with Glu92 from the adjacent subunit and the α-amino nitrogen atom of AORN. (3) Similarly to AOTCase, in the structure of SOTCase E92Z (Z = Ala, Ser, Pro, Val), mutant with N-succinyl-L-norvaline bound (22) the carboxylated Lys302 hydrogen bonds to the α-amino nitrogen atom and the succinyl carboxyl group of N- succinyl-L-norvaline via water molecules (Figure S1). To obtain direct, independent evidence for the carboxylation of Lys302, 13C NMR experiments were carried out with both wild-type protein and the K302A mutant. As observed for other proteins with carboxylated lysine (36,37), the strong 13C NMR signal at 164 ppm characteristic of a carboxyl group was clearly detectable in AOTCase wild-type protein labeled by 13C-bicarbonate, in contrast to the K302A mutant where the signal was weak (Figure 2). Since there are 17 other lysine residues in the protein, the weak signal seen for the K302A mutant might be caused by the adventitious carboxylation of another lysine with reduced pKa, as has been observed for the K392A mutant of the sensor domain of the BlaR protein (38). Functional and structural studies of Lys302 mutants To investigate the effect of lysine carboxylation on enzyme activity, Lys302 was mutated to alanine, glutamate or arginine. Each of these variants was expressed in E. coli and gave similar yields. Enzymatic assays demonstrated a significant decrease in enzymatic activity in all three mutants, reflecting the functional importance of Lys302 (Table 3). The level of enzymatic activity for the wild-type (WT) and three mutants was WT > K302A > K302E ≫ K302R. To determine the structural basis of these results, the WT and mutant enzymes bound with PALAO were crystallized and their structures were determined at 1.8–2.2 Å resolution. Only the K302R mutation had and appreciable effect on the structure of the protein. Since K302 is located near the AORN binding site, the mutations would weaken AORN binding to the active site. In the structure of the K302A mutant, three additional water molecules (labeled as w3, w4 and w5 in Figure 1B) replace the carboxylated lysine. The two water molecules (labeled w1 and w2 in Figure 1A–1D) that mediate the hydrogen bonding interaction of carboxylated Lys302 with PALAO and Glu92 from the adjacent subunit are also found in the K302A mutant structure. Furthermore, these water molecules maintain a similar hydrogen-bonding network to the wild-type enzyme. These results might explain why the K302A mutant retains significant catalytic activity (Table 3). To investigate whether adding short-chain carboxylic acids to the K302A mutant increases its activity as other enzymes (14, 15, 39, 40), the activity of the K302A mutant was measured in the presence of high formate and acetate concentration (0.5 M). Surprisingly, the activity of the K302A mutant was not significantly improved. The crystal structure of the K302A mutant soaking with the crystallization buffer in the presence of 0.5 M acetate was also determined (not shown) and it was observed that the same five water molecules were present in the cavity that replaced the side chain of the carboxylated lysine. This, the acetate’s inability to replace the water molecules in the crystal structure, is consistent with the unchanged activity assay results. Li et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The side-chain of Glu302 in the K302E mutant structure is well defined and anchored by hydrogen bonding interaction with the main-chain nitrogen atom of Arg298 and weakly hydrogen bonded to the main-chain nitrogen atom of Ser253 (Figure 1C). Two of three additional water molecules (w4 and w5) observed in the K302A mutant structure occupied the same position as the carboxyl oxygen atoms of Glu302 and form a similar hydrogen- bonding network. Relative to the PALAO-bound wild-type structure, there is only one more water molecule (w3) at the position of the carboxyl group of the carboxylated Lys302. This water molecule mediates a hydrogen bonding interaction between Glu302 and Lys252. Two common water molecules (w1 and w2) that interact with PALAO and Glu92 from the adjacent subunit, respectively, were also identified in the K302E structure. Our observation that the K302E mutant had lower enzymatic activity than that of the K302A mutant (Table 3) was surprising since the carboxyl group of the glutamate could conceivably function similarly to a carboxylated lysine. The explanation may be that, in the K302A mutant, the hydrogen bonding network is well maintained by water molecules in the cavity that replaces the carboxylated lysine. In particular, w3 is optimally located for strong hydrogen bonding to w1 (2.7 Å), which in turn binds AORN. The distances between w1 and the carboxyl oxygen of carboxylated Lys302 in all wild-type crystal structures are within 2.4–2.7 Å, but the distance between w1 and w2 in the K302E structure is significantly greater (3.2 Å). The weaker hydrogen bonding interaction may be a reason for lower enzymatic activity of the K302E mutant. In contrast to the K302A and K302E structures, the K302R structure shows a much larger reduction in enzyme activity relative to the wild-type enzyme. The electron density for the side-chain of Arg302 is weak and the temperature factor of its side-chain is 54.4 Å2, significantly higher than those of carboxylated Lys302 (44.7 Å2) and Glu302 (33.4 Å2), implying greater flexibility. Furthermore, the side-chain of Arg302 is oriented differently from the carboxyl group of carboxylated Lys302 and pushes the nearby residues His180, Pro181 and Lys182 outwards about 1.0 Å (Figure 1D). However, the water molecules involved in hydrogen bonding to the α-amino nitrogen atom of PALAO (w1) and the side- chains of Lys252 (w3) and Glu92 (w2) from the adjacent subunit are conserved. Consistent with the K302E structure, the distance between w1 and w2 is even greater (3.4 Å) than in the WT structure and the hydrogen bonding interaction between w2 and w3 is no longer observed. Thus, the almost undetectable enzymatic activity of the K302R mutant probably results from the changes at its active site, including the weakened hydrogen bonding network involved in substrate binding. DISCUSSION Several lines of evidence clearly indicate that Lys302 in AOTCase is carboxylated. First, the extra electron density indicates that the side-chain of Lys302 is modified. Second, the hydrogen bonding environment of Lys302 for hydrogen bonding interactions is compatible with a carboxyl group, but not for a positively charged lysine side-chain. Third, the modification is labile at low pH, since mass spectroscopy of samples prepared at low pH indicated that Lys302 was no longer modified. Fourth, the clear presence of the indicative 13C NMR signal at 164 ppm for wild-type protein and its absence in the K302A mutant confirms carboxylation of Lys302. It is well known that lysine carboxylation is non-enzymatic and reversible, while other post- translational modifications such as methylation, acetylation, and carbamylation are irreversible and detectable by mass spectroscopy. Furthermore, lysine methylation and acetylation usually require an enzyme-catalyzed reaction in vivo (41). Therefore, it is unlikely that such lysine modifications will be observed in recombinant proteins Li et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript overexpressed in a foreign host (e.g. E. coli). Lysine methylation can be achieved by using special chemicals in vitro, but these chemicals are not present in vivo. Lysine carbamylation and carboxylation use completely different mechanisms to form functionally different groups (Figure 3). Carbamylation can be achieved by cyanate produced from myeloperoxidase-catalyzed oxidation of thiocyanate, an anion abundant in plasma and increased in smokers, or from urea in the plasma. Lysine carboxylation, on the other hand, occurs readily in aqueous solution in the presence of carbon dioxide at a basic pH (32,42). Even though carbamylation and carboxylation use very different mechanisms, the two are confused in the literatures. Lysine carbamylation (or carbamoylation) is referred to in several publications (15,32,42–44), when the actual reaction is in fact carboxylation. The activity of the K302A mutant is almost half of that of the wild-type enzyme raising the question of why AOTCase retains a lysine in this position. Perhaps this lysine was maintained through evolution to distinguish AOTCase from SOTCase which uses N- succinyl-L-ornithine (SORN) rather than AORN (22), and OTCase which uses L-ornithine. An alternative explanation may be found in the very low activity of the K302R mutant. The side-chain of arginine has a positive charge while carboxylated lysine has a negative charge. The side chain of unmodified lysine is usually located in a similar position as that of arginine, as observed in the structure of UV damage endonuclease (14). It would be expected that the activity of AOTCase with an uncarboxylated lysine would be as low as the K302R mutant’s. It could further be surmised that, the respective organisms need to use carboxylation as a switch to turn “on” or “off” the arginine biosynthetic pathway. It has been well known that rubulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) in plant cells uses the carboxylation on Lys201 as a switch to turn the enzyme “on” during the day and “off” at night by removing the carboxyl group (45,46). Carbon dioxide and bicarbonate have been found to play an important biological role in modulating several biological processes including photosynthetic carbon fixation (47), pH homeostasis (48), carbon metabolism (49), activation of virulence in pathogenic organisms (50), sperm maturation (51), stimulation of mammalian G-protein-responsive adenylyl cyclase (52), and as an alarmone in Drosophila (53,54). Whether or not carboxylation of a key lysine in their related proteins is used as an underlying regulatory mechanism should be investigated further. There are 197 structures with carboxylated lysine residue (modified residue indicated as Kcx) in the Protein Data Bank (PDB). If structures with 90% identity are counted only once, there are still 52 unique structures remaining in this pool (Table 4). These proteins include hydantoinase (40,55), folylpolyglutamate synthase (43), UV damage endonuclease (14), OXA10, OXA-1 class D β-lactamase (38,56,57), urease (42), phosphotriesterase (58), dihydroorotase (59), dihydropyrimidinase (60), organophosphate hydrolase (61) and MurE and MurD ligases (44,62). In most of these proteins, the carboxylated lysine bridges two metal ions, similar to the role of glutamate or aspartate in proteins with two metal-binding sites (26 structures among 52). However, the urease apoenzyme can be activated in vitro only in the presence of carbon dioxide prior to nickel binding (63), suggesting that the carboxylated lysine may have other structural roles beyond binding metals. In some proteins such as β-lactamase, UV damage endonuclease, Rubisco, MurD and MurE ligase and BlaR signal transducer protein, a carboxylated lysine plays an essential catalytic role. More interestingly, in three structures (PDB ID: 1HL9, 1PU6 and 2UYN for fucosiadase, 3- methyladenine DNA glycosylase and TdcF protein of unknown function, respectively), the carboxylated lysines are located near the surface of proteins, presumably playing primarily a structure stabilizing role (64–66). Since the carboxyl group is labile at acidic pH, but easily formed in the presence of carbon dioxide at basic pH, the number of proteins with lysine carboxylation must be underestimated. Furthermore, the carboxylated lysine must be fixed in place by metal ions (either one or two) or hydrogen bonding with other protein residues (at least one). Therefore, any detection method involving denaturing the proteins will result Li et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript in release of the carboxyl group. With current technology, 13C NMR (38) and crystallography are the only methods that can detect this modification. However, these methods are not amenable to high-throughput investigations. The majority (49 out of 52 structures in the PDB) of known lysine carboxylation modifications were found to be located at or near the active site, probably because these sites receive the most attention. Revisiting the structures in PDB with more attention to surface lysines might reveal more structures with carboxylated lysines. In conclusion, we have shown that Lys302 in AOTCase is post-translationally modified by carboxylation and that this modification may be functionally important for enzymatic activity. Lysine carboxylation is likely to be a more common event than currently appreciated and may play a critical role in enzymatic activity and protein stability. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Abbreviations ACIT N-acteyl-L-citrulline ANOR N-acetyl-L-norvaline AORN N-acetyl-L-Ornithine AOTCase N-acetyl-L-ornithine transcarbamylase ATCase aspartate transcarbamlyase OTCase ornithine transcarbamylase CP carbamyl phosphate ORN L-ornithine PALAO Nδ-(phosphonacetyl)-Nα-acetyl-L-ornithine SORN N-succinyl-L-ornithine WT wild-type xc Xanthomonas campestris Acknowledgments We thank Dr. David Davies for facilitating our use of the diffraction equipment in the Molecular Structure Section of the National Institute of Health and Dr. Fred Dyda for help in data collection and processing, and Dr. Yui-Fai Lam in the University of Maryland for help in setting up NMR measurements. REFERENCES 1. Close P, Creppe C, Gillard M, Ladang A, Chapelle JP, Nguyen L, Chariot A. The emerging role of lysine acetylation of non-nuclear proteins. Cell Mol Life Sci. 2010; 67:1255–1264. [PubMed: 20082207] 2. Geiman TM, Robertson KD. Chromatin remodeling, histone modifications, and DNA methylation- how does it all fit together? J Cell Biochem. 2002; 87:117–125. [PubMed: 12244565] 3. An W. Histone acetylation and methylation: combinatorial players for transcriptional regulation. Subcell Biochem. 2007; 41:351–369. [PubMed: 17484136] Li et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 4. Yang F, Bian C, Zhu L, Zhao G, Huang Z, Huang M. Effect of human serum albumin on drug metabolism: structural evidence of esterase activity of human serum albumin. J Struct Biol. 2007; 157:348–355. [PubMed: 17067818] 5. Xu AS, Labotka RJ, London RE. Acetylation of human hemoglobin by methyl acetylphosphate. Evidence of broad regio-selectivity revealed by NMR studies. J Biol Chem. 1999; 274:26629– 26632. [PubMed: 10480863] 6. Ueno H, Pospischil MA, Manning JM. Methyl acetyl phosphate as a covalent probe for anion- binding sites in human and bovine hemoglobins. J Biol Chem. 1989; 264:12344–12351. [PubMed: 2745446] 7. Stavropoulos P, Nagy V, Blobel G, Hoelz A. Molecular basis for the autoregulation of the protein acetyl transferase Rtt109. Proc Natl Acad Sci U S A. 2008; 105:12236–12241. [PubMed: 18719104] 8. Bewley MC, Graziano V, Jiang J, Matz E, Studier FW, Pegg AE, Coleman CS, Flanagan JM. Structures of wild-type and mutant human spermidine/spermine N1-acetyltransferase, a potential therapeutic drug target. Proc Natl Acad Sci U S A. 2006; 103:2063–2068. [PubMed: 16455797] 9. Walter TS, Meier C, Assenberg R, Au KF, Ren J, Verma A, Nettleship JE, Owens RJ, Stuart DI, Grimes JM. Lysine methylation as a routine rescue strategy for protein crystallization. Structure. 2006; 14:1617–1622. [PubMed: 17098187] 10. Stark GRSW, Moore S. Reactions of the cyanante present in aqueous urea with amino acids and proteins. J Biol Chem. 1960; 235:3177–3181. 11. Bobb D, Hofstee BH. Gel isoelectric focusing for following the successive carbamylations of amino groups in chymotrypsinogen A. Anal Biochem. 1971; 40:209–217. [PubMed: 5550146] 12. Kraus LM, Kraus AP Jr. Carbamoylation of amino acids and proteins in uremia. Kidney Int Suppl. 2001; 78:S102–S107. [PubMed: 11168993] 13. Al-Dirbashi OY, Al-Hassnan ZN, Rashed MS. Determination of homocitrulline in urine of patients with HHH syndrome by liquid chromatography tandem mass spectrometry. Anal Bioanal Chem. 2006; 386:2013–2017. [PubMed: 17053917] 14. Meulenbroek EM, Paspaleva K, Thomassen EA, Abrahams JP, Goosen N, Pannu NS. Involvement of a carboxylated lysine in UV damage endonuclease. Protein Sci. 2009; 18:549–558. [PubMed: 19241382] 15. Dementin S, Bouhss A, Auger G, Parquet C, Mengin-Lecreulx D, Dideberg O, van Heijenoort J, Blanot D. Evidence of a functional requirement for a carbamoylated lysine residue in MurD, MurE and MurF synthetases as established by chemical rescue experiments. Eur J Biochem. 2001; 268:5800–5807. [PubMed: 11722566] 16. Cha J, Mobashery S. Lysine N(zeta)-decarboxylation in the BlaR1 protein from Staphylococcus aureus at the root of its function as an antibiotic sensor. J Am Chem Soc. 2007; 129:3834–3835. [PubMed: 17343387] 17. Shi D, Yu X, Roth L, Morizono H, Tuchman M, Allewell NM. Structures of N-acetylornithine transcarbamoylase from Xanthomonas campestris complexed with substrates and substrate analogs imply mechanisms for substrate binding and catalysis. Proteins. 2006; 64:532–542. [PubMed: 16741992] 18. Shi D, Morizono H, Yu X, Roth L, Caldovic L, Allewell NM, Malamy MH, Tuchman M. Crystal structure of N-acetylornithine transcarbamylase from Xanthomonas campestris: a novel enzyme in a new arginine biosynthetic pathway found in several eubacteria. J Biol Chem. 2005; 280:14366– 14369. [PubMed: 15731101] 19. Morizono H, Cabrera-Luque J, Shi D, Gallegos R, Yamaguchi S, Yu X, Allewell NM, Malamy MH, Tuchman M. Acetylornithine transcarbamylase: a novel enzyme in arginine biosynthesis. J Bacteriol. 2006; 188:2974–2982. [PubMed: 16585758] 20. da Silva FR, Vettore AL, Kemper EL, Leite A, Arruda P. Fastidian gum: the Xylella fastidiosa exopolysaccharide possibly involved in bacterial pathogenicity. FEMS Microbiol Lett. 2001; 203:165–171. [PubMed: 11583843] 21. da Silva AC, Ferro JA, Reinach FC, Farah CS, Furlan LR, Quaggio RB, Monteiro-Vitorello CB, Van Sluys MA, Almeida NF, Alves LM, do Amaral AM, Bertolini MC, Camargo LE, Camarotte G, Cannavan F, Cardozo J, Chambergo F, Ciapina LP, Cicarelli RM, Coutinho LL, Cursino-Santos Li et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript JR, El-Dorry H, Faria JB, Ferreira AJ, Ferreira RC, Ferro MI, Formighieri EF, Franco MC, Greggio CC, Gruber A, Katsuyama AM, Kishi LT, Leite RP, Lemos EG, Lemos MV, Locali EC, Machado MA, Madeira AM, Martinez-Rossi NM, Martins EC, Meidanis J, Menck CF, Miyaki CY, Moon DH, Moreira LM, Novo MT, Okura VK, Oliveira MC, Oliveira VR, Pereira HA, Rossi A, Sena JA, Silva C, de Souza RF, Spinola LA, Takita MA, Tamura RE, Teixeira EC, Tezza RI, Trindade dos Santos M, Truffi D, Tsai SM, White FF, Setubal JC, Kitajima JP. Comparison of the genomes of two Xanthomonas pathogens with differing host specificities. Nature. 2002; 417:459– 463. [PubMed: 12024217] 22. Shi D, Yu X, Cabrera-Luque J, Chen TY, Roth L, Morizono H, Allewell NM, Tuchman M. A single mutation in the active site swaps the substrate specificity of N-acetyl-L-ornithine transcarbamylase and N-succinyl-L-ornithine transcarbamylase. Protein Sci. 2007; 16:1689–1699. [PubMed: 17600144] 23. Shi D, Morizono H, Cabrera-Luque J, Yu X, Roth L, Malamy MH, Allewell NM, Tuchman M. Structure and catalytic mechanism of a novel N-succinyl-L-ornithine transcarbamylase in arginine biosynthesis of Bacteroides fragilis. J Biol Chem. 2006; 281:20623–20631. [PubMed: 16704984] 24. Pastra-Landis SC, Foote J, Kantrowitz ER. An improved colorimetric assay for aspartate and ornithine transcarbamylases. Anal Biochem. 1981; 118:358–363. [PubMed: 7337232] 25. Shi D, Yu X, Roth L, Morizono H, Hathout Y, Allewell NM, Tuchman M. Expression, purification, crystallization and preliminary X-ray crystallographic studies of a novel acetylcitrulline deacetylase from Xanthomonas campestris. Acta Crystallogr Sect F Struct Biol Cryst Commun. 2005; 61:676–679. 26. Otwinowski Z, Minor W. Processing of X-ray Diffraction Data Collected in Oscillation Mode. Methods in Enzymology. 1997; 276:307–326. 27. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr. 1994; 50:760–763. [PubMed: 15299374] 28. Jones TA, Zou JY, Cowan SW, Kjeldgaard M. Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr A. 1991; 47(Pt 2):110–119. [PubMed: 2025413] 29. Brunger AT, Adams PD, Clore GM, DeLano WL, Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges M, Pannu NS, Read RJ, Rice LM, Simonson T, Warren GL. Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr. 1998; 54:905–921. [PubMed: 9757107] 30. Brunger AT. Free R value: a novel statistical quantity for assessing the accuracy of crystal structures. Nature. 1992; 355:472–475. [PubMed: 18481394] 31. Laskowski RA, MacArthur MW, Moss DS, Thornton JM. PROCHECK: a program to check the stereochemical quality of protein structures. J Appl Crystallogr. 1993; 26:283–291. 32. Golemi D, Maveyraud L, Vakulenko S, Samama JP, Mobashery S. Critical involvement of a carbamylated lysine in catalytic function of class D beta-lactamases. Proc Natl Acad Sci U S A. 2001; 98:14280–14285. [PubMed: 11724923] 33. Lapko VN, Smith DL, Smith JB. In vivo carbamylation and acetylation of water-soluble human lens alphaB-crystallin lysine 92. Protein Sci. 2001; 10:1130–1136. [PubMed: 11369851] 34. Shi D, Morizono H, Ha Y, Aoyagi M, Tuchman M, Allewell NM. 1.85-A resolution crystal structure of human ornithine transcarbamoylase complexed with N-phosphonacetyl-L-ornithine. Catalytic mechanism and correlation with inherited deficiency. J Biol Chem. 1998; 273:34247– 34254. [PubMed: 9852088] 35. Langley DB, Templeton MD, Fields BA, Mitchell RE, Collyer CA. Mechanism of inactivation of ornithine transcarbamoylase by Ndelta -(N'-Sulfodiaminophosphinyl)-L-ornithine, a true transition state analogue? Crystal structure and implications for catalytic mechanism. J Biol Chem. 2000; 275:20012–20019. [PubMed: 10747936] 36. Cha J, Vakulenko SB, Mobashery S. Characterization of the beta-lactam antibiotic sensor domain of the MecR1 signal sensor/transducer protein from methicillin-resistant Staphylococcus aureus. Biochemistry. 2007; 46:7822–7831. [PubMed: 17550272] Li et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 37. O'Leary MH, Jaworski RJ, Hartman FC. C nuclear magnetic resonance study of the CO(2) activation of ribulosebisphosphate carboxylase from Rhodospirillum rubrum. Proc Natl Acad Sci U S A. 1979; 76:673–675. [PubMed: 16592618] 38. Golemi-Kotra D, Cha JY, Meroueh SO, Vakulenko SB, Mobashery S. Resistance to beta-lactam antibiotics and its mediation by the sensor domain of the transmembrane BlaR signaling pathway in Staphylococcus aureus. J Biol Chem. 2003; 278:18419–18425. [PubMed: 12591921] 39. Schneider KD, Bethel CR, Distler AM, Hujer AM, Bonomo RA, Leonard DA. Mutation of the active site carboxy-lysine (K70) of OXA-1 beta-lactamase results in a deacylation-deficient enzyme. Biochemistry. 2009; 48:6136–6145. [PubMed: 19485421] 40. Huang CY, Hsu CC, Chen MC, Yang YS. Effect of metal binding and posttranslational lysine carboxylation on the activity of recombinant hydantoinase. J Biol Inorg Chem. 2009; 14:111–121. [PubMed: 18781344] 41. Zhang Q, Wang Y. High mobility group proteins and their post-translational modifications. Biochim Biophys Acta. 2008; 1784:1159–1166. [PubMed: 18513496] 42. Jabri E, Carr MB, Hausinger RP, Karplus PA. The crystal structure of urease from Klebsiella aerogenes. Science. 1995; 268:998–1004. [PubMed: 7754395] 43. Young PG, Smith CA, Metcalf P, Baker EN. Structures of Mycobacterium tuberculosis folylpolyglutamate synthase complexed with ADP and AMPPCP. Acta Crystallogr D Biol Crystallogr D. 2008; 64:745–753. 44. Gordon E, Flouret B, Chantalat L, van Heijenoort J, Mengin-Lecreulx D, Dideberg O. Crystal structure of UDP-N-acetylmuramoyl-L-alanyl-D-glutamate: meso-diaminopimelate ligase from Escherichia coli. J Biol Chem. 2001; 276:10999–11006. [PubMed: 11124264] 45. Taylor TC, Andersson I. Structure of a product complex of spinach ribulose-1,5-bisphosphate carboxylase/oxygenase. Biochemistry. 1997; 36:4041–4046. [PubMed: 9092835] 46. Jensen R. Activation of Rubisco controls CO(2) assimilation in light: a perspective on its discovery. Photosynth Res. 2004; 82:187–193. [PubMed: 16151874] 47. Falkowski PG. Photosynthesis: the paradox of carbon dioxide efflux. Curr Biol. 1997; 7:R637– R639. [PubMed: 9368746] 48. Roos A, Boron WF. Intracellular pH. Physiol Rev. 1981; 61:296–434. [PubMed: 7012859] 49. Smith KS, Ferry JG. Prokaryotic carbonic anhydrases. FEMS Microbiol Rev. 2000; 24:335–366. [PubMed: 10978542] 50. Bahn YS, Muhlschlegel FA. CO2 sensing in fungi and beyond. Curr Opin Microbiol. 2006; 9:572– 578. [PubMed: 17045514] 51. Esposito G, Jaiswal BS, Xie F, Krajnc-Franken MA, Robben TJ, Strik AM, Kuil C, Philipsen RL, van Duin M, Conti M, Gossen JA. Mice deficient for soluble adenylyl cyclase are infertile because of a severe sperm-motility defect. Proc Natl Acad Sci U S A. 2004; 101:2993–2998. [PubMed: 14976244] 52. Townsend PD, Holliday PM, Fenyk S, Hess KC, Gray MA, Hodgson DR, Cann MJ. Stimulation of mammalian G-protein-responsive adenylyl cyclases by carbon dioxide. J Biol Chem. 2009; 284:784–791. [PubMed: 19008230] 53. Kwon JY, Dahanukar A, Weiss LA, Carlson JR. The molecular basis of CO2 reception in Drosophila. Proc Natl Acad Sci U S A. 2007; 104:3574–3578. [PubMed: 17360684] 54. Jones WD, Cayirlioglu P, Kadow IG, Vosshall LB. Two chemosensory receptors together mediate carbon dioxide detection in Drosophila. Nature. 2007; 445:86–90. [PubMed: 17167414] 55. Xu Z, Liu Y, Yang Y, Jiang W, Arnold E, Ding J. Crystal structure of D-Hydantoinase from Burkholderia pickettii at a resolution of 2.7 Angstroms: insights into the molecular basis of enzyme thermostability. J Bacteriol. 2003; 185:4038–4049. [PubMed: 12837777] 56. Sun T, Nukaga M, Mayama K, Braswell EH, Knox JR. Comparison of beta-lactamases of classes A and D: 1.5-A crystallographic structure of the class D OXA-1 oxacillinase. Protein Sci. 2003; 12:82–91. [PubMed: 12493831] 57. Maveyraud L, Golemi D, Kotra LP, Tranier S, Vakulenko S, Mobashery S, Samama JP. Insights into class D beta-lactamases are revealed by the crystal structure of the OXA10 enzyme from Pseudomonas aeruginosa. Structure. 2000; 8:1289–1298. [PubMed: 11188693] Li et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 58. Benning MM, Kuo JM, Raushel FM, Holden HM. Three-dimensional structure of the binuclear metal center of phosphotriesterase. Biochemistry. 1995; 34:7973–7978. [PubMed: 7794910] 59. Thoden JB, Phillips GN Jr, Neal TM, Raushel FM, Holden HM. Molecular structure of dihydroorotase: a paradigm for catalysis through the use of a binuclear metal center. Biochemistry. 2001; 40:6989–6997. [PubMed: 11401542] 60. Abendroth J, Niefind K, Schomburg D. X-ray structure of a dihydropyrimidinase from Thermus sp. at 1.3 A resolution. J Mol Biol. 2002; 320:143–156. [PubMed: 12079340] 61. Yang H, Carr PD, McLoughlin SY, Liu JW, Horne I, Qiu X, Jeffries CM, Russell RJ, Oakeshott JG, Ollis DL. Evolution of an organophosphate-degrading enzyme: a comparison of natural and directed evolution. Protein Eng. 2003; 16:135–145. [PubMed: 12676982] 62. Bertrand JA, Auger G, Martin L, Fanchon E, Blanot D, Le Beller D, van Heijenoort J, Dideberg O. Determination of the MurD mechanism through crystallographic analysis of enzyme complexes. J Mol Biol. 1999; 289:579–590. [PubMed: 10356330] 63. Park IS, Hausinger RP. Requirement of carbon dioxide for in vitro assembly of the urease nickel metallocenter. Science. 1995; 267:1156–1158. [PubMed: 7855593] 64. Sulzenbacher G, Bignon C, Nishimura T, Tarling CA, Withers SG, Henrissat B, Bourne Y. Crystal structure of Thermotoga maritima alpha-L-fucosidase. Insights into the catalytic mechanism and the molecular basis for fucosidosis. J Biol Chem. 2004; 279:13119–13128. [PubMed: 14715651] 65. Eichman BF, O'Rourke EJ, Radicella JP, Ellenberger T. Crystal structures of 3-methyladenine DNA glycosylase MagIII and the recognition of alkylated bases. Embo J. 2003; 22:4898–4909. [PubMed: 14517230] 66. Burman JD, Stevenson CE, Sawers RG, Lawson DM. The crystal structure of Escherichia coli TdcF, a member of the highly conserved YjgF/YER057c/UK114 family. BMC Struct Biol. 2007; 7:30. [PubMed: 17506874] Li et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Stereo view of the structure and hydrogen bonding network surrounding residue 302. A, PALAO bound wild-type AOTCase, B, PALAO bound K302A AOTCase C, PALAO bound K302E AOTCase, D, PALAO bound K302R AOTCase. Contours of the electron density maps (2Fo-Fc) around PALAO, residue 302 and water molecules are shown as a brown cage at 1.0σ. The final refined positions of the ligands and surrounding protein residues are represented as colored sticks. The predicted hydrogen bonding interactions are in pink dashed lines. The water molecules are represented as pink balls. The carbon of PALAO, residue 302 and other protein residues are shown in pink, light blue and green sticks, respectively. Li et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. 13C NMR spectra of wild-type (upper panel) and K302A mutant (lower panel) AOTCase (1 mM). Experiments were performed in 100 mM Tris HCl, 50 mM NaCl, 7% D2O, pH 8.0, supplemented with 20 mM NaH13CO3. The position of the resonance attributed to carboxylated lysine in the enzyme is around 164 ppm. Li et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Chemical structure of carbamylated vs. carboxylated lysine. Li et al. Page 16 Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 17 Table 1 Data collection and refinement statistics Dataset PALAO K302A K302E K302R Space group I213 I213 I213 I213 Resolution (Å) 2.2 1.9 1.85 2.2 Unit-cell parameters (Å) a = b = c =128.88 a = b = c =128.92 a = b = c =129.29 a = b = c =127.39 Measurements 219,475 305,757 390,128 246,817 Unique reflections 18,269 (1,832) a 28,236 (1,365) 30,622 (1,456) 17,635 (879) Redundancy 12.0 (11.8) 10.8(5.4) 12.8 (5.4) 14.0 (13.1) Completeness (%) 99.8 (100.0) 100.0 (100.0) 99.7 (95.1) 100.0 (100.0) <I/σ (I)> 15.0 (4.9) 16.4 (2.3) 19.8 (2.8) 8.7 (3.7) Rmerg b 7.4 (48.4) 6.5(64.9) 5.2 (55.3) 9.8 (79.1) Wilson B (Å2) 30.4 27.6 28.6 21.9 Refinement Resolution range (Å) 50.0-2.2 50-1.9 50-1.85 50-2.2 No. of protein atoms 2620 2613 2617 2619 No. of water atoms 90 219 193 146 No. of hetero atoms 24 24 24 24 Rmsd of bond lengths (Å) 0.006 0.005 0.005 0.005 Rmsd of bond angle (°) 1.1 1.2 1.2 1.2 Rwork (%)c 20.0 19.8 20.0 18.9 Rfree (%)d 24.3 23.2 23.2 22.2 Average B factor (Å2) 41.7 32.2 32.3 35.3 aFigures in brackets apply to the highest-resolution shell. bRmerg = ΣhΣi|I(h,i)-<I(h)>|/∑hΣiI(h,i), where I(h,i) is the intensity of the ith observation of reflection h, and < I(h)> is the average intensity of redundant measurements of reflection h. cRwork= Σh‖Fobs| – |Fcalc‖/Σh|Fobs|. dRfree = Σh‖Fobs| – |Fcalc‖/Σh|Fobs| for 5% of the reserved reflections. Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 18 Table 2 Interactions between carboxylated lysine and other residues at the active site of AOTCase Kcx302 Other residues Bound ligands PALAO CPa AORNb CP +ANORc SO4+ACITd OQ1 K252 NZ 2.6 2.6 2.7 2.6 2.6 OQ1 W1e 2.6 2.6 2.6 2.7 OQ2 S253 N 3.0 3.1 2.8 2.9 2.9 OQ2 H293 NE2 3.0 3.2 3.0 2.9 2.9 NZ W2f 3.1 2.9 3.0 3.0 aThe values were calculated based on PDB ID 3KZM. bThe values were calculated based on PDB ID 3KZN. cThe values were calculated based on PDB ID 3KZO. dThe values were calculated based on PDB ID 3KZK. eThis water molecule hydrogen bonds to N1 atom of PALAO, AORN or ANOR, and backbone O atom of Pro296 as well. fThis water molecule hydrogen bonds to OE1 atom of Glu92 from adjacent subunit as well. Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 19 Table 3 Specific activity of wild-type and mutant AOTCase in the presence of acids (0.5M). Compounds added Specific activity(µmol/min/mg) Wild-type K302A K302E K302R None 43.4 ± 0.4a 23.0 ± 0.5 7.1 ± 0.1 0.059±0.01 Formate 44.1 ± 1.2 26.4 ± 0.6 6.7 ± 0.2 0.093±0.01 Acetate 48.5 ± 1.1 21.2 ± 0.8 6.6 ± 0.5 0.104±0.03 aThe Mean ± S.D. are shown (n = 3). Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 20 Table 4 Protein structures with lysine carboxylation modification PDB ID Enzyme name Residue Organism source Funciton 2OGJ Dihydroorotase 175 A.tumefaciens Bridging two Zn(II) 2Z26 Dihydroorotase 102 E.coli Bridging two Zn(II) 3JZE Dihydroorotase 103 S.enterica Bridging two Zn(II) 2GWN Dihydroorotase 149 P. gingivalis Bridging two Zn(II) 3F4C Organophosphorus hydrolase 243 G. stearothermophilus Bridging two Co(II) 3ICJ Metal-dependent hydrolase 294 P. furiosus Bridging two Zn(II) 3GTX Organophosphorus hydrolase 243 D. radiodurans Bridging two Co(II) 2QPX Metal-dependent hydrolase 166 L. casei Bridging two Zn(II) 2FTW Dihydropyrimidinase 158 D. discoideum Bridging two Zn(II) 2FVK Dihydropyrimidinase 167 S. kluyveri Bridging two Zn(II) 3DC8 Dihydropyrimidinase 147 S. meliloti Bridging two Zn(II) 3GNH L-Lys/Arg carboxypeptidase 211 C. crescentus cb15 Bridging two Zn(II) 3DUG Arginine carboxypeptidase 182 Unidentified Bridging two Zn(II) 2VC7 Phosphotriesterase 137 S. solfataricus Bridging two Co(II) 2R1N Metallophosphotriesterases 169 A. tumefaciens Bridging two Co(II) 2OB3 Phosphotriesterase 169 B. diminuta Bridging two Zn(II) 3E74 Allantoinase 146 E. coli Bridging two Fe(III) 1EJX Urease 217 K. aerogenes Bridging two Ni(II) 1E9Z Urease 219 H. pylori Bridging two Ni(II) 4UBP Urease 220 B. pasteurii Bridging two Ni(II) 1ONW Isoaspartyl dipeptidase 162 E. coli Bridging two Zn(II) 1K1D D-hydanroinase 150 G. stearothermophilus Bridging two Zn(II) 1GKR L-hydanroinase 147 A. aurescens Bridging two Zn(II) 1GKP D-hydanroinase 150 Thermus sp. Bridging two Zn(II) 1NFG D-hydantoinase 148 R. pickettii Bridging two Zn(II) 2ICS Adenine deaminase 154 E. faecalis Bridging two Zn(II) 1RQB Transcarboxylase 184 P. freudenreichii Binding one Co(II) 2QF7 Pyruvate carboxylase 718 R. etli Binding one Zn(II) 3BG3 Pyruvate carboxylase 741 H. sapiens Binding one Mn(II) 2OEM Rubisco-like protein 173 G. kaustophilus Binding one Mg(II) 1WDD Rubisco 201 O. sativa Binding one Mg(II) 1GK8 Rubisco 201 C. reinhardtii Binding one Mg(II) 1BWV Rubisco 201 G. partita Binding one Mg(II) 2WTZ ATP-dependent MurE ligase 262 M. tuberculosis Binding one Mg(II) 2JFG MurD ligase 198 E. coli Catalytic role? 1E8C MurE ligase 224 E. coli Catalytic role? 1JBW Folypolyglutamate synthetase 185 L. casei Catalytic role? 1W78 FolC bifunctional protein 188 E. coli Binding one Mg(II) 3HBR OXA-48 β-lactamase 73 K. pneumoniae Catalytic role Biochemistry. Author manuscript; available in PMC 2011 August 17. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Li et al. Page 21 PDB ID Enzyme name Residue Organism source Funciton 3ISG Class D β-lactamase 70 E. coli Catalytic role 2P9V AmpC beta-lactamase 315 E. coli Catalytic role 1K55 OXA-10 β-lactamase 70 P. aeruginosa Catalytic role 1K38 β-lactamase OXA-2 70 S. typhimurium Catalytic role 1XQL Alanine racemase 129 G. stearothermophilus Binding substrate? 1VFS Alanine racemase 129 S. lavendulae Binding substrate? 1RCQ Alanine racemase 122 P. aeruginosa Binding substrate? 2J6V UV damage endonuclease 229 T. thermophilus Catalytic role 1H01 Cell division protein kinase 2 33 H. sapiens Catalytic role? 2UYN Protein TdcF A58 E. coli Structural role? 1HL9 Fucosidase 338 T. maritime Structural role? 1PU6 3-methyladenine DNA glycosylase 205 H. pylori Structural role? Biochemistry. Author manuscript; available in PMC 2011 August 17.
3M5G
Crystal structure of a H7 influenza virus hemagglutinin
Structures of Receptor Complexes of a North American H7N2 Influenza Hemagglutinin with a Loop Deletion in the Receptor Binding Site Hua Yang, Li-Mei Chen, Paul J. Carney, Ruben O. Donis, James Stevens* Influenza Division, Centers for Disease Control and Prevention, Atlanta, Georgia, United States of America Abstract Human infections with subtype H7 avian influenza viruses have been reported as early as 1979. In 1996, a genetically stable 24-nucleotide deletion emerged in North American H7 influenza virus hemagglutinins, resulting in an eight amino acid deletion in the receptor-binding site. The continuous circulation of these viruses in live bird markets, as well as its documented ability to infect humans, raises the question of how these viruses achieve structural stability and functionality. Here we report a detailed molecular analysis of the receptor binding site of the North American lineage subtype H7N2 virus A/New York/107/2003 (NY107), including complexes with an avian receptor analog (39-sialyl-N-acetyllactosamine, 39SLN) and two human receptor analogs (69-sialyl-N-acetyllactosamine, 69SLN; sialyllacto-N-tetraose b, LSTb). Structural results suggest a novel mechanism by which residues Arg220 and Arg229 (H3 numbering) are used to compensate for the deletion of the 220-loop and form interactions with the receptor analogs. Glycan microarray results reveal that NY107 maintains an avian-type (a2-3) receptor binding profile, with only moderate binding to human-type (a2-6) receptor. Thus despite its dramatically altered receptor binding site, this HA maintains functionality and confirms a need for continued influenza virus surveillance of avian and other animal reservoirs to define their zoonotic potential. Citation: Yang H, Chen L-M, Carney PJ, Donis RO, Stevens J (2010) Structures of Receptor Complexes of a North American H7N2 Influenza Hemagglutinin with a Loop Deletion in the Receptor Binding Site. PLoS Pathog 6(9): e1001081. doi:10.1371/journal.ppat.1001081 Editor: Fe´lix A. Rey, Institut Pasteur, France Received March 16, 2010; Accepted July 28, 2010; Published September 2, 2010 This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. Funding: This work was funded by the Centers for Disease Control and Prevention. Use of the Advanced Photon Source at Argonne National Laboratory was supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC02-06CH11357. Glycan microarrays as well as glycans for direct binding experiments were produced for the Centers for Disease Control by the CFG funded by National Institute of General Medical Sciences Grant GM62116. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: fwb4@cdc.gov Introduction Influenza is an acute respiratory virus that infects up to 20% of the population in the United States, resulting in ,36,000 deaths annually [1,2]. The two membrane glycoproteins on the surface of influenza A virus, hemagglutinin (HA), which functions as the receptor binding and membrane fusion glycoprotein in cell entry, and neuraminidase (NA), which functions as the receptor destroying enzyme in virus release, form the basis for defining subtypes [3]. To date, 16 HA (H1–H16) and 9 NA (N1–N9) have been identified in avian species [4], while in the last century, only three subtypes, H1N1 in 1918 and 2009, H2N2 in 1957, and H3N2 in 1968 [5,6,7], have successfully adapted to humans. Hemagglutinin binds to sialic acid (SA) glycans present on host cell surfaces. The receptors on epithelial cells of the human upper respiratory tract are mainly a2-6-linked SA moieties [8]. Since avian influenza viruses pre- dominately bind a2-3-linked SA, and human influenza viruses preferentially bind to a2-6-linked SA, human infection by avian influenza viruses is rare [9]. However, since 1997 a growing number of human cases of avian influenza infection have been reported [10], including H5N1, H7N2, H7N3, H7N7, and H9N2 strains [11]. Although the current situation with the pandemic H1N1 influenza virus dominates public health efforts, the prospect of a novel pandemic emerging from these isolated cases continues to be a major public health threat around the world. Early cases of human infection by H7 influenza viruses are reported as far back as 1979 [12,13]. Since 2002, multiple outbreaks and human infections of H7 subtype viruses; within both Eurasian and North American lineages have been reported. In the Netherlands in 2003, a highly pathogenic avian influenza (HPAI) H7N7 outbreak resulted in more than 80 cases of human infections, including one fatality [14,15]. In New York in 2003, a single case of human respiratory infection of H7N2 was reported [16] and in British Columbia in 2004, an H7N3 virus caused two cases of conjunctivitis [17,18]. More recently in 2007, the United Kingdom reported several cases of low pathogenic avian influenza (LPAI) H7N2 virus infections that caused influenza-like illness and conjunctivitis [19]. Since 1996, H7 viruses of the North American lineage have been circulating in regional live bird markets [20], containing a 24-nucleotide deletion resulting in an eight amino acid deletion in the receptor-binding site (RBS) of HA (Figure S1). The recent human infections with H7 in North America have raised public health concerns as to how these viruses adapt to such a dramatic structural change while remaining one of the predominant circulating viral strains. A recent study of H7 viruses isolated from previous outbreaks revealed efficient replication in both mouse and ferret animal models [21]. In particular, ferret studies with A/New York/107/2003 (NY107), an H7N2 virus isolated from a man in New York, not only showed efficient replication in the upper respiratory tract of the ferret but also the capacity for PLoS Pathogens | www.plospathogens.org 1 September 2010 | Volume 6 | Issue 9 | e1001081 intra-species transmission by direct contact [21,22]. Interestingly, both an increased preference for a2-6 and decreased preference for a2-3-linked sialosides of this virus compared to the other avian influenza viruses was shown by previous glycan microarray analysis but less so by a competitive solid-phase binding assay [22,23]. Here we report a detailed molecular analysis of the RBS of the HA from North American lineage H7N2 virus, NY107, including glycan microarray analyses and structural analyses of the HA in complex with an avian receptor analog (39-Sialyl-N-acetyllactosa- mine, 39SLN) and two human receptor analogs (69-Sialyl-N- acetyllactosamine, 69SLN; Sialyllacto-N-tetraose b, LSTb). These results provide important insight into the interaction of H7 HAs with both avian and human hosts. Results Overall structure By using x-ray crystallography, the structure of H7 HA from the NY107 virus was determined to 2.6 A˚ resolution (Table 1). In addition, we also report three H7 HA receptor complex structures, with avian receptor analog (39SLN) to 2.7 A˚ resolution and with human receptor analogs (69SLN and LSTb) to 3.0 A˚ and 2.6 A˚ resolution, respectively (Table 1). The overall structure of NY107 is similar to other reported HA structures with a globular head containing the RBS and vestigial esterase domain, and a membrane proximal domain with its distinctive, central helical stalk and HA1/HA2 cleavage site (Figure 1A). Although five asparagine-linked glycosylation sites are predicted in the NY107 HA monomer, interpretable electron density was observed at only two sites, Asn38 in HA1 and Asn82 in HA2 (all residue numbers are based on H3 numbering). At these sites, only one or two N- acetyl glucosamines could be interpreted. During viral replication, HA is synthesized as a single chain precursor (HA0) and cleaved by specific host proteases into the infectious HA1/HA2 form. In baculovirus expression systems, highly pathogenic HAs, with a polybasic cleavage site, are expressed as an HA1/HA2 form [24], whereas HAs with monobasic cleavage sites (single Arg) from low pathogenic viruses are expressed as the HA0 form [25]. NY107 is regarded as a low pathogenic virus, and as expected, was produced in the HA0 form (Figure S2). However, subsequent digestion with thrombin protease to remove the His-tag resulted in cleavage to a profile on SDS-PAGE comparable to that of an HA1/HA2 form (Figure S2). A comparison of the NY107 cleavage site with the consensus cleavage pattern in the MEROPS database (http://merops. sanger.ac.uk) suggests it to be a possible thrombin cleavage site. Based on their molecular phylogenies, HAs are divided into two groups and five clades: group 1 includes H8, H9, and H12; H1, H2, H5, and H6; H11, H13 and H16; group 2 includes H3, H4, and H14; H7, H10 and H15 [26]. Among all available HA structures, we selected ten representative HAs from both avian and human subtypes for structural analysis. As expected, NY107 HA is structurally very similar to the Avian-H7 in all comparisons and closely related to H3, the other group 2 members used in the analyses (Tables S1 and S2). The receptor binding site The RBS is at the membrane distal end of each HA monomer and its specificity for sialic acid and the nature of its linkage to a vicinal galactose residue is a major determinant of host range- restriction. The consensus RBS for all current HAs is composed of three major structural elements: a 190-helix (residues 188–194), a 220-loop (residues 221–228), and a 130-loop (residues 134–138). In addition, highly conserved residues (Tyr98, Trp153, His183, and Tyr195) form the base of the pocket. Although the NY107 RBS is similar to other subtypes (H1, H2, H3, H5, and H9), a previously observed specific feature of H7 HAs, is also observed in the NY107 150-loop region: two residues inserted at position 158 result in this loop protruding more than 6A˚ towards the binding site compared to other subtype HAs (Figure 1B and Table S2) [27]. More interestingly, the eight amino acid deletion, only found in the North American lineage H7s, from position 221 to 228 (Figure S1), resulted in a complete loss of the 220-loop (Figure 1B). Sequence alignment shows that Arg220 and Arg229 are conserved in all influenza A HA subtypes (Figure S1), but structural alignment of NY107 HA shows Arg220 occupying the Gly228 position, and the much shorter loop turns at residue Pro217 (Figure 1C). The Ca distance between NY107 Arg220 and its homolog in the Av-H7 structure (PDB: 1TI8) [27] is 5.8A˚ , and they point in opposite directions (Figure 1C). The side chain direction of Av-H7 Arg220 is almost parallel with the beta sheet after Arg229, whereas the NY107 Arg220 points downward to the binding pocket. The Ca position of Arg229 in both H7 structures remains the same, except the side chain in the NY107 swings away by about 5.9A˚ (Figure 1C) and could help to stabilize this region by forming a hydrogen bond to the mainchain carbonyl of Gln210 in the neighboring monomer. In the absence of the 220-loop in NY107 HA, upon glycan binding the long side chain of Arg220 compensates for its loss and is displaced 4A˚ upward to form hydrogen bonds with receptor analogs inside the binding pocket (Figure 1D). Effect of loop truncation on the receptor binding specificity of NY107 Previously, mutations in the HA receptor binding domains of H1N1 (Glu190Asp/Gly225Asp) and H2N2/H3N2 (Gln226Leu and Gly228Ser) subtypes were responsible for adaptation of these viruses to pandemic strains [24,28,29,30]. Due to missing residues 221–228 in the NY107 HA RBS, neither mechanism for adaptation is possible. Thus, in order to look more closely at the role of the missing loop and its effect on receptor specificity, we first subjected the recombinant HA (recHA) to glycan microarray analyses and compared it to a reverse genetics-derived NY107 virus, and a co-circulating Eurasian virus and recHA, A/ Author Summary Influenza virus adaptation to different hosts usually results in a switch in receptor specificity of the viral surface coat protein, hemagglutinin. Indeed, the hemagglutinin sub- types from the last two human influenza pandemics of the 20th Century (H2 in 1957 and H3 1968) both adapted successfully to human-type receptor specificity through only two amino acid mutations in the receptor binding pocket (Glutamine226RLeucine and Glycine228RSerine). The recent human infections reported with other avian subtypes such as H5, H7 and H9 have raised public health concerns and focused efforts on identifying potential subtypes from which a future pandemic strain may emerge. Since 1996, H7 viruses of the North American lineage have been circulating in regional live bird markets, containing an eight amino acid deletion in the receptor- binding site of HA. Here we report a detailed structural analysis of the receptor binding site of a hemagglutinin from the North American lineage of H7N2 viruses, in complex with avian and human receptor analogs, to understand how these viruses have adapted to such a dramatic structural change in the binding site while remaining one of the predominant circulating viral strains. Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 2 September 2010 | Volume 6 | Issue 9 | e1001081 Netherlands/219/2003 (NL219), that has the consensus avian sequence in the 220-loop and it also infected a human [15]. Glycan microarray analysis of recombinant NY107 (Figure 2A and Table 2) revealed a highly restricted binding profile with strong binding to only a2-3 sulfated (#4–8), a2-3 branched (#9– 11) and mixed a2-3/a2-6 branched sialosides (#60–64) as well as to the long linear sialyl di- and tri-lactosamines (#22, 24). Weak binding was also observed (above background) to other a2-3 glycans on the array. The recombinant NY107 also revealed a strict glycan binding preference to only one a2-6 glycan, the internal structure, Galb1-3(Neu5Aca2-6)GlcNAcb1-3Galb1-4Glc (#58; LSTb) (Figure 2A), a glycan highlighted in a previous study [22]. The virus with higher valency and avidity revealed stronger binding to all a2-3 groups, in addition to the branched di-sialyl a2- 6 biantennary structures (#46–48) as well the LSTb (#58) (Figure 2B and Table 2). In contrast, the NL219 recHA (Figure 2C and Table 2) bound well to only the avian a2-3 containing sialyl- glycans (sulfated, branched, linear and fucosylated). Its corre- sponding virus also reflected this specificity although it also revealed strong binding to a2-3 N-glycolylneuraminic acid (Neu5Gc) containing glycans (#66–70) (Figure 2D and Table 2). To further assess the effect of the missing 220 loop on HA structural stability and receptor specificity it was essential to evaluate these functions on the ancestral HA containing the full length 220-loop. To this end, we engineered an HA with an avian H7 consensus (PQVNGQSG) 220-loop re-introduced (NY107- 220ins) into the NY107 HA and recovered this virus by reverse genetics. Compared to the NY107 virus (Figure 2A) glycan microarray analyses of the resulting NY107-220ins virus (Figure 3A and Table 2) revealed a decrease in binding to branched (#9–11) and linear (#12–27) a2-3 sialosides and a loss of binding to the branched di-sialyl a2-6 biantennary structures (#46–48), LSTb (#58) as well as the mixed a2-3/a2-6 branched sialosides (#60– 64). In addition, sequence analysis of the NY107-220ins HA revealed the presence of quasispecies in the second position of the inserted loop, P(Q/K)VNGQSG, suggesting that re-introduction of the loop alone is not tolerated and does not create an avian-type binding profile. Thus other amino acid substitutions in the HA might have co-evolved with the deletion of the 220 loop to help stabilize the RBS/HA to maintain functionality. When viruses containing this 220-loop deletion emerged in North America in the mid 90’s, four additional amino acid substitutions, Gly114Arg, Asp119Gly, Gly186Glu and Gly205Arg, in the HA1 as well as an Asp19Asn in the HA2 chain were also introduced to most of the circulating isolates. Of these, Gly186Glu and Gly205Arg in the HA1 are close to the RBS, at the monomer interface, and could potentially modulate its structure and/or function. NY107 viruses with a restored consensus 220-loop and a single Glu186Gly (NY107- ins-186) or Arg205Gly (NY107-ins-205) substitution as well as the Glu186Gly/Arg205Gly double substitution (NY107-ins-186/205) were derived by reverse genetics and evaluated. Glycan microarray analysis for the three resulting viruses revealed similar glycan binding profiles with increased binding to a2-3 sialosides, including mixed a2-3/a2-6 branched sialosides (#60–64), a2-3 Neu5Gc (#66–70), but limited binding to the a2,6 sialosides (Figures 3B, 3C, 3D), resulting in a binding profile virtually identical to that of the NL219 virus and other avian influenza viruses (Figure 2D) [30]. Sequence analysis of the three reverse genetics derived viruses revealed no mutations/quasispecies in the HAs of either the NY107- ins-186 or the NY107-ins-186/205 virus stocks, indicative of replication fitness. For the NY107-ins-205 virus however, a Glu186Gly substitution emerged in the HA after only two passages in eggs following recovery from DNA transfection, indicating the importance of the co-variant position 186 with respect to HA functionality/glycan specificity. Altogether, the data indicates that the H7 subtype avian influenza viruses that were circulating in Table 1. Data collection and refinement statistics. NY107 NY107+39SLN NY107+69SLN NY107+LSTb Data collection Space group P212121 P212121 P212121 P212121 Cell dimensions (A˚) 66.96, 115.92, 251.61 67.80, 116.70, 249.84 66.60, 116.58, 250.68 67.08, 116.52, 251.95 Resolution (A˚) 50-2.6 (2.69-2.60)a 30-2.7 (2.80-2.70) 50-3.0 (3.11-3.0) 50-2.6 (2.69-2.60) Rsym or Rmerge 10.6 (41.3) 14.6 (48.6) 14.3 (35.4) 12.2 (31.5) I/s 39.6 (2.0) 24.3 (1.7) 34.2 (8.2) 40.5 (9.9) Completeness (%) 99.2 (99.0) 99.3 (94.6) 92.3 (75.6) 91.3 (86.2) Redundancy 7.2 (6.2) 5.8 (5.5) 4.9 (4.4) 10.9 (11.2) Refinement Resolution (A˚) 50-2.6 (2.67-2.60) 30-2.7 (2.77-2.70) 50-3.0 (3.08-3.00) 50-2.6 (2.67-2.60) No. of reflections (total) 57285 51770 33421 53603 No. of reflections (test) 3053 2769 1779 2842 Rwork/Rfree 21.7/25.6 21.4/26.4 20.5/26.0 20.4/24.7 No. of atoms 11795 11878 11648 12108 r.m.s.d.- bond length (A˚) 0.006 0.006 0.008 0.006 r.m.s.d.- bond angle (u) 0.905 0.974 1.085 0.859 MolProbityb scores Favored (%) 96.9 96.5 94.3 97.1 Outliers (%) (No. of residues) 0.1 (1/1434) 0.0 (0/1429) 0.1 (2/1433) 0.1 (2/1435) aNumbers in parentheses refer to the highest resolution shell. bReference [51]. doi:10.1371/journal.ppat.1001081.t001 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 3 September 2010 | Volume 6 | Issue 9 | e1001081 aquatic birds and poultry in North America before 1996 exhibited a classic avian a2-3 sialoside binding preference. In order for the 220- loop deletion to be tolerated, concurrent Gly186Glu and Gly205Arg substitutions in the vicinity of RBS of HA emerged to achieve a restricted a2-3 binding profile and only a moderate/limited increase in binding to branched di-sialyl a2-6 biantennary structures (#46– 48) as well the a2,6 internal sialoside, LSTb (#58). NY107 avian receptor complex To understand from a structural perspective how NY107 interacts with host receptors, we solved the structure of NY107 in complex with an avian and two human receptor analogs. For the avian receptor analog, 39SLN, the electron density maps revealed well-ordered features for the Sia-1, Gal-2, and GlcNAc-3 in the NY107 HA complex structure (Figure 4A). Structural comparison of NY107 HA binding to other, H1, H2, H3, H5, and H9 subtypes (Figure S2A) revealed that 39SLN binding to NY107 resembled binding of the other published HAs. Indeed, the terminal Sia-1 moiety is positioned almost identically in all structures, and forms the majority of hydrogen bonds and contacts with residues in the RBS (Figure 4A and Table S3). Figure 1. NY107 HA monomer and comparison of its RBS to other HA structures. (A) One monomer is shown with the HA1 chain colored in green and the HA2 chain in cyan. The location of the receptor binding site and the HA1/HA2 cleavage site are circled. (B) The superposition of receptor binding domains of NY107 (green), Av-H7 (marine), 1918-Hu-H1 (magenta), Hu-H5 (yellow), Hu-H3 (orange), and Sw-H9 (grey). The proximity of Arg220 and Gln226 are highlighted. Three structural elements comprising this binding site are labeled. The two major differences are the extended 150-loop and the deletion of 220-loop of NY107. (C) Overlap of NY107 (green) and Av-H7 (marine) (PDB: 1TI8) illustrates the compensatory effect of R220 bringing it close to the position occupied by G228 in the avian HA. (D) Overlap of the NY107 (green), NY107- 39SLN (orange), NY107-69SLN (red), and NY107-LSTb (magenta) structures. All the figures were generated and rendered with the use of MacPyMOL [56]. doi:10.1371/journal.ppat.1001081.g001 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 4 September 2010 | Volume 6 | Issue 9 | e1001081 Published avian HA structures with an intact 220-loop form very close interactions with Gal-2 of 39SLN via residue Gln226 which is important in receptor specificity and host adaptation. For example, in the avian H7/39SLN HA structure it interacts with Gal-2 O4 [31]. In the NY107 HA structure, although Gln226 is absent and no other residue occupies the same space as Gln226 (Figure 1B), Arg220 does forms a hydrogen bond between Arg220 NH2 and Gal-2 O4 (Figure 4A). Interestingly, although there was interpretable density for the GlcNAc-3 (Figure 4A and Figure S4B), no hydrogen bonding was apparent between the HA and the GlcNAc-3, which is consistent with other reported structures [32]. Thus, for binding to avian receptors, the trans conformation of a2-3 linkages is essential and perhaps only the first two saccharides are required. Indeed, due to the absence of 220-loop in the NY107 HA structure, the ‘‘aperture’’ of the RBS formed by 220-loop and 130-loop in regular HAs is increased by ,10 A˚ , so that the branched, internal, and perhaps more complicated glycans might be accommodated more efficiently. NY107 human receptor complexes In the NY107/69SLN complex, only Sia-1 and Gal-2 are ordered (Figure 4B). The Sia-1 remains in the same position as previously analyzed glycan/HA complexes from H1, H2, H3, H5, and H9 (Figure S3B), whereas the Av-H7 complex structure with Sialyllacto-N-tetraose c (LSTc) did not reveal any density for the Sia-1 in the receptor binding site [31]. The Gal-2 position varies significantly among different subtypes. Compared to the human- adapted H1 HA [32], Gal-2 in the NY107 HA is 3A˚ higher, and thus is further from the protein (Figure S3B). In NY107, the Gal-2 only forms an intramolecular, saccharide-saccharide interaction with Sia-1. The poor electron density map and fewer interactions with protein residues suggest that the cis conformation of a2-6 Figure 2. Receptor specificity of NY107 recHA and virus. Glycan microarray analysis of recombinant NY107 HA (A) and NY107 virus (B) compared to the recHA (C) and virus (D) from a Eurasian lineage A/Netherlands/219/2003 H7 influenza virus that was circulating in the same year and also infected a human. Colored bars highlight glycans that contain a2-3 SA (blue) and a2-6 SA (red), a2-6/a2-3 mixed SA (purple), N-glycolyl SA (green), a2-8 SA (brown), b2-6 and 9-O-acetyl SA, and non-SA (grey). Error bars reflect the standard error in the signal for six independent replicates on the array. Structures of each of the numbered glycans are found in Table S4. doi:10.1371/journal.ppat.1001081.g002 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 5 September 2010 | Volume 6 | Issue 9 | e1001081 linkages in 69SLN trisaccharides show a reduced binding affinity with NY107. Glycan array results with NY107 revealed a strong binding signal for the internal a2-6 sialoside, LSTb. To further investigate this interaction, we solved the structure of the NY107/LSTb complex. The final model contained Sia-1, NAG-2, Gal-3, and Gal-5 in the RBS. Although glycan microarray data indicated NY107 to have a specific affinity for LSTb, few interactions were apparent from the crystal structure. Sia-1 still forms multiple hydrogen bonds with residues in the RBS (Table S3 & Figure 4C). The branched Gal-5 interacts with Ser137, to help stabilize the LSTb binding. However, Arg220 and Lys193, the two residues showing close binding with 39SLN, did not form any hydrogen bonds with LSTb. In the structure, Gal-5 also interacts with a crystal packing symmetry mate and thus the flexibility of whole LSTb may be restricted. In solution, with more freedom, the LSTb should be able to tilt closer to the RBS, and thus Glc-4 may have more interactions with the 190-helix than seen in the crystal structure. Discussion Human infections by avian influenza viruses, including H7 subtypes, continue to pose a major public health threat. Although the species barrier prevents avian influenza viruses from widespread infection of the human population, the molecular determinants of efficient interspecies transmission and pathoge- nicity are still poorly understood. The viral coat protein HA however, is perhaps a critical molecule since previous pandemic viruses modified their receptor specificity and overcame the interspecies barrier to spread in the human population. Although HA structures alone and in complex with receptor analogs provide considerable insight into receptor binding, it is clear that HAs from different species and subtypes have significant structural variation. Indeed, low-pathogenic H7N2 avian influenza viruses with an 8 amino acid deletion within its RBS started to circulate in live-bird markets in the northeast United States in 1996. Despite what one would consider a debilitating mutation, these viruses have been reported as the predominant isolate [33]. Whether such a deletion contributed to their evolutionary success and how are an important questions, especially in light of NY107’s ability to produce respiratory illness in humans [16], as well as its reported increased affinity for human-type receptors and ability for contact transmission in ferrets [21]. To try to help answer these questions, we have analyzed the molecular structures of NY107 and its complexes with receptor analogs to explain receptor specificity at the molecular level. The crystal structures of NY107 and its complexes with both avian and human receptor analogs describe a mechanism as to how an influenza virus might adapt by dramatically altering its RBS, and still be functional. Arg220 of the HA1 chain of NY107 compensates for the loss of the 220-loop, by forming hydrogen bonds with Gal-2 from the avian analog (binding was not observed in either of the structures complexes with the human analogs). However, in the LSTb complex, branched Gal-5 forms extra interactions with the 130-loop, thus improving the binding preference for this particular glycan. Consistent with the structural evidence, glycan microarray analyses of NY107 revealed a strong binding preference for the branched a2-6 sialoside, LSTb. Except for the absence of the 220-loop, other key residues within the RBS are conserved in NY107 and thus, direct interactions with sialic acid are maintained. The 220-loop is recognized as one of the three crucial structural elements in the RBS. Aside from the North American lineage H7N2 viruses, which have been circulating with a deletion (221– 228) in this loop, there has been one other report describing a seven amino acid deletion (224–230) in a laboratory generated H3N2 escape mutant which was reported to have a slightly Table 2. Comparison of the sialoside receptor specificity of the HAs from H7 influenza viruses. Glycan Group Graph Numbera NY107 RecHA NY107 Virus NY107-ins Virus NY107-ins E186G Virus NY107-ins R205G Virus NY107-ins E186G/ R205G Virus NL219 RecHA NL219 Virus a2-3 Sulfated 4–8 +++b +++ +++ +++ +++ +++ +++ +++ Branched 9–11 +++ +++ + +++ +++ +++ +++ +++ Linear 12–27 + +++ + +++ +++ +++ +++ +++ Fucosylated 28–34 2 +++ +++ +++ +++ +++ +++ +++ a2-6 Sulfated 41 2 2 2 2 2 2 2 2 Branched mono-sialyl 42–45, 49 2 2 2 2 2 2 2 2 Branched di-sialyl 46–48 2 +++ 2 2 2 2 2 2 Linear 50–56 2 2 2 2 2 2 2 2 Internal 58–59 +++ +++ 2 2 2 2 2 2 Other Sialic acid 1–2 2 +++ + 2 2 2 2 2 a2-3/a2-6 Branched 60–64 2 2 2 +++ +++ +++ +++ +++ Neu5Gcc 65–72 2 2 2 +++ +++ +++ 2 +++ aMembers of each group are identified according to the graph number used in the microarray data in Figures 2 and 3 and correspond to numbers in the complete glycan list (Table S4). bBinding of samples to glycan subclasses are qualitatively estimated based on relative strength of the signal for the data shown in Figures 2 and 3: strong (+++), weak (+), absent (2). cN-glycolylneuraminic acid. doi:10.1371/journal.ppat.1001081.t002 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 6 September 2010 | Volume 6 | Issue 9 | e1001081 increased affinity for a2-3-linked glycans by hemagglutination assay [34]. Meanwhile, the equivalent region in the hemaggluti- nin-esterase-fusion (HEF) protein of influenza C virus reveals a rearrangement resulting in a truncated 260-loop in its RBS (Figure S5) [35]. However, without structural data with appropriate receptor analogs, it is not possible to compare the role of these loop variants in receptor binding to the H7 HA structure described here. When compared to NL219, another co-circulating H7 avian virus HA (Figure 2C and D), overall binding to a2-3-linked glycans was markedly reduced, while increased binding to a2-6- linked receptors was only marginal. However, these results focus attention on only 2 sub-classes of human-type receptors that may be important for infection (and transmission in ferrets). The NY107 virus interaction with biantennary glycans (Figure 2B), although weak (not seen in Figure 2A with recHA), is a possible route for virus entry as biantennary structures are common on tissues, i.e. glycan profiling data from human lung tissue on the Consortium for Functional Glycomics (CFG) web site. In addition, the internal sialoside, LSTb, was observed in both virus and recHA microarray data, suggesting this type of glycan has good affinity for this HA. The significance of this is unknown since LSTb has only been described in human milk [36]. Interestingly, NY107 and NL219 virus receptor binding and specificity has been addressed previously using glycan microarray analysis that reported a significantly increased preference for a2-6 and decreased preference for a2-3-linked sialosides [22]. In addition, the same viruses were also included in a recent study from Gambaryan et al. using a competitive solid-phase binding assay [23]. Our findings confirm and extend the receptor binding specificity reported by these authors in that they reported both viruses binding to sulfated sialylglycans with a lactosamine (Galb1- 4GlcNAc core and reported only a moderate binding affinity for a2-6-sialyllactosamine, the human-type receptor analog used in their assay. The 220-loop is an integral feature of the receptor binding site, and thus one would predict that such a deletion might have compromised this strain to be deleted from the population of circulating viruses. However, this was not the case [33] and its existence appears to be in part due to the additional mutations at Figure 3. Effect of 220-loop deletion and additional RBS mutations on NY107 receptor specificity. NY107 was engineered to restore the 220-loop to a consensus full-length HA from 1996 (A) and additional co-variant amino acid substitutions, Glu186Gly (B), Arg205Gly (C) and the double mutant Glu186Gly/Arg205Gly (D) to restore, on the NY107 framework, an HA RBS found in viruses prior to the introduction of the deletion in North American viruses. Colored bars group glycans as described in Figure 3. Error bars reflect the standard error in the signal for six independent replicates on the array. Structures of each of the numbered glycans are found in Table S4. doi:10.1371/journal.ppat.1001081.g003 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 7 September 2010 | Volume 6 | Issue 9 | e1001081 positions 186 and 205. Restoration of the loop with either or both residues mutated back to the pre-1994 consensus sequence resulted in a classic avian influenza virus binding profile. The emergence of the Glu186Gly mutation in the HA of the NY107- ins-205 mutant after only two passages of the rescued virus in eggs, also indicates the importance of these positions for HA functionality/glycan specificity. Analysis of the structural data reveals that positions 186 and 205 are on opposite sides of a monomer but are both close to the 220-loop deletion region in the trimeric form. The Glu at position 186 is close to Arg220 and may interact with Arg220 when binding avian receptors. Position 205 in the neighboring monomer may be important in trimer stability and maintaining RBS functionality. If one models the pre- 1996 220-loop restored into the NY107 structure, Arg205, Glu186 and the loop all clash, thus explaining the Glu186Gly mutation that emerged in the NY107-ins-205 virus HA after limited egg passage. The NY107 RBS with its more restricted a2-3 glycan binding preference and weak/moderate increase in a2-6 binding may have given the virus a selective advantage to be maintained in poultry at live bird markets and supplying farms. Certain terrestrial birds, such as quails and chickens, have recently been shown to present both human and avian types of receptors in the trachea and intestine [37,38,39]. Although it is not known what specific glycans are presented in these animals, it is conceivable that a virus with mixed specificity might have a distinct advantage over avian viruses that have specific avian receptor requirements, particularly in bird markets where multiple species coalesce. Previous results with H7N2, H9N2 and H5N1 viruses all highlight the fact that an increase in a2-6-binding preference is not sufficient for efficient transmission of avian influenza viruses to humans [22,40,41]. Although it remains to be seen whether prolonged circulation of viruses in terrestrial birds, such as domestic chickens, can provide a possible route for viruses to adapt for efficient human infection [11], continued surveillance of influenza viruses from avian and other animal reservoirs is urgently needed to define their zoonotic potential. Materials and Methods Cloning Based on H3 numbering [42], cDNA corresponding to residues 11–329 (HA1) and 1–176 (HA2) of the ectodomain of the hemagglutinin (HA) from A/New York/107/2003 (H7N2; Genbank:ACC55270) and A/Netherlands/219/2003 (H7N7; Genebank: AAR02640) was cloned into the baculovirus transfer vector, pAcGP67-A (BD Biosciences), incorporating a C-terminal thrombin cleavage site, a ‘‘foldon’’ sequence [43] and a His-tag at the extreme C-terminus of the construct to enable protein purification [25,44]. Transfection and virus amplification were carried out according to the baculovirus expression system manual (BD Biosciences Pharmingen). Figure 4. Glycan interactions within the NY107 RBS. The top panel shows the interactions of NY107 with (A) 39SLN, (B) 69SLN and (C) LSTb. NY107 is shown in orange/red/magenta cartoon respectively. The interacting HA residues are shown as green sticks. The bottom panel shows the electron density map of the ligands. The NY107 is shown in the same colors as above, and the ligands are shown as green sticks, the 2fo-fc electron density maps (contoured at 1s) are shown in grey. Simulated annealing omit maps are shown in supplementary Figure S4. doi:10.1371/journal.ppat.1001081.g004 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 8 September 2010 | Volume 6 | Issue 9 | e1001081 Protein expression and purification Soluble NY107 was recovered from the cell supernatant by metal affinity chromatography using Ni-NTA resin (Qiagen Inc.). Frac- tions containing NY107 were pooled and dialyzed against 10 mM Tris-HCl, 50 mM NaCl, pH 8.0, then subjected to ion-exchange chromatography (IEX) using a Mono-Q HR 10/10 column (GE Healthcare). IEX purified NY107 was subjected to thrombin digest (3 units/mg protein; overnight at 4uC) and purified by gel filtra- tion chromatography using a Superdex-200 16/60 column (GE Healthcare) and 50 mM Tris-HCl, 100 mM NaCl, pH 8.0 as running buffer. Protein eluting as a trimer was buffer exchanged into 10 mM Tris-HCl, 50 mM NaCl, pH 8.0 and concentrated to 14.5 mg/ml for crystallization trials. At this stage, the protein sample still contained the additional plasmid-encoded residues at both the N (ADPG) and C terminus (SGRLVPR). Crystallization, ligand soaking and data collection Initial crystallization trials were set up using a Topaz Free Interface Diffusion (FID) Crystallizer system (Fluidigm Corpora- tion, San Francisco, CA). Crystals were observed in several conditions containing PEG 3350 or PEG 4000. Following opti- mization, diffraction quality crystals for NY107 were obtained at room temperature using a modified method for microbath under oil [45], by mixing the protein with reservoir solution containing 20% PEG 3350, 0.2 M magnesium chloride at pH 7.2. For receptor analog complexes, crystals were soaked for 3 hours in the crystallization buffer containing 10 mM 39SLN or 69SLN (V-labs Inc., Covington, LA), or overnight in 10mM LSTb (Sigma, St. Louis, MO). All crystals were flash-cooled at 100K using 20% glycerol as the cryo-protectant. Datasets were collected at Advanced Photon Source (APS) beamlines 22 ID and BM at 100K. Data were processed with the DENZO-SACLEPACK suite [46]. Statistics for data collection are presented in Table 1. Structure determination and refinement The structure of NY107 was determined by molecular replacement with Phaser [47] using the structure of the avian H7 (Av-H7) from A/turkey/Italy/2002, pdb:1TI8 (HA1, 78% identity; HA2, 90% identity) as the searching model. One HA trimer occupies the asymmetric unit with an estimated solvent content of 58% based on a Matthews’ coefficient (Vm) of 2.9 A˚ 3/ Da. Rigid body refinement of the trimer led to an overall R/Rfree of 28.6%/37.4%. The model was then ‘‘mutated’’ to the correct sequence and rebuilt by Coot [48], then the protein structures were refined with REFMAC [49] using TLS refinement [50]. The final models were assessed using MolProbity [51]. The three complex structures were refined and evaluated using the same strategy. All statistics for data processing and refinement are presented in Table 1. Electron density maps (2fo-fc) were generated in Refmac [49] while simulated annealing omit maps were generated by sa-omit-map, a part of the Crystallography and NMR System (CNS) software [52]. Virus generation Wild type and mutant viruses of NY107 (H7N2) and A/ Netherland/219/2003 (H7N7) were generated from plasmids by a reverse genetics approach [53]. To generate viruses with amino acid insertion or substitution in the HA, mutations were introduced into plasmid DNA with an overlap extension PCR approach [54]. Viruses derived by plasmid transfection of HK293 cells were propagated in eggs. The genomes of resulting virus stocks were sequenced to detect the emergence of possible variants during amplification. Glycan binding analyses Glycan microarray printing and recHA analyses have been described previously [24,30,44,55] (see Table 2 for glycans used for analyses in these experiments). Virus were analyzed on the microarray as described previously [30]. PDB accession codes The atomic coordinates and structure factors of NY107 are available from the RCSB PDB under accession codes 3M5G for the unliganded NY107, 3M5H for the NY107 with 39-SLN and 3M5I and 3M5J for NY107 with 69SLN and LSTb, respectively. Accession/ID numbers for genes/proteins used in this work A/New York/107/03 (H7N2), Genbank: ACC55270; A/ Netherlands/219/03 (H7N7), Genbank: AAR02640; A/Hong Kong/1-9/68 (H3N2), 2HMG; A/Duck/Ukraine/1/63 (H3N8), PDB: 1MQL; A/South Carolina/1/18 (H1N1), PDB: 1RD8; A/ Puerto Rico/8/34 (H1N1), PDB: 1RU7; A/Swine/Iowa/15/30 (H1N1), PDB: 1RUY; A/Singapore/1/1957 (H2N2), PDB: 2WRC; A/Viet Nam/1203/04 (H5N1), PDB: 2FK0; A/Duck/ Singapore/3/97 (H5N3), PDB: 1JSM; A/Swine/Hong Kong/9/ 98 (H9N2), PDB: 1JSD; A/Turkey/Italy/8000/02 (H7N3), PDB: 1TI8; C/Johannesburg/1/66, 1FLC. Supporting Information Figure S1 Sequence alignment of selected structurally available HAs. Human H3 (PDB: 2HMG), Avian H3 (PDB: 1MQL), 1918- Human H1 (PDB: 1RD8), 1934-Human H1 (PDB: 1RU7), Swine H1 (PDB: 1RUY), 1957-Huamn H2 (PDB: 2WRC), Human H5 (PDB: 2FK0), Avian H5 (PDB: 1JSM), Swine H9 (PDB: 1JSD), and Avian H7 (PDB: 1TI8) were used in the alignments. The fusion domain of HA1 is highlighted in magenta, the vestigial esterase domain is highlighted in green, the receptor binding domain is highlighted in blue, and the fusion domain of HA2 is highlighted in red. Residue numbering is based on the H3 HA sequence. Found at: doi:10.1371/journal.ppat.1001081.s001 (2.84 MB TIF) Figure S2 Expression and purification of NY107. SDS-PAGE reveals that NY107 was expressed as the HA0 form with a mass approximately 60kDa (middle lane). Thrombin cleavage resulted in an unexpected reduction in band size to a HA1/HA2 profile (right lane) with possible multiple glycoforms for the HA2 clearly present. Found at: doi:10.1371/journal.ppat.1001081.s002 (0.23 MB TIF) Figure S3 Comparison of glycan binding to NY107 with other HAs. A. Overlap of a2-3 ligands binding in the receptor binding site from NY-H7 (green), Av-H3 (orange), 1930-Hu-H1 (magenta), 1957-Hu-H2 (cyan), Av-H5 (yellow), and Sw-H9 (grey). B. Overlap of a2-6 linkage ligands binding in the receptor binding site from NY-H7 (green), Av-H3 (orange), 1930-Hu-H1 (magenta), 1957-Hu-H2 (cyan), Av-H5 (yellow), and Sw-H9 (grey). Found at: doi:10.1371/journal.ppat.1001081.s003 (2.55 MB TIF) Figure S4 Simulated annealing omit maps of the receptor binding site (contoured at 1s). A. NY107 (blue), B. NY107-39SLN (orange), C. NY107-69SLN (red), and D. NY107-LSTb (magenta). The protein model is shown in cartoon, and the residues involved in the binding to receptor analogs were shown in sticks. Maps were generated using version 1.2 of the Crystallography and NMR System (CNS) software. Found at: doi:10.1371/journal.ppat.1001081.s004 (1.93 MB TIF) Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 9 September 2010 | Volume 6 | Issue 9 | e1001081 Figure S5 Comparison of NY107 RBS to HEF. Overlap of RBS from NY107 (green), Av-H7 (marine) and HEF (magenta). Found at: doi:10.1371/journal.ppat.1001081.s005 (1.12 MB TIF) Table S1 Comparison of r.m.s.d. (A˚ ) for different HA domains. For analyzing differences in the overall structure, r.m.s.d. values were calculated between monomers or domains of different HA’s, after the Ca atoms of the HA2 domains were superposed by sequence and structural alignment onto the equivalent domains of NY107. Found at: doi:10.1371/journal.ppat.1001081.s006 (0.04 MB DOC) Table S2 Comparison of r.m.s.d. (A˚ ) for individual domains. Each domain was superimposed separately to determine how the individual NY107 domains compared to equivalent domains in the other structures. Found at: doi:10.1371/journal.ppat.1001081.s007 (0.04 MB DOC) Table S3 Molecular interactions between NY107 and receptor analogs. The hydrogen bond cutoff is 3.8 A˚ for the listing interactions. Found at: doi:10.1371/journal.ppat.1001081.s008 (0.07 MB DOC) Table S4 Glycan array differences between NY107, the fully restored NY107-ins, and NL219 (virus and rHA). The color coding in the left hand column reflects the same coloring scheme used in Figures 2 and 3. Significant binding of samples to glycans are qualitatively estimated based on relative strength of the signal for the data shown in Figures 2 and 3 Strong (+++), weak (+). Found at: doi:10.1371/journal.ppat.1001081.s009 (0.19 MB DOC) Acknowledgments The authors would like to thank the staff of SER-CAT sector 22 for their help in data collection. We also thank WHO Global Influenza Surveillance Network for providing NY107 and NL219 viruses from which the reverse genetics viruses were generated. Glycan microarray data presented here will be made available on-line through the CFG web site upon publication (www.functionalglycomics.org). The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention or the Agency for Toxic Substances and Disease Registry. Author Contributions Conceived and designed the experiments: HY LMC PJC ROD JS. Performed the experiments: HY LMC PJC JS. Analyzed the data: HY LMC PJC ROD JS. Wrote the paper: HY LMC PJC ROD JS. References 1. Thompson WW, Shay DK, Weintraub E, Brammer L, Cox N, et al. (2003) Mortality associated with influenza and respiratory syncytial virus in the United States. JAMA 289: 179–186. 2. Thompson WW, Shay DK, Weintraub E, Brammer L, Bridges CB, et al. (2004) Influenza-associated hospitalizations in the United States. JAMA 292: 1333–1340. 3. WHO (1980) A revision of the system of nomenclature for influenza viruses: a WHO memorandum. Bull, WHO 58: 585–591. 4. Fouchier RA, Munster V, Wallensten A, Bestebroer TM, Herfst S, et al. (2005) Characterization of a novel influenza A virus hemagglutinin subtype (H16) obtained from black-headed gulls. J Virol 79: 2814–2822. 5. Scholtissek C, Rohde W, Von Hoyningen V, Rott R (1978) On the origin of the human influenza virus subtypes H2N2 and H3N2. Virology 87: 13–20. 6. Kawaoka Y, Bean WJ, Webster RG (1989) Evolution of the hemagglutinin of equine H3 influenza viruses. Virology 169: 283–292. 7. Garten RJ, Davis CT, Russell CA, Shu B, Lindstrom S, et al. (2009) Antigenic and genetic characteristics of swine-origin 2009 A(H1N1) influenza viruses circulating in humans. Science 325: 197–201. 8. Shinya K, Ebina M, Yamada S, Ono M, Kasai N, et al. (2006) Avian flu: influenza virus receptors in the human airway. Nature 440: 435–436. 9. Matrosovich MN, Gambaryan AS, Teneberg S, Piskarev VE, Yamnikova SS, et al. (1997) Avian influenza A viruses differ from human viruses by recognition of sialyloligosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology 233: 224–234. 10. de Jong JC, Claas EC, Osterhaus AD, Webster RG, Lim WL (1997) A pandemic warning? Nature 389: 554. 11. Taubenberger JK, Morens DM, Fauci AS (2007) The next influenza pandemic: can it be predicted? JAMA 297: 2025–2027. 12. Webster RG, Geraci J, Petursson G, Skirnisson K (1981) Conjunctivitis in human beings caused by influenza A virus of seals. N Engl J Med 304: 911. 13. Kurtz J, Manvell RJ, Banks J (1996) Avian influenza virus isolated from a woman with conjunctivitis. Lancet 348: 901–902. 14. Koopmans M, Wilbrink B, Conyn M, Natrop G, van der Nat H, et al. (2004) Transmission of H7N7 avian influenza A virus to human beings during a large outbreak in commercial poultry farms in the Netherlands. Lancet 363: 587– 593. 15. Fouchier RA, Schneeberger PM, Rozendaal FW, Broekman JM, Kemink SA, et al. (2004) Avian influenza A virus (H7N7) associated with human conjunctivitis and a fatal case of acute respiratory distress syndrome. Proc Natl Acad Sci U S A 101: 1356–1361. 16. CDC (2004) Update: influenza activity–United States and worldwide, 2003–04 season, and composition of the 2004–05 influenza vaccine. MMWR Morb Mortal Wkly Rep: Centers for Disease Control. pp 547–552. 17. Hirst M, Astell CR, Griffith M, Coughlin SM, Moksa M, et al. (2004) Novel avian influenza H7N3 strain outbreak, British Columbia. Emerg Infect Dis 10: 2192–2195. 18. Tweed SA, Skowronski DM, David ST, Larder A, Petric M, et al. (2004) Human illness from avian influenza H7N3, British Columbia. Emerg Infect Dis 10: 2196–2199. 19. EditorialTeam (2007) Avian influenza A/H7N2 outbreak in the United Kingdom. Euro Surveill 12: 2. 20. Suarez DL, Garcia M, Latimer J, Senne D, Perdue M (1999) Phylogenetic analysis of H7 avian influenza viruses isolated from the live bird markets of the Northeast United States. J Virol 73: 3567–3573. 21. Belser JA, Lu X, Maines TR, Smith C, Li Y, et al. (2007) Pathogenesis of avian influenza (H7) virus infection in mice and ferrets: enhanced virulence of Eurasian H7N7 viruses isolated from humans. J Virol 81: 11139–11147. 22. Belser JA, Blixt O, Chen LM, Pappas C, Maines TR, et al. (2008) Contemporary North American influenza H7 viruses possess human receptor specificity: Implications for virus transmissibility. Proc Natl Acad Sci U S A 105: 7558–7563. 23. Gambaryan AS, Tuzikov AB, Pazynina GV, Desheva JA, Bovin NV, et al. (2008) 6-sulfo sialyl Lewis X is the common receptor determinant recognized by H5, H6, H7 and H9 influenza viruses of terrestrial poultry. Virol J 5: 85. 24. Stevens J, Blixt O, Glaser L, Taubenberger JK, Palese P, et al. (2006) Glycan microarray analysis of the hemagglutinins from modern and pandemic influenza viruses reveals different receptor specificities. J Mol Biol 355: 1143–1155. 25. Stevens J, Corper AL, Basler CF, Taubenberger JK, Palese P, et al. (2004) Structure of the uncleaved human H1 hemagglutinin from the extinct 1918 influenza virus. Science 303: 1866–1870. 26. Russell RJ, Kerry PS, Stevens DJ, Steinhauer DA, Martin SR, et al. (2008) Structure of influenza hemagglutinin in complex with an inhibitor of membrane fusion. Proc Natl Acad Sci U S A 105: 17736–17741. 27. Russell RJ, Gamblin SJ, Haire LF, Stevens DJ, Xiao B, et al. (2004) H1 and H7 influenza haemagglutinin structures extend a structural classification of haemagglutinin subtypes. Virology 325: 287–296. 28. Matrosovich M, Tuzikov A, Bovin N, Gambaryan A, Klimov A, et al. (2000) Early alterations of the receptor-binding properties of H1, H2, and H3 avian influenza virus hemagglutinins after their introduction into mammals. J Virol 74: 8502–8512. 29. Nobusawa E, Ishihara H, Morishita T, Sato K, Nakajima K (2000) Change in receptor-binding specificity of recent human influenza A viruses (H3N2): a single amino acid change in hemagglutinin altered its recognition of sialyloligosacchar- ides. Virology 278: 587–596. 30. Stevens J, Blixt O, Chen LM, Donis RO, Paulson JC, et al. (2008) Recent avian H5N1 viruses exhibit increased propensity for acquiring human receptor specificity. J Mol Biol 381: 1382–1394. 31. Russell RJ, Stevens DJ, Haire LF, Gamblin SJ, Skehel JJ (2006) Avian and human receptor binding by hemagglutinins of influenza A viruses. Glycoconj J 23: 85–92. 32. Gamblin SJ, Haire LF, Russell RJ, Stevens DJ, Xiao B, et al. (2004) The structure and receptor binding properties of the 1918 influenza hemagglutinin. Science 303: 1838–1842. 33. Suarez DL, Spackman E, Senne DA (2003) Update on molecular epidemiology of H1, H5, and H7 influenza virus infections in poultry in North America. Avian Dis 47: 888–897. 34. Daniels PS, Jeffries S, Yates P, Schild GC, Rogers GN, et al. (1987) The receptor-binding and membrane-fusion properties of influenza virus variants Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 10 September 2010 | Volume 6 | Issue 9 | e1001081 selected using anti-haemagglutinin monoclonal antibodies. EMBO J 6: 1459–1465. 35. Rosenthal PB, Zhang X, Formanowski F, Fitz W, Wong CH, et al. (1998) Structure of the haemagglutinin-esterase-fusion glycoprotein of influenza C virus. Nature 396: 92–96. 36. Weinstein J, de Souza-e-Silva U, Paulson JC (1982) Purification of a Gal beta 1 to 4GlcNAc alpha 2 to 6 sialyltransferase and a Gal beta 1 to 3(4)GlcNAc alpha 2 to 3 sialyltransferase to homogeneity from rat liver. J Biol Chem 257: 13835–13844. 37. Gambaryan A, Webster R, Matrosovich M (2002) Differences between influenza virus receptors on target cells of duck and chicken. Arch Virol 147: 1197–1208. 38. Wan H, Perez DR (2006) Quail carry sialic acid receptors compatible with binding of avian and human influenza viruses. Virology 346: 278–286. 39. Guo CT, Takahashi N, Yagi H, Kato K, Takahashi T, et al. (2007) The quail and chicken intestine have sialyl-galactose sugar chains responsible for the binding of influenza A viruses to human type receptors. Glycobiology 17: 713–724. 40. Maines TR, Chen LM, Matsuoka Y, Chen H, Rowe T, et al. (2006) Lack of transmission of H5N1 avian-human reassortant influenza viruses in a ferret model. Proc Natl Acad Sci U S A 103: 12121–12126. 41. Wan H, Sorrell EM, Song H, Hossain MJ, Ramirez-Nieto G, et al. (2008) Replication and transmission of H9N2 influenza viruses in ferrets: evaluation of pandemic potential. PLoS One 3: e2923. 42. Weis WI, Brunger AT, Skehel JJ, Wiley DC (1990) Refinement of the influenza virus hemagglutinin by simulated annealing. J Mol Biol 212: 737–761. 43. Frank S, Kammerer RA, Mechling D, Schulthess T, Landwehr R, et al. (2001) Stabilization of short collagen-like triple helices by protein engineering. J Mol Biol 308: 1081–1089. 44. Stevens J, Blixt O, Tumpey TM, Taubenberger JK, Paulson JC, et al. (2006) Structure and receptor specificity of the hemagglutinin from an H5N1 influenza virus. Science 312: 404–410. 45. Chayen NE, Shaw-Steward PD, Blow DM (1992) Microbatch crystallization under oil – a new technique allowing many small volume crystallization experiments. J Cryst Growth 122: 176–180. 46. Otwinowski A, Minor W (1997) Processing of X-ray diffraction data collected in oscillation mode. Meothds in Enzymology 276: 307–326. 47. McCoy AJ, Grosse-Kunstleve RW, Storoni LC, Read RJ (2005) Likelihood- enhanced fast translation functions. Acta Crystallogr D Biol Crystallogr 61: 458–464. 48. Emsley P, Cowtan K (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132. 49. CCP4 (1994) The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 50: 760–763. 50. Winn MD, Isupov MN, Murshudov GN (2001) Use of TLS parameters to model anisotropic displacements in macromolecular refinement. Acta Crystallogr D Biol Crystallogr 57: 122–133. 51. Davis IW, Leaver-Fay A, Chen VB, Block JN, Kapral GJ, et al. (2007) MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res 35: W375–383. 52. Brunger AT (2007) Version 1.2 of the Crystallography and NMR system. Nat Protoc 2: 2728–2733. 53. Hoffmann E, Webster RG (2000) Unidirectional RNA polymerase I-polymerase II transcription system for the generation of influenza A virus from eight plasmids. J Gen Virol 81: 2843–2847. 54. Higuchi R, Krummel B, Saiki RK (1988) A general method of in vitro preparation and specific mutagenesis of DNA fragments: study of protein and DNA interactions. Nucleic Acids Res 16: 7351–7367. 55. Blixt O, Head S, Mondala T, Scanlan C, Huflejt ME, et al. (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc Natl Acad Sci U S A 101: 17033–17038. 56. DeLano WL (2002) The PyMol Molecular Graphics Systems. wwwpymolorg. Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 11 September 2010 | Volume 6 | Issue 9 | e1001081
3M5H
Crystal structure of a H7 influenza virus hemagglutinin complexed with 3SLN
Structures of Receptor Complexes of a North American H7N2 Influenza Hemagglutinin with a Loop Deletion in the Receptor Binding Site Hua Yang, Li-Mei Chen, Paul J. Carney, Ruben O. Donis, James Stevens* Influenza Division, Centers for Disease Control and Prevention, Atlanta, Georgia, United States of America Abstract Human infections with subtype H7 avian influenza viruses have been reported as early as 1979. In 1996, a genetically stable 24-nucleotide deletion emerged in North American H7 influenza virus hemagglutinins, resulting in an eight amino acid deletion in the receptor-binding site. The continuous circulation of these viruses in live bird markets, as well as its documented ability to infect humans, raises the question of how these viruses achieve structural stability and functionality. Here we report a detailed molecular analysis of the receptor binding site of the North American lineage subtype H7N2 virus A/New York/107/2003 (NY107), including complexes with an avian receptor analog (39-sialyl-N-acetyllactosamine, 39SLN) and two human receptor analogs (69-sialyl-N-acetyllactosamine, 69SLN; sialyllacto-N-tetraose b, LSTb). Structural results suggest a novel mechanism by which residues Arg220 and Arg229 (H3 numbering) are used to compensate for the deletion of the 220-loop and form interactions with the receptor analogs. Glycan microarray results reveal that NY107 maintains an avian-type (a2-3) receptor binding profile, with only moderate binding to human-type (a2-6) receptor. Thus despite its dramatically altered receptor binding site, this HA maintains functionality and confirms a need for continued influenza virus surveillance of avian and other animal reservoirs to define their zoonotic potential. Citation: Yang H, Chen L-M, Carney PJ, Donis RO, Stevens J (2010) Structures of Receptor Complexes of a North American H7N2 Influenza Hemagglutinin with a Loop Deletion in the Receptor Binding Site. PLoS Pathog 6(9): e1001081. doi:10.1371/journal.ppat.1001081 Editor: Fe´lix A. Rey, Institut Pasteur, France Received March 16, 2010; Accepted July 28, 2010; Published September 2, 2010 This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. Funding: This work was funded by the Centers for Disease Control and Prevention. Use of the Advanced Photon Source at Argonne National Laboratory was supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC02-06CH11357. Glycan microarrays as well as glycans for direct binding experiments were produced for the Centers for Disease Control by the CFG funded by National Institute of General Medical Sciences Grant GM62116. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: fwb4@cdc.gov Introduction Influenza is an acute respiratory virus that infects up to 20% of the population in the United States, resulting in ,36,000 deaths annually [1,2]. The two membrane glycoproteins on the surface of influenza A virus, hemagglutinin (HA), which functions as the receptor binding and membrane fusion glycoprotein in cell entry, and neuraminidase (NA), which functions as the receptor destroying enzyme in virus release, form the basis for defining subtypes [3]. To date, 16 HA (H1–H16) and 9 NA (N1–N9) have been identified in avian species [4], while in the last century, only three subtypes, H1N1 in 1918 and 2009, H2N2 in 1957, and H3N2 in 1968 [5,6,7], have successfully adapted to humans. Hemagglutinin binds to sialic acid (SA) glycans present on host cell surfaces. The receptors on epithelial cells of the human upper respiratory tract are mainly a2-6-linked SA moieties [8]. Since avian influenza viruses pre- dominately bind a2-3-linked SA, and human influenza viruses preferentially bind to a2-6-linked SA, human infection by avian influenza viruses is rare [9]. However, since 1997 a growing number of human cases of avian influenza infection have been reported [10], including H5N1, H7N2, H7N3, H7N7, and H9N2 strains [11]. Although the current situation with the pandemic H1N1 influenza virus dominates public health efforts, the prospect of a novel pandemic emerging from these isolated cases continues to be a major public health threat around the world. Early cases of human infection by H7 influenza viruses are reported as far back as 1979 [12,13]. Since 2002, multiple outbreaks and human infections of H7 subtype viruses; within both Eurasian and North American lineages have been reported. In the Netherlands in 2003, a highly pathogenic avian influenza (HPAI) H7N7 outbreak resulted in more than 80 cases of human infections, including one fatality [14,15]. In New York in 2003, a single case of human respiratory infection of H7N2 was reported [16] and in British Columbia in 2004, an H7N3 virus caused two cases of conjunctivitis [17,18]. More recently in 2007, the United Kingdom reported several cases of low pathogenic avian influenza (LPAI) H7N2 virus infections that caused influenza-like illness and conjunctivitis [19]. Since 1996, H7 viruses of the North American lineage have been circulating in regional live bird markets [20], containing a 24-nucleotide deletion resulting in an eight amino acid deletion in the receptor-binding site (RBS) of HA (Figure S1). The recent human infections with H7 in North America have raised public health concerns as to how these viruses adapt to such a dramatic structural change while remaining one of the predominant circulating viral strains. A recent study of H7 viruses isolated from previous outbreaks revealed efficient replication in both mouse and ferret animal models [21]. In particular, ferret studies with A/New York/107/2003 (NY107), an H7N2 virus isolated from a man in New York, not only showed efficient replication in the upper respiratory tract of the ferret but also the capacity for PLoS Pathogens | www.plospathogens.org 1 September 2010 | Volume 6 | Issue 9 | e1001081 intra-species transmission by direct contact [21,22]. Interestingly, both an increased preference for a2-6 and decreased preference for a2-3-linked sialosides of this virus compared to the other avian influenza viruses was shown by previous glycan microarray analysis but less so by a competitive solid-phase binding assay [22,23]. Here we report a detailed molecular analysis of the RBS of the HA from North American lineage H7N2 virus, NY107, including glycan microarray analyses and structural analyses of the HA in complex with an avian receptor analog (39-Sialyl-N-acetyllactosa- mine, 39SLN) and two human receptor analogs (69-Sialyl-N- acetyllactosamine, 69SLN; Sialyllacto-N-tetraose b, LSTb). These results provide important insight into the interaction of H7 HAs with both avian and human hosts. Results Overall structure By using x-ray crystallography, the structure of H7 HA from the NY107 virus was determined to 2.6 A˚ resolution (Table 1). In addition, we also report three H7 HA receptor complex structures, with avian receptor analog (39SLN) to 2.7 A˚ resolution and with human receptor analogs (69SLN and LSTb) to 3.0 A˚ and 2.6 A˚ resolution, respectively (Table 1). The overall structure of NY107 is similar to other reported HA structures with a globular head containing the RBS and vestigial esterase domain, and a membrane proximal domain with its distinctive, central helical stalk and HA1/HA2 cleavage site (Figure 1A). Although five asparagine-linked glycosylation sites are predicted in the NY107 HA monomer, interpretable electron density was observed at only two sites, Asn38 in HA1 and Asn82 in HA2 (all residue numbers are based on H3 numbering). At these sites, only one or two N- acetyl glucosamines could be interpreted. During viral replication, HA is synthesized as a single chain precursor (HA0) and cleaved by specific host proteases into the infectious HA1/HA2 form. In baculovirus expression systems, highly pathogenic HAs, with a polybasic cleavage site, are expressed as an HA1/HA2 form [24], whereas HAs with monobasic cleavage sites (single Arg) from low pathogenic viruses are expressed as the HA0 form [25]. NY107 is regarded as a low pathogenic virus, and as expected, was produced in the HA0 form (Figure S2). However, subsequent digestion with thrombin protease to remove the His-tag resulted in cleavage to a profile on SDS-PAGE comparable to that of an HA1/HA2 form (Figure S2). A comparison of the NY107 cleavage site with the consensus cleavage pattern in the MEROPS database (http://merops. sanger.ac.uk) suggests it to be a possible thrombin cleavage site. Based on their molecular phylogenies, HAs are divided into two groups and five clades: group 1 includes H8, H9, and H12; H1, H2, H5, and H6; H11, H13 and H16; group 2 includes H3, H4, and H14; H7, H10 and H15 [26]. Among all available HA structures, we selected ten representative HAs from both avian and human subtypes for structural analysis. As expected, NY107 HA is structurally very similar to the Avian-H7 in all comparisons and closely related to H3, the other group 2 members used in the analyses (Tables S1 and S2). The receptor binding site The RBS is at the membrane distal end of each HA monomer and its specificity for sialic acid and the nature of its linkage to a vicinal galactose residue is a major determinant of host range- restriction. The consensus RBS for all current HAs is composed of three major structural elements: a 190-helix (residues 188–194), a 220-loop (residues 221–228), and a 130-loop (residues 134–138). In addition, highly conserved residues (Tyr98, Trp153, His183, and Tyr195) form the base of the pocket. Although the NY107 RBS is similar to other subtypes (H1, H2, H3, H5, and H9), a previously observed specific feature of H7 HAs, is also observed in the NY107 150-loop region: two residues inserted at position 158 result in this loop protruding more than 6A˚ towards the binding site compared to other subtype HAs (Figure 1B and Table S2) [27]. More interestingly, the eight amino acid deletion, only found in the North American lineage H7s, from position 221 to 228 (Figure S1), resulted in a complete loss of the 220-loop (Figure 1B). Sequence alignment shows that Arg220 and Arg229 are conserved in all influenza A HA subtypes (Figure S1), but structural alignment of NY107 HA shows Arg220 occupying the Gly228 position, and the much shorter loop turns at residue Pro217 (Figure 1C). The Ca distance between NY107 Arg220 and its homolog in the Av-H7 structure (PDB: 1TI8) [27] is 5.8A˚ , and they point in opposite directions (Figure 1C). The side chain direction of Av-H7 Arg220 is almost parallel with the beta sheet after Arg229, whereas the NY107 Arg220 points downward to the binding pocket. The Ca position of Arg229 in both H7 structures remains the same, except the side chain in the NY107 swings away by about 5.9A˚ (Figure 1C) and could help to stabilize this region by forming a hydrogen bond to the mainchain carbonyl of Gln210 in the neighboring monomer. In the absence of the 220-loop in NY107 HA, upon glycan binding the long side chain of Arg220 compensates for its loss and is displaced 4A˚ upward to form hydrogen bonds with receptor analogs inside the binding pocket (Figure 1D). Effect of loop truncation on the receptor binding specificity of NY107 Previously, mutations in the HA receptor binding domains of H1N1 (Glu190Asp/Gly225Asp) and H2N2/H3N2 (Gln226Leu and Gly228Ser) subtypes were responsible for adaptation of these viruses to pandemic strains [24,28,29,30]. Due to missing residues 221–228 in the NY107 HA RBS, neither mechanism for adaptation is possible. Thus, in order to look more closely at the role of the missing loop and its effect on receptor specificity, we first subjected the recombinant HA (recHA) to glycan microarray analyses and compared it to a reverse genetics-derived NY107 virus, and a co-circulating Eurasian virus and recHA, A/ Author Summary Influenza virus adaptation to different hosts usually results in a switch in receptor specificity of the viral surface coat protein, hemagglutinin. Indeed, the hemagglutinin sub- types from the last two human influenza pandemics of the 20th Century (H2 in 1957 and H3 1968) both adapted successfully to human-type receptor specificity through only two amino acid mutations in the receptor binding pocket (Glutamine226RLeucine and Glycine228RSerine). The recent human infections reported with other avian subtypes such as H5, H7 and H9 have raised public health concerns and focused efforts on identifying potential subtypes from which a future pandemic strain may emerge. Since 1996, H7 viruses of the North American lineage have been circulating in regional live bird markets, containing an eight amino acid deletion in the receptor- binding site of HA. Here we report a detailed structural analysis of the receptor binding site of a hemagglutinin from the North American lineage of H7N2 viruses, in complex with avian and human receptor analogs, to understand how these viruses have adapted to such a dramatic structural change in the binding site while remaining one of the predominant circulating viral strains. Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 2 September 2010 | Volume 6 | Issue 9 | e1001081 Netherlands/219/2003 (NL219), that has the consensus avian sequence in the 220-loop and it also infected a human [15]. Glycan microarray analysis of recombinant NY107 (Figure 2A and Table 2) revealed a highly restricted binding profile with strong binding to only a2-3 sulfated (#4–8), a2-3 branched (#9– 11) and mixed a2-3/a2-6 branched sialosides (#60–64) as well as to the long linear sialyl di- and tri-lactosamines (#22, 24). Weak binding was also observed (above background) to other a2-3 glycans on the array. The recombinant NY107 also revealed a strict glycan binding preference to only one a2-6 glycan, the internal structure, Galb1-3(Neu5Aca2-6)GlcNAcb1-3Galb1-4Glc (#58; LSTb) (Figure 2A), a glycan highlighted in a previous study [22]. The virus with higher valency and avidity revealed stronger binding to all a2-3 groups, in addition to the branched di-sialyl a2- 6 biantennary structures (#46–48) as well the LSTb (#58) (Figure 2B and Table 2). In contrast, the NL219 recHA (Figure 2C and Table 2) bound well to only the avian a2-3 containing sialyl- glycans (sulfated, branched, linear and fucosylated). Its corre- sponding virus also reflected this specificity although it also revealed strong binding to a2-3 N-glycolylneuraminic acid (Neu5Gc) containing glycans (#66–70) (Figure 2D and Table 2). To further assess the effect of the missing 220 loop on HA structural stability and receptor specificity it was essential to evaluate these functions on the ancestral HA containing the full length 220-loop. To this end, we engineered an HA with an avian H7 consensus (PQVNGQSG) 220-loop re-introduced (NY107- 220ins) into the NY107 HA and recovered this virus by reverse genetics. Compared to the NY107 virus (Figure 2A) glycan microarray analyses of the resulting NY107-220ins virus (Figure 3A and Table 2) revealed a decrease in binding to branched (#9–11) and linear (#12–27) a2-3 sialosides and a loss of binding to the branched di-sialyl a2-6 biantennary structures (#46–48), LSTb (#58) as well as the mixed a2-3/a2-6 branched sialosides (#60– 64). In addition, sequence analysis of the NY107-220ins HA revealed the presence of quasispecies in the second position of the inserted loop, P(Q/K)VNGQSG, suggesting that re-introduction of the loop alone is not tolerated and does not create an avian-type binding profile. Thus other amino acid substitutions in the HA might have co-evolved with the deletion of the 220 loop to help stabilize the RBS/HA to maintain functionality. When viruses containing this 220-loop deletion emerged in North America in the mid 90’s, four additional amino acid substitutions, Gly114Arg, Asp119Gly, Gly186Glu and Gly205Arg, in the HA1 as well as an Asp19Asn in the HA2 chain were also introduced to most of the circulating isolates. Of these, Gly186Glu and Gly205Arg in the HA1 are close to the RBS, at the monomer interface, and could potentially modulate its structure and/or function. NY107 viruses with a restored consensus 220-loop and a single Glu186Gly (NY107- ins-186) or Arg205Gly (NY107-ins-205) substitution as well as the Glu186Gly/Arg205Gly double substitution (NY107-ins-186/205) were derived by reverse genetics and evaluated. Glycan microarray analysis for the three resulting viruses revealed similar glycan binding profiles with increased binding to a2-3 sialosides, including mixed a2-3/a2-6 branched sialosides (#60–64), a2-3 Neu5Gc (#66–70), but limited binding to the a2,6 sialosides (Figures 3B, 3C, 3D), resulting in a binding profile virtually identical to that of the NL219 virus and other avian influenza viruses (Figure 2D) [30]. Sequence analysis of the three reverse genetics derived viruses revealed no mutations/quasispecies in the HAs of either the NY107- ins-186 or the NY107-ins-186/205 virus stocks, indicative of replication fitness. For the NY107-ins-205 virus however, a Glu186Gly substitution emerged in the HA after only two passages in eggs following recovery from DNA transfection, indicating the importance of the co-variant position 186 with respect to HA functionality/glycan specificity. Altogether, the data indicates that the H7 subtype avian influenza viruses that were circulating in Table 1. Data collection and refinement statistics. NY107 NY107+39SLN NY107+69SLN NY107+LSTb Data collection Space group P212121 P212121 P212121 P212121 Cell dimensions (A˚) 66.96, 115.92, 251.61 67.80, 116.70, 249.84 66.60, 116.58, 250.68 67.08, 116.52, 251.95 Resolution (A˚) 50-2.6 (2.69-2.60)a 30-2.7 (2.80-2.70) 50-3.0 (3.11-3.0) 50-2.6 (2.69-2.60) Rsym or Rmerge 10.6 (41.3) 14.6 (48.6) 14.3 (35.4) 12.2 (31.5) I/s 39.6 (2.0) 24.3 (1.7) 34.2 (8.2) 40.5 (9.9) Completeness (%) 99.2 (99.0) 99.3 (94.6) 92.3 (75.6) 91.3 (86.2) Redundancy 7.2 (6.2) 5.8 (5.5) 4.9 (4.4) 10.9 (11.2) Refinement Resolution (A˚) 50-2.6 (2.67-2.60) 30-2.7 (2.77-2.70) 50-3.0 (3.08-3.00) 50-2.6 (2.67-2.60) No. of reflections (total) 57285 51770 33421 53603 No. of reflections (test) 3053 2769 1779 2842 Rwork/Rfree 21.7/25.6 21.4/26.4 20.5/26.0 20.4/24.7 No. of atoms 11795 11878 11648 12108 r.m.s.d.- bond length (A˚) 0.006 0.006 0.008 0.006 r.m.s.d.- bond angle (u) 0.905 0.974 1.085 0.859 MolProbityb scores Favored (%) 96.9 96.5 94.3 97.1 Outliers (%) (No. of residues) 0.1 (1/1434) 0.0 (0/1429) 0.1 (2/1433) 0.1 (2/1435) aNumbers in parentheses refer to the highest resolution shell. bReference [51]. doi:10.1371/journal.ppat.1001081.t001 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 3 September 2010 | Volume 6 | Issue 9 | e1001081 aquatic birds and poultry in North America before 1996 exhibited a classic avian a2-3 sialoside binding preference. In order for the 220- loop deletion to be tolerated, concurrent Gly186Glu and Gly205Arg substitutions in the vicinity of RBS of HA emerged to achieve a restricted a2-3 binding profile and only a moderate/limited increase in binding to branched di-sialyl a2-6 biantennary structures (#46– 48) as well the a2,6 internal sialoside, LSTb (#58). NY107 avian receptor complex To understand from a structural perspective how NY107 interacts with host receptors, we solved the structure of NY107 in complex with an avian and two human receptor analogs. For the avian receptor analog, 39SLN, the electron density maps revealed well-ordered features for the Sia-1, Gal-2, and GlcNAc-3 in the NY107 HA complex structure (Figure 4A). Structural comparison of NY107 HA binding to other, H1, H2, H3, H5, and H9 subtypes (Figure S2A) revealed that 39SLN binding to NY107 resembled binding of the other published HAs. Indeed, the terminal Sia-1 moiety is positioned almost identically in all structures, and forms the majority of hydrogen bonds and contacts with residues in the RBS (Figure 4A and Table S3). Figure 1. NY107 HA monomer and comparison of its RBS to other HA structures. (A) One monomer is shown with the HA1 chain colored in green and the HA2 chain in cyan. The location of the receptor binding site and the HA1/HA2 cleavage site are circled. (B) The superposition of receptor binding domains of NY107 (green), Av-H7 (marine), 1918-Hu-H1 (magenta), Hu-H5 (yellow), Hu-H3 (orange), and Sw-H9 (grey). The proximity of Arg220 and Gln226 are highlighted. Three structural elements comprising this binding site are labeled. The two major differences are the extended 150-loop and the deletion of 220-loop of NY107. (C) Overlap of NY107 (green) and Av-H7 (marine) (PDB: 1TI8) illustrates the compensatory effect of R220 bringing it close to the position occupied by G228 in the avian HA. (D) Overlap of the NY107 (green), NY107- 39SLN (orange), NY107-69SLN (red), and NY107-LSTb (magenta) structures. All the figures were generated and rendered with the use of MacPyMOL [56]. doi:10.1371/journal.ppat.1001081.g001 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 4 September 2010 | Volume 6 | Issue 9 | e1001081 Published avian HA structures with an intact 220-loop form very close interactions with Gal-2 of 39SLN via residue Gln226 which is important in receptor specificity and host adaptation. For example, in the avian H7/39SLN HA structure it interacts with Gal-2 O4 [31]. In the NY107 HA structure, although Gln226 is absent and no other residue occupies the same space as Gln226 (Figure 1B), Arg220 does forms a hydrogen bond between Arg220 NH2 and Gal-2 O4 (Figure 4A). Interestingly, although there was interpretable density for the GlcNAc-3 (Figure 4A and Figure S4B), no hydrogen bonding was apparent between the HA and the GlcNAc-3, which is consistent with other reported structures [32]. Thus, for binding to avian receptors, the trans conformation of a2-3 linkages is essential and perhaps only the first two saccharides are required. Indeed, due to the absence of 220-loop in the NY107 HA structure, the ‘‘aperture’’ of the RBS formed by 220-loop and 130-loop in regular HAs is increased by ,10 A˚ , so that the branched, internal, and perhaps more complicated glycans might be accommodated more efficiently. NY107 human receptor complexes In the NY107/69SLN complex, only Sia-1 and Gal-2 are ordered (Figure 4B). The Sia-1 remains in the same position as previously analyzed glycan/HA complexes from H1, H2, H3, H5, and H9 (Figure S3B), whereas the Av-H7 complex structure with Sialyllacto-N-tetraose c (LSTc) did not reveal any density for the Sia-1 in the receptor binding site [31]. The Gal-2 position varies significantly among different subtypes. Compared to the human- adapted H1 HA [32], Gal-2 in the NY107 HA is 3A˚ higher, and thus is further from the protein (Figure S3B). In NY107, the Gal-2 only forms an intramolecular, saccharide-saccharide interaction with Sia-1. The poor electron density map and fewer interactions with protein residues suggest that the cis conformation of a2-6 Figure 2. Receptor specificity of NY107 recHA and virus. Glycan microarray analysis of recombinant NY107 HA (A) and NY107 virus (B) compared to the recHA (C) and virus (D) from a Eurasian lineage A/Netherlands/219/2003 H7 influenza virus that was circulating in the same year and also infected a human. Colored bars highlight glycans that contain a2-3 SA (blue) and a2-6 SA (red), a2-6/a2-3 mixed SA (purple), N-glycolyl SA (green), a2-8 SA (brown), b2-6 and 9-O-acetyl SA, and non-SA (grey). Error bars reflect the standard error in the signal for six independent replicates on the array. Structures of each of the numbered glycans are found in Table S4. doi:10.1371/journal.ppat.1001081.g002 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 5 September 2010 | Volume 6 | Issue 9 | e1001081 linkages in 69SLN trisaccharides show a reduced binding affinity with NY107. Glycan array results with NY107 revealed a strong binding signal for the internal a2-6 sialoside, LSTb. To further investigate this interaction, we solved the structure of the NY107/LSTb complex. The final model contained Sia-1, NAG-2, Gal-3, and Gal-5 in the RBS. Although glycan microarray data indicated NY107 to have a specific affinity for LSTb, few interactions were apparent from the crystal structure. Sia-1 still forms multiple hydrogen bonds with residues in the RBS (Table S3 & Figure 4C). The branched Gal-5 interacts with Ser137, to help stabilize the LSTb binding. However, Arg220 and Lys193, the two residues showing close binding with 39SLN, did not form any hydrogen bonds with LSTb. In the structure, Gal-5 also interacts with a crystal packing symmetry mate and thus the flexibility of whole LSTb may be restricted. In solution, with more freedom, the LSTb should be able to tilt closer to the RBS, and thus Glc-4 may have more interactions with the 190-helix than seen in the crystal structure. Discussion Human infections by avian influenza viruses, including H7 subtypes, continue to pose a major public health threat. Although the species barrier prevents avian influenza viruses from widespread infection of the human population, the molecular determinants of efficient interspecies transmission and pathoge- nicity are still poorly understood. The viral coat protein HA however, is perhaps a critical molecule since previous pandemic viruses modified their receptor specificity and overcame the interspecies barrier to spread in the human population. Although HA structures alone and in complex with receptor analogs provide considerable insight into receptor binding, it is clear that HAs from different species and subtypes have significant structural variation. Indeed, low-pathogenic H7N2 avian influenza viruses with an 8 amino acid deletion within its RBS started to circulate in live-bird markets in the northeast United States in 1996. Despite what one would consider a debilitating mutation, these viruses have been reported as the predominant isolate [33]. Whether such a deletion contributed to their evolutionary success and how are an important questions, especially in light of NY107’s ability to produce respiratory illness in humans [16], as well as its reported increased affinity for human-type receptors and ability for contact transmission in ferrets [21]. To try to help answer these questions, we have analyzed the molecular structures of NY107 and its complexes with receptor analogs to explain receptor specificity at the molecular level. The crystal structures of NY107 and its complexes with both avian and human receptor analogs describe a mechanism as to how an influenza virus might adapt by dramatically altering its RBS, and still be functional. Arg220 of the HA1 chain of NY107 compensates for the loss of the 220-loop, by forming hydrogen bonds with Gal-2 from the avian analog (binding was not observed in either of the structures complexes with the human analogs). However, in the LSTb complex, branched Gal-5 forms extra interactions with the 130-loop, thus improving the binding preference for this particular glycan. Consistent with the structural evidence, glycan microarray analyses of NY107 revealed a strong binding preference for the branched a2-6 sialoside, LSTb. Except for the absence of the 220-loop, other key residues within the RBS are conserved in NY107 and thus, direct interactions with sialic acid are maintained. The 220-loop is recognized as one of the three crucial structural elements in the RBS. Aside from the North American lineage H7N2 viruses, which have been circulating with a deletion (221– 228) in this loop, there has been one other report describing a seven amino acid deletion (224–230) in a laboratory generated H3N2 escape mutant which was reported to have a slightly Table 2. Comparison of the sialoside receptor specificity of the HAs from H7 influenza viruses. Glycan Group Graph Numbera NY107 RecHA NY107 Virus NY107-ins Virus NY107-ins E186G Virus NY107-ins R205G Virus NY107-ins E186G/ R205G Virus NL219 RecHA NL219 Virus a2-3 Sulfated 4–8 +++b +++ +++ +++ +++ +++ +++ +++ Branched 9–11 +++ +++ + +++ +++ +++ +++ +++ Linear 12–27 + +++ + +++ +++ +++ +++ +++ Fucosylated 28–34 2 +++ +++ +++ +++ +++ +++ +++ a2-6 Sulfated 41 2 2 2 2 2 2 2 2 Branched mono-sialyl 42–45, 49 2 2 2 2 2 2 2 2 Branched di-sialyl 46–48 2 +++ 2 2 2 2 2 2 Linear 50–56 2 2 2 2 2 2 2 2 Internal 58–59 +++ +++ 2 2 2 2 2 2 Other Sialic acid 1–2 2 +++ + 2 2 2 2 2 a2-3/a2-6 Branched 60–64 2 2 2 +++ +++ +++ +++ +++ Neu5Gcc 65–72 2 2 2 +++ +++ +++ 2 +++ aMembers of each group are identified according to the graph number used in the microarray data in Figures 2 and 3 and correspond to numbers in the complete glycan list (Table S4). bBinding of samples to glycan subclasses are qualitatively estimated based on relative strength of the signal for the data shown in Figures 2 and 3: strong (+++), weak (+), absent (2). cN-glycolylneuraminic acid. doi:10.1371/journal.ppat.1001081.t002 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 6 September 2010 | Volume 6 | Issue 9 | e1001081 increased affinity for a2-3-linked glycans by hemagglutination assay [34]. Meanwhile, the equivalent region in the hemaggluti- nin-esterase-fusion (HEF) protein of influenza C virus reveals a rearrangement resulting in a truncated 260-loop in its RBS (Figure S5) [35]. However, without structural data with appropriate receptor analogs, it is not possible to compare the role of these loop variants in receptor binding to the H7 HA structure described here. When compared to NL219, another co-circulating H7 avian virus HA (Figure 2C and D), overall binding to a2-3-linked glycans was markedly reduced, while increased binding to a2-6- linked receptors was only marginal. However, these results focus attention on only 2 sub-classes of human-type receptors that may be important for infection (and transmission in ferrets). The NY107 virus interaction with biantennary glycans (Figure 2B), although weak (not seen in Figure 2A with recHA), is a possible route for virus entry as biantennary structures are common on tissues, i.e. glycan profiling data from human lung tissue on the Consortium for Functional Glycomics (CFG) web site. In addition, the internal sialoside, LSTb, was observed in both virus and recHA microarray data, suggesting this type of glycan has good affinity for this HA. The significance of this is unknown since LSTb has only been described in human milk [36]. Interestingly, NY107 and NL219 virus receptor binding and specificity has been addressed previously using glycan microarray analysis that reported a significantly increased preference for a2-6 and decreased preference for a2-3-linked sialosides [22]. In addition, the same viruses were also included in a recent study from Gambaryan et al. using a competitive solid-phase binding assay [23]. Our findings confirm and extend the receptor binding specificity reported by these authors in that they reported both viruses binding to sulfated sialylglycans with a lactosamine (Galb1- 4GlcNAc core and reported only a moderate binding affinity for a2-6-sialyllactosamine, the human-type receptor analog used in their assay. The 220-loop is an integral feature of the receptor binding site, and thus one would predict that such a deletion might have compromised this strain to be deleted from the population of circulating viruses. However, this was not the case [33] and its existence appears to be in part due to the additional mutations at Figure 3. Effect of 220-loop deletion and additional RBS mutations on NY107 receptor specificity. NY107 was engineered to restore the 220-loop to a consensus full-length HA from 1996 (A) and additional co-variant amino acid substitutions, Glu186Gly (B), Arg205Gly (C) and the double mutant Glu186Gly/Arg205Gly (D) to restore, on the NY107 framework, an HA RBS found in viruses prior to the introduction of the deletion in North American viruses. Colored bars group glycans as described in Figure 3. Error bars reflect the standard error in the signal for six independent replicates on the array. Structures of each of the numbered glycans are found in Table S4. doi:10.1371/journal.ppat.1001081.g003 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 7 September 2010 | Volume 6 | Issue 9 | e1001081 positions 186 and 205. Restoration of the loop with either or both residues mutated back to the pre-1994 consensus sequence resulted in a classic avian influenza virus binding profile. The emergence of the Glu186Gly mutation in the HA of the NY107- ins-205 mutant after only two passages of the rescued virus in eggs, also indicates the importance of these positions for HA functionality/glycan specificity. Analysis of the structural data reveals that positions 186 and 205 are on opposite sides of a monomer but are both close to the 220-loop deletion region in the trimeric form. The Glu at position 186 is close to Arg220 and may interact with Arg220 when binding avian receptors. Position 205 in the neighboring monomer may be important in trimer stability and maintaining RBS functionality. If one models the pre- 1996 220-loop restored into the NY107 structure, Arg205, Glu186 and the loop all clash, thus explaining the Glu186Gly mutation that emerged in the NY107-ins-205 virus HA after limited egg passage. The NY107 RBS with its more restricted a2-3 glycan binding preference and weak/moderate increase in a2-6 binding may have given the virus a selective advantage to be maintained in poultry at live bird markets and supplying farms. Certain terrestrial birds, such as quails and chickens, have recently been shown to present both human and avian types of receptors in the trachea and intestine [37,38,39]. Although it is not known what specific glycans are presented in these animals, it is conceivable that a virus with mixed specificity might have a distinct advantage over avian viruses that have specific avian receptor requirements, particularly in bird markets where multiple species coalesce. Previous results with H7N2, H9N2 and H5N1 viruses all highlight the fact that an increase in a2-6-binding preference is not sufficient for efficient transmission of avian influenza viruses to humans [22,40,41]. Although it remains to be seen whether prolonged circulation of viruses in terrestrial birds, such as domestic chickens, can provide a possible route for viruses to adapt for efficient human infection [11], continued surveillance of influenza viruses from avian and other animal reservoirs is urgently needed to define their zoonotic potential. Materials and Methods Cloning Based on H3 numbering [42], cDNA corresponding to residues 11–329 (HA1) and 1–176 (HA2) of the ectodomain of the hemagglutinin (HA) from A/New York/107/2003 (H7N2; Genbank:ACC55270) and A/Netherlands/219/2003 (H7N7; Genebank: AAR02640) was cloned into the baculovirus transfer vector, pAcGP67-A (BD Biosciences), incorporating a C-terminal thrombin cleavage site, a ‘‘foldon’’ sequence [43] and a His-tag at the extreme C-terminus of the construct to enable protein purification [25,44]. Transfection and virus amplification were carried out according to the baculovirus expression system manual (BD Biosciences Pharmingen). Figure 4. Glycan interactions within the NY107 RBS. The top panel shows the interactions of NY107 with (A) 39SLN, (B) 69SLN and (C) LSTb. NY107 is shown in orange/red/magenta cartoon respectively. The interacting HA residues are shown as green sticks. The bottom panel shows the electron density map of the ligands. The NY107 is shown in the same colors as above, and the ligands are shown as green sticks, the 2fo-fc electron density maps (contoured at 1s) are shown in grey. Simulated annealing omit maps are shown in supplementary Figure S4. doi:10.1371/journal.ppat.1001081.g004 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 8 September 2010 | Volume 6 | Issue 9 | e1001081 Protein expression and purification Soluble NY107 was recovered from the cell supernatant by metal affinity chromatography using Ni-NTA resin (Qiagen Inc.). Frac- tions containing NY107 were pooled and dialyzed against 10 mM Tris-HCl, 50 mM NaCl, pH 8.0, then subjected to ion-exchange chromatography (IEX) using a Mono-Q HR 10/10 column (GE Healthcare). IEX purified NY107 was subjected to thrombin digest (3 units/mg protein; overnight at 4uC) and purified by gel filtra- tion chromatography using a Superdex-200 16/60 column (GE Healthcare) and 50 mM Tris-HCl, 100 mM NaCl, pH 8.0 as running buffer. Protein eluting as a trimer was buffer exchanged into 10 mM Tris-HCl, 50 mM NaCl, pH 8.0 and concentrated to 14.5 mg/ml for crystallization trials. At this stage, the protein sample still contained the additional plasmid-encoded residues at both the N (ADPG) and C terminus (SGRLVPR). Crystallization, ligand soaking and data collection Initial crystallization trials were set up using a Topaz Free Interface Diffusion (FID) Crystallizer system (Fluidigm Corpora- tion, San Francisco, CA). Crystals were observed in several conditions containing PEG 3350 or PEG 4000. Following opti- mization, diffraction quality crystals for NY107 were obtained at room temperature using a modified method for microbath under oil [45], by mixing the protein with reservoir solution containing 20% PEG 3350, 0.2 M magnesium chloride at pH 7.2. For receptor analog complexes, crystals were soaked for 3 hours in the crystallization buffer containing 10 mM 39SLN or 69SLN (V-labs Inc., Covington, LA), or overnight in 10mM LSTb (Sigma, St. Louis, MO). All crystals were flash-cooled at 100K using 20% glycerol as the cryo-protectant. Datasets were collected at Advanced Photon Source (APS) beamlines 22 ID and BM at 100K. Data were processed with the DENZO-SACLEPACK suite [46]. Statistics for data collection are presented in Table 1. Structure determination and refinement The structure of NY107 was determined by molecular replacement with Phaser [47] using the structure of the avian H7 (Av-H7) from A/turkey/Italy/2002, pdb:1TI8 (HA1, 78% identity; HA2, 90% identity) as the searching model. One HA trimer occupies the asymmetric unit with an estimated solvent content of 58% based on a Matthews’ coefficient (Vm) of 2.9 A˚ 3/ Da. Rigid body refinement of the trimer led to an overall R/Rfree of 28.6%/37.4%. The model was then ‘‘mutated’’ to the correct sequence and rebuilt by Coot [48], then the protein structures were refined with REFMAC [49] using TLS refinement [50]. The final models were assessed using MolProbity [51]. The three complex structures were refined and evaluated using the same strategy. All statistics for data processing and refinement are presented in Table 1. Electron density maps (2fo-fc) were generated in Refmac [49] while simulated annealing omit maps were generated by sa-omit-map, a part of the Crystallography and NMR System (CNS) software [52]. Virus generation Wild type and mutant viruses of NY107 (H7N2) and A/ Netherland/219/2003 (H7N7) were generated from plasmids by a reverse genetics approach [53]. To generate viruses with amino acid insertion or substitution in the HA, mutations were introduced into plasmid DNA with an overlap extension PCR approach [54]. Viruses derived by plasmid transfection of HK293 cells were propagated in eggs. The genomes of resulting virus stocks were sequenced to detect the emergence of possible variants during amplification. Glycan binding analyses Glycan microarray printing and recHA analyses have been described previously [24,30,44,55] (see Table 2 for glycans used for analyses in these experiments). Virus were analyzed on the microarray as described previously [30]. PDB accession codes The atomic coordinates and structure factors of NY107 are available from the RCSB PDB under accession codes 3M5G for the unliganded NY107, 3M5H for the NY107 with 39-SLN and 3M5I and 3M5J for NY107 with 69SLN and LSTb, respectively. Accession/ID numbers for genes/proteins used in this work A/New York/107/03 (H7N2), Genbank: ACC55270; A/ Netherlands/219/03 (H7N7), Genbank: AAR02640; A/Hong Kong/1-9/68 (H3N2), 2HMG; A/Duck/Ukraine/1/63 (H3N8), PDB: 1MQL; A/South Carolina/1/18 (H1N1), PDB: 1RD8; A/ Puerto Rico/8/34 (H1N1), PDB: 1RU7; A/Swine/Iowa/15/30 (H1N1), PDB: 1RUY; A/Singapore/1/1957 (H2N2), PDB: 2WRC; A/Viet Nam/1203/04 (H5N1), PDB: 2FK0; A/Duck/ Singapore/3/97 (H5N3), PDB: 1JSM; A/Swine/Hong Kong/9/ 98 (H9N2), PDB: 1JSD; A/Turkey/Italy/8000/02 (H7N3), PDB: 1TI8; C/Johannesburg/1/66, 1FLC. Supporting Information Figure S1 Sequence alignment of selected structurally available HAs. Human H3 (PDB: 2HMG), Avian H3 (PDB: 1MQL), 1918- Human H1 (PDB: 1RD8), 1934-Human H1 (PDB: 1RU7), Swine H1 (PDB: 1RUY), 1957-Huamn H2 (PDB: 2WRC), Human H5 (PDB: 2FK0), Avian H5 (PDB: 1JSM), Swine H9 (PDB: 1JSD), and Avian H7 (PDB: 1TI8) were used in the alignments. The fusion domain of HA1 is highlighted in magenta, the vestigial esterase domain is highlighted in green, the receptor binding domain is highlighted in blue, and the fusion domain of HA2 is highlighted in red. Residue numbering is based on the H3 HA sequence. Found at: doi:10.1371/journal.ppat.1001081.s001 (2.84 MB TIF) Figure S2 Expression and purification of NY107. SDS-PAGE reveals that NY107 was expressed as the HA0 form with a mass approximately 60kDa (middle lane). Thrombin cleavage resulted in an unexpected reduction in band size to a HA1/HA2 profile (right lane) with possible multiple glycoforms for the HA2 clearly present. Found at: doi:10.1371/journal.ppat.1001081.s002 (0.23 MB TIF) Figure S3 Comparison of glycan binding to NY107 with other HAs. A. Overlap of a2-3 ligands binding in the receptor binding site from NY-H7 (green), Av-H3 (orange), 1930-Hu-H1 (magenta), 1957-Hu-H2 (cyan), Av-H5 (yellow), and Sw-H9 (grey). B. Overlap of a2-6 linkage ligands binding in the receptor binding site from NY-H7 (green), Av-H3 (orange), 1930-Hu-H1 (magenta), 1957-Hu-H2 (cyan), Av-H5 (yellow), and Sw-H9 (grey). Found at: doi:10.1371/journal.ppat.1001081.s003 (2.55 MB TIF) Figure S4 Simulated annealing omit maps of the receptor binding site (contoured at 1s). A. NY107 (blue), B. NY107-39SLN (orange), C. NY107-69SLN (red), and D. NY107-LSTb (magenta). The protein model is shown in cartoon, and the residues involved in the binding to receptor analogs were shown in sticks. Maps were generated using version 1.2 of the Crystallography and NMR System (CNS) software. Found at: doi:10.1371/journal.ppat.1001081.s004 (1.93 MB TIF) Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 9 September 2010 | Volume 6 | Issue 9 | e1001081 Figure S5 Comparison of NY107 RBS to HEF. Overlap of RBS from NY107 (green), Av-H7 (marine) and HEF (magenta). Found at: doi:10.1371/journal.ppat.1001081.s005 (1.12 MB TIF) Table S1 Comparison of r.m.s.d. (A˚ ) for different HA domains. For analyzing differences in the overall structure, r.m.s.d. values were calculated between monomers or domains of different HA’s, after the Ca atoms of the HA2 domains were superposed by sequence and structural alignment onto the equivalent domains of NY107. Found at: doi:10.1371/journal.ppat.1001081.s006 (0.04 MB DOC) Table S2 Comparison of r.m.s.d. (A˚ ) for individual domains. Each domain was superimposed separately to determine how the individual NY107 domains compared to equivalent domains in the other structures. Found at: doi:10.1371/journal.ppat.1001081.s007 (0.04 MB DOC) Table S3 Molecular interactions between NY107 and receptor analogs. The hydrogen bond cutoff is 3.8 A˚ for the listing interactions. Found at: doi:10.1371/journal.ppat.1001081.s008 (0.07 MB DOC) Table S4 Glycan array differences between NY107, the fully restored NY107-ins, and NL219 (virus and rHA). The color coding in the left hand column reflects the same coloring scheme used in Figures 2 and 3. Significant binding of samples to glycans are qualitatively estimated based on relative strength of the signal for the data shown in Figures 2 and 3 Strong (+++), weak (+). Found at: doi:10.1371/journal.ppat.1001081.s009 (0.19 MB DOC) Acknowledgments The authors would like to thank the staff of SER-CAT sector 22 for their help in data collection. We also thank WHO Global Influenza Surveillance Network for providing NY107 and NL219 viruses from which the reverse genetics viruses were generated. Glycan microarray data presented here will be made available on-line through the CFG web site upon publication (www.functionalglycomics.org). The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention or the Agency for Toxic Substances and Disease Registry. Author Contributions Conceived and designed the experiments: HY LMC PJC ROD JS. Performed the experiments: HY LMC PJC JS. Analyzed the data: HY LMC PJC ROD JS. Wrote the paper: HY LMC PJC ROD JS. References 1. Thompson WW, Shay DK, Weintraub E, Brammer L, Cox N, et al. (2003) Mortality associated with influenza and respiratory syncytial virus in the United States. JAMA 289: 179–186. 2. Thompson WW, Shay DK, Weintraub E, Brammer L, Bridges CB, et al. (2004) Influenza-associated hospitalizations in the United States. JAMA 292: 1333–1340. 3. WHO (1980) A revision of the system of nomenclature for influenza viruses: a WHO memorandum. Bull, WHO 58: 585–591. 4. Fouchier RA, Munster V, Wallensten A, Bestebroer TM, Herfst S, et al. (2005) Characterization of a novel influenza A virus hemagglutinin subtype (H16) obtained from black-headed gulls. J Virol 79: 2814–2822. 5. Scholtissek C, Rohde W, Von Hoyningen V, Rott R (1978) On the origin of the human influenza virus subtypes H2N2 and H3N2. Virology 87: 13–20. 6. Kawaoka Y, Bean WJ, Webster RG (1989) Evolution of the hemagglutinin of equine H3 influenza viruses. Virology 169: 283–292. 7. Garten RJ, Davis CT, Russell CA, Shu B, Lindstrom S, et al. (2009) Antigenic and genetic characteristics of swine-origin 2009 A(H1N1) influenza viruses circulating in humans. Science 325: 197–201. 8. Shinya K, Ebina M, Yamada S, Ono M, Kasai N, et al. (2006) Avian flu: influenza virus receptors in the human airway. Nature 440: 435–436. 9. Matrosovich MN, Gambaryan AS, Teneberg S, Piskarev VE, Yamnikova SS, et al. (1997) Avian influenza A viruses differ from human viruses by recognition of sialyloligosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology 233: 224–234. 10. de Jong JC, Claas EC, Osterhaus AD, Webster RG, Lim WL (1997) A pandemic warning? Nature 389: 554. 11. Taubenberger JK, Morens DM, Fauci AS (2007) The next influenza pandemic: can it be predicted? JAMA 297: 2025–2027. 12. Webster RG, Geraci J, Petursson G, Skirnisson K (1981) Conjunctivitis in human beings caused by influenza A virus of seals. N Engl J Med 304: 911. 13. Kurtz J, Manvell RJ, Banks J (1996) Avian influenza virus isolated from a woman with conjunctivitis. Lancet 348: 901–902. 14. Koopmans M, Wilbrink B, Conyn M, Natrop G, van der Nat H, et al. (2004) Transmission of H7N7 avian influenza A virus to human beings during a large outbreak in commercial poultry farms in the Netherlands. Lancet 363: 587– 593. 15. Fouchier RA, Schneeberger PM, Rozendaal FW, Broekman JM, Kemink SA, et al. (2004) Avian influenza A virus (H7N7) associated with human conjunctivitis and a fatal case of acute respiratory distress syndrome. Proc Natl Acad Sci U S A 101: 1356–1361. 16. CDC (2004) Update: influenza activity–United States and worldwide, 2003–04 season, and composition of the 2004–05 influenza vaccine. MMWR Morb Mortal Wkly Rep: Centers for Disease Control. pp 547–552. 17. Hirst M, Astell CR, Griffith M, Coughlin SM, Moksa M, et al. (2004) Novel avian influenza H7N3 strain outbreak, British Columbia. Emerg Infect Dis 10: 2192–2195. 18. Tweed SA, Skowronski DM, David ST, Larder A, Petric M, et al. (2004) Human illness from avian influenza H7N3, British Columbia. Emerg Infect Dis 10: 2196–2199. 19. EditorialTeam (2007) Avian influenza A/H7N2 outbreak in the United Kingdom. Euro Surveill 12: 2. 20. Suarez DL, Garcia M, Latimer J, Senne D, Perdue M (1999) Phylogenetic analysis of H7 avian influenza viruses isolated from the live bird markets of the Northeast United States. J Virol 73: 3567–3573. 21. Belser JA, Lu X, Maines TR, Smith C, Li Y, et al. (2007) Pathogenesis of avian influenza (H7) virus infection in mice and ferrets: enhanced virulence of Eurasian H7N7 viruses isolated from humans. J Virol 81: 11139–11147. 22. Belser JA, Blixt O, Chen LM, Pappas C, Maines TR, et al. (2008) Contemporary North American influenza H7 viruses possess human receptor specificity: Implications for virus transmissibility. Proc Natl Acad Sci U S A 105: 7558–7563. 23. Gambaryan AS, Tuzikov AB, Pazynina GV, Desheva JA, Bovin NV, et al. (2008) 6-sulfo sialyl Lewis X is the common receptor determinant recognized by H5, H6, H7 and H9 influenza viruses of terrestrial poultry. Virol J 5: 85. 24. Stevens J, Blixt O, Glaser L, Taubenberger JK, Palese P, et al. (2006) Glycan microarray analysis of the hemagglutinins from modern and pandemic influenza viruses reveals different receptor specificities. J Mol Biol 355: 1143–1155. 25. Stevens J, Corper AL, Basler CF, Taubenberger JK, Palese P, et al. (2004) Structure of the uncleaved human H1 hemagglutinin from the extinct 1918 influenza virus. Science 303: 1866–1870. 26. Russell RJ, Kerry PS, Stevens DJ, Steinhauer DA, Martin SR, et al. (2008) Structure of influenza hemagglutinin in complex with an inhibitor of membrane fusion. Proc Natl Acad Sci U S A 105: 17736–17741. 27. Russell RJ, Gamblin SJ, Haire LF, Stevens DJ, Xiao B, et al. (2004) H1 and H7 influenza haemagglutinin structures extend a structural classification of haemagglutinin subtypes. Virology 325: 287–296. 28. Matrosovich M, Tuzikov A, Bovin N, Gambaryan A, Klimov A, et al. (2000) Early alterations of the receptor-binding properties of H1, H2, and H3 avian influenza virus hemagglutinins after their introduction into mammals. J Virol 74: 8502–8512. 29. Nobusawa E, Ishihara H, Morishita T, Sato K, Nakajima K (2000) Change in receptor-binding specificity of recent human influenza A viruses (H3N2): a single amino acid change in hemagglutinin altered its recognition of sialyloligosacchar- ides. Virology 278: 587–596. 30. Stevens J, Blixt O, Chen LM, Donis RO, Paulson JC, et al. (2008) Recent avian H5N1 viruses exhibit increased propensity for acquiring human receptor specificity. J Mol Biol 381: 1382–1394. 31. Russell RJ, Stevens DJ, Haire LF, Gamblin SJ, Skehel JJ (2006) Avian and human receptor binding by hemagglutinins of influenza A viruses. Glycoconj J 23: 85–92. 32. Gamblin SJ, Haire LF, Russell RJ, Stevens DJ, Xiao B, et al. (2004) The structure and receptor binding properties of the 1918 influenza hemagglutinin. Science 303: 1838–1842. 33. Suarez DL, Spackman E, Senne DA (2003) Update on molecular epidemiology of H1, H5, and H7 influenza virus infections in poultry in North America. Avian Dis 47: 888–897. 34. Daniels PS, Jeffries S, Yates P, Schild GC, Rogers GN, et al. (1987) The receptor-binding and membrane-fusion properties of influenza virus variants Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 10 September 2010 | Volume 6 | Issue 9 | e1001081 selected using anti-haemagglutinin monoclonal antibodies. EMBO J 6: 1459–1465. 35. Rosenthal PB, Zhang X, Formanowski F, Fitz W, Wong CH, et al. (1998) Structure of the haemagglutinin-esterase-fusion glycoprotein of influenza C virus. Nature 396: 92–96. 36. Weinstein J, de Souza-e-Silva U, Paulson JC (1982) Purification of a Gal beta 1 to 4GlcNAc alpha 2 to 6 sialyltransferase and a Gal beta 1 to 3(4)GlcNAc alpha 2 to 3 sialyltransferase to homogeneity from rat liver. J Biol Chem 257: 13835–13844. 37. Gambaryan A, Webster R, Matrosovich M (2002) Differences between influenza virus receptors on target cells of duck and chicken. Arch Virol 147: 1197–1208. 38. Wan H, Perez DR (2006) Quail carry sialic acid receptors compatible with binding of avian and human influenza viruses. Virology 346: 278–286. 39. Guo CT, Takahashi N, Yagi H, Kato K, Takahashi T, et al. (2007) The quail and chicken intestine have sialyl-galactose sugar chains responsible for the binding of influenza A viruses to human type receptors. Glycobiology 17: 713–724. 40. Maines TR, Chen LM, Matsuoka Y, Chen H, Rowe T, et al. (2006) Lack of transmission of H5N1 avian-human reassortant influenza viruses in a ferret model. Proc Natl Acad Sci U S A 103: 12121–12126. 41. Wan H, Sorrell EM, Song H, Hossain MJ, Ramirez-Nieto G, et al. (2008) Replication and transmission of H9N2 influenza viruses in ferrets: evaluation of pandemic potential. PLoS One 3: e2923. 42. Weis WI, Brunger AT, Skehel JJ, Wiley DC (1990) Refinement of the influenza virus hemagglutinin by simulated annealing. J Mol Biol 212: 737–761. 43. Frank S, Kammerer RA, Mechling D, Schulthess T, Landwehr R, et al. (2001) Stabilization of short collagen-like triple helices by protein engineering. J Mol Biol 308: 1081–1089. 44. Stevens J, Blixt O, Tumpey TM, Taubenberger JK, Paulson JC, et al. (2006) Structure and receptor specificity of the hemagglutinin from an H5N1 influenza virus. Science 312: 404–410. 45. Chayen NE, Shaw-Steward PD, Blow DM (1992) Microbatch crystallization under oil – a new technique allowing many small volume crystallization experiments. J Cryst Growth 122: 176–180. 46. Otwinowski A, Minor W (1997) Processing of X-ray diffraction data collected in oscillation mode. Meothds in Enzymology 276: 307–326. 47. McCoy AJ, Grosse-Kunstleve RW, Storoni LC, Read RJ (2005) Likelihood- enhanced fast translation functions. Acta Crystallogr D Biol Crystallogr 61: 458–464. 48. Emsley P, Cowtan K (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132. 49. CCP4 (1994) The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 50: 760–763. 50. Winn MD, Isupov MN, Murshudov GN (2001) Use of TLS parameters to model anisotropic displacements in macromolecular refinement. Acta Crystallogr D Biol Crystallogr 57: 122–133. 51. Davis IW, Leaver-Fay A, Chen VB, Block JN, Kapral GJ, et al. (2007) MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res 35: W375–383. 52. Brunger AT (2007) Version 1.2 of the Crystallography and NMR system. Nat Protoc 2: 2728–2733. 53. Hoffmann E, Webster RG (2000) Unidirectional RNA polymerase I-polymerase II transcription system for the generation of influenza A virus from eight plasmids. J Gen Virol 81: 2843–2847. 54. Higuchi R, Krummel B, Saiki RK (1988) A general method of in vitro preparation and specific mutagenesis of DNA fragments: study of protein and DNA interactions. Nucleic Acids Res 16: 7351–7367. 55. Blixt O, Head S, Mondala T, Scanlan C, Huflejt ME, et al. (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc Natl Acad Sci U S A 101: 17033–17038. 56. DeLano WL (2002) The PyMol Molecular Graphics Systems. wwwpymolorg. Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 11 September 2010 | Volume 6 | Issue 9 | e1001081
3M5I
Crystal structure of a H7 influenza virus hemagglutinin complexed with 6SLN
Structures of Receptor Complexes of a North American H7N2 Influenza Hemagglutinin with a Loop Deletion in the Receptor Binding Site Hua Yang, Li-Mei Chen, Paul J. Carney, Ruben O. Donis, James Stevens* Influenza Division, Centers for Disease Control and Prevention, Atlanta, Georgia, United States of America Abstract Human infections with subtype H7 avian influenza viruses have been reported as early as 1979. In 1996, a genetically stable 24-nucleotide deletion emerged in North American H7 influenza virus hemagglutinins, resulting in an eight amino acid deletion in the receptor-binding site. The continuous circulation of these viruses in live bird markets, as well as its documented ability to infect humans, raises the question of how these viruses achieve structural stability and functionality. Here we report a detailed molecular analysis of the receptor binding site of the North American lineage subtype H7N2 virus A/New York/107/2003 (NY107), including complexes with an avian receptor analog (39-sialyl-N-acetyllactosamine, 39SLN) and two human receptor analogs (69-sialyl-N-acetyllactosamine, 69SLN; sialyllacto-N-tetraose b, LSTb). Structural results suggest a novel mechanism by which residues Arg220 and Arg229 (H3 numbering) are used to compensate for the deletion of the 220-loop and form interactions with the receptor analogs. Glycan microarray results reveal that NY107 maintains an avian-type (a2-3) receptor binding profile, with only moderate binding to human-type (a2-6) receptor. Thus despite its dramatically altered receptor binding site, this HA maintains functionality and confirms a need for continued influenza virus surveillance of avian and other animal reservoirs to define their zoonotic potential. Citation: Yang H, Chen L-M, Carney PJ, Donis RO, Stevens J (2010) Structures of Receptor Complexes of a North American H7N2 Influenza Hemagglutinin with a Loop Deletion in the Receptor Binding Site. PLoS Pathog 6(9): e1001081. doi:10.1371/journal.ppat.1001081 Editor: Fe´lix A. Rey, Institut Pasteur, France Received March 16, 2010; Accepted July 28, 2010; Published September 2, 2010 This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. Funding: This work was funded by the Centers for Disease Control and Prevention. Use of the Advanced Photon Source at Argonne National Laboratory was supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC02-06CH11357. Glycan microarrays as well as glycans for direct binding experiments were produced for the Centers for Disease Control by the CFG funded by National Institute of General Medical Sciences Grant GM62116. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: fwb4@cdc.gov Introduction Influenza is an acute respiratory virus that infects up to 20% of the population in the United States, resulting in ,36,000 deaths annually [1,2]. The two membrane glycoproteins on the surface of influenza A virus, hemagglutinin (HA), which functions as the receptor binding and membrane fusion glycoprotein in cell entry, and neuraminidase (NA), which functions as the receptor destroying enzyme in virus release, form the basis for defining subtypes [3]. To date, 16 HA (H1–H16) and 9 NA (N1–N9) have been identified in avian species [4], while in the last century, only three subtypes, H1N1 in 1918 and 2009, H2N2 in 1957, and H3N2 in 1968 [5,6,7], have successfully adapted to humans. Hemagglutinin binds to sialic acid (SA) glycans present on host cell surfaces. The receptors on epithelial cells of the human upper respiratory tract are mainly a2-6-linked SA moieties [8]. Since avian influenza viruses pre- dominately bind a2-3-linked SA, and human influenza viruses preferentially bind to a2-6-linked SA, human infection by avian influenza viruses is rare [9]. However, since 1997 a growing number of human cases of avian influenza infection have been reported [10], including H5N1, H7N2, H7N3, H7N7, and H9N2 strains [11]. Although the current situation with the pandemic H1N1 influenza virus dominates public health efforts, the prospect of a novel pandemic emerging from these isolated cases continues to be a major public health threat around the world. Early cases of human infection by H7 influenza viruses are reported as far back as 1979 [12,13]. Since 2002, multiple outbreaks and human infections of H7 subtype viruses; within both Eurasian and North American lineages have been reported. In the Netherlands in 2003, a highly pathogenic avian influenza (HPAI) H7N7 outbreak resulted in more than 80 cases of human infections, including one fatality [14,15]. In New York in 2003, a single case of human respiratory infection of H7N2 was reported [16] and in British Columbia in 2004, an H7N3 virus caused two cases of conjunctivitis [17,18]. More recently in 2007, the United Kingdom reported several cases of low pathogenic avian influenza (LPAI) H7N2 virus infections that caused influenza-like illness and conjunctivitis [19]. Since 1996, H7 viruses of the North American lineage have been circulating in regional live bird markets [20], containing a 24-nucleotide deletion resulting in an eight amino acid deletion in the receptor-binding site (RBS) of HA (Figure S1). The recent human infections with H7 in North America have raised public health concerns as to how these viruses adapt to such a dramatic structural change while remaining one of the predominant circulating viral strains. A recent study of H7 viruses isolated from previous outbreaks revealed efficient replication in both mouse and ferret animal models [21]. In particular, ferret studies with A/New York/107/2003 (NY107), an H7N2 virus isolated from a man in New York, not only showed efficient replication in the upper respiratory tract of the ferret but also the capacity for PLoS Pathogens | www.plospathogens.org 1 September 2010 | Volume 6 | Issue 9 | e1001081 intra-species transmission by direct contact [21,22]. Interestingly, both an increased preference for a2-6 and decreased preference for a2-3-linked sialosides of this virus compared to the other avian influenza viruses was shown by previous glycan microarray analysis but less so by a competitive solid-phase binding assay [22,23]. Here we report a detailed molecular analysis of the RBS of the HA from North American lineage H7N2 virus, NY107, including glycan microarray analyses and structural analyses of the HA in complex with an avian receptor analog (39-Sialyl-N-acetyllactosa- mine, 39SLN) and two human receptor analogs (69-Sialyl-N- acetyllactosamine, 69SLN; Sialyllacto-N-tetraose b, LSTb). These results provide important insight into the interaction of H7 HAs with both avian and human hosts. Results Overall structure By using x-ray crystallography, the structure of H7 HA from the NY107 virus was determined to 2.6 A˚ resolution (Table 1). In addition, we also report three H7 HA receptor complex structures, with avian receptor analog (39SLN) to 2.7 A˚ resolution and with human receptor analogs (69SLN and LSTb) to 3.0 A˚ and 2.6 A˚ resolution, respectively (Table 1). The overall structure of NY107 is similar to other reported HA structures with a globular head containing the RBS and vestigial esterase domain, and a membrane proximal domain with its distinctive, central helical stalk and HA1/HA2 cleavage site (Figure 1A). Although five asparagine-linked glycosylation sites are predicted in the NY107 HA monomer, interpretable electron density was observed at only two sites, Asn38 in HA1 and Asn82 in HA2 (all residue numbers are based on H3 numbering). At these sites, only one or two N- acetyl glucosamines could be interpreted. During viral replication, HA is synthesized as a single chain precursor (HA0) and cleaved by specific host proteases into the infectious HA1/HA2 form. In baculovirus expression systems, highly pathogenic HAs, with a polybasic cleavage site, are expressed as an HA1/HA2 form [24], whereas HAs with monobasic cleavage sites (single Arg) from low pathogenic viruses are expressed as the HA0 form [25]. NY107 is regarded as a low pathogenic virus, and as expected, was produced in the HA0 form (Figure S2). However, subsequent digestion with thrombin protease to remove the His-tag resulted in cleavage to a profile on SDS-PAGE comparable to that of an HA1/HA2 form (Figure S2). A comparison of the NY107 cleavage site with the consensus cleavage pattern in the MEROPS database (http://merops. sanger.ac.uk) suggests it to be a possible thrombin cleavage site. Based on their molecular phylogenies, HAs are divided into two groups and five clades: group 1 includes H8, H9, and H12; H1, H2, H5, and H6; H11, H13 and H16; group 2 includes H3, H4, and H14; H7, H10 and H15 [26]. Among all available HA structures, we selected ten representative HAs from both avian and human subtypes for structural analysis. As expected, NY107 HA is structurally very similar to the Avian-H7 in all comparisons and closely related to H3, the other group 2 members used in the analyses (Tables S1 and S2). The receptor binding site The RBS is at the membrane distal end of each HA monomer and its specificity for sialic acid and the nature of its linkage to a vicinal galactose residue is a major determinant of host range- restriction. The consensus RBS for all current HAs is composed of three major structural elements: a 190-helix (residues 188–194), a 220-loop (residues 221–228), and a 130-loop (residues 134–138). In addition, highly conserved residues (Tyr98, Trp153, His183, and Tyr195) form the base of the pocket. Although the NY107 RBS is similar to other subtypes (H1, H2, H3, H5, and H9), a previously observed specific feature of H7 HAs, is also observed in the NY107 150-loop region: two residues inserted at position 158 result in this loop protruding more than 6A˚ towards the binding site compared to other subtype HAs (Figure 1B and Table S2) [27]. More interestingly, the eight amino acid deletion, only found in the North American lineage H7s, from position 221 to 228 (Figure S1), resulted in a complete loss of the 220-loop (Figure 1B). Sequence alignment shows that Arg220 and Arg229 are conserved in all influenza A HA subtypes (Figure S1), but structural alignment of NY107 HA shows Arg220 occupying the Gly228 position, and the much shorter loop turns at residue Pro217 (Figure 1C). The Ca distance between NY107 Arg220 and its homolog in the Av-H7 structure (PDB: 1TI8) [27] is 5.8A˚ , and they point in opposite directions (Figure 1C). The side chain direction of Av-H7 Arg220 is almost parallel with the beta sheet after Arg229, whereas the NY107 Arg220 points downward to the binding pocket. The Ca position of Arg229 in both H7 structures remains the same, except the side chain in the NY107 swings away by about 5.9A˚ (Figure 1C) and could help to stabilize this region by forming a hydrogen bond to the mainchain carbonyl of Gln210 in the neighboring monomer. In the absence of the 220-loop in NY107 HA, upon glycan binding the long side chain of Arg220 compensates for its loss and is displaced 4A˚ upward to form hydrogen bonds with receptor analogs inside the binding pocket (Figure 1D). Effect of loop truncation on the receptor binding specificity of NY107 Previously, mutations in the HA receptor binding domains of H1N1 (Glu190Asp/Gly225Asp) and H2N2/H3N2 (Gln226Leu and Gly228Ser) subtypes were responsible for adaptation of these viruses to pandemic strains [24,28,29,30]. Due to missing residues 221–228 in the NY107 HA RBS, neither mechanism for adaptation is possible. Thus, in order to look more closely at the role of the missing loop and its effect on receptor specificity, we first subjected the recombinant HA (recHA) to glycan microarray analyses and compared it to a reverse genetics-derived NY107 virus, and a co-circulating Eurasian virus and recHA, A/ Author Summary Influenza virus adaptation to different hosts usually results in a switch in receptor specificity of the viral surface coat protein, hemagglutinin. Indeed, the hemagglutinin sub- types from the last two human influenza pandemics of the 20th Century (H2 in 1957 and H3 1968) both adapted successfully to human-type receptor specificity through only two amino acid mutations in the receptor binding pocket (Glutamine226RLeucine and Glycine228RSerine). The recent human infections reported with other avian subtypes such as H5, H7 and H9 have raised public health concerns and focused efforts on identifying potential subtypes from which a future pandemic strain may emerge. Since 1996, H7 viruses of the North American lineage have been circulating in regional live bird markets, containing an eight amino acid deletion in the receptor- binding site of HA. Here we report a detailed structural analysis of the receptor binding site of a hemagglutinin from the North American lineage of H7N2 viruses, in complex with avian and human receptor analogs, to understand how these viruses have adapted to such a dramatic structural change in the binding site while remaining one of the predominant circulating viral strains. Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 2 September 2010 | Volume 6 | Issue 9 | e1001081 Netherlands/219/2003 (NL219), that has the consensus avian sequence in the 220-loop and it also infected a human [15]. Glycan microarray analysis of recombinant NY107 (Figure 2A and Table 2) revealed a highly restricted binding profile with strong binding to only a2-3 sulfated (#4–8), a2-3 branched (#9– 11) and mixed a2-3/a2-6 branched sialosides (#60–64) as well as to the long linear sialyl di- and tri-lactosamines (#22, 24). Weak binding was also observed (above background) to other a2-3 glycans on the array. The recombinant NY107 also revealed a strict glycan binding preference to only one a2-6 glycan, the internal structure, Galb1-3(Neu5Aca2-6)GlcNAcb1-3Galb1-4Glc (#58; LSTb) (Figure 2A), a glycan highlighted in a previous study [22]. The virus with higher valency and avidity revealed stronger binding to all a2-3 groups, in addition to the branched di-sialyl a2- 6 biantennary structures (#46–48) as well the LSTb (#58) (Figure 2B and Table 2). In contrast, the NL219 recHA (Figure 2C and Table 2) bound well to only the avian a2-3 containing sialyl- glycans (sulfated, branched, linear and fucosylated). Its corre- sponding virus also reflected this specificity although it also revealed strong binding to a2-3 N-glycolylneuraminic acid (Neu5Gc) containing glycans (#66–70) (Figure 2D and Table 2). To further assess the effect of the missing 220 loop on HA structural stability and receptor specificity it was essential to evaluate these functions on the ancestral HA containing the full length 220-loop. To this end, we engineered an HA with an avian H7 consensus (PQVNGQSG) 220-loop re-introduced (NY107- 220ins) into the NY107 HA and recovered this virus by reverse genetics. Compared to the NY107 virus (Figure 2A) glycan microarray analyses of the resulting NY107-220ins virus (Figure 3A and Table 2) revealed a decrease in binding to branched (#9–11) and linear (#12–27) a2-3 sialosides and a loss of binding to the branched di-sialyl a2-6 biantennary structures (#46–48), LSTb (#58) as well as the mixed a2-3/a2-6 branched sialosides (#60– 64). In addition, sequence analysis of the NY107-220ins HA revealed the presence of quasispecies in the second position of the inserted loop, P(Q/K)VNGQSG, suggesting that re-introduction of the loop alone is not tolerated and does not create an avian-type binding profile. Thus other amino acid substitutions in the HA might have co-evolved with the deletion of the 220 loop to help stabilize the RBS/HA to maintain functionality. When viruses containing this 220-loop deletion emerged in North America in the mid 90’s, four additional amino acid substitutions, Gly114Arg, Asp119Gly, Gly186Glu and Gly205Arg, in the HA1 as well as an Asp19Asn in the HA2 chain were also introduced to most of the circulating isolates. Of these, Gly186Glu and Gly205Arg in the HA1 are close to the RBS, at the monomer interface, and could potentially modulate its structure and/or function. NY107 viruses with a restored consensus 220-loop and a single Glu186Gly (NY107- ins-186) or Arg205Gly (NY107-ins-205) substitution as well as the Glu186Gly/Arg205Gly double substitution (NY107-ins-186/205) were derived by reverse genetics and evaluated. Glycan microarray analysis for the three resulting viruses revealed similar glycan binding profiles with increased binding to a2-3 sialosides, including mixed a2-3/a2-6 branched sialosides (#60–64), a2-3 Neu5Gc (#66–70), but limited binding to the a2,6 sialosides (Figures 3B, 3C, 3D), resulting in a binding profile virtually identical to that of the NL219 virus and other avian influenza viruses (Figure 2D) [30]. Sequence analysis of the three reverse genetics derived viruses revealed no mutations/quasispecies in the HAs of either the NY107- ins-186 or the NY107-ins-186/205 virus stocks, indicative of replication fitness. For the NY107-ins-205 virus however, a Glu186Gly substitution emerged in the HA after only two passages in eggs following recovery from DNA transfection, indicating the importance of the co-variant position 186 with respect to HA functionality/glycan specificity. Altogether, the data indicates that the H7 subtype avian influenza viruses that were circulating in Table 1. Data collection and refinement statistics. NY107 NY107+39SLN NY107+69SLN NY107+LSTb Data collection Space group P212121 P212121 P212121 P212121 Cell dimensions (A˚) 66.96, 115.92, 251.61 67.80, 116.70, 249.84 66.60, 116.58, 250.68 67.08, 116.52, 251.95 Resolution (A˚) 50-2.6 (2.69-2.60)a 30-2.7 (2.80-2.70) 50-3.0 (3.11-3.0) 50-2.6 (2.69-2.60) Rsym or Rmerge 10.6 (41.3) 14.6 (48.6) 14.3 (35.4) 12.2 (31.5) I/s 39.6 (2.0) 24.3 (1.7) 34.2 (8.2) 40.5 (9.9) Completeness (%) 99.2 (99.0) 99.3 (94.6) 92.3 (75.6) 91.3 (86.2) Redundancy 7.2 (6.2) 5.8 (5.5) 4.9 (4.4) 10.9 (11.2) Refinement Resolution (A˚) 50-2.6 (2.67-2.60) 30-2.7 (2.77-2.70) 50-3.0 (3.08-3.00) 50-2.6 (2.67-2.60) No. of reflections (total) 57285 51770 33421 53603 No. of reflections (test) 3053 2769 1779 2842 Rwork/Rfree 21.7/25.6 21.4/26.4 20.5/26.0 20.4/24.7 No. of atoms 11795 11878 11648 12108 r.m.s.d.- bond length (A˚) 0.006 0.006 0.008 0.006 r.m.s.d.- bond angle (u) 0.905 0.974 1.085 0.859 MolProbityb scores Favored (%) 96.9 96.5 94.3 97.1 Outliers (%) (No. of residues) 0.1 (1/1434) 0.0 (0/1429) 0.1 (2/1433) 0.1 (2/1435) aNumbers in parentheses refer to the highest resolution shell. bReference [51]. doi:10.1371/journal.ppat.1001081.t001 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 3 September 2010 | Volume 6 | Issue 9 | e1001081 aquatic birds and poultry in North America before 1996 exhibited a classic avian a2-3 sialoside binding preference. In order for the 220- loop deletion to be tolerated, concurrent Gly186Glu and Gly205Arg substitutions in the vicinity of RBS of HA emerged to achieve a restricted a2-3 binding profile and only a moderate/limited increase in binding to branched di-sialyl a2-6 biantennary structures (#46– 48) as well the a2,6 internal sialoside, LSTb (#58). NY107 avian receptor complex To understand from a structural perspective how NY107 interacts with host receptors, we solved the structure of NY107 in complex with an avian and two human receptor analogs. For the avian receptor analog, 39SLN, the electron density maps revealed well-ordered features for the Sia-1, Gal-2, and GlcNAc-3 in the NY107 HA complex structure (Figure 4A). Structural comparison of NY107 HA binding to other, H1, H2, H3, H5, and H9 subtypes (Figure S2A) revealed that 39SLN binding to NY107 resembled binding of the other published HAs. Indeed, the terminal Sia-1 moiety is positioned almost identically in all structures, and forms the majority of hydrogen bonds and contacts with residues in the RBS (Figure 4A and Table S3). Figure 1. NY107 HA monomer and comparison of its RBS to other HA structures. (A) One monomer is shown with the HA1 chain colored in green and the HA2 chain in cyan. The location of the receptor binding site and the HA1/HA2 cleavage site are circled. (B) The superposition of receptor binding domains of NY107 (green), Av-H7 (marine), 1918-Hu-H1 (magenta), Hu-H5 (yellow), Hu-H3 (orange), and Sw-H9 (grey). The proximity of Arg220 and Gln226 are highlighted. Three structural elements comprising this binding site are labeled. The two major differences are the extended 150-loop and the deletion of 220-loop of NY107. (C) Overlap of NY107 (green) and Av-H7 (marine) (PDB: 1TI8) illustrates the compensatory effect of R220 bringing it close to the position occupied by G228 in the avian HA. (D) Overlap of the NY107 (green), NY107- 39SLN (orange), NY107-69SLN (red), and NY107-LSTb (magenta) structures. All the figures were generated and rendered with the use of MacPyMOL [56]. doi:10.1371/journal.ppat.1001081.g001 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 4 September 2010 | Volume 6 | Issue 9 | e1001081 Published avian HA structures with an intact 220-loop form very close interactions with Gal-2 of 39SLN via residue Gln226 which is important in receptor specificity and host adaptation. For example, in the avian H7/39SLN HA structure it interacts with Gal-2 O4 [31]. In the NY107 HA structure, although Gln226 is absent and no other residue occupies the same space as Gln226 (Figure 1B), Arg220 does forms a hydrogen bond between Arg220 NH2 and Gal-2 O4 (Figure 4A). Interestingly, although there was interpretable density for the GlcNAc-3 (Figure 4A and Figure S4B), no hydrogen bonding was apparent between the HA and the GlcNAc-3, which is consistent with other reported structures [32]. Thus, for binding to avian receptors, the trans conformation of a2-3 linkages is essential and perhaps only the first two saccharides are required. Indeed, due to the absence of 220-loop in the NY107 HA structure, the ‘‘aperture’’ of the RBS formed by 220-loop and 130-loop in regular HAs is increased by ,10 A˚ , so that the branched, internal, and perhaps more complicated glycans might be accommodated more efficiently. NY107 human receptor complexes In the NY107/69SLN complex, only Sia-1 and Gal-2 are ordered (Figure 4B). The Sia-1 remains in the same position as previously analyzed glycan/HA complexes from H1, H2, H3, H5, and H9 (Figure S3B), whereas the Av-H7 complex structure with Sialyllacto-N-tetraose c (LSTc) did not reveal any density for the Sia-1 in the receptor binding site [31]. The Gal-2 position varies significantly among different subtypes. Compared to the human- adapted H1 HA [32], Gal-2 in the NY107 HA is 3A˚ higher, and thus is further from the protein (Figure S3B). In NY107, the Gal-2 only forms an intramolecular, saccharide-saccharide interaction with Sia-1. The poor electron density map and fewer interactions with protein residues suggest that the cis conformation of a2-6 Figure 2. Receptor specificity of NY107 recHA and virus. Glycan microarray analysis of recombinant NY107 HA (A) and NY107 virus (B) compared to the recHA (C) and virus (D) from a Eurasian lineage A/Netherlands/219/2003 H7 influenza virus that was circulating in the same year and also infected a human. Colored bars highlight glycans that contain a2-3 SA (blue) and a2-6 SA (red), a2-6/a2-3 mixed SA (purple), N-glycolyl SA (green), a2-8 SA (brown), b2-6 and 9-O-acetyl SA, and non-SA (grey). Error bars reflect the standard error in the signal for six independent replicates on the array. Structures of each of the numbered glycans are found in Table S4. doi:10.1371/journal.ppat.1001081.g002 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 5 September 2010 | Volume 6 | Issue 9 | e1001081 linkages in 69SLN trisaccharides show a reduced binding affinity with NY107. Glycan array results with NY107 revealed a strong binding signal for the internal a2-6 sialoside, LSTb. To further investigate this interaction, we solved the structure of the NY107/LSTb complex. The final model contained Sia-1, NAG-2, Gal-3, and Gal-5 in the RBS. Although glycan microarray data indicated NY107 to have a specific affinity for LSTb, few interactions were apparent from the crystal structure. Sia-1 still forms multiple hydrogen bonds with residues in the RBS (Table S3 & Figure 4C). The branched Gal-5 interacts with Ser137, to help stabilize the LSTb binding. However, Arg220 and Lys193, the two residues showing close binding with 39SLN, did not form any hydrogen bonds with LSTb. In the structure, Gal-5 also interacts with a crystal packing symmetry mate and thus the flexibility of whole LSTb may be restricted. In solution, with more freedom, the LSTb should be able to tilt closer to the RBS, and thus Glc-4 may have more interactions with the 190-helix than seen in the crystal structure. Discussion Human infections by avian influenza viruses, including H7 subtypes, continue to pose a major public health threat. Although the species barrier prevents avian influenza viruses from widespread infection of the human population, the molecular determinants of efficient interspecies transmission and pathoge- nicity are still poorly understood. The viral coat protein HA however, is perhaps a critical molecule since previous pandemic viruses modified their receptor specificity and overcame the interspecies barrier to spread in the human population. Although HA structures alone and in complex with receptor analogs provide considerable insight into receptor binding, it is clear that HAs from different species and subtypes have significant structural variation. Indeed, low-pathogenic H7N2 avian influenza viruses with an 8 amino acid deletion within its RBS started to circulate in live-bird markets in the northeast United States in 1996. Despite what one would consider a debilitating mutation, these viruses have been reported as the predominant isolate [33]. Whether such a deletion contributed to their evolutionary success and how are an important questions, especially in light of NY107’s ability to produce respiratory illness in humans [16], as well as its reported increased affinity for human-type receptors and ability for contact transmission in ferrets [21]. To try to help answer these questions, we have analyzed the molecular structures of NY107 and its complexes with receptor analogs to explain receptor specificity at the molecular level. The crystal structures of NY107 and its complexes with both avian and human receptor analogs describe a mechanism as to how an influenza virus might adapt by dramatically altering its RBS, and still be functional. Arg220 of the HA1 chain of NY107 compensates for the loss of the 220-loop, by forming hydrogen bonds with Gal-2 from the avian analog (binding was not observed in either of the structures complexes with the human analogs). However, in the LSTb complex, branched Gal-5 forms extra interactions with the 130-loop, thus improving the binding preference for this particular glycan. Consistent with the structural evidence, glycan microarray analyses of NY107 revealed a strong binding preference for the branched a2-6 sialoside, LSTb. Except for the absence of the 220-loop, other key residues within the RBS are conserved in NY107 and thus, direct interactions with sialic acid are maintained. The 220-loop is recognized as one of the three crucial structural elements in the RBS. Aside from the North American lineage H7N2 viruses, which have been circulating with a deletion (221– 228) in this loop, there has been one other report describing a seven amino acid deletion (224–230) in a laboratory generated H3N2 escape mutant which was reported to have a slightly Table 2. Comparison of the sialoside receptor specificity of the HAs from H7 influenza viruses. Glycan Group Graph Numbera NY107 RecHA NY107 Virus NY107-ins Virus NY107-ins E186G Virus NY107-ins R205G Virus NY107-ins E186G/ R205G Virus NL219 RecHA NL219 Virus a2-3 Sulfated 4–8 +++b +++ +++ +++ +++ +++ +++ +++ Branched 9–11 +++ +++ + +++ +++ +++ +++ +++ Linear 12–27 + +++ + +++ +++ +++ +++ +++ Fucosylated 28–34 2 +++ +++ +++ +++ +++ +++ +++ a2-6 Sulfated 41 2 2 2 2 2 2 2 2 Branched mono-sialyl 42–45, 49 2 2 2 2 2 2 2 2 Branched di-sialyl 46–48 2 +++ 2 2 2 2 2 2 Linear 50–56 2 2 2 2 2 2 2 2 Internal 58–59 +++ +++ 2 2 2 2 2 2 Other Sialic acid 1–2 2 +++ + 2 2 2 2 2 a2-3/a2-6 Branched 60–64 2 2 2 +++ +++ +++ +++ +++ Neu5Gcc 65–72 2 2 2 +++ +++ +++ 2 +++ aMembers of each group are identified according to the graph number used in the microarray data in Figures 2 and 3 and correspond to numbers in the complete glycan list (Table S4). bBinding of samples to glycan subclasses are qualitatively estimated based on relative strength of the signal for the data shown in Figures 2 and 3: strong (+++), weak (+), absent (2). cN-glycolylneuraminic acid. doi:10.1371/journal.ppat.1001081.t002 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 6 September 2010 | Volume 6 | Issue 9 | e1001081 increased affinity for a2-3-linked glycans by hemagglutination assay [34]. Meanwhile, the equivalent region in the hemaggluti- nin-esterase-fusion (HEF) protein of influenza C virus reveals a rearrangement resulting in a truncated 260-loop in its RBS (Figure S5) [35]. However, without structural data with appropriate receptor analogs, it is not possible to compare the role of these loop variants in receptor binding to the H7 HA structure described here. When compared to NL219, another co-circulating H7 avian virus HA (Figure 2C and D), overall binding to a2-3-linked glycans was markedly reduced, while increased binding to a2-6- linked receptors was only marginal. However, these results focus attention on only 2 sub-classes of human-type receptors that may be important for infection (and transmission in ferrets). The NY107 virus interaction with biantennary glycans (Figure 2B), although weak (not seen in Figure 2A with recHA), is a possible route for virus entry as biantennary structures are common on tissues, i.e. glycan profiling data from human lung tissue on the Consortium for Functional Glycomics (CFG) web site. In addition, the internal sialoside, LSTb, was observed in both virus and recHA microarray data, suggesting this type of glycan has good affinity for this HA. The significance of this is unknown since LSTb has only been described in human milk [36]. Interestingly, NY107 and NL219 virus receptor binding and specificity has been addressed previously using glycan microarray analysis that reported a significantly increased preference for a2-6 and decreased preference for a2-3-linked sialosides [22]. In addition, the same viruses were also included in a recent study from Gambaryan et al. using a competitive solid-phase binding assay [23]. Our findings confirm and extend the receptor binding specificity reported by these authors in that they reported both viruses binding to sulfated sialylglycans with a lactosamine (Galb1- 4GlcNAc core and reported only a moderate binding affinity for a2-6-sialyllactosamine, the human-type receptor analog used in their assay. The 220-loop is an integral feature of the receptor binding site, and thus one would predict that such a deletion might have compromised this strain to be deleted from the population of circulating viruses. However, this was not the case [33] and its existence appears to be in part due to the additional mutations at Figure 3. Effect of 220-loop deletion and additional RBS mutations on NY107 receptor specificity. NY107 was engineered to restore the 220-loop to a consensus full-length HA from 1996 (A) and additional co-variant amino acid substitutions, Glu186Gly (B), Arg205Gly (C) and the double mutant Glu186Gly/Arg205Gly (D) to restore, on the NY107 framework, an HA RBS found in viruses prior to the introduction of the deletion in North American viruses. Colored bars group glycans as described in Figure 3. Error bars reflect the standard error in the signal for six independent replicates on the array. Structures of each of the numbered glycans are found in Table S4. doi:10.1371/journal.ppat.1001081.g003 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 7 September 2010 | Volume 6 | Issue 9 | e1001081 positions 186 and 205. Restoration of the loop with either or both residues mutated back to the pre-1994 consensus sequence resulted in a classic avian influenza virus binding profile. The emergence of the Glu186Gly mutation in the HA of the NY107- ins-205 mutant after only two passages of the rescued virus in eggs, also indicates the importance of these positions for HA functionality/glycan specificity. Analysis of the structural data reveals that positions 186 and 205 are on opposite sides of a monomer but are both close to the 220-loop deletion region in the trimeric form. The Glu at position 186 is close to Arg220 and may interact with Arg220 when binding avian receptors. Position 205 in the neighboring monomer may be important in trimer stability and maintaining RBS functionality. If one models the pre- 1996 220-loop restored into the NY107 structure, Arg205, Glu186 and the loop all clash, thus explaining the Glu186Gly mutation that emerged in the NY107-ins-205 virus HA after limited egg passage. The NY107 RBS with its more restricted a2-3 glycan binding preference and weak/moderate increase in a2-6 binding may have given the virus a selective advantage to be maintained in poultry at live bird markets and supplying farms. Certain terrestrial birds, such as quails and chickens, have recently been shown to present both human and avian types of receptors in the trachea and intestine [37,38,39]. Although it is not known what specific glycans are presented in these animals, it is conceivable that a virus with mixed specificity might have a distinct advantage over avian viruses that have specific avian receptor requirements, particularly in bird markets where multiple species coalesce. Previous results with H7N2, H9N2 and H5N1 viruses all highlight the fact that an increase in a2-6-binding preference is not sufficient for efficient transmission of avian influenza viruses to humans [22,40,41]. Although it remains to be seen whether prolonged circulation of viruses in terrestrial birds, such as domestic chickens, can provide a possible route for viruses to adapt for efficient human infection [11], continued surveillance of influenza viruses from avian and other animal reservoirs is urgently needed to define their zoonotic potential. Materials and Methods Cloning Based on H3 numbering [42], cDNA corresponding to residues 11–329 (HA1) and 1–176 (HA2) of the ectodomain of the hemagglutinin (HA) from A/New York/107/2003 (H7N2; Genbank:ACC55270) and A/Netherlands/219/2003 (H7N7; Genebank: AAR02640) was cloned into the baculovirus transfer vector, pAcGP67-A (BD Biosciences), incorporating a C-terminal thrombin cleavage site, a ‘‘foldon’’ sequence [43] and a His-tag at the extreme C-terminus of the construct to enable protein purification [25,44]. Transfection and virus amplification were carried out according to the baculovirus expression system manual (BD Biosciences Pharmingen). Figure 4. Glycan interactions within the NY107 RBS. The top panel shows the interactions of NY107 with (A) 39SLN, (B) 69SLN and (C) LSTb. NY107 is shown in orange/red/magenta cartoon respectively. The interacting HA residues are shown as green sticks. The bottom panel shows the electron density map of the ligands. The NY107 is shown in the same colors as above, and the ligands are shown as green sticks, the 2fo-fc electron density maps (contoured at 1s) are shown in grey. Simulated annealing omit maps are shown in supplementary Figure S4. doi:10.1371/journal.ppat.1001081.g004 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 8 September 2010 | Volume 6 | Issue 9 | e1001081 Protein expression and purification Soluble NY107 was recovered from the cell supernatant by metal affinity chromatography using Ni-NTA resin (Qiagen Inc.). Frac- tions containing NY107 were pooled and dialyzed against 10 mM Tris-HCl, 50 mM NaCl, pH 8.0, then subjected to ion-exchange chromatography (IEX) using a Mono-Q HR 10/10 column (GE Healthcare). IEX purified NY107 was subjected to thrombin digest (3 units/mg protein; overnight at 4uC) and purified by gel filtra- tion chromatography using a Superdex-200 16/60 column (GE Healthcare) and 50 mM Tris-HCl, 100 mM NaCl, pH 8.0 as running buffer. Protein eluting as a trimer was buffer exchanged into 10 mM Tris-HCl, 50 mM NaCl, pH 8.0 and concentrated to 14.5 mg/ml for crystallization trials. At this stage, the protein sample still contained the additional plasmid-encoded residues at both the N (ADPG) and C terminus (SGRLVPR). Crystallization, ligand soaking and data collection Initial crystallization trials were set up using a Topaz Free Interface Diffusion (FID) Crystallizer system (Fluidigm Corpora- tion, San Francisco, CA). Crystals were observed in several conditions containing PEG 3350 or PEG 4000. Following opti- mization, diffraction quality crystals for NY107 were obtained at room temperature using a modified method for microbath under oil [45], by mixing the protein with reservoir solution containing 20% PEG 3350, 0.2 M magnesium chloride at pH 7.2. For receptor analog complexes, crystals were soaked for 3 hours in the crystallization buffer containing 10 mM 39SLN or 69SLN (V-labs Inc., Covington, LA), or overnight in 10mM LSTb (Sigma, St. Louis, MO). All crystals were flash-cooled at 100K using 20% glycerol as the cryo-protectant. Datasets were collected at Advanced Photon Source (APS) beamlines 22 ID and BM at 100K. Data were processed with the DENZO-SACLEPACK suite [46]. Statistics for data collection are presented in Table 1. Structure determination and refinement The structure of NY107 was determined by molecular replacement with Phaser [47] using the structure of the avian H7 (Av-H7) from A/turkey/Italy/2002, pdb:1TI8 (HA1, 78% identity; HA2, 90% identity) as the searching model. One HA trimer occupies the asymmetric unit with an estimated solvent content of 58% based on a Matthews’ coefficient (Vm) of 2.9 A˚ 3/ Da. Rigid body refinement of the trimer led to an overall R/Rfree of 28.6%/37.4%. The model was then ‘‘mutated’’ to the correct sequence and rebuilt by Coot [48], then the protein structures were refined with REFMAC [49] using TLS refinement [50]. The final models were assessed using MolProbity [51]. The three complex structures were refined and evaluated using the same strategy. All statistics for data processing and refinement are presented in Table 1. Electron density maps (2fo-fc) were generated in Refmac [49] while simulated annealing omit maps were generated by sa-omit-map, a part of the Crystallography and NMR System (CNS) software [52]. Virus generation Wild type and mutant viruses of NY107 (H7N2) and A/ Netherland/219/2003 (H7N7) were generated from plasmids by a reverse genetics approach [53]. To generate viruses with amino acid insertion or substitution in the HA, mutations were introduced into plasmid DNA with an overlap extension PCR approach [54]. Viruses derived by plasmid transfection of HK293 cells were propagated in eggs. The genomes of resulting virus stocks were sequenced to detect the emergence of possible variants during amplification. Glycan binding analyses Glycan microarray printing and recHA analyses have been described previously [24,30,44,55] (see Table 2 for glycans used for analyses in these experiments). Virus were analyzed on the microarray as described previously [30]. PDB accession codes The atomic coordinates and structure factors of NY107 are available from the RCSB PDB under accession codes 3M5G for the unliganded NY107, 3M5H for the NY107 with 39-SLN and 3M5I and 3M5J for NY107 with 69SLN and LSTb, respectively. Accession/ID numbers for genes/proteins used in this work A/New York/107/03 (H7N2), Genbank: ACC55270; A/ Netherlands/219/03 (H7N7), Genbank: AAR02640; A/Hong Kong/1-9/68 (H3N2), 2HMG; A/Duck/Ukraine/1/63 (H3N8), PDB: 1MQL; A/South Carolina/1/18 (H1N1), PDB: 1RD8; A/ Puerto Rico/8/34 (H1N1), PDB: 1RU7; A/Swine/Iowa/15/30 (H1N1), PDB: 1RUY; A/Singapore/1/1957 (H2N2), PDB: 2WRC; A/Viet Nam/1203/04 (H5N1), PDB: 2FK0; A/Duck/ Singapore/3/97 (H5N3), PDB: 1JSM; A/Swine/Hong Kong/9/ 98 (H9N2), PDB: 1JSD; A/Turkey/Italy/8000/02 (H7N3), PDB: 1TI8; C/Johannesburg/1/66, 1FLC. Supporting Information Figure S1 Sequence alignment of selected structurally available HAs. Human H3 (PDB: 2HMG), Avian H3 (PDB: 1MQL), 1918- Human H1 (PDB: 1RD8), 1934-Human H1 (PDB: 1RU7), Swine H1 (PDB: 1RUY), 1957-Huamn H2 (PDB: 2WRC), Human H5 (PDB: 2FK0), Avian H5 (PDB: 1JSM), Swine H9 (PDB: 1JSD), and Avian H7 (PDB: 1TI8) were used in the alignments. The fusion domain of HA1 is highlighted in magenta, the vestigial esterase domain is highlighted in green, the receptor binding domain is highlighted in blue, and the fusion domain of HA2 is highlighted in red. Residue numbering is based on the H3 HA sequence. Found at: doi:10.1371/journal.ppat.1001081.s001 (2.84 MB TIF) Figure S2 Expression and purification of NY107. SDS-PAGE reveals that NY107 was expressed as the HA0 form with a mass approximately 60kDa (middle lane). Thrombin cleavage resulted in an unexpected reduction in band size to a HA1/HA2 profile (right lane) with possible multiple glycoforms for the HA2 clearly present. Found at: doi:10.1371/journal.ppat.1001081.s002 (0.23 MB TIF) Figure S3 Comparison of glycan binding to NY107 with other HAs. A. Overlap of a2-3 ligands binding in the receptor binding site from NY-H7 (green), Av-H3 (orange), 1930-Hu-H1 (magenta), 1957-Hu-H2 (cyan), Av-H5 (yellow), and Sw-H9 (grey). B. Overlap of a2-6 linkage ligands binding in the receptor binding site from NY-H7 (green), Av-H3 (orange), 1930-Hu-H1 (magenta), 1957-Hu-H2 (cyan), Av-H5 (yellow), and Sw-H9 (grey). Found at: doi:10.1371/journal.ppat.1001081.s003 (2.55 MB TIF) Figure S4 Simulated annealing omit maps of the receptor binding site (contoured at 1s). A. NY107 (blue), B. NY107-39SLN (orange), C. NY107-69SLN (red), and D. NY107-LSTb (magenta). The protein model is shown in cartoon, and the residues involved in the binding to receptor analogs were shown in sticks. Maps were generated using version 1.2 of the Crystallography and NMR System (CNS) software. Found at: doi:10.1371/journal.ppat.1001081.s004 (1.93 MB TIF) Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 9 September 2010 | Volume 6 | Issue 9 | e1001081 Figure S5 Comparison of NY107 RBS to HEF. Overlap of RBS from NY107 (green), Av-H7 (marine) and HEF (magenta). Found at: doi:10.1371/journal.ppat.1001081.s005 (1.12 MB TIF) Table S1 Comparison of r.m.s.d. (A˚ ) for different HA domains. For analyzing differences in the overall structure, r.m.s.d. values were calculated between monomers or domains of different HA’s, after the Ca atoms of the HA2 domains were superposed by sequence and structural alignment onto the equivalent domains of NY107. Found at: doi:10.1371/journal.ppat.1001081.s006 (0.04 MB DOC) Table S2 Comparison of r.m.s.d. (A˚ ) for individual domains. Each domain was superimposed separately to determine how the individual NY107 domains compared to equivalent domains in the other structures. Found at: doi:10.1371/journal.ppat.1001081.s007 (0.04 MB DOC) Table S3 Molecular interactions between NY107 and receptor analogs. The hydrogen bond cutoff is 3.8 A˚ for the listing interactions. Found at: doi:10.1371/journal.ppat.1001081.s008 (0.07 MB DOC) Table S4 Glycan array differences between NY107, the fully restored NY107-ins, and NL219 (virus and rHA). The color coding in the left hand column reflects the same coloring scheme used in Figures 2 and 3. Significant binding of samples to glycans are qualitatively estimated based on relative strength of the signal for the data shown in Figures 2 and 3 Strong (+++), weak (+). Found at: doi:10.1371/journal.ppat.1001081.s009 (0.19 MB DOC) Acknowledgments The authors would like to thank the staff of SER-CAT sector 22 for their help in data collection. We also thank WHO Global Influenza Surveillance Network for providing NY107 and NL219 viruses from which the reverse genetics viruses were generated. Glycan microarray data presented here will be made available on-line through the CFG web site upon publication (www.functionalglycomics.org). The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention or the Agency for Toxic Substances and Disease Registry. Author Contributions Conceived and designed the experiments: HY LMC PJC ROD JS. Performed the experiments: HY LMC PJC JS. Analyzed the data: HY LMC PJC ROD JS. Wrote the paper: HY LMC PJC ROD JS. References 1. Thompson WW, Shay DK, Weintraub E, Brammer L, Cox N, et al. (2003) Mortality associated with influenza and respiratory syncytial virus in the United States. JAMA 289: 179–186. 2. Thompson WW, Shay DK, Weintraub E, Brammer L, Bridges CB, et al. (2004) Influenza-associated hospitalizations in the United States. JAMA 292: 1333–1340. 3. WHO (1980) A revision of the system of nomenclature for influenza viruses: a WHO memorandum. Bull, WHO 58: 585–591. 4. Fouchier RA, Munster V, Wallensten A, Bestebroer TM, Herfst S, et al. (2005) Characterization of a novel influenza A virus hemagglutinin subtype (H16) obtained from black-headed gulls. J Virol 79: 2814–2822. 5. Scholtissek C, Rohde W, Von Hoyningen V, Rott R (1978) On the origin of the human influenza virus subtypes H2N2 and H3N2. Virology 87: 13–20. 6. Kawaoka Y, Bean WJ, Webster RG (1989) Evolution of the hemagglutinin of equine H3 influenza viruses. Virology 169: 283–292. 7. Garten RJ, Davis CT, Russell CA, Shu B, Lindstrom S, et al. (2009) Antigenic and genetic characteristics of swine-origin 2009 A(H1N1) influenza viruses circulating in humans. Science 325: 197–201. 8. Shinya K, Ebina M, Yamada S, Ono M, Kasai N, et al. (2006) Avian flu: influenza virus receptors in the human airway. Nature 440: 435–436. 9. Matrosovich MN, Gambaryan AS, Teneberg S, Piskarev VE, Yamnikova SS, et al. (1997) Avian influenza A viruses differ from human viruses by recognition of sialyloligosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology 233: 224–234. 10. de Jong JC, Claas EC, Osterhaus AD, Webster RG, Lim WL (1997) A pandemic warning? Nature 389: 554. 11. Taubenberger JK, Morens DM, Fauci AS (2007) The next influenza pandemic: can it be predicted? JAMA 297: 2025–2027. 12. Webster RG, Geraci J, Petursson G, Skirnisson K (1981) Conjunctivitis in human beings caused by influenza A virus of seals. N Engl J Med 304: 911. 13. Kurtz J, Manvell RJ, Banks J (1996) Avian influenza virus isolated from a woman with conjunctivitis. Lancet 348: 901–902. 14. Koopmans M, Wilbrink B, Conyn M, Natrop G, van der Nat H, et al. (2004) Transmission of H7N7 avian influenza A virus to human beings during a large outbreak in commercial poultry farms in the Netherlands. Lancet 363: 587– 593. 15. Fouchier RA, Schneeberger PM, Rozendaal FW, Broekman JM, Kemink SA, et al. (2004) Avian influenza A virus (H7N7) associated with human conjunctivitis and a fatal case of acute respiratory distress syndrome. Proc Natl Acad Sci U S A 101: 1356–1361. 16. CDC (2004) Update: influenza activity–United States and worldwide, 2003–04 season, and composition of the 2004–05 influenza vaccine. MMWR Morb Mortal Wkly Rep: Centers for Disease Control. pp 547–552. 17. Hirst M, Astell CR, Griffith M, Coughlin SM, Moksa M, et al. (2004) Novel avian influenza H7N3 strain outbreak, British Columbia. Emerg Infect Dis 10: 2192–2195. 18. Tweed SA, Skowronski DM, David ST, Larder A, Petric M, et al. (2004) Human illness from avian influenza H7N3, British Columbia. Emerg Infect Dis 10: 2196–2199. 19. EditorialTeam (2007) Avian influenza A/H7N2 outbreak in the United Kingdom. Euro Surveill 12: 2. 20. Suarez DL, Garcia M, Latimer J, Senne D, Perdue M (1999) Phylogenetic analysis of H7 avian influenza viruses isolated from the live bird markets of the Northeast United States. J Virol 73: 3567–3573. 21. Belser JA, Lu X, Maines TR, Smith C, Li Y, et al. (2007) Pathogenesis of avian influenza (H7) virus infection in mice and ferrets: enhanced virulence of Eurasian H7N7 viruses isolated from humans. J Virol 81: 11139–11147. 22. Belser JA, Blixt O, Chen LM, Pappas C, Maines TR, et al. (2008) Contemporary North American influenza H7 viruses possess human receptor specificity: Implications for virus transmissibility. Proc Natl Acad Sci U S A 105: 7558–7563. 23. Gambaryan AS, Tuzikov AB, Pazynina GV, Desheva JA, Bovin NV, et al. (2008) 6-sulfo sialyl Lewis X is the common receptor determinant recognized by H5, H6, H7 and H9 influenza viruses of terrestrial poultry. Virol J 5: 85. 24. Stevens J, Blixt O, Glaser L, Taubenberger JK, Palese P, et al. (2006) Glycan microarray analysis of the hemagglutinins from modern and pandemic influenza viruses reveals different receptor specificities. J Mol Biol 355: 1143–1155. 25. Stevens J, Corper AL, Basler CF, Taubenberger JK, Palese P, et al. (2004) Structure of the uncleaved human H1 hemagglutinin from the extinct 1918 influenza virus. Science 303: 1866–1870. 26. Russell RJ, Kerry PS, Stevens DJ, Steinhauer DA, Martin SR, et al. (2008) Structure of influenza hemagglutinin in complex with an inhibitor of membrane fusion. Proc Natl Acad Sci U S A 105: 17736–17741. 27. Russell RJ, Gamblin SJ, Haire LF, Stevens DJ, Xiao B, et al. (2004) H1 and H7 influenza haemagglutinin structures extend a structural classification of haemagglutinin subtypes. Virology 325: 287–296. 28. Matrosovich M, Tuzikov A, Bovin N, Gambaryan A, Klimov A, et al. (2000) Early alterations of the receptor-binding properties of H1, H2, and H3 avian influenza virus hemagglutinins after their introduction into mammals. J Virol 74: 8502–8512. 29. Nobusawa E, Ishihara H, Morishita T, Sato K, Nakajima K (2000) Change in receptor-binding specificity of recent human influenza A viruses (H3N2): a single amino acid change in hemagglutinin altered its recognition of sialyloligosacchar- ides. Virology 278: 587–596. 30. Stevens J, Blixt O, Chen LM, Donis RO, Paulson JC, et al. (2008) Recent avian H5N1 viruses exhibit increased propensity for acquiring human receptor specificity. J Mol Biol 381: 1382–1394. 31. Russell RJ, Stevens DJ, Haire LF, Gamblin SJ, Skehel JJ (2006) Avian and human receptor binding by hemagglutinins of influenza A viruses. Glycoconj J 23: 85–92. 32. Gamblin SJ, Haire LF, Russell RJ, Stevens DJ, Xiao B, et al. (2004) The structure and receptor binding properties of the 1918 influenza hemagglutinin. Science 303: 1838–1842. 33. Suarez DL, Spackman E, Senne DA (2003) Update on molecular epidemiology of H1, H5, and H7 influenza virus infections in poultry in North America. Avian Dis 47: 888–897. 34. Daniels PS, Jeffries S, Yates P, Schild GC, Rogers GN, et al. (1987) The receptor-binding and membrane-fusion properties of influenza virus variants Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 10 September 2010 | Volume 6 | Issue 9 | e1001081 selected using anti-haemagglutinin monoclonal antibodies. EMBO J 6: 1459–1465. 35. Rosenthal PB, Zhang X, Formanowski F, Fitz W, Wong CH, et al. (1998) Structure of the haemagglutinin-esterase-fusion glycoprotein of influenza C virus. Nature 396: 92–96. 36. Weinstein J, de Souza-e-Silva U, Paulson JC (1982) Purification of a Gal beta 1 to 4GlcNAc alpha 2 to 6 sialyltransferase and a Gal beta 1 to 3(4)GlcNAc alpha 2 to 3 sialyltransferase to homogeneity from rat liver. J Biol Chem 257: 13835–13844. 37. Gambaryan A, Webster R, Matrosovich M (2002) Differences between influenza virus receptors on target cells of duck and chicken. Arch Virol 147: 1197–1208. 38. Wan H, Perez DR (2006) Quail carry sialic acid receptors compatible with binding of avian and human influenza viruses. Virology 346: 278–286. 39. Guo CT, Takahashi N, Yagi H, Kato K, Takahashi T, et al. (2007) The quail and chicken intestine have sialyl-galactose sugar chains responsible for the binding of influenza A viruses to human type receptors. Glycobiology 17: 713–724. 40. Maines TR, Chen LM, Matsuoka Y, Chen H, Rowe T, et al. (2006) Lack of transmission of H5N1 avian-human reassortant influenza viruses in a ferret model. Proc Natl Acad Sci U S A 103: 12121–12126. 41. Wan H, Sorrell EM, Song H, Hossain MJ, Ramirez-Nieto G, et al. (2008) Replication and transmission of H9N2 influenza viruses in ferrets: evaluation of pandemic potential. PLoS One 3: e2923. 42. Weis WI, Brunger AT, Skehel JJ, Wiley DC (1990) Refinement of the influenza virus hemagglutinin by simulated annealing. J Mol Biol 212: 737–761. 43. Frank S, Kammerer RA, Mechling D, Schulthess T, Landwehr R, et al. (2001) Stabilization of short collagen-like triple helices by protein engineering. J Mol Biol 308: 1081–1089. 44. Stevens J, Blixt O, Tumpey TM, Taubenberger JK, Paulson JC, et al. (2006) Structure and receptor specificity of the hemagglutinin from an H5N1 influenza virus. Science 312: 404–410. 45. Chayen NE, Shaw-Steward PD, Blow DM (1992) Microbatch crystallization under oil – a new technique allowing many small volume crystallization experiments. J Cryst Growth 122: 176–180. 46. Otwinowski A, Minor W (1997) Processing of X-ray diffraction data collected in oscillation mode. Meothds in Enzymology 276: 307–326. 47. McCoy AJ, Grosse-Kunstleve RW, Storoni LC, Read RJ (2005) Likelihood- enhanced fast translation functions. Acta Crystallogr D Biol Crystallogr 61: 458–464. 48. Emsley P, Cowtan K (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132. 49. CCP4 (1994) The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 50: 760–763. 50. Winn MD, Isupov MN, Murshudov GN (2001) Use of TLS parameters to model anisotropic displacements in macromolecular refinement. Acta Crystallogr D Biol Crystallogr 57: 122–133. 51. Davis IW, Leaver-Fay A, Chen VB, Block JN, Kapral GJ, et al. (2007) MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res 35: W375–383. 52. Brunger AT (2007) Version 1.2 of the Crystallography and NMR system. Nat Protoc 2: 2728–2733. 53. Hoffmann E, Webster RG (2000) Unidirectional RNA polymerase I-polymerase II transcription system for the generation of influenza A virus from eight plasmids. J Gen Virol 81: 2843–2847. 54. Higuchi R, Krummel B, Saiki RK (1988) A general method of in vitro preparation and specific mutagenesis of DNA fragments: study of protein and DNA interactions. Nucleic Acids Res 16: 7351–7367. 55. Blixt O, Head S, Mondala T, Scanlan C, Huflejt ME, et al. (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc Natl Acad Sci U S A 101: 17033–17038. 56. DeLano WL (2002) The PyMol Molecular Graphics Systems. wwwpymolorg. Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 11 September 2010 | Volume 6 | Issue 9 | e1001081
3M5J
Crystal structure of a H7 influenza virus hemagglutinin complexed with LSTb
Structures of Receptor Complexes of a North American H7N2 Influenza Hemagglutinin with a Loop Deletion in the Receptor Binding Site Hua Yang, Li-Mei Chen, Paul J. Carney, Ruben O. Donis, James Stevens* Influenza Division, Centers for Disease Control and Prevention, Atlanta, Georgia, United States of America Abstract Human infections with subtype H7 avian influenza viruses have been reported as early as 1979. In 1996, a genetically stable 24-nucleotide deletion emerged in North American H7 influenza virus hemagglutinins, resulting in an eight amino acid deletion in the receptor-binding site. The continuous circulation of these viruses in live bird markets, as well as its documented ability to infect humans, raises the question of how these viruses achieve structural stability and functionality. Here we report a detailed molecular analysis of the receptor binding site of the North American lineage subtype H7N2 virus A/New York/107/2003 (NY107), including complexes with an avian receptor analog (39-sialyl-N-acetyllactosamine, 39SLN) and two human receptor analogs (69-sialyl-N-acetyllactosamine, 69SLN; sialyllacto-N-tetraose b, LSTb). Structural results suggest a novel mechanism by which residues Arg220 and Arg229 (H3 numbering) are used to compensate for the deletion of the 220-loop and form interactions with the receptor analogs. Glycan microarray results reveal that NY107 maintains an avian-type (a2-3) receptor binding profile, with only moderate binding to human-type (a2-6) receptor. Thus despite its dramatically altered receptor binding site, this HA maintains functionality and confirms a need for continued influenza virus surveillance of avian and other animal reservoirs to define their zoonotic potential. Citation: Yang H, Chen L-M, Carney PJ, Donis RO, Stevens J (2010) Structures of Receptor Complexes of a North American H7N2 Influenza Hemagglutinin with a Loop Deletion in the Receptor Binding Site. PLoS Pathog 6(9): e1001081. doi:10.1371/journal.ppat.1001081 Editor: Fe´lix A. Rey, Institut Pasteur, France Received March 16, 2010; Accepted July 28, 2010; Published September 2, 2010 This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. Funding: This work was funded by the Centers for Disease Control and Prevention. Use of the Advanced Photon Source at Argonne National Laboratory was supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC02-06CH11357. Glycan microarrays as well as glycans for direct binding experiments were produced for the Centers for Disease Control by the CFG funded by National Institute of General Medical Sciences Grant GM62116. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: fwb4@cdc.gov Introduction Influenza is an acute respiratory virus that infects up to 20% of the population in the United States, resulting in ,36,000 deaths annually [1,2]. The two membrane glycoproteins on the surface of influenza A virus, hemagglutinin (HA), which functions as the receptor binding and membrane fusion glycoprotein in cell entry, and neuraminidase (NA), which functions as the receptor destroying enzyme in virus release, form the basis for defining subtypes [3]. To date, 16 HA (H1–H16) and 9 NA (N1–N9) have been identified in avian species [4], while in the last century, only three subtypes, H1N1 in 1918 and 2009, H2N2 in 1957, and H3N2 in 1968 [5,6,7], have successfully adapted to humans. Hemagglutinin binds to sialic acid (SA) glycans present on host cell surfaces. The receptors on epithelial cells of the human upper respiratory tract are mainly a2-6-linked SA moieties [8]. Since avian influenza viruses pre- dominately bind a2-3-linked SA, and human influenza viruses preferentially bind to a2-6-linked SA, human infection by avian influenza viruses is rare [9]. However, since 1997 a growing number of human cases of avian influenza infection have been reported [10], including H5N1, H7N2, H7N3, H7N7, and H9N2 strains [11]. Although the current situation with the pandemic H1N1 influenza virus dominates public health efforts, the prospect of a novel pandemic emerging from these isolated cases continues to be a major public health threat around the world. Early cases of human infection by H7 influenza viruses are reported as far back as 1979 [12,13]. Since 2002, multiple outbreaks and human infections of H7 subtype viruses; within both Eurasian and North American lineages have been reported. In the Netherlands in 2003, a highly pathogenic avian influenza (HPAI) H7N7 outbreak resulted in more than 80 cases of human infections, including one fatality [14,15]. In New York in 2003, a single case of human respiratory infection of H7N2 was reported [16] and in British Columbia in 2004, an H7N3 virus caused two cases of conjunctivitis [17,18]. More recently in 2007, the United Kingdom reported several cases of low pathogenic avian influenza (LPAI) H7N2 virus infections that caused influenza-like illness and conjunctivitis [19]. Since 1996, H7 viruses of the North American lineage have been circulating in regional live bird markets [20], containing a 24-nucleotide deletion resulting in an eight amino acid deletion in the receptor-binding site (RBS) of HA (Figure S1). The recent human infections with H7 in North America have raised public health concerns as to how these viruses adapt to such a dramatic structural change while remaining one of the predominant circulating viral strains. A recent study of H7 viruses isolated from previous outbreaks revealed efficient replication in both mouse and ferret animal models [21]. In particular, ferret studies with A/New York/107/2003 (NY107), an H7N2 virus isolated from a man in New York, not only showed efficient replication in the upper respiratory tract of the ferret but also the capacity for PLoS Pathogens | www.plospathogens.org 1 September 2010 | Volume 6 | Issue 9 | e1001081 intra-species transmission by direct contact [21,22]. Interestingly, both an increased preference for a2-6 and decreased preference for a2-3-linked sialosides of this virus compared to the other avian influenza viruses was shown by previous glycan microarray analysis but less so by a competitive solid-phase binding assay [22,23]. Here we report a detailed molecular analysis of the RBS of the HA from North American lineage H7N2 virus, NY107, including glycan microarray analyses and structural analyses of the HA in complex with an avian receptor analog (39-Sialyl-N-acetyllactosa- mine, 39SLN) and two human receptor analogs (69-Sialyl-N- acetyllactosamine, 69SLN; Sialyllacto-N-tetraose b, LSTb). These results provide important insight into the interaction of H7 HAs with both avian and human hosts. Results Overall structure By using x-ray crystallography, the structure of H7 HA from the NY107 virus was determined to 2.6 A˚ resolution (Table 1). In addition, we also report three H7 HA receptor complex structures, with avian receptor analog (39SLN) to 2.7 A˚ resolution and with human receptor analogs (69SLN and LSTb) to 3.0 A˚ and 2.6 A˚ resolution, respectively (Table 1). The overall structure of NY107 is similar to other reported HA structures with a globular head containing the RBS and vestigial esterase domain, and a membrane proximal domain with its distinctive, central helical stalk and HA1/HA2 cleavage site (Figure 1A). Although five asparagine-linked glycosylation sites are predicted in the NY107 HA monomer, interpretable electron density was observed at only two sites, Asn38 in HA1 and Asn82 in HA2 (all residue numbers are based on H3 numbering). At these sites, only one or two N- acetyl glucosamines could be interpreted. During viral replication, HA is synthesized as a single chain precursor (HA0) and cleaved by specific host proteases into the infectious HA1/HA2 form. In baculovirus expression systems, highly pathogenic HAs, with a polybasic cleavage site, are expressed as an HA1/HA2 form [24], whereas HAs with monobasic cleavage sites (single Arg) from low pathogenic viruses are expressed as the HA0 form [25]. NY107 is regarded as a low pathogenic virus, and as expected, was produced in the HA0 form (Figure S2). However, subsequent digestion with thrombin protease to remove the His-tag resulted in cleavage to a profile on SDS-PAGE comparable to that of an HA1/HA2 form (Figure S2). A comparison of the NY107 cleavage site with the consensus cleavage pattern in the MEROPS database (http://merops. sanger.ac.uk) suggests it to be a possible thrombin cleavage site. Based on their molecular phylogenies, HAs are divided into two groups and five clades: group 1 includes H8, H9, and H12; H1, H2, H5, and H6; H11, H13 and H16; group 2 includes H3, H4, and H14; H7, H10 and H15 [26]. Among all available HA structures, we selected ten representative HAs from both avian and human subtypes for structural analysis. As expected, NY107 HA is structurally very similar to the Avian-H7 in all comparisons and closely related to H3, the other group 2 members used in the analyses (Tables S1 and S2). The receptor binding site The RBS is at the membrane distal end of each HA monomer and its specificity for sialic acid and the nature of its linkage to a vicinal galactose residue is a major determinant of host range- restriction. The consensus RBS for all current HAs is composed of three major structural elements: a 190-helix (residues 188–194), a 220-loop (residues 221–228), and a 130-loop (residues 134–138). In addition, highly conserved residues (Tyr98, Trp153, His183, and Tyr195) form the base of the pocket. Although the NY107 RBS is similar to other subtypes (H1, H2, H3, H5, and H9), a previously observed specific feature of H7 HAs, is also observed in the NY107 150-loop region: two residues inserted at position 158 result in this loop protruding more than 6A˚ towards the binding site compared to other subtype HAs (Figure 1B and Table S2) [27]. More interestingly, the eight amino acid deletion, only found in the North American lineage H7s, from position 221 to 228 (Figure S1), resulted in a complete loss of the 220-loop (Figure 1B). Sequence alignment shows that Arg220 and Arg229 are conserved in all influenza A HA subtypes (Figure S1), but structural alignment of NY107 HA shows Arg220 occupying the Gly228 position, and the much shorter loop turns at residue Pro217 (Figure 1C). The Ca distance between NY107 Arg220 and its homolog in the Av-H7 structure (PDB: 1TI8) [27] is 5.8A˚ , and they point in opposite directions (Figure 1C). The side chain direction of Av-H7 Arg220 is almost parallel with the beta sheet after Arg229, whereas the NY107 Arg220 points downward to the binding pocket. The Ca position of Arg229 in both H7 structures remains the same, except the side chain in the NY107 swings away by about 5.9A˚ (Figure 1C) and could help to stabilize this region by forming a hydrogen bond to the mainchain carbonyl of Gln210 in the neighboring monomer. In the absence of the 220-loop in NY107 HA, upon glycan binding the long side chain of Arg220 compensates for its loss and is displaced 4A˚ upward to form hydrogen bonds with receptor analogs inside the binding pocket (Figure 1D). Effect of loop truncation on the receptor binding specificity of NY107 Previously, mutations in the HA receptor binding domains of H1N1 (Glu190Asp/Gly225Asp) and H2N2/H3N2 (Gln226Leu and Gly228Ser) subtypes were responsible for adaptation of these viruses to pandemic strains [24,28,29,30]. Due to missing residues 221–228 in the NY107 HA RBS, neither mechanism for adaptation is possible. Thus, in order to look more closely at the role of the missing loop and its effect on receptor specificity, we first subjected the recombinant HA (recHA) to glycan microarray analyses and compared it to a reverse genetics-derived NY107 virus, and a co-circulating Eurasian virus and recHA, A/ Author Summary Influenza virus adaptation to different hosts usually results in a switch in receptor specificity of the viral surface coat protein, hemagglutinin. Indeed, the hemagglutinin sub- types from the last two human influenza pandemics of the 20th Century (H2 in 1957 and H3 1968) both adapted successfully to human-type receptor specificity through only two amino acid mutations in the receptor binding pocket (Glutamine226RLeucine and Glycine228RSerine). The recent human infections reported with other avian subtypes such as H5, H7 and H9 have raised public health concerns and focused efforts on identifying potential subtypes from which a future pandemic strain may emerge. Since 1996, H7 viruses of the North American lineage have been circulating in regional live bird markets, containing an eight amino acid deletion in the receptor- binding site of HA. Here we report a detailed structural analysis of the receptor binding site of a hemagglutinin from the North American lineage of H7N2 viruses, in complex with avian and human receptor analogs, to understand how these viruses have adapted to such a dramatic structural change in the binding site while remaining one of the predominant circulating viral strains. Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 2 September 2010 | Volume 6 | Issue 9 | e1001081 Netherlands/219/2003 (NL219), that has the consensus avian sequence in the 220-loop and it also infected a human [15]. Glycan microarray analysis of recombinant NY107 (Figure 2A and Table 2) revealed a highly restricted binding profile with strong binding to only a2-3 sulfated (#4–8), a2-3 branched (#9– 11) and mixed a2-3/a2-6 branched sialosides (#60–64) as well as to the long linear sialyl di- and tri-lactosamines (#22, 24). Weak binding was also observed (above background) to other a2-3 glycans on the array. The recombinant NY107 also revealed a strict glycan binding preference to only one a2-6 glycan, the internal structure, Galb1-3(Neu5Aca2-6)GlcNAcb1-3Galb1-4Glc (#58; LSTb) (Figure 2A), a glycan highlighted in a previous study [22]. The virus with higher valency and avidity revealed stronger binding to all a2-3 groups, in addition to the branched di-sialyl a2- 6 biantennary structures (#46–48) as well the LSTb (#58) (Figure 2B and Table 2). In contrast, the NL219 recHA (Figure 2C and Table 2) bound well to only the avian a2-3 containing sialyl- glycans (sulfated, branched, linear and fucosylated). Its corre- sponding virus also reflected this specificity although it also revealed strong binding to a2-3 N-glycolylneuraminic acid (Neu5Gc) containing glycans (#66–70) (Figure 2D and Table 2). To further assess the effect of the missing 220 loop on HA structural stability and receptor specificity it was essential to evaluate these functions on the ancestral HA containing the full length 220-loop. To this end, we engineered an HA with an avian H7 consensus (PQVNGQSG) 220-loop re-introduced (NY107- 220ins) into the NY107 HA and recovered this virus by reverse genetics. Compared to the NY107 virus (Figure 2A) glycan microarray analyses of the resulting NY107-220ins virus (Figure 3A and Table 2) revealed a decrease in binding to branched (#9–11) and linear (#12–27) a2-3 sialosides and a loss of binding to the branched di-sialyl a2-6 biantennary structures (#46–48), LSTb (#58) as well as the mixed a2-3/a2-6 branched sialosides (#60– 64). In addition, sequence analysis of the NY107-220ins HA revealed the presence of quasispecies in the second position of the inserted loop, P(Q/K)VNGQSG, suggesting that re-introduction of the loop alone is not tolerated and does not create an avian-type binding profile. Thus other amino acid substitutions in the HA might have co-evolved with the deletion of the 220 loop to help stabilize the RBS/HA to maintain functionality. When viruses containing this 220-loop deletion emerged in North America in the mid 90’s, four additional amino acid substitutions, Gly114Arg, Asp119Gly, Gly186Glu and Gly205Arg, in the HA1 as well as an Asp19Asn in the HA2 chain were also introduced to most of the circulating isolates. Of these, Gly186Glu and Gly205Arg in the HA1 are close to the RBS, at the monomer interface, and could potentially modulate its structure and/or function. NY107 viruses with a restored consensus 220-loop and a single Glu186Gly (NY107- ins-186) or Arg205Gly (NY107-ins-205) substitution as well as the Glu186Gly/Arg205Gly double substitution (NY107-ins-186/205) were derived by reverse genetics and evaluated. Glycan microarray analysis for the three resulting viruses revealed similar glycan binding profiles with increased binding to a2-3 sialosides, including mixed a2-3/a2-6 branched sialosides (#60–64), a2-3 Neu5Gc (#66–70), but limited binding to the a2,6 sialosides (Figures 3B, 3C, 3D), resulting in a binding profile virtually identical to that of the NL219 virus and other avian influenza viruses (Figure 2D) [30]. Sequence analysis of the three reverse genetics derived viruses revealed no mutations/quasispecies in the HAs of either the NY107- ins-186 or the NY107-ins-186/205 virus stocks, indicative of replication fitness. For the NY107-ins-205 virus however, a Glu186Gly substitution emerged in the HA after only two passages in eggs following recovery from DNA transfection, indicating the importance of the co-variant position 186 with respect to HA functionality/glycan specificity. Altogether, the data indicates that the H7 subtype avian influenza viruses that were circulating in Table 1. Data collection and refinement statistics. NY107 NY107+39SLN NY107+69SLN NY107+LSTb Data collection Space group P212121 P212121 P212121 P212121 Cell dimensions (A˚) 66.96, 115.92, 251.61 67.80, 116.70, 249.84 66.60, 116.58, 250.68 67.08, 116.52, 251.95 Resolution (A˚) 50-2.6 (2.69-2.60)a 30-2.7 (2.80-2.70) 50-3.0 (3.11-3.0) 50-2.6 (2.69-2.60) Rsym or Rmerge 10.6 (41.3) 14.6 (48.6) 14.3 (35.4) 12.2 (31.5) I/s 39.6 (2.0) 24.3 (1.7) 34.2 (8.2) 40.5 (9.9) Completeness (%) 99.2 (99.0) 99.3 (94.6) 92.3 (75.6) 91.3 (86.2) Redundancy 7.2 (6.2) 5.8 (5.5) 4.9 (4.4) 10.9 (11.2) Refinement Resolution (A˚) 50-2.6 (2.67-2.60) 30-2.7 (2.77-2.70) 50-3.0 (3.08-3.00) 50-2.6 (2.67-2.60) No. of reflections (total) 57285 51770 33421 53603 No. of reflections (test) 3053 2769 1779 2842 Rwork/Rfree 21.7/25.6 21.4/26.4 20.5/26.0 20.4/24.7 No. of atoms 11795 11878 11648 12108 r.m.s.d.- bond length (A˚) 0.006 0.006 0.008 0.006 r.m.s.d.- bond angle (u) 0.905 0.974 1.085 0.859 MolProbityb scores Favored (%) 96.9 96.5 94.3 97.1 Outliers (%) (No. of residues) 0.1 (1/1434) 0.0 (0/1429) 0.1 (2/1433) 0.1 (2/1435) aNumbers in parentheses refer to the highest resolution shell. bReference [51]. doi:10.1371/journal.ppat.1001081.t001 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 3 September 2010 | Volume 6 | Issue 9 | e1001081 aquatic birds and poultry in North America before 1996 exhibited a classic avian a2-3 sialoside binding preference. In order for the 220- loop deletion to be tolerated, concurrent Gly186Glu and Gly205Arg substitutions in the vicinity of RBS of HA emerged to achieve a restricted a2-3 binding profile and only a moderate/limited increase in binding to branched di-sialyl a2-6 biantennary structures (#46– 48) as well the a2,6 internal sialoside, LSTb (#58). NY107 avian receptor complex To understand from a structural perspective how NY107 interacts with host receptors, we solved the structure of NY107 in complex with an avian and two human receptor analogs. For the avian receptor analog, 39SLN, the electron density maps revealed well-ordered features for the Sia-1, Gal-2, and GlcNAc-3 in the NY107 HA complex structure (Figure 4A). Structural comparison of NY107 HA binding to other, H1, H2, H3, H5, and H9 subtypes (Figure S2A) revealed that 39SLN binding to NY107 resembled binding of the other published HAs. Indeed, the terminal Sia-1 moiety is positioned almost identically in all structures, and forms the majority of hydrogen bonds and contacts with residues in the RBS (Figure 4A and Table S3). Figure 1. NY107 HA monomer and comparison of its RBS to other HA structures. (A) One monomer is shown with the HA1 chain colored in green and the HA2 chain in cyan. The location of the receptor binding site and the HA1/HA2 cleavage site are circled. (B) The superposition of receptor binding domains of NY107 (green), Av-H7 (marine), 1918-Hu-H1 (magenta), Hu-H5 (yellow), Hu-H3 (orange), and Sw-H9 (grey). The proximity of Arg220 and Gln226 are highlighted. Three structural elements comprising this binding site are labeled. The two major differences are the extended 150-loop and the deletion of 220-loop of NY107. (C) Overlap of NY107 (green) and Av-H7 (marine) (PDB: 1TI8) illustrates the compensatory effect of R220 bringing it close to the position occupied by G228 in the avian HA. (D) Overlap of the NY107 (green), NY107- 39SLN (orange), NY107-69SLN (red), and NY107-LSTb (magenta) structures. All the figures were generated and rendered with the use of MacPyMOL [56]. doi:10.1371/journal.ppat.1001081.g001 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 4 September 2010 | Volume 6 | Issue 9 | e1001081 Published avian HA structures with an intact 220-loop form very close interactions with Gal-2 of 39SLN via residue Gln226 which is important in receptor specificity and host adaptation. For example, in the avian H7/39SLN HA structure it interacts with Gal-2 O4 [31]. In the NY107 HA structure, although Gln226 is absent and no other residue occupies the same space as Gln226 (Figure 1B), Arg220 does forms a hydrogen bond between Arg220 NH2 and Gal-2 O4 (Figure 4A). Interestingly, although there was interpretable density for the GlcNAc-3 (Figure 4A and Figure S4B), no hydrogen bonding was apparent between the HA and the GlcNAc-3, which is consistent with other reported structures [32]. Thus, for binding to avian receptors, the trans conformation of a2-3 linkages is essential and perhaps only the first two saccharides are required. Indeed, due to the absence of 220-loop in the NY107 HA structure, the ‘‘aperture’’ of the RBS formed by 220-loop and 130-loop in regular HAs is increased by ,10 A˚ , so that the branched, internal, and perhaps more complicated glycans might be accommodated more efficiently. NY107 human receptor complexes In the NY107/69SLN complex, only Sia-1 and Gal-2 are ordered (Figure 4B). The Sia-1 remains in the same position as previously analyzed glycan/HA complexes from H1, H2, H3, H5, and H9 (Figure S3B), whereas the Av-H7 complex structure with Sialyllacto-N-tetraose c (LSTc) did not reveal any density for the Sia-1 in the receptor binding site [31]. The Gal-2 position varies significantly among different subtypes. Compared to the human- adapted H1 HA [32], Gal-2 in the NY107 HA is 3A˚ higher, and thus is further from the protein (Figure S3B). In NY107, the Gal-2 only forms an intramolecular, saccharide-saccharide interaction with Sia-1. The poor electron density map and fewer interactions with protein residues suggest that the cis conformation of a2-6 Figure 2. Receptor specificity of NY107 recHA and virus. Glycan microarray analysis of recombinant NY107 HA (A) and NY107 virus (B) compared to the recHA (C) and virus (D) from a Eurasian lineage A/Netherlands/219/2003 H7 influenza virus that was circulating in the same year and also infected a human. Colored bars highlight glycans that contain a2-3 SA (blue) and a2-6 SA (red), a2-6/a2-3 mixed SA (purple), N-glycolyl SA (green), a2-8 SA (brown), b2-6 and 9-O-acetyl SA, and non-SA (grey). Error bars reflect the standard error in the signal for six independent replicates on the array. Structures of each of the numbered glycans are found in Table S4. doi:10.1371/journal.ppat.1001081.g002 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 5 September 2010 | Volume 6 | Issue 9 | e1001081 linkages in 69SLN trisaccharides show a reduced binding affinity with NY107. Glycan array results with NY107 revealed a strong binding signal for the internal a2-6 sialoside, LSTb. To further investigate this interaction, we solved the structure of the NY107/LSTb complex. The final model contained Sia-1, NAG-2, Gal-3, and Gal-5 in the RBS. Although glycan microarray data indicated NY107 to have a specific affinity for LSTb, few interactions were apparent from the crystal structure. Sia-1 still forms multiple hydrogen bonds with residues in the RBS (Table S3 & Figure 4C). The branched Gal-5 interacts with Ser137, to help stabilize the LSTb binding. However, Arg220 and Lys193, the two residues showing close binding with 39SLN, did not form any hydrogen bonds with LSTb. In the structure, Gal-5 also interacts with a crystal packing symmetry mate and thus the flexibility of whole LSTb may be restricted. In solution, with more freedom, the LSTb should be able to tilt closer to the RBS, and thus Glc-4 may have more interactions with the 190-helix than seen in the crystal structure. Discussion Human infections by avian influenza viruses, including H7 subtypes, continue to pose a major public health threat. Although the species barrier prevents avian influenza viruses from widespread infection of the human population, the molecular determinants of efficient interspecies transmission and pathoge- nicity are still poorly understood. The viral coat protein HA however, is perhaps a critical molecule since previous pandemic viruses modified their receptor specificity and overcame the interspecies barrier to spread in the human population. Although HA structures alone and in complex with receptor analogs provide considerable insight into receptor binding, it is clear that HAs from different species and subtypes have significant structural variation. Indeed, low-pathogenic H7N2 avian influenza viruses with an 8 amino acid deletion within its RBS started to circulate in live-bird markets in the northeast United States in 1996. Despite what one would consider a debilitating mutation, these viruses have been reported as the predominant isolate [33]. Whether such a deletion contributed to their evolutionary success and how are an important questions, especially in light of NY107’s ability to produce respiratory illness in humans [16], as well as its reported increased affinity for human-type receptors and ability for contact transmission in ferrets [21]. To try to help answer these questions, we have analyzed the molecular structures of NY107 and its complexes with receptor analogs to explain receptor specificity at the molecular level. The crystal structures of NY107 and its complexes with both avian and human receptor analogs describe a mechanism as to how an influenza virus might adapt by dramatically altering its RBS, and still be functional. Arg220 of the HA1 chain of NY107 compensates for the loss of the 220-loop, by forming hydrogen bonds with Gal-2 from the avian analog (binding was not observed in either of the structures complexes with the human analogs). However, in the LSTb complex, branched Gal-5 forms extra interactions with the 130-loop, thus improving the binding preference for this particular glycan. Consistent with the structural evidence, glycan microarray analyses of NY107 revealed a strong binding preference for the branched a2-6 sialoside, LSTb. Except for the absence of the 220-loop, other key residues within the RBS are conserved in NY107 and thus, direct interactions with sialic acid are maintained. The 220-loop is recognized as one of the three crucial structural elements in the RBS. Aside from the North American lineage H7N2 viruses, which have been circulating with a deletion (221– 228) in this loop, there has been one other report describing a seven amino acid deletion (224–230) in a laboratory generated H3N2 escape mutant which was reported to have a slightly Table 2. Comparison of the sialoside receptor specificity of the HAs from H7 influenza viruses. Glycan Group Graph Numbera NY107 RecHA NY107 Virus NY107-ins Virus NY107-ins E186G Virus NY107-ins R205G Virus NY107-ins E186G/ R205G Virus NL219 RecHA NL219 Virus a2-3 Sulfated 4–8 +++b +++ +++ +++ +++ +++ +++ +++ Branched 9–11 +++ +++ + +++ +++ +++ +++ +++ Linear 12–27 + +++ + +++ +++ +++ +++ +++ Fucosylated 28–34 2 +++ +++ +++ +++ +++ +++ +++ a2-6 Sulfated 41 2 2 2 2 2 2 2 2 Branched mono-sialyl 42–45, 49 2 2 2 2 2 2 2 2 Branched di-sialyl 46–48 2 +++ 2 2 2 2 2 2 Linear 50–56 2 2 2 2 2 2 2 2 Internal 58–59 +++ +++ 2 2 2 2 2 2 Other Sialic acid 1–2 2 +++ + 2 2 2 2 2 a2-3/a2-6 Branched 60–64 2 2 2 +++ +++ +++ +++ +++ Neu5Gcc 65–72 2 2 2 +++ +++ +++ 2 +++ aMembers of each group are identified according to the graph number used in the microarray data in Figures 2 and 3 and correspond to numbers in the complete glycan list (Table S4). bBinding of samples to glycan subclasses are qualitatively estimated based on relative strength of the signal for the data shown in Figures 2 and 3: strong (+++), weak (+), absent (2). cN-glycolylneuraminic acid. doi:10.1371/journal.ppat.1001081.t002 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 6 September 2010 | Volume 6 | Issue 9 | e1001081 increased affinity for a2-3-linked glycans by hemagglutination assay [34]. Meanwhile, the equivalent region in the hemaggluti- nin-esterase-fusion (HEF) protein of influenza C virus reveals a rearrangement resulting in a truncated 260-loop in its RBS (Figure S5) [35]. However, without structural data with appropriate receptor analogs, it is not possible to compare the role of these loop variants in receptor binding to the H7 HA structure described here. When compared to NL219, another co-circulating H7 avian virus HA (Figure 2C and D), overall binding to a2-3-linked glycans was markedly reduced, while increased binding to a2-6- linked receptors was only marginal. However, these results focus attention on only 2 sub-classes of human-type receptors that may be important for infection (and transmission in ferrets). The NY107 virus interaction with biantennary glycans (Figure 2B), although weak (not seen in Figure 2A with recHA), is a possible route for virus entry as biantennary structures are common on tissues, i.e. glycan profiling data from human lung tissue on the Consortium for Functional Glycomics (CFG) web site. In addition, the internal sialoside, LSTb, was observed in both virus and recHA microarray data, suggesting this type of glycan has good affinity for this HA. The significance of this is unknown since LSTb has only been described in human milk [36]. Interestingly, NY107 and NL219 virus receptor binding and specificity has been addressed previously using glycan microarray analysis that reported a significantly increased preference for a2-6 and decreased preference for a2-3-linked sialosides [22]. In addition, the same viruses were also included in a recent study from Gambaryan et al. using a competitive solid-phase binding assay [23]. Our findings confirm and extend the receptor binding specificity reported by these authors in that they reported both viruses binding to sulfated sialylglycans with a lactosamine (Galb1- 4GlcNAc core and reported only a moderate binding affinity for a2-6-sialyllactosamine, the human-type receptor analog used in their assay. The 220-loop is an integral feature of the receptor binding site, and thus one would predict that such a deletion might have compromised this strain to be deleted from the population of circulating viruses. However, this was not the case [33] and its existence appears to be in part due to the additional mutations at Figure 3. Effect of 220-loop deletion and additional RBS mutations on NY107 receptor specificity. NY107 was engineered to restore the 220-loop to a consensus full-length HA from 1996 (A) and additional co-variant amino acid substitutions, Glu186Gly (B), Arg205Gly (C) and the double mutant Glu186Gly/Arg205Gly (D) to restore, on the NY107 framework, an HA RBS found in viruses prior to the introduction of the deletion in North American viruses. Colored bars group glycans as described in Figure 3. Error bars reflect the standard error in the signal for six independent replicates on the array. Structures of each of the numbered glycans are found in Table S4. doi:10.1371/journal.ppat.1001081.g003 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 7 September 2010 | Volume 6 | Issue 9 | e1001081 positions 186 and 205. Restoration of the loop with either or both residues mutated back to the pre-1994 consensus sequence resulted in a classic avian influenza virus binding profile. The emergence of the Glu186Gly mutation in the HA of the NY107- ins-205 mutant after only two passages of the rescued virus in eggs, also indicates the importance of these positions for HA functionality/glycan specificity. Analysis of the structural data reveals that positions 186 and 205 are on opposite sides of a monomer but are both close to the 220-loop deletion region in the trimeric form. The Glu at position 186 is close to Arg220 and may interact with Arg220 when binding avian receptors. Position 205 in the neighboring monomer may be important in trimer stability and maintaining RBS functionality. If one models the pre- 1996 220-loop restored into the NY107 structure, Arg205, Glu186 and the loop all clash, thus explaining the Glu186Gly mutation that emerged in the NY107-ins-205 virus HA after limited egg passage. The NY107 RBS with its more restricted a2-3 glycan binding preference and weak/moderate increase in a2-6 binding may have given the virus a selective advantage to be maintained in poultry at live bird markets and supplying farms. Certain terrestrial birds, such as quails and chickens, have recently been shown to present both human and avian types of receptors in the trachea and intestine [37,38,39]. Although it is not known what specific glycans are presented in these animals, it is conceivable that a virus with mixed specificity might have a distinct advantage over avian viruses that have specific avian receptor requirements, particularly in bird markets where multiple species coalesce. Previous results with H7N2, H9N2 and H5N1 viruses all highlight the fact that an increase in a2-6-binding preference is not sufficient for efficient transmission of avian influenza viruses to humans [22,40,41]. Although it remains to be seen whether prolonged circulation of viruses in terrestrial birds, such as domestic chickens, can provide a possible route for viruses to adapt for efficient human infection [11], continued surveillance of influenza viruses from avian and other animal reservoirs is urgently needed to define their zoonotic potential. Materials and Methods Cloning Based on H3 numbering [42], cDNA corresponding to residues 11–329 (HA1) and 1–176 (HA2) of the ectodomain of the hemagglutinin (HA) from A/New York/107/2003 (H7N2; Genbank:ACC55270) and A/Netherlands/219/2003 (H7N7; Genebank: AAR02640) was cloned into the baculovirus transfer vector, pAcGP67-A (BD Biosciences), incorporating a C-terminal thrombin cleavage site, a ‘‘foldon’’ sequence [43] and a His-tag at the extreme C-terminus of the construct to enable protein purification [25,44]. Transfection and virus amplification were carried out according to the baculovirus expression system manual (BD Biosciences Pharmingen). Figure 4. Glycan interactions within the NY107 RBS. The top panel shows the interactions of NY107 with (A) 39SLN, (B) 69SLN and (C) LSTb. NY107 is shown in orange/red/magenta cartoon respectively. The interacting HA residues are shown as green sticks. The bottom panel shows the electron density map of the ligands. The NY107 is shown in the same colors as above, and the ligands are shown as green sticks, the 2fo-fc electron density maps (contoured at 1s) are shown in grey. Simulated annealing omit maps are shown in supplementary Figure S4. doi:10.1371/journal.ppat.1001081.g004 Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 8 September 2010 | Volume 6 | Issue 9 | e1001081 Protein expression and purification Soluble NY107 was recovered from the cell supernatant by metal affinity chromatography using Ni-NTA resin (Qiagen Inc.). Frac- tions containing NY107 were pooled and dialyzed against 10 mM Tris-HCl, 50 mM NaCl, pH 8.0, then subjected to ion-exchange chromatography (IEX) using a Mono-Q HR 10/10 column (GE Healthcare). IEX purified NY107 was subjected to thrombin digest (3 units/mg protein; overnight at 4uC) and purified by gel filtra- tion chromatography using a Superdex-200 16/60 column (GE Healthcare) and 50 mM Tris-HCl, 100 mM NaCl, pH 8.0 as running buffer. Protein eluting as a trimer was buffer exchanged into 10 mM Tris-HCl, 50 mM NaCl, pH 8.0 and concentrated to 14.5 mg/ml for crystallization trials. At this stage, the protein sample still contained the additional plasmid-encoded residues at both the N (ADPG) and C terminus (SGRLVPR). Crystallization, ligand soaking and data collection Initial crystallization trials were set up using a Topaz Free Interface Diffusion (FID) Crystallizer system (Fluidigm Corpora- tion, San Francisco, CA). Crystals were observed in several conditions containing PEG 3350 or PEG 4000. Following opti- mization, diffraction quality crystals for NY107 were obtained at room temperature using a modified method for microbath under oil [45], by mixing the protein with reservoir solution containing 20% PEG 3350, 0.2 M magnesium chloride at pH 7.2. For receptor analog complexes, crystals were soaked for 3 hours in the crystallization buffer containing 10 mM 39SLN or 69SLN (V-labs Inc., Covington, LA), or overnight in 10mM LSTb (Sigma, St. Louis, MO). All crystals were flash-cooled at 100K using 20% glycerol as the cryo-protectant. Datasets were collected at Advanced Photon Source (APS) beamlines 22 ID and BM at 100K. Data were processed with the DENZO-SACLEPACK suite [46]. Statistics for data collection are presented in Table 1. Structure determination and refinement The structure of NY107 was determined by molecular replacement with Phaser [47] using the structure of the avian H7 (Av-H7) from A/turkey/Italy/2002, pdb:1TI8 (HA1, 78% identity; HA2, 90% identity) as the searching model. One HA trimer occupies the asymmetric unit with an estimated solvent content of 58% based on a Matthews’ coefficient (Vm) of 2.9 A˚ 3/ Da. Rigid body refinement of the trimer led to an overall R/Rfree of 28.6%/37.4%. The model was then ‘‘mutated’’ to the correct sequence and rebuilt by Coot [48], then the protein structures were refined with REFMAC [49] using TLS refinement [50]. The final models were assessed using MolProbity [51]. The three complex structures were refined and evaluated using the same strategy. All statistics for data processing and refinement are presented in Table 1. Electron density maps (2fo-fc) were generated in Refmac [49] while simulated annealing omit maps were generated by sa-omit-map, a part of the Crystallography and NMR System (CNS) software [52]. Virus generation Wild type and mutant viruses of NY107 (H7N2) and A/ Netherland/219/2003 (H7N7) were generated from plasmids by a reverse genetics approach [53]. To generate viruses with amino acid insertion or substitution in the HA, mutations were introduced into plasmid DNA with an overlap extension PCR approach [54]. Viruses derived by plasmid transfection of HK293 cells were propagated in eggs. The genomes of resulting virus stocks were sequenced to detect the emergence of possible variants during amplification. Glycan binding analyses Glycan microarray printing and recHA analyses have been described previously [24,30,44,55] (see Table 2 for glycans used for analyses in these experiments). Virus were analyzed on the microarray as described previously [30]. PDB accession codes The atomic coordinates and structure factors of NY107 are available from the RCSB PDB under accession codes 3M5G for the unliganded NY107, 3M5H for the NY107 with 39-SLN and 3M5I and 3M5J for NY107 with 69SLN and LSTb, respectively. Accession/ID numbers for genes/proteins used in this work A/New York/107/03 (H7N2), Genbank: ACC55270; A/ Netherlands/219/03 (H7N7), Genbank: AAR02640; A/Hong Kong/1-9/68 (H3N2), 2HMG; A/Duck/Ukraine/1/63 (H3N8), PDB: 1MQL; A/South Carolina/1/18 (H1N1), PDB: 1RD8; A/ Puerto Rico/8/34 (H1N1), PDB: 1RU7; A/Swine/Iowa/15/30 (H1N1), PDB: 1RUY; A/Singapore/1/1957 (H2N2), PDB: 2WRC; A/Viet Nam/1203/04 (H5N1), PDB: 2FK0; A/Duck/ Singapore/3/97 (H5N3), PDB: 1JSM; A/Swine/Hong Kong/9/ 98 (H9N2), PDB: 1JSD; A/Turkey/Italy/8000/02 (H7N3), PDB: 1TI8; C/Johannesburg/1/66, 1FLC. Supporting Information Figure S1 Sequence alignment of selected structurally available HAs. Human H3 (PDB: 2HMG), Avian H3 (PDB: 1MQL), 1918- Human H1 (PDB: 1RD8), 1934-Human H1 (PDB: 1RU7), Swine H1 (PDB: 1RUY), 1957-Huamn H2 (PDB: 2WRC), Human H5 (PDB: 2FK0), Avian H5 (PDB: 1JSM), Swine H9 (PDB: 1JSD), and Avian H7 (PDB: 1TI8) were used in the alignments. The fusion domain of HA1 is highlighted in magenta, the vestigial esterase domain is highlighted in green, the receptor binding domain is highlighted in blue, and the fusion domain of HA2 is highlighted in red. Residue numbering is based on the H3 HA sequence. Found at: doi:10.1371/journal.ppat.1001081.s001 (2.84 MB TIF) Figure S2 Expression and purification of NY107. SDS-PAGE reveals that NY107 was expressed as the HA0 form with a mass approximately 60kDa (middle lane). Thrombin cleavage resulted in an unexpected reduction in band size to a HA1/HA2 profile (right lane) with possible multiple glycoforms for the HA2 clearly present. Found at: doi:10.1371/journal.ppat.1001081.s002 (0.23 MB TIF) Figure S3 Comparison of glycan binding to NY107 with other HAs. A. Overlap of a2-3 ligands binding in the receptor binding site from NY-H7 (green), Av-H3 (orange), 1930-Hu-H1 (magenta), 1957-Hu-H2 (cyan), Av-H5 (yellow), and Sw-H9 (grey). B. Overlap of a2-6 linkage ligands binding in the receptor binding site from NY-H7 (green), Av-H3 (orange), 1930-Hu-H1 (magenta), 1957-Hu-H2 (cyan), Av-H5 (yellow), and Sw-H9 (grey). Found at: doi:10.1371/journal.ppat.1001081.s003 (2.55 MB TIF) Figure S4 Simulated annealing omit maps of the receptor binding site (contoured at 1s). A. NY107 (blue), B. NY107-39SLN (orange), C. NY107-69SLN (red), and D. NY107-LSTb (magenta). The protein model is shown in cartoon, and the residues involved in the binding to receptor analogs were shown in sticks. Maps were generated using version 1.2 of the Crystallography and NMR System (CNS) software. Found at: doi:10.1371/journal.ppat.1001081.s004 (1.93 MB TIF) Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 9 September 2010 | Volume 6 | Issue 9 | e1001081 Figure S5 Comparison of NY107 RBS to HEF. Overlap of RBS from NY107 (green), Av-H7 (marine) and HEF (magenta). Found at: doi:10.1371/journal.ppat.1001081.s005 (1.12 MB TIF) Table S1 Comparison of r.m.s.d. (A˚ ) for different HA domains. For analyzing differences in the overall structure, r.m.s.d. values were calculated between monomers or domains of different HA’s, after the Ca atoms of the HA2 domains were superposed by sequence and structural alignment onto the equivalent domains of NY107. Found at: doi:10.1371/journal.ppat.1001081.s006 (0.04 MB DOC) Table S2 Comparison of r.m.s.d. (A˚ ) for individual domains. Each domain was superimposed separately to determine how the individual NY107 domains compared to equivalent domains in the other structures. Found at: doi:10.1371/journal.ppat.1001081.s007 (0.04 MB DOC) Table S3 Molecular interactions between NY107 and receptor analogs. The hydrogen bond cutoff is 3.8 A˚ for the listing interactions. Found at: doi:10.1371/journal.ppat.1001081.s008 (0.07 MB DOC) Table S4 Glycan array differences between NY107, the fully restored NY107-ins, and NL219 (virus and rHA). The color coding in the left hand column reflects the same coloring scheme used in Figures 2 and 3. Significant binding of samples to glycans are qualitatively estimated based on relative strength of the signal for the data shown in Figures 2 and 3 Strong (+++), weak (+). Found at: doi:10.1371/journal.ppat.1001081.s009 (0.19 MB DOC) Acknowledgments The authors would like to thank the staff of SER-CAT sector 22 for their help in data collection. We also thank WHO Global Influenza Surveillance Network for providing NY107 and NL219 viruses from which the reverse genetics viruses were generated. Glycan microarray data presented here will be made available on-line through the CFG web site upon publication (www.functionalglycomics.org). The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention or the Agency for Toxic Substances and Disease Registry. Author Contributions Conceived and designed the experiments: HY LMC PJC ROD JS. Performed the experiments: HY LMC PJC JS. Analyzed the data: HY LMC PJC ROD JS. Wrote the paper: HY LMC PJC ROD JS. References 1. Thompson WW, Shay DK, Weintraub E, Brammer L, Cox N, et al. (2003) Mortality associated with influenza and respiratory syncytial virus in the United States. JAMA 289: 179–186. 2. Thompson WW, Shay DK, Weintraub E, Brammer L, Bridges CB, et al. (2004) Influenza-associated hospitalizations in the United States. JAMA 292: 1333–1340. 3. WHO (1980) A revision of the system of nomenclature for influenza viruses: a WHO memorandum. Bull, WHO 58: 585–591. 4. Fouchier RA, Munster V, Wallensten A, Bestebroer TM, Herfst S, et al. (2005) Characterization of a novel influenza A virus hemagglutinin subtype (H16) obtained from black-headed gulls. J Virol 79: 2814–2822. 5. Scholtissek C, Rohde W, Von Hoyningen V, Rott R (1978) On the origin of the human influenza virus subtypes H2N2 and H3N2. Virology 87: 13–20. 6. Kawaoka Y, Bean WJ, Webster RG (1989) Evolution of the hemagglutinin of equine H3 influenza viruses. Virology 169: 283–292. 7. Garten RJ, Davis CT, Russell CA, Shu B, Lindstrom S, et al. (2009) Antigenic and genetic characteristics of swine-origin 2009 A(H1N1) influenza viruses circulating in humans. Science 325: 197–201. 8. Shinya K, Ebina M, Yamada S, Ono M, Kasai N, et al. (2006) Avian flu: influenza virus receptors in the human airway. Nature 440: 435–436. 9. Matrosovich MN, Gambaryan AS, Teneberg S, Piskarev VE, Yamnikova SS, et al. (1997) Avian influenza A viruses differ from human viruses by recognition of sialyloligosaccharides and gangliosides and by a higher conservation of the HA receptor-binding site. Virology 233: 224–234. 10. de Jong JC, Claas EC, Osterhaus AD, Webster RG, Lim WL (1997) A pandemic warning? Nature 389: 554. 11. Taubenberger JK, Morens DM, Fauci AS (2007) The next influenza pandemic: can it be predicted? JAMA 297: 2025–2027. 12. Webster RG, Geraci J, Petursson G, Skirnisson K (1981) Conjunctivitis in human beings caused by influenza A virus of seals. N Engl J Med 304: 911. 13. Kurtz J, Manvell RJ, Banks J (1996) Avian influenza virus isolated from a woman with conjunctivitis. Lancet 348: 901–902. 14. Koopmans M, Wilbrink B, Conyn M, Natrop G, van der Nat H, et al. (2004) Transmission of H7N7 avian influenza A virus to human beings during a large outbreak in commercial poultry farms in the Netherlands. Lancet 363: 587– 593. 15. Fouchier RA, Schneeberger PM, Rozendaal FW, Broekman JM, Kemink SA, et al. (2004) Avian influenza A virus (H7N7) associated with human conjunctivitis and a fatal case of acute respiratory distress syndrome. Proc Natl Acad Sci U S A 101: 1356–1361. 16. CDC (2004) Update: influenza activity–United States and worldwide, 2003–04 season, and composition of the 2004–05 influenza vaccine. MMWR Morb Mortal Wkly Rep: Centers for Disease Control. pp 547–552. 17. Hirst M, Astell CR, Griffith M, Coughlin SM, Moksa M, et al. (2004) Novel avian influenza H7N3 strain outbreak, British Columbia. Emerg Infect Dis 10: 2192–2195. 18. Tweed SA, Skowronski DM, David ST, Larder A, Petric M, et al. (2004) Human illness from avian influenza H7N3, British Columbia. Emerg Infect Dis 10: 2196–2199. 19. EditorialTeam (2007) Avian influenza A/H7N2 outbreak in the United Kingdom. Euro Surveill 12: 2. 20. Suarez DL, Garcia M, Latimer J, Senne D, Perdue M (1999) Phylogenetic analysis of H7 avian influenza viruses isolated from the live bird markets of the Northeast United States. J Virol 73: 3567–3573. 21. Belser JA, Lu X, Maines TR, Smith C, Li Y, et al. (2007) Pathogenesis of avian influenza (H7) virus infection in mice and ferrets: enhanced virulence of Eurasian H7N7 viruses isolated from humans. J Virol 81: 11139–11147. 22. Belser JA, Blixt O, Chen LM, Pappas C, Maines TR, et al. (2008) Contemporary North American influenza H7 viruses possess human receptor specificity: Implications for virus transmissibility. Proc Natl Acad Sci U S A 105: 7558–7563. 23. Gambaryan AS, Tuzikov AB, Pazynina GV, Desheva JA, Bovin NV, et al. (2008) 6-sulfo sialyl Lewis X is the common receptor determinant recognized by H5, H6, H7 and H9 influenza viruses of terrestrial poultry. Virol J 5: 85. 24. Stevens J, Blixt O, Glaser L, Taubenberger JK, Palese P, et al. (2006) Glycan microarray analysis of the hemagglutinins from modern and pandemic influenza viruses reveals different receptor specificities. J Mol Biol 355: 1143–1155. 25. Stevens J, Corper AL, Basler CF, Taubenberger JK, Palese P, et al. (2004) Structure of the uncleaved human H1 hemagglutinin from the extinct 1918 influenza virus. Science 303: 1866–1870. 26. Russell RJ, Kerry PS, Stevens DJ, Steinhauer DA, Martin SR, et al. (2008) Structure of influenza hemagglutinin in complex with an inhibitor of membrane fusion. Proc Natl Acad Sci U S A 105: 17736–17741. 27. Russell RJ, Gamblin SJ, Haire LF, Stevens DJ, Xiao B, et al. (2004) H1 and H7 influenza haemagglutinin structures extend a structural classification of haemagglutinin subtypes. Virology 325: 287–296. 28. Matrosovich M, Tuzikov A, Bovin N, Gambaryan A, Klimov A, et al. (2000) Early alterations of the receptor-binding properties of H1, H2, and H3 avian influenza virus hemagglutinins after their introduction into mammals. J Virol 74: 8502–8512. 29. Nobusawa E, Ishihara H, Morishita T, Sato K, Nakajima K (2000) Change in receptor-binding specificity of recent human influenza A viruses (H3N2): a single amino acid change in hemagglutinin altered its recognition of sialyloligosacchar- ides. Virology 278: 587–596. 30. Stevens J, Blixt O, Chen LM, Donis RO, Paulson JC, et al. (2008) Recent avian H5N1 viruses exhibit increased propensity for acquiring human receptor specificity. J Mol Biol 381: 1382–1394. 31. Russell RJ, Stevens DJ, Haire LF, Gamblin SJ, Skehel JJ (2006) Avian and human receptor binding by hemagglutinins of influenza A viruses. Glycoconj J 23: 85–92. 32. Gamblin SJ, Haire LF, Russell RJ, Stevens DJ, Xiao B, et al. (2004) The structure and receptor binding properties of the 1918 influenza hemagglutinin. Science 303: 1838–1842. 33. Suarez DL, Spackman E, Senne DA (2003) Update on molecular epidemiology of H1, H5, and H7 influenza virus infections in poultry in North America. Avian Dis 47: 888–897. 34. Daniels PS, Jeffries S, Yates P, Schild GC, Rogers GN, et al. (1987) The receptor-binding and membrane-fusion properties of influenza virus variants Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 10 September 2010 | Volume 6 | Issue 9 | e1001081 selected using anti-haemagglutinin monoclonal antibodies. EMBO J 6: 1459–1465. 35. Rosenthal PB, Zhang X, Formanowski F, Fitz W, Wong CH, et al. (1998) Structure of the haemagglutinin-esterase-fusion glycoprotein of influenza C virus. Nature 396: 92–96. 36. Weinstein J, de Souza-e-Silva U, Paulson JC (1982) Purification of a Gal beta 1 to 4GlcNAc alpha 2 to 6 sialyltransferase and a Gal beta 1 to 3(4)GlcNAc alpha 2 to 3 sialyltransferase to homogeneity from rat liver. J Biol Chem 257: 13835–13844. 37. Gambaryan A, Webster R, Matrosovich M (2002) Differences between influenza virus receptors on target cells of duck and chicken. Arch Virol 147: 1197–1208. 38. Wan H, Perez DR (2006) Quail carry sialic acid receptors compatible with binding of avian and human influenza viruses. Virology 346: 278–286. 39. Guo CT, Takahashi N, Yagi H, Kato K, Takahashi T, et al. (2007) The quail and chicken intestine have sialyl-galactose sugar chains responsible for the binding of influenza A viruses to human type receptors. Glycobiology 17: 713–724. 40. Maines TR, Chen LM, Matsuoka Y, Chen H, Rowe T, et al. (2006) Lack of transmission of H5N1 avian-human reassortant influenza viruses in a ferret model. Proc Natl Acad Sci U S A 103: 12121–12126. 41. Wan H, Sorrell EM, Song H, Hossain MJ, Ramirez-Nieto G, et al. (2008) Replication and transmission of H9N2 influenza viruses in ferrets: evaluation of pandemic potential. PLoS One 3: e2923. 42. Weis WI, Brunger AT, Skehel JJ, Wiley DC (1990) Refinement of the influenza virus hemagglutinin by simulated annealing. J Mol Biol 212: 737–761. 43. Frank S, Kammerer RA, Mechling D, Schulthess T, Landwehr R, et al. (2001) Stabilization of short collagen-like triple helices by protein engineering. J Mol Biol 308: 1081–1089. 44. Stevens J, Blixt O, Tumpey TM, Taubenberger JK, Paulson JC, et al. (2006) Structure and receptor specificity of the hemagglutinin from an H5N1 influenza virus. Science 312: 404–410. 45. Chayen NE, Shaw-Steward PD, Blow DM (1992) Microbatch crystallization under oil – a new technique allowing many small volume crystallization experiments. J Cryst Growth 122: 176–180. 46. Otwinowski A, Minor W (1997) Processing of X-ray diffraction data collected in oscillation mode. Meothds in Enzymology 276: 307–326. 47. McCoy AJ, Grosse-Kunstleve RW, Storoni LC, Read RJ (2005) Likelihood- enhanced fast translation functions. Acta Crystallogr D Biol Crystallogr 61: 458–464. 48. Emsley P, Cowtan K (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132. 49. CCP4 (1994) The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 50: 760–763. 50. Winn MD, Isupov MN, Murshudov GN (2001) Use of TLS parameters to model anisotropic displacements in macromolecular refinement. Acta Crystallogr D Biol Crystallogr 57: 122–133. 51. Davis IW, Leaver-Fay A, Chen VB, Block JN, Kapral GJ, et al. (2007) MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res 35: W375–383. 52. Brunger AT (2007) Version 1.2 of the Crystallography and NMR system. Nat Protoc 2: 2728–2733. 53. Hoffmann E, Webster RG (2000) Unidirectional RNA polymerase I-polymerase II transcription system for the generation of influenza A virus from eight plasmids. J Gen Virol 81: 2843–2847. 54. Higuchi R, Krummel B, Saiki RK (1988) A general method of in vitro preparation and specific mutagenesis of DNA fragments: study of protein and DNA interactions. Nucleic Acids Res 16: 7351–7367. 55. Blixt O, Head S, Mondala T, Scanlan C, Huflejt ME, et al. (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc Natl Acad Sci U S A 101: 17033–17038. 56. DeLano WL (2002) The PyMol Molecular Graphics Systems. wwwpymolorg. Structure of a North American H7 Hemagglutinin PLoS Pathogens | www.plospathogens.org 11 September 2010 | Volume 6 | Issue 9 | e1001081
3M5L
Crystal structure of HCV NS3/4A protease in complex with ITMN-191
Drug resistance against HCV NS3/4A inhibitors is defined by the balance of substrate recognition versus inhibitor binding Keith P. Romano1, Akbar Ali1, William E. Royer, and Celia A. Schiffer2 Department of Biochemistry and Molecular Pharmacology, University of Massachusetts Medical School, Worcester, MA 01605 Edited by John M. Coffin, Tufts University School of Medicine, Boston, MA, and approved September 14, 2010 (received for review May 13, 2010) Hepatitis C virus infects an estimated 180 million people world- wide, prompting enormous efforts to develop inhibitors targeting the essential NS3/4A protease. Resistance against the most promis- ing protease inhibitors, telaprevir, boceprevir, and ITMN-191, has emerged in clinical trials. In this study, crystal structures of the NS3/4A protease domain reveal that viral substrates bind to the protease active site in a conserved manner defining a consensus volume, or substrate envelope. Mutations that confer the most severe resistance in the clinic occur where the inhibitors protrude from the substrate envelope, as these changes selectively weaken inhibitor binding without compromising the binding of substrates. These findings suggest a general model for predicting the suscept- ibility of protease inhibitors to resistance: drugs designed to fit within the substrate envelope will be less susceptible to resistance, as mutations affecting inhibitor binding would simultaneously interfere with the recognition of viral substrates. drug design ∣hepatitis C ∣substrate envelope D rug resistance is a major obstacle in the treatment of quickly evolving diseases. Hepatitis C virus (HCV) is a genetically diverse Hepacivirus of the Flaviviridae family infecting an esti- mated 180 million people worldwide (1). The viral RNA genome is translated as a single polyprotein and subsequently processed by host-cell and viral proteases into structural (C, E1, E2, and p7) and nonstructural (NS2, NS3, NS4A, NS4B, NS5A, and NS5B) proteins (2). The viral RNA-dependent RNA polymerase, NS5B, is inherently inaccurate and misincorporation of bases accounts for a very high mutation rate (3). While some mutations are neu- tral, others will alter the viability of the virus and propagate with varying efficiencies in each patient. Thus HCV infected indivi- duals will develop a heterogeneous population of virus variants known as quasispecies (4). As patients begin treatment, the selec- tive pressures of antiviral drugs will favor drug resistant variants (5). Therefore, an inhibitor must not only recognize one protein variant, but an ensemble of related enzymes. A detailed under- standing of the atomic mechanisms of resistance is essential to effectively combat drug resistance against HCV antivirals. The essential HCV NS3/4A protease is an attractive therapeu- tic target responsible for cleaving at least four sites along the viral polyprotein. These sites share little sequence homology except for an acid at position P6, Cys or Thr at P1, and Ser or Ala at P1′ (Table S1). The first cleavage event at the 3-4A junction occurs in cis as a unimolecular process, while the remaining sub- strates are processed bimolecularly in trans. The NS3/4A protease also cleaves the human cellular targets TRIF and MAVS, which confounds the innate immune response to viral infection (6–8). Early drug design efforts were hampered by the relatively shallow, featureless architecture of the protease active site. The eventual observation of N-terminal product inhibition served as a stepping stone for the discovery of more potent peptidomimetic inhibitors (9, 10). Over the past decade, pharmaceutical companies have further developed these lead compounds. Many structure-activ- ity-relationship (SAR) studies have been performed to evaluate the effect of different functional moieties on protease inhibition at positions P4-P1′ (11–17). Crystal structures have been deter- mined of the NS3/4A protease domain bound to a variety of inhibitors as well as of several drug resistant protease variants, such as R155K and V36M (18, 19). These data elucidate the mo- lecular interactions of NS3/4A with inhibitors and the effect of specific drug resistance mutations on binding. These efforts, con- ducted in parallel by several pharmaceutical companies, led to the discovery of many protease inhibitors. Proof-of-concept for the successful clinical activity of this drug class was first demon- strated by the macrocyclic inhibitor BILN-2061 (Boehringer Ingelheim) (20, 21), which was later dropped from clinical trials in 2006 due to cardiotoxicity (22). Many other NS3/4A protease inhibitors are currently in development, and telaprevir (Vertex), boceprevir (Schering-Plough), and ITMN-191 (Intermune) lead the way in advanced phases of human clinical trials (Fig. 1A). Despite these successes, the rapid acquisition of drug resis- tance has limited the efficacy of the most potent NS3/4A protease inhibitors in both replicon studies and human clinical trials (Fig. 1B and Table 1). In this study, we show that mutations con- ferring the most severe resistance occur where the protease extensively contacts the inhibitors but not the natural viral sub- strates. Four crystal structures of the NS3/4A protease domain in complex with the N-terminal products of viral substrates reveal a conserved mode of substrate binding, with the consensus volume defining the substrate envelope. The protease inhibitors ITMN- 191 (3M5L), TMC435 (3KEE) (23), and boceprevir (2OC8) (24) protrude extensively from the substrate envelope in regions that correlate with known sites of resistance mutations. Most notably, the P2 moieties of all three drugs protrude to contact A156 and R155, which mutate to confer high-level resistance against nearly all drugs reported in the literature (25–30). These findings sug- gest that drug resistance results from a change in molecular recognition and imply that drugs designed to fit within the sub- strate envelope will be less susceptible to resistance, as mutations altering inhibitor binding will simultaneously interfere with the binding of substrates. Results Synthesis of ITMN-191. We synthesized the macrocyclic inhibitor ITMN-191 using a convergent reaction sequence described in SI Text. Briefly, the P2 and P1-P1′ fragments were preassembled Author contributions: K.P.R., A.A., and C.A.S. designed research; K.P.R. and A.A. performed research; A.A. and W.E.R. contributed new reagents/analytic tools; K.P.R., W.E.R., and C.A.S. analyzed data; and K.P.R., A.A., W.E.R., and C.A.S. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Data deposition: The atomic coordinates and structure factors have been deposited in the Protein Data Bank, www.pdb.org (PDB ID codes 3M5L, 3M5M, 3M5N, and 3M5O). 1K.P.R. and A.A. contributed equally to this work. 2To whom correspondence should be addressed. E-mail: Celia.Schiffer@umassmed.edu. This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1006370107/-/DCSupplemental. 20986–20991 ∣PNAS ∣December 7, 2010 ∣vol. 107 ∣no. 49 www.pnas.org/cgi/doi/10.1073/pnas.1006370107 and the macrocyclic drug compound was generated by a four- step reaction sequence, including P2-P3 amide coupling, ester hydrolysis, coupling with the P1-P1′ fragment, and ring-closing metathesis. The P2-P3 fragment was assembled by coupling the commercially available Boc-protected amino acid (S)-2-(tert- butoxycarbonylamino)non-8-enoic acid (Acme Biosciences, Inc) with the preassembled P2 fragment, (3R, 5S)-5-(methoxy- carbonyl)pyrrolidin-3-yl 4-fluoroisoindoline-2-carboxylate (31), using O-(7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexa- fluorophosphate (HATU)/diisopropylethylamine (DIPEA). Hy- drolysis of the P2-P3 methyl ester with LiOH:H2O in a mixture of THF-MeOH-H2O followed by coupling of the resulting acid under HATU/DIPEA conditions with the preassembled P1-P1′ fragment, (1R, 2S)-1-amino-N-(cyclopropylsulfonyl)-2-vinylcyclo- propanecarboxamide (32), provided the bis-olefin precursor for ring-closing metathesis. Cyclization of the bis-olefin intermediate was accomplished using a highly efficient ring-closing metathesis catalyst Zhan 1B and provided the protease inhibitor ITMN-191. Structure Determination of Inhibitor and Substrate Complexes. Although NS3/4A cleaves the viral polyprotein of over 3,000 residues at four specific sites in vivo, we focused on the local interactions of the protease domain with short peptide sequences corresponding to the immediate cleavage sites. All structural studies were carried out with the highly soluble, single-chain con- struct of the NS3/4A protease domain described previously (33), which contains a fragment of the essential cofactor NS4A cova- lently linked at the N terminus by a flexible linker. A similar pro- tease construct was shown to retain comparable catalytic activity to the authentic protein complex (34). Crystallization trials were initially carried out using the inactive (S139A) protease variant in complex with substrate peptides spanning P7-P5′. The 4A4B substrate complex revealed cleavage of the scissile bond and no ordered regions for the C-terminal fragment of the substrate. Similar observations were previously described for two other serine proteases where catalytic activity was observed, presum- ably facilitated by water, despite Ala substitutions of the catalytic Ser (35, 36). Thus all subsequent crystallization trials with the NS3/4A protease were performed using N-terminal cleavage products of the viral substrates spanning P7-P1. NS3/4A crystal structures in complex with ITMN-191 and peptide products 4A4B, 4B5A, and 5A5B were determined and refined at 1.25 Å, 1.70 Å, 1.90 Å, and 1.60 Å resolution, respec- tively (Table S2). The complexes crystallized in the space groups P212121 and P21 with one, two, or four molecules in the asym- metric unit. The average B factors range from 16.8–29.7 Å2 and there are no outliers in the Ramachandran plots. These structures represent the highest resolution crystal structures of NS3/4A protease reported to date. Overall Structure Analysis. The NS3/4A protease domain adopts a tertiary fold characteristic of serine proteases of the chymotrypsin family (37, 38). A total of nine protease molecules were modeled in the four crystal structures solved in this study with an overall rms deviation (rmsd) of 0.28 Å. The rmsds reveal the five most variable regions of the protease to be (Fig. S1): (i) the linker con- necting cofactor 4A at the N terminus, (ii) the loop containing residues 65–70, (iii) the zinc-binding site containing residues 95–105, (iv) the 310 helix region spanning residues 128–136, and (v) the active site antiparallel β-sheet containing residues 156–168. These structural differences likely indicate inherent flexibility in the protease and do not appear to correlate with ligand type or active site occupancy. Analysis of Product Complexes. Product complexes 4A4B, 4B5A, and 5A5B were further analyzed with the C terminus of the full-length NS3/4A structure (1CU1), which contains the N-term- inal cleavage product of viral substrate 3-4A (39). All four products bind to the protease active site in a conserved manner (Fig. 2), forming an antiparallel β-sheet with residues 154–160 Fig. 1. NS3/4A protease inhibitors and reported sites of drug resistance. (A) The leading protease inhibitors in development mimic the N-terminal side of the viral substrates. (B) The majority of reported drug resistance mutations cluster around the protease active site with the catalytic triad depicted in yellow. Table 1. Drug resistance mutations reported in replicon studies and clinical trials* Residue Mutation Drug V36 A, M, L, G Boceprevir, telaprevir Q41 R Boceprevir, ITMN-191 F43 S, C, V, I Boceprevir, telaprevir, ITMN-191, TMC435 V55 A Boceprevir T54 A, S Boceprevir, telaprevir Q80 K, R, H, G, L TMC435 S138 T ITMN-191, TMC435† R155 K, T, I, M, G, L, S, Q Boceprevir, telaprevir, ITMN-191, BILN-2061, TMC435 A156 V, T, S, I, G Boceprevir, telaprevir, ITMN-191, BILN-2061, TMC435 V158 I Boceprevir D168 A, V, E, G, N, T, Y, H, I ITMN-191, BILN-2061, TMC435 V170 A Boceprevir, telaprevir M175 L Boceprevir *References (18, 25, 26, 28, 30–37). †TMC435 displays reduced activity against S138T, but the mutation was not observed in selection experiments. Romano et al. PNAS ∣ December 7, 2010 ∣ vol. 107 ∣ no. 49 ∣ 20987 BIOPHYSICS AND COMPUTATIONAL BIOLOGY CHEMISTRY and burying 500–600 Å2 of solvent accessible surface area as cal- culated by PISA (40). The peptide backbone torsions are very similar, being most conserved at position P1 and deviating slightly toward position P4. Eight hydrogen bonds between backbone amide and carbonyl groups are completely conserved, involving protease residues S159 (C159 in product 3-4A), A157, R155, S139A, S138, and G137. S159 (C159 in product 3-4A), and A157 each contribute two hydrogen bonds with the P5 and P3 peptide residues, respectively. All P1 terminal carboxyl groups sit in the oxyanion hole, hydrogen bonding with the Nϵ atom of H57 and the amide nitrogens of residues 137–139. Although only product 4B5A contains a proline at P2, the other substrate sequences still adopt constrained P2 φ torsion angles. Thus products bind similarly despite their high sequence diversity. The P1 and P6 residues are most conserved among the substrate sequences, as are most of their interactions with the protease. The P1 side chains interact with the aromatic ring of F154. In all structures but product complex 4B5A, K165 forms salt-bridges with the P6 acids, while residues R123, D168, R155, and the catalytic D81 form an ionic network along one surface of the bound products (Fig. 2). In complex 4B5A, R123 interacts directly with the P6 acid, while D168 reorients and no longer contacts R155. Other molecular interactions in the product complexes are more diverse. Notably, K136 interacts differently with the cleavage products, forming: (i) a hydrogen bond with the P2 backbone carbonyl oxygen of 3-4A; (ii) a salt-bridge with the P3 glutamate of product 4A4B; and (iii) non- specific van der Waals interactions with the P2 and P3 side chains of products 4B5A and 5A5B. Also, in product complex 4A4B, an intramolecular hydrogen bond forms between the P3 and P5 glu- tamate residues, while the unique P4 acid of product 3-4A forms salt-bridges with the guanidinium groups of R123 and R155. Thus distinct patterns of side chain interactions underlie the set of con- served features involved in NS3/4A cleavage product binding. The Substrate Envelope. To further analyze the structural similari- ties of the four NS3/4A product complexes, the active sites were superposed on the Cα atoms of residues 137–139 and 154–160, revealing that both the active site residues and substrate products spanning P6-P1 align closely with an average Cα rmsd of 0.24 Å and 0.35 Å, respectively. The consensus van der Waals volume shared by any three of the four cleavage products was then cal- culated to generate the NS3/4A substrate envelope (Fig. 3A). This shape could not be predicted by the primary sequences alone and highlights the conserved mode of viral substrate recognition despite their high sequence diversity. Analysis of Inhibitor Complexes. ITMN-191, TMC435, and bocepre- vir are all peptidomimetic NS3/4A protease inhibitors. Active site superpositions of these drug complexes reveal that the inhibitors interact with many of the same protease residues as the cleavage products. Despite the P3-P1 cyclization of ITMN-191 and TMC435, the functional groups are positioned similarly in all three inhibitor complexes. The P1 cysteine surrogates interact with the aromatic ring of F154, while the P2 and P3 moieties over- lap closely. Although TMC435 does not contain a P4 substituent, the P4 tert-butyl groups of ITMN-191 and boceprevir also align closely. In addition, the P1 and P3 backbone atoms of all inhibi- tors hydrogen bond with the carbonyl oxygens of R155 and A157, respectively. These observations verify the peptidomimetic nat- ure of these drugs and support their observed mechanism as competitive active site inhibitors. The largest variation between these three protease inhibitors occurs at P2 where the aromatic rings of ITMN-191 and TMC435 stack against the guanidinium group of R155 (Fig. 3). This molecular interaction alters the electrostatic network involving R123, D168, R155, and D81. R155 rotates nearly 180° around Cδ relative to its conformation observed in product complexes, losing its hydrogen bond with D81 but maintaining interaction with D168. Mutations at R155 or D168 would disrupt the elec- trostatic network and destabilize this packing thereby lowering the affinity of these macrocyclic drugs. This observation provides a structural rationale for the drug resistance mutations R155K, as previously proposed (19), and D168A/V, which both confer a selective advantage in vitro in the presence of ITMN-191 or TMC435 (26, 30). In addition, the TMC435 complex reveals that R155 is stabilized by a hydrogen bond with Q80, which also mutates to confer resistance to TMC435 (30). Thus many of the primary drug resistance mutations can be explained by the disrup- tion of atomic interactions involving the P2 functional groups of the drugs. Insights into Drug Resistance. To determine the locations where the inhibitors protrude from the substrate envelope, the inhibitor and product complexes were also superposed using residues 137–139 and 154–160. The van der Waals volumes of inhibitor protrusion from the substrate envelope (V out) (41, 42) were calculated for each drug and compared with published EC50 fold-change data for drug resistance variants (30). The magnitudes of the EC50 fold-change data determined for each NS3/4A mutant generally trend with the V out values for the three drugs. The P2 moieties of boceprevir, ITMN-191 and TMC435 protrude most extensively from the substrate envelope with V out values of 105, 294, and Fig. 2. Stereo view of N-terminal cleavage product binding to NS3/4A pro- tease. N-terminal protease cleavage products (A) 3-4A, (B) 4A4B, (C) 4B5A, and (D) 5A5B are depicted as they bind to the protease active site. All conserved interactions are indicated by black dashes, while red lines depict interactions that are not present in all structures. The electrostatic network involving residues R123, D168, R155, and D81 is indicated by blue dashes. 20988 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1006370107 Romano et al. 496 Å3, respectively (Table 2). However, the precise level of drug resistance observed is also determined by the particular change in molecular interaction occurring for a given mutation. For ex- ample, A156 and R155 pack with the P2 moieties of these three inhibitors where they protrude beyond the substrate envelope. Mutations of A156 to bulkier side chains would result in a steric clash with the P2 drug moieties. Indeed, the rigid dimethylcyclo- propane group of boceprevir protrudes from the substrate envel- ope at the P2 subsite, and A156V or A156T confer 65 and 75 fold-changes in EC50, respectively (Table 2). Similarly, molecular changes at R155 and D168 would result in a substantial loss of interactions with P2. The most extensive protrusions of ITMN- 191 and TMC435 at P2 trend with their greatest fold-change in potency of nearly 450 and 600, respectively, from mutations in this subsite. Thus the extent by which an inhibitor protrudes from the substrate envelope in a given subsite is indicative of its vulnerability to resistance. Further structural analyses with the substrate envelope provide insights into other NS3/4A drug resistance mutations. The P1′ sulfonamide groups of ITMN-191 and TMC435, as well as the P1′ ketoamide of boceprevir, protrude from the substrate envel- ope near residues Q41 and F43, which both mutate to confer low-level resistance to these drugs (25, 30, 43). The keto group of boceprevir also projects outside the substrate envelope near T54 and V55. T54A/S confers low-level resistance to boceprevir, while V55A was recently identified in patient isolates after treat- ment with boceprevir (44). The analogous carbonyl groups of ITMN-191 and TMC435, however, are orientated in the opposite direction and protrude toward S138. In fact, in vitro studies reveal reduced activity for ITMN-191 and TMC435 against S138T variants, while boceprevir remains fully active (30, 43). The bulky P4 tert-butyl group of boceprevir extends outside the substrate envelope contacting V158; the V158I variant has lower affinity for this drug, likely due to a steric clash (45). This variant may also impact the affinity of ITMN-191, as its P4 tert-butyl also pro- trudes at the same location. These findings demonstrate that in regions outside the P2 subsite, positions where ITMN-191, TMC435, and boceprevir protrude from the substrate envelope also correlate with many other known sites of drug resistance mutations. Discussion The emergence of drug resistance is a major obstacle in modern medicine that limits the long-term usefulness of the most promis- ing therapeutics. By considering how HCV NS3/4A protease in- hibitors bind relative to natural viral substrates, we discovered that primary sites of resistance occur in regions of the protease where drugs protrude from the substrate envelope. In particular, R155 and A156, which mutate to confer severe resistance against ITMN-191, TMC435, and boceprevir, interact closely with the P2 drug moieties where they protrude most extensively from the substrate envelope. Molecular changes at these residues confer resistance by selectively weakening inhibitor binding without compromising the binding of viral substrates. We further specu- late that these mutations will not considerably affect the binding of the host cellular substrates TRIF and MAVS, which likely fit well within the substrate envelope as they share many features with the viral substrates. However, TRIF contains a track of eight proline residues instead of an acidic residue at position P6, which may modulate its binding. Further structural studies are warranted to better ascertain the molecular details of how these cellular substrates are recognized by the NS3/4A protease. Although this study focuses on ITMN-191, TMC435, and boceprevir, other NS3/4A protease inhibitors in clinical trials, in- cluding telaprevir, narlaprevir, and vaniprevir (Fig. 1A), contain similar functional groups that likely protrude from the substrate envelope. Most notably, all these drug candidates contain bulky P2 moieties and are therefore susceptible to cross-resistance against mutations at R155 and A156. R155 and A156 mutations have been shown to confer telaprevir resistance in treated patients (46). Cross-resistance studies have also shown that nar- Table 2. EC50 fold-change (FC) data * for several NS3/4A drug resistant variants tabulated with Vout, the van der Waals volume of protrusion from the substrate envelope, at each subsite of the enzyme Boceprevir ITMN-191 TMC435 Subsite Resistance mutation Vout (Å3) EC50 FC Vout (Å3) EC50 FC Vout (Å3) EC50 FC Total 292 500 649 P1 76 67 64 P2 105 294 496 Q80R 0.5 3.5 6.9 Q80K 0.8 2.3 7.7 R155K 4.7 447 30 A156V 75 63 177 A156T 65 41 44 D168A 0.7 153 594 D168E 0.8 75 40 P3 34 70 67 P4 76 69 0 V158I 3.3† ND ND *Antiviral activity was reported previously by Lenz et al., 2010 (30). †Fold-change in EC50 reported in replicon assay by Qiu et al., 2009 (45). Fig. 3. Stereo view of the NS3/4A substrate envelope and protease inhibitors. (A) After active site superpositions, the overlapping van der Waals volume shared by any three of the four cleavage products defines the substrate envelope, depicted in blue. NS3/4A protease residues which mutate to confer drug resistance are shown in brown. (B) ITMN-191, (C) boceprevir and (D) TMC435 protrude from the substrate envelope at several locations, which correlate with known sites of drug-resistant mutations to each inhibi- tor, shown in red. Romano et al. PNAS ∣ December 7, 2010 ∣ vol. 107 ∣ no. 49 ∣ 20989 BIOPHYSICS AND COMPUTATIONAL BIOLOGY CHEMISTRY laprevir displays similar fold losses in activity against most of the known drug resistance mutations for telaprevir and boceprevir (47). Ultimately, to slow the emergence of multidrug resistant viral strains, inhibitors should be confined within the substrate envelope, particularly at the P2 position. To compensate for the loss of binding affinity that will likely accompany these changes, additional interactions could potentially be optimized spanning the S4-S6 subsites of the protease and the catalytic triad. Our findings further suggest a general model for using the sub- strate envelope to predict patterns of drug resistance in other quickly evolving diseases. For drug resistance to occur, mutations must selectively weaken a target’s affinity for an inhibitor without significantly altering its natural biological function. Mutations occurring outside the substrate envelope are better able to achieve this effect, as these molecular changes can selectively alter inhibitor binding without compromising the binding of natural substrates. Whenever the interaction of a drug target with its biological substrates can be structurally characterized, we pre- dict that drugs designed to fit within the substrate envelope will be less susceptible to resistance. Structure-based design strategies can utilize this model as an added constraint to develop inhibitors that fit within the substrate envelope. In fact, previous work in our laboratory provides proof-of-concept for the successful incor- poration of the substrate envelope in the design of unique HIV protease inhibitors, which maintain high affinities against a panel of multidrug resistant variants of HIV-1 protease (42, 48–53). As a general paradigm, design efforts incorporating the substrate envelope would facilitate a more rationale evaluation of drug candidates and lead to the development of more robust inhibitors that are less susceptible to resistance. Materials and Methods Protein Crystallization. The NS3/4A protease construct was expressed and purified as reported previously (33, 54), detailed in SI Text. Purified protein was concentrated to ∼3 mg∕mL and loaded on a HiLoad Superdex75 16/60 column equilibrated with 25 mM 2-(N-morpholino)ethanesulfonic acid (MES) at pH 6.5, 500 mM NaCl, 10% glycerol, 30 μM zinc chloride, and 2 mM DTT. The protease fractions were pooled and concentrated to 20–25 mg∕mL with an Amicon Ultra-15 device (10 kD; Millipore). The concentrated samples were then incubated at 4 °C for 1 h with 2–20 M excess of viral substrate 4A4B, peptide products 4B5A or 5A5B, or ITMN-191. Information about the synth- esis of viral peptides and ITMN-191 is provided in SI Text. Diffraction-quality crystals were obtained overnight for all ligands by mixing equal volume of concentrated protein solution with precipitant solution (20–26% PEG-3350, 0.1 M sodium MES buffer at pH 6.5, and 4% ammonium sulfate) in 24-well VDX hanging drop trays. Data Collection and Structure Solution. Crystals were flash-frozen in liquid nitrogen and mounted under constant cryostream. X-ray diffraction data were collected at Advanced Photon Source BioCARS 14-IDB, 14-BMC, and LS-CAT 21-ID-F. Diffraction intensities of the complexes were indexed, integrated, and scaled using the programs HKL2000 (55) and XDS (56). 5% of the data was used to calculate R-free (57). All structure solutions were generated using isomorphous molecular replacement with PHASER (58) or AMORE. The NS3/4A protease domain (PDB code 2A4G) (59) was used for molecular replacement in solving the product 4A4B structure, and this struc- ture was subsequently used for solving the other complexes. In all cases, initial refinement was carried out in the absence of modeled ligand, which was subsequently built in. Phases were improved using ARP/wARP (60). Itera- tive rounds of translation, libration, and screw (TLS) and restrained refine- ment with CCP4 (61) and graphical model building with COOT (62) until convergence was achieved. The final structures were evaluated with MolProbity (63) prior to deposition in the protein data bank. Structural Analysis. Double-difference plots (64) were used to determine the structurally invariant regions of the protease, consisting of residues 32–36, 42–47, 50–54, 84–86, and 140–143. Structures were superposed with PyMOL (65) using the Cα atoms of these residues for all protease molecules from the solved structures (nine total). The B chain of product complex 4A4B was used as the reference structure in all alignments. Fit of individual inhibitors into the substrate envelope was quantified by mapping the substrate envelope and the van der Waals volume of each inhibitor on a three-dimensional grid with spacing of 0.5 Å. Vout for each drug moiety was computed by counting the grid cells, which were occupied by any inhi- bitor atom of that site but not the substrate envelope, and multiplied by the grid cell size, 0.125 Å3 (41, 42). Substrate Envelope and Inhibitor Analyses. NS3/4A substrate envelope was computed using product complexes 4A4B (B chain), 4B5A (D chain), and 5A5B (A chain). In structures with multiple protease molecules in the asym- metric unit, the one containing the most ordered peptide product was used for the alignment. The protease domain of the full-length NS3/4A structure (A chain; PDB code 1CU1) (39), including the C-terminal six amino acids, was included as a product complex 3-4A. All active site alignments were performed in PyMOL using Cα atoms of protease residues 137–139 and 154–160. After superposition, Gaussian object maps were generated in Py- MOL for each cleavage product. Four consensus Gaussian maps were then calculated, representing the intersecting volume of a group of three object maps. Finally, the summation of these four consensus maps was generated to construct the substrate envelope, depicting the van der Waals volume shared by any three of the four products. The previously determined boceprevir complex (PDB code 2OC8) (24) and TMC435 complex (PDB code 3KEE) (23) were used in this study (66). ACKNOWLEDGMENTS. We thank H. Klei for helpful discussions. We also thank Z. Wawrzak, M. Bolbat, and K. Brister of the LS-CAT beamline at Argonne National Laboratory for data collection of the ITMN-191 complex; M. Nalam and R. Bandaranayake for assistance with structural refinement; A. Ozen for providing V out calculations; and S. Shandilya and Y. Cai for computational support. The National Institute of Health (NIH) Grants R01-GM65347 and R01-AI085051 supported this work. Use of Advanced Photon Source (APS) was supported by the Department of Energy (DOE), Basic Energy Sciences, Office of Science, under Contract No. DE-AC02-06CH11357. Use of the Bio- CARS Sector 14 was supported by NIH-NCRR RR007707. Use of the LS-CAT Sector 21 was supported by the Michigan Economic Development Corpora- tion and the Michigan Technology TriCorridor under Grant 085P1000817. 1. World Health Organization. Barnes E, ed. (2010) Vaccine research: hepatitis C. Hepa- titis C Virus: Disease Burden. Available at http://www.who.int/vaccine_research/ diseases/viral_cancers/en/index2.html. March 15, 2010. 2. Major ME, Feinstone SM (1997) The molecular virology of hepatitis C. Hepatology 25:1527–1538. 3. Qureshi SA (2007) Hepatitis C virus—biology, host evasion strategies, and promising new therapies on the horizon. Med Res Rev 27:353–373. 4. Martell M, et al. (1992) Hepatitis C virus (HCV) circulates as a population of different but closely related genomes: quasispecies nature of HCV genome distribution. J Virol 66:3225–3229. 5. Paolucci S, et al. (2001) Analysis of HIV drug-resistant quasispecies in plasma, peripheral blood mononuclear cells and viral isolates from treatment-naive and HAART patients. J Med Virol 65:207–217. 6. Chen Z, et al. (2007) GB virus B disrupts RIG-I signaling by NS3/4A-mediated cleavage of the adaptor protein MAVS. J Virol 81:964–976. 7. Li XD, Sun L, Seth RB, Pineda G, Chen ZJ (2005) Hepatitis C virus protease NS3/4A cleaves mitochondrial antiviral signaling protein off the mitochondria to evade innate immunity. Proc Natl Acad Sci USA 102:17717–17722. 8. Li K, et al. (2005) Immune evasion by hepatitis C virus NS3/4A protease-mediated cleavage of the Toll-like receptor 3 adaptor protein TRIF. Proc Natl Acad Sci USA 102:2992–2997. 9. Llinas-Brunet M, et al. (1998) Peptide-based inhibitors of the hepatitis C virus serine protease. Bioorg Med Chem Lett 8:1713–1718. 10. Steinkuhler C, et al. (1998) Product inhibition of the hepatitis C virus NS3 protease. Biochemistry 37:8899–8905. 11. Arasappan A, et al. (2006) P2-P4 macrocyclic inhibitors of hepatitis C virus NS3-4A serine protease. Bioorg Med Chem Lett 16:3960–3965. 12. Bogen S, et al. (2008) Hepatitis C virus NS3-4A serine protease inhibitors: SAR of new P1 derivatives of SCH 503034. Bioorg Med Chem Lett 18:4219–4223. 13. Malancona S, et al. (2004) SAR and pharmacokinetic studies on phenethylamide inhibitors of the hepatitis C virus NS3/NS4A serine protease. Bioorg Med Chem Lett 14:4575–4579. 14. Nilsson M, et al. (2010) Synthesis and SAR of potent inhibitors of the hepatitis C virus NS3/4A protease: exploration of P2 quinazoline substituents. Bioorg Med Chem Lett 20:4004–4011. 15. Venkatraman S, et al. (2009) Discovery and structure-activity relationship of P1-P3 ketoamide derived macrocyclic inhibitors of hepatitis C virus NS3 protease. J Med Chem 52:336–346. 16. Raboisson P, et al. (2008) Structure-activity relationship study on a novel series of cyclopentane-containing macrocyclic inhibitors of the hepatitis C virus NS3/4A pro- tease leading to the discovery of TMC435350. Bioorg Med Chem Lett 18:4853–4858. 17. Perni RB, et al. (2004) Inhibitors of hepatitis C virus NS3.4A protease 2. Warhead SAR and optimization. Bioorg Med Chem Lett 14:1441–1446. 20990 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1006370107 Romano et al. 18. Zhou Y, et al. (2008) Phenotypic characterization of resistant Val36 variants of hepatitis C virus NS3-4A serine protease. Antimicrob Agents Chemother 52:110–120. 19. Zhou Y, et al. (2007) Phenotypic and structural analyses of hepatitis C virus NS3 protease Arg155 variants: sensitivity to telaprevir (VX-950) and interferon alpha. J Biol Chem 282:22619–22628. 20. Lamarre D, et al. (2003) An NS3 protease inhibitor with antiviral effects in humans infected with hepatitis C virus. Nature 426:186–189. 21. Hinrichsen H, et al. (2004) Short-term antiviral efficacy of BILN 2061, a hepatitis C virus serine protease inhibitor, in hepatitis C genotype 1 patients. Gastroenterology 127:1347–1355. 22. Vanwolleghem T, et al. (2007) Ultra-rapid cardiotoxicity of the hepatitis C virus protease inhibitor BILN 2061 in the urokinase-type plasminogen activator mouse. Gastroenterology 133:1144–1155. 23. Cummings MD, et al. (2010) Induced-fit binding of the macrocyclic noncovalent inhi- bitor TMC435 to its HCV NS3/NS4A protease target. Angew Chem Int Ed Engl 49:1652–1655. 24. Prongay AJ, et al. (2007) Discovery of the HCV NS3/4A protease inhibitor (1R,5S)-N- [3-amino-1-(cyclobutylmethyl)-2,3-dioxopropyl]-3- [2(S)-[[[(1,1-dimethylethyl)amino] carbonyl]amino]-3,3-dimethyl-1-oxobutyl]–6,6-dimethyl-3-azabicyclo[3.1.0]hexan-2(S)- carboxamide (Sch 503034) II. Key steps in structure-based optimization. J Med Chem 50:2310–2318. 25. Tong X, et al. (2008) Characterization of resistance mutations against HCV ketoamide protease inhibitors. Antiviral Res 77:177–185. 26. He Y, et al. (2008) Relative replication capacity and selective advantage profiles of protease inhibitor-resistant hepatitis C virus (HCV) NS3 protease mutants in the HCV genotype 1b replicon system. Antimicrob Agents Chemother 52:1101–1110. 27. Sarrazin C, et al. (2007) SCH 503034, a novel hepatitis C virus protease inhibitor, plus pegylated interferon alpha-2b for genotype 1 nonresponders. Gastroenterology 132:1270–1278. 28. Kieffer TL, et al. (2007) Telaprevir and pegylated interferon-alpha-2a inhibit wild-type and resistant genotype 1 hepatitis C virus replication in patients. Hepatology 46:631–639. 29. Yi M, Ma Y, Yates J, Lemon SM (2009) Transcomplementation of an NS2 defect in a late step in hepatitis C virus (HCV) particle assembly and maturation. PLoS Pathog 5:e1000403. 30. Lenz O, et al. (2010) In vitro resistance profile of the HCV NS3/4A protease inhibitor TMC435. Antimicrob Agents Chemother 54:1878–1887. 31. Arasappan A, Njoroge FG, Girijavallabhan VM (2005) PCT Int.Appl. WO 2005/113581 Patent Application. 32. Wang XA, et al. (2006) US Patent 6995174. 33. Wittekind M, Weinheirner S, Zhang Y, Goldfarb V (2002) US Patent 6333186. 34. Taremi SS, et al. (1998) Construction, expression, and characterization of a novel fully activated recombinant single-chain hepatitis C virus protease. Protein Sci 7:2143–2149. 35. Carter P, Wells JA (1988) Dissecting the catalytic triad of a serine protease. Nature 332:564–568. 36. Krishnan R, Sadler JE, Tulinsky A (2000) Structure of the Ser195Ala mutant of human alpha—thrombin complexed with fibrinopeptide A(7—16): evidence for residual catalytic activity. Acta Crystallogr D 56:406–410. 37. Kim JL, et al. (1996) Crystal structure of the hepatitis C virus NS3 protease domain complexed with a synthetic NS4A cofactor peptide. Cell 87:343–355. 38. Bode W, Huber R (1978) Crystal structure analysis and refinement of two variants of trigonal trypsinogen: trigonal trypsin and PEG (polyethylene glycol) trypsinogen and their comparison with orthorhombic trypsin and trigonal trypsinogen. FEBS Lett 90:265–269. 39. Yao N, Reichert P, Taremi SS, Prosise WW, Weber PC (1999) Molecular views of viral polyprotein processing revealed by the crystal structure of the hepatitis C virus bifunctional protease-helicase. Structure 7:1353–1363. 40. Krissinel E, Henrick K (2007) Inference of macromolecular assemblies from crystalline state. J Mol Biol 372:774–797. 41. Nalam MN, et al. (2010) Evaluating the substrate-envelope hypothesis: structural analysis of novel HIV-1 protease inhibitors designed to be robust against drug resis- tance. J Virol 84:5368–5378. 42. Chellappan S, Kairys V, Fernandes MX, Schiffer C, Gilson MK (2007) Evaluation of the substrate envelope hypothesis for inhibitors of HIV-1 protease. Proteins 68:561–567. 43. Kuntzen T, et al. (2008) Naturally occurring dominant resistance mutations to hepatitis C virus protease and polymerase inhibitors in treatment-naive patients. Hepatology 48:1769–1778. 44. Susser S, et al. (2009) Characterization of resistance to the protease inhibitor bocepre- vir in hepatitis C virus-infected patients. Hepatology 50:1709–1718. 45. Qiu P, et al. (2009) Identification of HCV protease inhibitor resistance mutations by selection pressure-based method. Nucleic Acids Res 37:e74. 46. Sarrazin C, et al. (2007) Dynamic hepatitis C virus genotypic and phenotypic changes in patients treated with the protease inhibitor telaprevir. Gastroenterology 132:1767–1777. 47. Tong X, et al. (2010) Preclinical characterization of the antiviral activity of SCH 900518 (Narlaprevir), a novel mechanism-based inhibitor of hepatitis C virus NS3 protease. Antimicrob Agents Chemother 54:2365–2370. 48. Chellappan S, et al. (2007) Design of mutation-resistant HIV protease inhibitors with the substrate envelope hypothesis. Chem Biol Drug Des 69:298–313. 49. Ali A, et al. (2009) Substrate envelope based design of new HIV-1 protease inhibitors active against drug-resistant HIV-1. 238th ACS National Meeting, Washington, DC, United States (American Chemical Society, Washington, DC), MEDI-102. 50. King NM, Prabu-Jeyabalan M, Nalivaika EA, Schiffer CA (2004) Combating susceptibil- ity to drug resistance: lessons from HIV-1 protease. Chem Biol 11:1333–1338. 51. Prabu-Jeyabalan M, et al. (2006) Substrate envelope and drug resistance: crystal struc- ture of RO1 in complex with wild-type human immunodeficiency virus type 1 protease. Antimicrob Agents Chemother 50:1518–1521. 52. Altman MD, Nalivaika EA, Prabu-Jeyabalan M, Schiffer CA, Tidor B (2008) Computa- tional design and experimental study of tighter binding peptides to an inactivated mutant of HIV-1 protease. Proteins 70:678–694. 53. Altman MD, et al. (2008) HIV-1 protease inhibitors from inverse design in the substrate envelope exhibit subnanomolar binding to drug-resistant variants. J Am Chem Soc 130:6099–6113. 54. Gallinari P, et al. (1998) Multiple enzymatic activities associated with recombinant NS3 protein of hepatitis C virus. J Virol 72:6758–6769. 55. Otwinowski Z, Minor W (1997) Processing of X-ray diffraction data collected in oscillation mode. Methods in Enzymology, 276 pp:307–326. 56. Kabsch W (1993) Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Crystallogr 26:795–800. 57. Brunger AT (1992) Free R value: a novel statistical quantity for assessing the accuracy of crystal structures. Nature 355:472–475. 58. McCoy AJ, et al. (2007) Phaser crystallographic software. J Appl Crystallogr 40:658–674. 59. Arasappan A, et al. (2005) Hepatitis C virus NS3-4A serine protease inhibitors: SAR of P′ 2 moiety with improved potency. Bioorg Med Chem Lett 15:4180–4184. 60. Morris RJ, Perrakis A, Lamzin VS (2002) ARP/wARP’s model-building algorithms. I. The main chain. Acta Crystallogr D 58:968–975. 61. Collaborative Computational Project Number 4 (1994) The CCP4 Suite: programs for protein crystallography. Acta Cryst, D50 pp:760–763. 62. Emsley P, Cowtan K (2004) COOT: model-building tools for molecular graphics. Acta Crystallogr D 60:2126–2132. 63. Davis IW, et al. (2007) MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res 35(Web Server issue):W375–383. 64. Prabu-Jeyabalan M, Nalivaika EA, Romano K, Schiffer CA (2006) Mechanism of sub- strate recognition by drug-resistant human immunodeficiency virus type 1 protease variants revealed by a novel structural intermediate. J Virol 80:3607–3616. 65. DeLano WL (2008) The PyMOL Molecular Graphics System (DeLano Scientific LLC, San Carlos, CA). 66. Berman HM, et al. (2000) The protein data bank. Nucleic Acids Res 28:235–242. Romano et al. PNAS ∣ December 7, 2010 ∣ vol. 107 ∣ no. 49 ∣ 20991 BIOPHYSICS AND COMPUTATIONAL BIOLOGY CHEMISTRY
3M5M
Avoiding drug resistance against HCV NS3/4A protease inhibitors
Drug resistance against HCV NS3/4A inhibitors is defined by the balance of substrate recognition versus inhibitor binding Keith P. Romano1, Akbar Ali1, William E. Royer, and Celia A. Schiffer2 Department of Biochemistry and Molecular Pharmacology, University of Massachusetts Medical School, Worcester, MA 01605 Edited by John M. Coffin, Tufts University School of Medicine, Boston, MA, and approved September 14, 2010 (received for review May 13, 2010) Hepatitis C virus infects an estimated 180 million people world- wide, prompting enormous efforts to develop inhibitors targeting the essential NS3/4A protease. Resistance against the most promis- ing protease inhibitors, telaprevir, boceprevir, and ITMN-191, has emerged in clinical trials. In this study, crystal structures of the NS3/4A protease domain reveal that viral substrates bind to the protease active site in a conserved manner defining a consensus volume, or substrate envelope. Mutations that confer the most severe resistance in the clinic occur where the inhibitors protrude from the substrate envelope, as these changes selectively weaken inhibitor binding without compromising the binding of substrates. These findings suggest a general model for predicting the suscept- ibility of protease inhibitors to resistance: drugs designed to fit within the substrate envelope will be less susceptible to resistance, as mutations affecting inhibitor binding would simultaneously interfere with the recognition of viral substrates. drug design ∣hepatitis C ∣substrate envelope D rug resistance is a major obstacle in the treatment of quickly evolving diseases. Hepatitis C virus (HCV) is a genetically diverse Hepacivirus of the Flaviviridae family infecting an esti- mated 180 million people worldwide (1). The viral RNA genome is translated as a single polyprotein and subsequently processed by host-cell and viral proteases into structural (C, E1, E2, and p7) and nonstructural (NS2, NS3, NS4A, NS4B, NS5A, and NS5B) proteins (2). The viral RNA-dependent RNA polymerase, NS5B, is inherently inaccurate and misincorporation of bases accounts for a very high mutation rate (3). While some mutations are neu- tral, others will alter the viability of the virus and propagate with varying efficiencies in each patient. Thus HCV infected indivi- duals will develop a heterogeneous population of virus variants known as quasispecies (4). As patients begin treatment, the selec- tive pressures of antiviral drugs will favor drug resistant variants (5). Therefore, an inhibitor must not only recognize one protein variant, but an ensemble of related enzymes. A detailed under- standing of the atomic mechanisms of resistance is essential to effectively combat drug resistance against HCV antivirals. The essential HCV NS3/4A protease is an attractive therapeu- tic target responsible for cleaving at least four sites along the viral polyprotein. These sites share little sequence homology except for an acid at position P6, Cys or Thr at P1, and Ser or Ala at P1′ (Table S1). The first cleavage event at the 3-4A junction occurs in cis as a unimolecular process, while the remaining sub- strates are processed bimolecularly in trans. The NS3/4A protease also cleaves the human cellular targets TRIF and MAVS, which confounds the innate immune response to viral infection (6–8). Early drug design efforts were hampered by the relatively shallow, featureless architecture of the protease active site. The eventual observation of N-terminal product inhibition served as a stepping stone for the discovery of more potent peptidomimetic inhibitors (9, 10). Over the past decade, pharmaceutical companies have further developed these lead compounds. Many structure-activ- ity-relationship (SAR) studies have been performed to evaluate the effect of different functional moieties on protease inhibition at positions P4-P1′ (11–17). Crystal structures have been deter- mined of the NS3/4A protease domain bound to a variety of inhibitors as well as of several drug resistant protease variants, such as R155K and V36M (18, 19). These data elucidate the mo- lecular interactions of NS3/4A with inhibitors and the effect of specific drug resistance mutations on binding. These efforts, con- ducted in parallel by several pharmaceutical companies, led to the discovery of many protease inhibitors. Proof-of-concept for the successful clinical activity of this drug class was first demon- strated by the macrocyclic inhibitor BILN-2061 (Boehringer Ingelheim) (20, 21), which was later dropped from clinical trials in 2006 due to cardiotoxicity (22). Many other NS3/4A protease inhibitors are currently in development, and telaprevir (Vertex), boceprevir (Schering-Plough), and ITMN-191 (Intermune) lead the way in advanced phases of human clinical trials (Fig. 1A). Despite these successes, the rapid acquisition of drug resis- tance has limited the efficacy of the most potent NS3/4A protease inhibitors in both replicon studies and human clinical trials (Fig. 1B and Table 1). In this study, we show that mutations con- ferring the most severe resistance occur where the protease extensively contacts the inhibitors but not the natural viral sub- strates. Four crystal structures of the NS3/4A protease domain in complex with the N-terminal products of viral substrates reveal a conserved mode of substrate binding, with the consensus volume defining the substrate envelope. The protease inhibitors ITMN- 191 (3M5L), TMC435 (3KEE) (23), and boceprevir (2OC8) (24) protrude extensively from the substrate envelope in regions that correlate with known sites of resistance mutations. Most notably, the P2 moieties of all three drugs protrude to contact A156 and R155, which mutate to confer high-level resistance against nearly all drugs reported in the literature (25–30). These findings sug- gest that drug resistance results from a change in molecular recognition and imply that drugs designed to fit within the sub- strate envelope will be less susceptible to resistance, as mutations altering inhibitor binding will simultaneously interfere with the binding of substrates. Results Synthesis of ITMN-191. We synthesized the macrocyclic inhibitor ITMN-191 using a convergent reaction sequence described in SI Text. Briefly, the P2 and P1-P1′ fragments were preassembled Author contributions: K.P.R., A.A., and C.A.S. designed research; K.P.R. and A.A. performed research; A.A. and W.E.R. contributed new reagents/analytic tools; K.P.R., W.E.R., and C.A.S. analyzed data; and K.P.R., A.A., W.E.R., and C.A.S. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Data deposition: The atomic coordinates and structure factors have been deposited in the Protein Data Bank, www.pdb.org (PDB ID codes 3M5L, 3M5M, 3M5N, and 3M5O). 1K.P.R. and A.A. contributed equally to this work. 2To whom correspondence should be addressed. E-mail: Celia.Schiffer@umassmed.edu. This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1006370107/-/DCSupplemental. 20986–20991 ∣PNAS ∣December 7, 2010 ∣vol. 107 ∣no. 49 www.pnas.org/cgi/doi/10.1073/pnas.1006370107 and the macrocyclic drug compound was generated by a four- step reaction sequence, including P2-P3 amide coupling, ester hydrolysis, coupling with the P1-P1′ fragment, and ring-closing metathesis. The P2-P3 fragment was assembled by coupling the commercially available Boc-protected amino acid (S)-2-(tert- butoxycarbonylamino)non-8-enoic acid (Acme Biosciences, Inc) with the preassembled P2 fragment, (3R, 5S)-5-(methoxy- carbonyl)pyrrolidin-3-yl 4-fluoroisoindoline-2-carboxylate (31), using O-(7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexa- fluorophosphate (HATU)/diisopropylethylamine (DIPEA). Hy- drolysis of the P2-P3 methyl ester with LiOH:H2O in a mixture of THF-MeOH-H2O followed by coupling of the resulting acid under HATU/DIPEA conditions with the preassembled P1-P1′ fragment, (1R, 2S)-1-amino-N-(cyclopropylsulfonyl)-2-vinylcyclo- propanecarboxamide (32), provided the bis-olefin precursor for ring-closing metathesis. Cyclization of the bis-olefin intermediate was accomplished using a highly efficient ring-closing metathesis catalyst Zhan 1B and provided the protease inhibitor ITMN-191. Structure Determination of Inhibitor and Substrate Complexes. Although NS3/4A cleaves the viral polyprotein of over 3,000 residues at four specific sites in vivo, we focused on the local interactions of the protease domain with short peptide sequences corresponding to the immediate cleavage sites. All structural studies were carried out with the highly soluble, single-chain con- struct of the NS3/4A protease domain described previously (33), which contains a fragment of the essential cofactor NS4A cova- lently linked at the N terminus by a flexible linker. A similar pro- tease construct was shown to retain comparable catalytic activity to the authentic protein complex (34). Crystallization trials were initially carried out using the inactive (S139A) protease variant in complex with substrate peptides spanning P7-P5′. The 4A4B substrate complex revealed cleavage of the scissile bond and no ordered regions for the C-terminal fragment of the substrate. Similar observations were previously described for two other serine proteases where catalytic activity was observed, presum- ably facilitated by water, despite Ala substitutions of the catalytic Ser (35, 36). Thus all subsequent crystallization trials with the NS3/4A protease were performed using N-terminal cleavage products of the viral substrates spanning P7-P1. NS3/4A crystal structures in complex with ITMN-191 and peptide products 4A4B, 4B5A, and 5A5B were determined and refined at 1.25 Å, 1.70 Å, 1.90 Å, and 1.60 Å resolution, respec- tively (Table S2). The complexes crystallized in the space groups P212121 and P21 with one, two, or four molecules in the asym- metric unit. The average B factors range from 16.8–29.7 Å2 and there are no outliers in the Ramachandran plots. These structures represent the highest resolution crystal structures of NS3/4A protease reported to date. Overall Structure Analysis. The NS3/4A protease domain adopts a tertiary fold characteristic of serine proteases of the chymotrypsin family (37, 38). A total of nine protease molecules were modeled in the four crystal structures solved in this study with an overall rms deviation (rmsd) of 0.28 Å. The rmsds reveal the five most variable regions of the protease to be (Fig. S1): (i) the linker con- necting cofactor 4A at the N terminus, (ii) the loop containing residues 65–70, (iii) the zinc-binding site containing residues 95–105, (iv) the 310 helix region spanning residues 128–136, and (v) the active site antiparallel β-sheet containing residues 156–168. These structural differences likely indicate inherent flexibility in the protease and do not appear to correlate with ligand type or active site occupancy. Analysis of Product Complexes. Product complexes 4A4B, 4B5A, and 5A5B were further analyzed with the C terminus of the full-length NS3/4A structure (1CU1), which contains the N-term- inal cleavage product of viral substrate 3-4A (39). All four products bind to the protease active site in a conserved manner (Fig. 2), forming an antiparallel β-sheet with residues 154–160 Fig. 1. NS3/4A protease inhibitors and reported sites of drug resistance. (A) The leading protease inhibitors in development mimic the N-terminal side of the viral substrates. (B) The majority of reported drug resistance mutations cluster around the protease active site with the catalytic triad depicted in yellow. Table 1. Drug resistance mutations reported in replicon studies and clinical trials* Residue Mutation Drug V36 A, M, L, G Boceprevir, telaprevir Q41 R Boceprevir, ITMN-191 F43 S, C, V, I Boceprevir, telaprevir, ITMN-191, TMC435 V55 A Boceprevir T54 A, S Boceprevir, telaprevir Q80 K, R, H, G, L TMC435 S138 T ITMN-191, TMC435† R155 K, T, I, M, G, L, S, Q Boceprevir, telaprevir, ITMN-191, BILN-2061, TMC435 A156 V, T, S, I, G Boceprevir, telaprevir, ITMN-191, BILN-2061, TMC435 V158 I Boceprevir D168 A, V, E, G, N, T, Y, H, I ITMN-191, BILN-2061, TMC435 V170 A Boceprevir, telaprevir M175 L Boceprevir *References (18, 25, 26, 28, 30–37). †TMC435 displays reduced activity against S138T, but the mutation was not observed in selection experiments. Romano et al. PNAS ∣ December 7, 2010 ∣ vol. 107 ∣ no. 49 ∣ 20987 BIOPHYSICS AND COMPUTATIONAL BIOLOGY CHEMISTRY and burying 500–600 Å2 of solvent accessible surface area as cal- culated by PISA (40). The peptide backbone torsions are very similar, being most conserved at position P1 and deviating slightly toward position P4. Eight hydrogen bonds between backbone amide and carbonyl groups are completely conserved, involving protease residues S159 (C159 in product 3-4A), A157, R155, S139A, S138, and G137. S159 (C159 in product 3-4A), and A157 each contribute two hydrogen bonds with the P5 and P3 peptide residues, respectively. All P1 terminal carboxyl groups sit in the oxyanion hole, hydrogen bonding with the Nϵ atom of H57 and the amide nitrogens of residues 137–139. Although only product 4B5A contains a proline at P2, the other substrate sequences still adopt constrained P2 φ torsion angles. Thus products bind similarly despite their high sequence diversity. The P1 and P6 residues are most conserved among the substrate sequences, as are most of their interactions with the protease. The P1 side chains interact with the aromatic ring of F154. In all structures but product complex 4B5A, K165 forms salt-bridges with the P6 acids, while residues R123, D168, R155, and the catalytic D81 form an ionic network along one surface of the bound products (Fig. 2). In complex 4B5A, R123 interacts directly with the P6 acid, while D168 reorients and no longer contacts R155. Other molecular interactions in the product complexes are more diverse. Notably, K136 interacts differently with the cleavage products, forming: (i) a hydrogen bond with the P2 backbone carbonyl oxygen of 3-4A; (ii) a salt-bridge with the P3 glutamate of product 4A4B; and (iii) non- specific van der Waals interactions with the P2 and P3 side chains of products 4B5A and 5A5B. Also, in product complex 4A4B, an intramolecular hydrogen bond forms between the P3 and P5 glu- tamate residues, while the unique P4 acid of product 3-4A forms salt-bridges with the guanidinium groups of R123 and R155. Thus distinct patterns of side chain interactions underlie the set of con- served features involved in NS3/4A cleavage product binding. The Substrate Envelope. To further analyze the structural similari- ties of the four NS3/4A product complexes, the active sites were superposed on the Cα atoms of residues 137–139 and 154–160, revealing that both the active site residues and substrate products spanning P6-P1 align closely with an average Cα rmsd of 0.24 Å and 0.35 Å, respectively. The consensus van der Waals volume shared by any three of the four cleavage products was then cal- culated to generate the NS3/4A substrate envelope (Fig. 3A). This shape could not be predicted by the primary sequences alone and highlights the conserved mode of viral substrate recognition despite their high sequence diversity. Analysis of Inhibitor Complexes. ITMN-191, TMC435, and bocepre- vir are all peptidomimetic NS3/4A protease inhibitors. Active site superpositions of these drug complexes reveal that the inhibitors interact with many of the same protease residues as the cleavage products. Despite the P3-P1 cyclization of ITMN-191 and TMC435, the functional groups are positioned similarly in all three inhibitor complexes. The P1 cysteine surrogates interact with the aromatic ring of F154, while the P2 and P3 moieties over- lap closely. Although TMC435 does not contain a P4 substituent, the P4 tert-butyl groups of ITMN-191 and boceprevir also align closely. In addition, the P1 and P3 backbone atoms of all inhibi- tors hydrogen bond with the carbonyl oxygens of R155 and A157, respectively. These observations verify the peptidomimetic nat- ure of these drugs and support their observed mechanism as competitive active site inhibitors. The largest variation between these three protease inhibitors occurs at P2 where the aromatic rings of ITMN-191 and TMC435 stack against the guanidinium group of R155 (Fig. 3). This molecular interaction alters the electrostatic network involving R123, D168, R155, and D81. R155 rotates nearly 180° around Cδ relative to its conformation observed in product complexes, losing its hydrogen bond with D81 but maintaining interaction with D168. Mutations at R155 or D168 would disrupt the elec- trostatic network and destabilize this packing thereby lowering the affinity of these macrocyclic drugs. This observation provides a structural rationale for the drug resistance mutations R155K, as previously proposed (19), and D168A/V, which both confer a selective advantage in vitro in the presence of ITMN-191 or TMC435 (26, 30). In addition, the TMC435 complex reveals that R155 is stabilized by a hydrogen bond with Q80, which also mutates to confer resistance to TMC435 (30). Thus many of the primary drug resistance mutations can be explained by the disrup- tion of atomic interactions involving the P2 functional groups of the drugs. Insights into Drug Resistance. To determine the locations where the inhibitors protrude from the substrate envelope, the inhibitor and product complexes were also superposed using residues 137–139 and 154–160. The van der Waals volumes of inhibitor protrusion from the substrate envelope (V out) (41, 42) were calculated for each drug and compared with published EC50 fold-change data for drug resistance variants (30). The magnitudes of the EC50 fold-change data determined for each NS3/4A mutant generally trend with the V out values for the three drugs. The P2 moieties of boceprevir, ITMN-191 and TMC435 protrude most extensively from the substrate envelope with V out values of 105, 294, and Fig. 2. Stereo view of N-terminal cleavage product binding to NS3/4A pro- tease. N-terminal protease cleavage products (A) 3-4A, (B) 4A4B, (C) 4B5A, and (D) 5A5B are depicted as they bind to the protease active site. All conserved interactions are indicated by black dashes, while red lines depict interactions that are not present in all structures. The electrostatic network involving residues R123, D168, R155, and D81 is indicated by blue dashes. 20988 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1006370107 Romano et al. 496 Å3, respectively (Table 2). However, the precise level of drug resistance observed is also determined by the particular change in molecular interaction occurring for a given mutation. For ex- ample, A156 and R155 pack with the P2 moieties of these three inhibitors where they protrude beyond the substrate envelope. Mutations of A156 to bulkier side chains would result in a steric clash with the P2 drug moieties. Indeed, the rigid dimethylcyclo- propane group of boceprevir protrudes from the substrate envel- ope at the P2 subsite, and A156V or A156T confer 65 and 75 fold-changes in EC50, respectively (Table 2). Similarly, molecular changes at R155 and D168 would result in a substantial loss of interactions with P2. The most extensive protrusions of ITMN- 191 and TMC435 at P2 trend with their greatest fold-change in potency of nearly 450 and 600, respectively, from mutations in this subsite. Thus the extent by which an inhibitor protrudes from the substrate envelope in a given subsite is indicative of its vulnerability to resistance. Further structural analyses with the substrate envelope provide insights into other NS3/4A drug resistance mutations. The P1′ sulfonamide groups of ITMN-191 and TMC435, as well as the P1′ ketoamide of boceprevir, protrude from the substrate envel- ope near residues Q41 and F43, which both mutate to confer low-level resistance to these drugs (25, 30, 43). The keto group of boceprevir also projects outside the substrate envelope near T54 and V55. T54A/S confers low-level resistance to boceprevir, while V55A was recently identified in patient isolates after treat- ment with boceprevir (44). The analogous carbonyl groups of ITMN-191 and TMC435, however, are orientated in the opposite direction and protrude toward S138. In fact, in vitro studies reveal reduced activity for ITMN-191 and TMC435 against S138T variants, while boceprevir remains fully active (30, 43). The bulky P4 tert-butyl group of boceprevir extends outside the substrate envelope contacting V158; the V158I variant has lower affinity for this drug, likely due to a steric clash (45). This variant may also impact the affinity of ITMN-191, as its P4 tert-butyl also pro- trudes at the same location. These findings demonstrate that in regions outside the P2 subsite, positions where ITMN-191, TMC435, and boceprevir protrude from the substrate envelope also correlate with many other known sites of drug resistance mutations. Discussion The emergence of drug resistance is a major obstacle in modern medicine that limits the long-term usefulness of the most promis- ing therapeutics. By considering how HCV NS3/4A protease in- hibitors bind relative to natural viral substrates, we discovered that primary sites of resistance occur in regions of the protease where drugs protrude from the substrate envelope. In particular, R155 and A156, which mutate to confer severe resistance against ITMN-191, TMC435, and boceprevir, interact closely with the P2 drug moieties where they protrude most extensively from the substrate envelope. Molecular changes at these residues confer resistance by selectively weakening inhibitor binding without compromising the binding of viral substrates. We further specu- late that these mutations will not considerably affect the binding of the host cellular substrates TRIF and MAVS, which likely fit well within the substrate envelope as they share many features with the viral substrates. However, TRIF contains a track of eight proline residues instead of an acidic residue at position P6, which may modulate its binding. Further structural studies are warranted to better ascertain the molecular details of how these cellular substrates are recognized by the NS3/4A protease. Although this study focuses on ITMN-191, TMC435, and boceprevir, other NS3/4A protease inhibitors in clinical trials, in- cluding telaprevir, narlaprevir, and vaniprevir (Fig. 1A), contain similar functional groups that likely protrude from the substrate envelope. Most notably, all these drug candidates contain bulky P2 moieties and are therefore susceptible to cross-resistance against mutations at R155 and A156. R155 and A156 mutations have been shown to confer telaprevir resistance in treated patients (46). Cross-resistance studies have also shown that nar- Table 2. EC50 fold-change (FC) data * for several NS3/4A drug resistant variants tabulated with Vout, the van der Waals volume of protrusion from the substrate envelope, at each subsite of the enzyme Boceprevir ITMN-191 TMC435 Subsite Resistance mutation Vout (Å3) EC50 FC Vout (Å3) EC50 FC Vout (Å3) EC50 FC Total 292 500 649 P1 76 67 64 P2 105 294 496 Q80R 0.5 3.5 6.9 Q80K 0.8 2.3 7.7 R155K 4.7 447 30 A156V 75 63 177 A156T 65 41 44 D168A 0.7 153 594 D168E 0.8 75 40 P3 34 70 67 P4 76 69 0 V158I 3.3† ND ND *Antiviral activity was reported previously by Lenz et al., 2010 (30). †Fold-change in EC50 reported in replicon assay by Qiu et al., 2009 (45). Fig. 3. Stereo view of the NS3/4A substrate envelope and protease inhibitors. (A) After active site superpositions, the overlapping van der Waals volume shared by any three of the four cleavage products defines the substrate envelope, depicted in blue. NS3/4A protease residues which mutate to confer drug resistance are shown in brown. (B) ITMN-191, (C) boceprevir and (D) TMC435 protrude from the substrate envelope at several locations, which correlate with known sites of drug-resistant mutations to each inhibi- tor, shown in red. Romano et al. PNAS ∣ December 7, 2010 ∣ vol. 107 ∣ no. 49 ∣ 20989 BIOPHYSICS AND COMPUTATIONAL BIOLOGY CHEMISTRY laprevir displays similar fold losses in activity against most of the known drug resistance mutations for telaprevir and boceprevir (47). Ultimately, to slow the emergence of multidrug resistant viral strains, inhibitors should be confined within the substrate envelope, particularly at the P2 position. To compensate for the loss of binding affinity that will likely accompany these changes, additional interactions could potentially be optimized spanning the S4-S6 subsites of the protease and the catalytic triad. Our findings further suggest a general model for using the sub- strate envelope to predict patterns of drug resistance in other quickly evolving diseases. For drug resistance to occur, mutations must selectively weaken a target’s affinity for an inhibitor without significantly altering its natural biological function. Mutations occurring outside the substrate envelope are better able to achieve this effect, as these molecular changes can selectively alter inhibitor binding without compromising the binding of natural substrates. Whenever the interaction of a drug target with its biological substrates can be structurally characterized, we pre- dict that drugs designed to fit within the substrate envelope will be less susceptible to resistance. Structure-based design strategies can utilize this model as an added constraint to develop inhibitors that fit within the substrate envelope. In fact, previous work in our laboratory provides proof-of-concept for the successful incor- poration of the substrate envelope in the design of unique HIV protease inhibitors, which maintain high affinities against a panel of multidrug resistant variants of HIV-1 protease (42, 48–53). As a general paradigm, design efforts incorporating the substrate envelope would facilitate a more rationale evaluation of drug candidates and lead to the development of more robust inhibitors that are less susceptible to resistance. Materials and Methods Protein Crystallization. The NS3/4A protease construct was expressed and purified as reported previously (33, 54), detailed in SI Text. Purified protein was concentrated to ∼3 mg∕mL and loaded on a HiLoad Superdex75 16/60 column equilibrated with 25 mM 2-(N-morpholino)ethanesulfonic acid (MES) at pH 6.5, 500 mM NaCl, 10% glycerol, 30 μM zinc chloride, and 2 mM DTT. The protease fractions were pooled and concentrated to 20–25 mg∕mL with an Amicon Ultra-15 device (10 kD; Millipore). The concentrated samples were then incubated at 4 °C for 1 h with 2–20 M excess of viral substrate 4A4B, peptide products 4B5A or 5A5B, or ITMN-191. Information about the synth- esis of viral peptides and ITMN-191 is provided in SI Text. Diffraction-quality crystals were obtained overnight for all ligands by mixing equal volume of concentrated protein solution with precipitant solution (20–26% PEG-3350, 0.1 M sodium MES buffer at pH 6.5, and 4% ammonium sulfate) in 24-well VDX hanging drop trays. Data Collection and Structure Solution. Crystals were flash-frozen in liquid nitrogen and mounted under constant cryostream. X-ray diffraction data were collected at Advanced Photon Source BioCARS 14-IDB, 14-BMC, and LS-CAT 21-ID-F. Diffraction intensities of the complexes were indexed, integrated, and scaled using the programs HKL2000 (55) and XDS (56). 5% of the data was used to calculate R-free (57). All structure solutions were generated using isomorphous molecular replacement with PHASER (58) or AMORE. The NS3/4A protease domain (PDB code 2A4G) (59) was used for molecular replacement in solving the product 4A4B structure, and this struc- ture was subsequently used for solving the other complexes. In all cases, initial refinement was carried out in the absence of modeled ligand, which was subsequently built in. Phases were improved using ARP/wARP (60). Itera- tive rounds of translation, libration, and screw (TLS) and restrained refine- ment with CCP4 (61) and graphical model building with COOT (62) until convergence was achieved. The final structures were evaluated with MolProbity (63) prior to deposition in the protein data bank. Structural Analysis. Double-difference plots (64) were used to determine the structurally invariant regions of the protease, consisting of residues 32–36, 42–47, 50–54, 84–86, and 140–143. Structures were superposed with PyMOL (65) using the Cα atoms of these residues for all protease molecules from the solved structures (nine total). The B chain of product complex 4A4B was used as the reference structure in all alignments. Fit of individual inhibitors into the substrate envelope was quantified by mapping the substrate envelope and the van der Waals volume of each inhibitor on a three-dimensional grid with spacing of 0.5 Å. Vout for each drug moiety was computed by counting the grid cells, which were occupied by any inhi- bitor atom of that site but not the substrate envelope, and multiplied by the grid cell size, 0.125 Å3 (41, 42). Substrate Envelope and Inhibitor Analyses. NS3/4A substrate envelope was computed using product complexes 4A4B (B chain), 4B5A (D chain), and 5A5B (A chain). In structures with multiple protease molecules in the asym- metric unit, the one containing the most ordered peptide product was used for the alignment. The protease domain of the full-length NS3/4A structure (A chain; PDB code 1CU1) (39), including the C-terminal six amino acids, was included as a product complex 3-4A. All active site alignments were performed in PyMOL using Cα atoms of protease residues 137–139 and 154–160. After superposition, Gaussian object maps were generated in Py- MOL for each cleavage product. Four consensus Gaussian maps were then calculated, representing the intersecting volume of a group of three object maps. Finally, the summation of these four consensus maps was generated to construct the substrate envelope, depicting the van der Waals volume shared by any three of the four products. The previously determined boceprevir complex (PDB code 2OC8) (24) and TMC435 complex (PDB code 3KEE) (23) were used in this study (66). ACKNOWLEDGMENTS. We thank H. Klei for helpful discussions. We also thank Z. Wawrzak, M. Bolbat, and K. Brister of the LS-CAT beamline at Argonne National Laboratory for data collection of the ITMN-191 complex; M. Nalam and R. Bandaranayake for assistance with structural refinement; A. Ozen for providing V out calculations; and S. Shandilya and Y. Cai for computational support. The National Institute of Health (NIH) Grants R01-GM65347 and R01-AI085051 supported this work. Use of Advanced Photon Source (APS) was supported by the Department of Energy (DOE), Basic Energy Sciences, Office of Science, under Contract No. DE-AC02-06CH11357. Use of the Bio- CARS Sector 14 was supported by NIH-NCRR RR007707. Use of the LS-CAT Sector 21 was supported by the Michigan Economic Development Corpora- tion and the Michigan Technology TriCorridor under Grant 085P1000817. 1. World Health Organization. Barnes E, ed. (2010) Vaccine research: hepatitis C. Hepa- titis C Virus: Disease Burden. Available at http://www.who.int/vaccine_research/ diseases/viral_cancers/en/index2.html. March 15, 2010. 2. Major ME, Feinstone SM (1997) The molecular virology of hepatitis C. Hepatology 25:1527–1538. 3. Qureshi SA (2007) Hepatitis C virus—biology, host evasion strategies, and promising new therapies on the horizon. Med Res Rev 27:353–373. 4. Martell M, et al. (1992) Hepatitis C virus (HCV) circulates as a population of different but closely related genomes: quasispecies nature of HCV genome distribution. J Virol 66:3225–3229. 5. Paolucci S, et al. (2001) Analysis of HIV drug-resistant quasispecies in plasma, peripheral blood mononuclear cells and viral isolates from treatment-naive and HAART patients. J Med Virol 65:207–217. 6. Chen Z, et al. (2007) GB virus B disrupts RIG-I signaling by NS3/4A-mediated cleavage of the adaptor protein MAVS. J Virol 81:964–976. 7. Li XD, Sun L, Seth RB, Pineda G, Chen ZJ (2005) Hepatitis C virus protease NS3/4A cleaves mitochondrial antiviral signaling protein off the mitochondria to evade innate immunity. Proc Natl Acad Sci USA 102:17717–17722. 8. Li K, et al. (2005) Immune evasion by hepatitis C virus NS3/4A protease-mediated cleavage of the Toll-like receptor 3 adaptor protein TRIF. Proc Natl Acad Sci USA 102:2992–2997. 9. Llinas-Brunet M, et al. (1998) Peptide-based inhibitors of the hepatitis C virus serine protease. Bioorg Med Chem Lett 8:1713–1718. 10. Steinkuhler C, et al. (1998) Product inhibition of the hepatitis C virus NS3 protease. Biochemistry 37:8899–8905. 11. Arasappan A, et al. (2006) P2-P4 macrocyclic inhibitors of hepatitis C virus NS3-4A serine protease. Bioorg Med Chem Lett 16:3960–3965. 12. Bogen S, et al. (2008) Hepatitis C virus NS3-4A serine protease inhibitors: SAR of new P1 derivatives of SCH 503034. Bioorg Med Chem Lett 18:4219–4223. 13. Malancona S, et al. (2004) SAR and pharmacokinetic studies on phenethylamide inhibitors of the hepatitis C virus NS3/NS4A serine protease. Bioorg Med Chem Lett 14:4575–4579. 14. Nilsson M, et al. (2010) Synthesis and SAR of potent inhibitors of the hepatitis C virus NS3/4A protease: exploration of P2 quinazoline substituents. Bioorg Med Chem Lett 20:4004–4011. 15. Venkatraman S, et al. (2009) Discovery and structure-activity relationship of P1-P3 ketoamide derived macrocyclic inhibitors of hepatitis C virus NS3 protease. J Med Chem 52:336–346. 16. Raboisson P, et al. (2008) Structure-activity relationship study on a novel series of cyclopentane-containing macrocyclic inhibitors of the hepatitis C virus NS3/4A pro- tease leading to the discovery of TMC435350. Bioorg Med Chem Lett 18:4853–4858. 17. Perni RB, et al. (2004) Inhibitors of hepatitis C virus NS3.4A protease 2. Warhead SAR and optimization. Bioorg Med Chem Lett 14:1441–1446. 20990 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1006370107 Romano et al. 18. Zhou Y, et al. (2008) Phenotypic characterization of resistant Val36 variants of hepatitis C virus NS3-4A serine protease. Antimicrob Agents Chemother 52:110–120. 19. Zhou Y, et al. (2007) Phenotypic and structural analyses of hepatitis C virus NS3 protease Arg155 variants: sensitivity to telaprevir (VX-950) and interferon alpha. J Biol Chem 282:22619–22628. 20. Lamarre D, et al. (2003) An NS3 protease inhibitor with antiviral effects in humans infected with hepatitis C virus. Nature 426:186–189. 21. Hinrichsen H, et al. (2004) Short-term antiviral efficacy of BILN 2061, a hepatitis C virus serine protease inhibitor, in hepatitis C genotype 1 patients. Gastroenterology 127:1347–1355. 22. Vanwolleghem T, et al. (2007) Ultra-rapid cardiotoxicity of the hepatitis C virus protease inhibitor BILN 2061 in the urokinase-type plasminogen activator mouse. Gastroenterology 133:1144–1155. 23. Cummings MD, et al. (2010) Induced-fit binding of the macrocyclic noncovalent inhi- bitor TMC435 to its HCV NS3/NS4A protease target. Angew Chem Int Ed Engl 49:1652–1655. 24. Prongay AJ, et al. (2007) Discovery of the HCV NS3/4A protease inhibitor (1R,5S)-N- [3-amino-1-(cyclobutylmethyl)-2,3-dioxopropyl]-3- [2(S)-[[[(1,1-dimethylethyl)amino] carbonyl]amino]-3,3-dimethyl-1-oxobutyl]–6,6-dimethyl-3-azabicyclo[3.1.0]hexan-2(S)- carboxamide (Sch 503034) II. Key steps in structure-based optimization. J Med Chem 50:2310–2318. 25. Tong X, et al. (2008) Characterization of resistance mutations against HCV ketoamide protease inhibitors. Antiviral Res 77:177–185. 26. He Y, et al. (2008) Relative replication capacity and selective advantage profiles of protease inhibitor-resistant hepatitis C virus (HCV) NS3 protease mutants in the HCV genotype 1b replicon system. Antimicrob Agents Chemother 52:1101–1110. 27. Sarrazin C, et al. (2007) SCH 503034, a novel hepatitis C virus protease inhibitor, plus pegylated interferon alpha-2b for genotype 1 nonresponders. Gastroenterology 132:1270–1278. 28. Kieffer TL, et al. (2007) Telaprevir and pegylated interferon-alpha-2a inhibit wild-type and resistant genotype 1 hepatitis C virus replication in patients. Hepatology 46:631–639. 29. Yi M, Ma Y, Yates J, Lemon SM (2009) Transcomplementation of an NS2 defect in a late step in hepatitis C virus (HCV) particle assembly and maturation. PLoS Pathog 5:e1000403. 30. Lenz O, et al. (2010) In vitro resistance profile of the HCV NS3/4A protease inhibitor TMC435. Antimicrob Agents Chemother 54:1878–1887. 31. Arasappan A, Njoroge FG, Girijavallabhan VM (2005) PCT Int.Appl. WO 2005/113581 Patent Application. 32. Wang XA, et al. (2006) US Patent 6995174. 33. Wittekind M, Weinheirner S, Zhang Y, Goldfarb V (2002) US Patent 6333186. 34. Taremi SS, et al. (1998) Construction, expression, and characterization of a novel fully activated recombinant single-chain hepatitis C virus protease. Protein Sci 7:2143–2149. 35. Carter P, Wells JA (1988) Dissecting the catalytic triad of a serine protease. Nature 332:564–568. 36. Krishnan R, Sadler JE, Tulinsky A (2000) Structure of the Ser195Ala mutant of human alpha—thrombin complexed with fibrinopeptide A(7—16): evidence for residual catalytic activity. Acta Crystallogr D 56:406–410. 37. Kim JL, et al. (1996) Crystal structure of the hepatitis C virus NS3 protease domain complexed with a synthetic NS4A cofactor peptide. Cell 87:343–355. 38. Bode W, Huber R (1978) Crystal structure analysis and refinement of two variants of trigonal trypsinogen: trigonal trypsin and PEG (polyethylene glycol) trypsinogen and their comparison with orthorhombic trypsin and trigonal trypsinogen. FEBS Lett 90:265–269. 39. Yao N, Reichert P, Taremi SS, Prosise WW, Weber PC (1999) Molecular views of viral polyprotein processing revealed by the crystal structure of the hepatitis C virus bifunctional protease-helicase. Structure 7:1353–1363. 40. Krissinel E, Henrick K (2007) Inference of macromolecular assemblies from crystalline state. J Mol Biol 372:774–797. 41. Nalam MN, et al. (2010) Evaluating the substrate-envelope hypothesis: structural analysis of novel HIV-1 protease inhibitors designed to be robust against drug resis- tance. J Virol 84:5368–5378. 42. Chellappan S, Kairys V, Fernandes MX, Schiffer C, Gilson MK (2007) Evaluation of the substrate envelope hypothesis for inhibitors of HIV-1 protease. Proteins 68:561–567. 43. Kuntzen T, et al. (2008) Naturally occurring dominant resistance mutations to hepatitis C virus protease and polymerase inhibitors in treatment-naive patients. Hepatology 48:1769–1778. 44. Susser S, et al. (2009) Characterization of resistance to the protease inhibitor bocepre- vir in hepatitis C virus-infected patients. Hepatology 50:1709–1718. 45. Qiu P, et al. (2009) Identification of HCV protease inhibitor resistance mutations by selection pressure-based method. Nucleic Acids Res 37:e74. 46. Sarrazin C, et al. (2007) Dynamic hepatitis C virus genotypic and phenotypic changes in patients treated with the protease inhibitor telaprevir. Gastroenterology 132:1767–1777. 47. Tong X, et al. (2010) Preclinical characterization of the antiviral activity of SCH 900518 (Narlaprevir), a novel mechanism-based inhibitor of hepatitis C virus NS3 protease. Antimicrob Agents Chemother 54:2365–2370. 48. Chellappan S, et al. (2007) Design of mutation-resistant HIV protease inhibitors with the substrate envelope hypothesis. Chem Biol Drug Des 69:298–313. 49. Ali A, et al. (2009) Substrate envelope based design of new HIV-1 protease inhibitors active against drug-resistant HIV-1. 238th ACS National Meeting, Washington, DC, United States (American Chemical Society, Washington, DC), MEDI-102. 50. King NM, Prabu-Jeyabalan M, Nalivaika EA, Schiffer CA (2004) Combating susceptibil- ity to drug resistance: lessons from HIV-1 protease. Chem Biol 11:1333–1338. 51. Prabu-Jeyabalan M, et al. (2006) Substrate envelope and drug resistance: crystal struc- ture of RO1 in complex with wild-type human immunodeficiency virus type 1 protease. Antimicrob Agents Chemother 50:1518–1521. 52. Altman MD, Nalivaika EA, Prabu-Jeyabalan M, Schiffer CA, Tidor B (2008) Computa- tional design and experimental study of tighter binding peptides to an inactivated mutant of HIV-1 protease. Proteins 70:678–694. 53. Altman MD, et al. (2008) HIV-1 protease inhibitors from inverse design in the substrate envelope exhibit subnanomolar binding to drug-resistant variants. J Am Chem Soc 130:6099–6113. 54. Gallinari P, et al. (1998) Multiple enzymatic activities associated with recombinant NS3 protein of hepatitis C virus. J Virol 72:6758–6769. 55. Otwinowski Z, Minor W (1997) Processing of X-ray diffraction data collected in oscillation mode. Methods in Enzymology, 276 pp:307–326. 56. Kabsch W (1993) Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Crystallogr 26:795–800. 57. Brunger AT (1992) Free R value: a novel statistical quantity for assessing the accuracy of crystal structures. Nature 355:472–475. 58. McCoy AJ, et al. (2007) Phaser crystallographic software. J Appl Crystallogr 40:658–674. 59. Arasappan A, et al. (2005) Hepatitis C virus NS3-4A serine protease inhibitors: SAR of P′ 2 moiety with improved potency. Bioorg Med Chem Lett 15:4180–4184. 60. Morris RJ, Perrakis A, Lamzin VS (2002) ARP/wARP’s model-building algorithms. I. The main chain. Acta Crystallogr D 58:968–975. 61. Collaborative Computational Project Number 4 (1994) The CCP4 Suite: programs for protein crystallography. Acta Cryst, D50 pp:760–763. 62. Emsley P, Cowtan K (2004) COOT: model-building tools for molecular graphics. Acta Crystallogr D 60:2126–2132. 63. Davis IW, et al. (2007) MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res 35(Web Server issue):W375–383. 64. Prabu-Jeyabalan M, Nalivaika EA, Romano K, Schiffer CA (2006) Mechanism of sub- strate recognition by drug-resistant human immunodeficiency virus type 1 protease variants revealed by a novel structural intermediate. J Virol 80:3607–3616. 65. DeLano WL (2008) The PyMOL Molecular Graphics System (DeLano Scientific LLC, San Carlos, CA). 66. Berman HM, et al. (2000) The protein data bank. Nucleic Acids Res 28:235–242. Romano et al. PNAS ∣ December 7, 2010 ∣ vol. 107 ∣ no. 49 ∣ 20991 BIOPHYSICS AND COMPUTATIONAL BIOLOGY CHEMISTRY
3M5N
Crystal structure of HCV NS3/4A protease in complex with N-terminal product 4B5A
Drug resistance against HCV NS3/4A inhibitors is defined by the balance of substrate recognition versus inhibitor binding Keith P. Romano1, Akbar Ali1, William E. Royer, and Celia A. Schiffer2 Department of Biochemistry and Molecular Pharmacology, University of Massachusetts Medical School, Worcester, MA 01605 Edited by John M. Coffin, Tufts University School of Medicine, Boston, MA, and approved September 14, 2010 (received for review May 13, 2010) Hepatitis C virus infects an estimated 180 million people world- wide, prompting enormous efforts to develop inhibitors targeting the essential NS3/4A protease. Resistance against the most promis- ing protease inhibitors, telaprevir, boceprevir, and ITMN-191, has emerged in clinical trials. In this study, crystal structures of the NS3/4A protease domain reveal that viral substrates bind to the protease active site in a conserved manner defining a consensus volume, or substrate envelope. Mutations that confer the most severe resistance in the clinic occur where the inhibitors protrude from the substrate envelope, as these changes selectively weaken inhibitor binding without compromising the binding of substrates. These findings suggest a general model for predicting the suscept- ibility of protease inhibitors to resistance: drugs designed to fit within the substrate envelope will be less susceptible to resistance, as mutations affecting inhibitor binding would simultaneously interfere with the recognition of viral substrates. drug design ∣hepatitis C ∣substrate envelope D rug resistance is a major obstacle in the treatment of quickly evolving diseases. Hepatitis C virus (HCV) is a genetically diverse Hepacivirus of the Flaviviridae family infecting an esti- mated 180 million people worldwide (1). The viral RNA genome is translated as a single polyprotein and subsequently processed by host-cell and viral proteases into structural (C, E1, E2, and p7) and nonstructural (NS2, NS3, NS4A, NS4B, NS5A, and NS5B) proteins (2). The viral RNA-dependent RNA polymerase, NS5B, is inherently inaccurate and misincorporation of bases accounts for a very high mutation rate (3). While some mutations are neu- tral, others will alter the viability of the virus and propagate with varying efficiencies in each patient. Thus HCV infected indivi- duals will develop a heterogeneous population of virus variants known as quasispecies (4). As patients begin treatment, the selec- tive pressures of antiviral drugs will favor drug resistant variants (5). Therefore, an inhibitor must not only recognize one protein variant, but an ensemble of related enzymes. A detailed under- standing of the atomic mechanisms of resistance is essential to effectively combat drug resistance against HCV antivirals. The essential HCV NS3/4A protease is an attractive therapeu- tic target responsible for cleaving at least four sites along the viral polyprotein. These sites share little sequence homology except for an acid at position P6, Cys or Thr at P1, and Ser or Ala at P1′ (Table S1). The first cleavage event at the 3-4A junction occurs in cis as a unimolecular process, while the remaining sub- strates are processed bimolecularly in trans. The NS3/4A protease also cleaves the human cellular targets TRIF and MAVS, which confounds the innate immune response to viral infection (6–8). Early drug design efforts were hampered by the relatively shallow, featureless architecture of the protease active site. The eventual observation of N-terminal product inhibition served as a stepping stone for the discovery of more potent peptidomimetic inhibitors (9, 10). Over the past decade, pharmaceutical companies have further developed these lead compounds. Many structure-activ- ity-relationship (SAR) studies have been performed to evaluate the effect of different functional moieties on protease inhibition at positions P4-P1′ (11–17). Crystal structures have been deter- mined of the NS3/4A protease domain bound to a variety of inhibitors as well as of several drug resistant protease variants, such as R155K and V36M (18, 19). These data elucidate the mo- lecular interactions of NS3/4A with inhibitors and the effect of specific drug resistance mutations on binding. These efforts, con- ducted in parallel by several pharmaceutical companies, led to the discovery of many protease inhibitors. Proof-of-concept for the successful clinical activity of this drug class was first demon- strated by the macrocyclic inhibitor BILN-2061 (Boehringer Ingelheim) (20, 21), which was later dropped from clinical trials in 2006 due to cardiotoxicity (22). Many other NS3/4A protease inhibitors are currently in development, and telaprevir (Vertex), boceprevir (Schering-Plough), and ITMN-191 (Intermune) lead the way in advanced phases of human clinical trials (Fig. 1A). Despite these successes, the rapid acquisition of drug resis- tance has limited the efficacy of the most potent NS3/4A protease inhibitors in both replicon studies and human clinical trials (Fig. 1B and Table 1). In this study, we show that mutations con- ferring the most severe resistance occur where the protease extensively contacts the inhibitors but not the natural viral sub- strates. Four crystal structures of the NS3/4A protease domain in complex with the N-terminal products of viral substrates reveal a conserved mode of substrate binding, with the consensus volume defining the substrate envelope. The protease inhibitors ITMN- 191 (3M5L), TMC435 (3KEE) (23), and boceprevir (2OC8) (24) protrude extensively from the substrate envelope in regions that correlate with known sites of resistance mutations. Most notably, the P2 moieties of all three drugs protrude to contact A156 and R155, which mutate to confer high-level resistance against nearly all drugs reported in the literature (25–30). These findings sug- gest that drug resistance results from a change in molecular recognition and imply that drugs designed to fit within the sub- strate envelope will be less susceptible to resistance, as mutations altering inhibitor binding will simultaneously interfere with the binding of substrates. Results Synthesis of ITMN-191. We synthesized the macrocyclic inhibitor ITMN-191 using a convergent reaction sequence described in SI Text. Briefly, the P2 and P1-P1′ fragments were preassembled Author contributions: K.P.R., A.A., and C.A.S. designed research; K.P.R. and A.A. performed research; A.A. and W.E.R. contributed new reagents/analytic tools; K.P.R., W.E.R., and C.A.S. analyzed data; and K.P.R., A.A., W.E.R., and C.A.S. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Data deposition: The atomic coordinates and structure factors have been deposited in the Protein Data Bank, www.pdb.org (PDB ID codes 3M5L, 3M5M, 3M5N, and 3M5O). 1K.P.R. and A.A. contributed equally to this work. 2To whom correspondence should be addressed. E-mail: Celia.Schiffer@umassmed.edu. This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1006370107/-/DCSupplemental. 20986–20991 ∣PNAS ∣December 7, 2010 ∣vol. 107 ∣no. 49 www.pnas.org/cgi/doi/10.1073/pnas.1006370107 and the macrocyclic drug compound was generated by a four- step reaction sequence, including P2-P3 amide coupling, ester hydrolysis, coupling with the P1-P1′ fragment, and ring-closing metathesis. The P2-P3 fragment was assembled by coupling the commercially available Boc-protected amino acid (S)-2-(tert- butoxycarbonylamino)non-8-enoic acid (Acme Biosciences, Inc) with the preassembled P2 fragment, (3R, 5S)-5-(methoxy- carbonyl)pyrrolidin-3-yl 4-fluoroisoindoline-2-carboxylate (31), using O-(7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexa- fluorophosphate (HATU)/diisopropylethylamine (DIPEA). Hy- drolysis of the P2-P3 methyl ester with LiOH:H2O in a mixture of THF-MeOH-H2O followed by coupling of the resulting acid under HATU/DIPEA conditions with the preassembled P1-P1′ fragment, (1R, 2S)-1-amino-N-(cyclopropylsulfonyl)-2-vinylcyclo- propanecarboxamide (32), provided the bis-olefin precursor for ring-closing metathesis. Cyclization of the bis-olefin intermediate was accomplished using a highly efficient ring-closing metathesis catalyst Zhan 1B and provided the protease inhibitor ITMN-191. Structure Determination of Inhibitor and Substrate Complexes. Although NS3/4A cleaves the viral polyprotein of over 3,000 residues at four specific sites in vivo, we focused on the local interactions of the protease domain with short peptide sequences corresponding to the immediate cleavage sites. All structural studies were carried out with the highly soluble, single-chain con- struct of the NS3/4A protease domain described previously (33), which contains a fragment of the essential cofactor NS4A cova- lently linked at the N terminus by a flexible linker. A similar pro- tease construct was shown to retain comparable catalytic activity to the authentic protein complex (34). Crystallization trials were initially carried out using the inactive (S139A) protease variant in complex with substrate peptides spanning P7-P5′. The 4A4B substrate complex revealed cleavage of the scissile bond and no ordered regions for the C-terminal fragment of the substrate. Similar observations were previously described for two other serine proteases where catalytic activity was observed, presum- ably facilitated by water, despite Ala substitutions of the catalytic Ser (35, 36). Thus all subsequent crystallization trials with the NS3/4A protease were performed using N-terminal cleavage products of the viral substrates spanning P7-P1. NS3/4A crystal structures in complex with ITMN-191 and peptide products 4A4B, 4B5A, and 5A5B were determined and refined at 1.25 Å, 1.70 Å, 1.90 Å, and 1.60 Å resolution, respec- tively (Table S2). The complexes crystallized in the space groups P212121 and P21 with one, two, or four molecules in the asym- metric unit. The average B factors range from 16.8–29.7 Å2 and there are no outliers in the Ramachandran plots. These structures represent the highest resolution crystal structures of NS3/4A protease reported to date. Overall Structure Analysis. The NS3/4A protease domain adopts a tertiary fold characteristic of serine proteases of the chymotrypsin family (37, 38). A total of nine protease molecules were modeled in the four crystal structures solved in this study with an overall rms deviation (rmsd) of 0.28 Å. The rmsds reveal the five most variable regions of the protease to be (Fig. S1): (i) the linker con- necting cofactor 4A at the N terminus, (ii) the loop containing residues 65–70, (iii) the zinc-binding site containing residues 95–105, (iv) the 310 helix region spanning residues 128–136, and (v) the active site antiparallel β-sheet containing residues 156–168. These structural differences likely indicate inherent flexibility in the protease and do not appear to correlate with ligand type or active site occupancy. Analysis of Product Complexes. Product complexes 4A4B, 4B5A, and 5A5B were further analyzed with the C terminus of the full-length NS3/4A structure (1CU1), which contains the N-term- inal cleavage product of viral substrate 3-4A (39). All four products bind to the protease active site in a conserved manner (Fig. 2), forming an antiparallel β-sheet with residues 154–160 Fig. 1. NS3/4A protease inhibitors and reported sites of drug resistance. (A) The leading protease inhibitors in development mimic the N-terminal side of the viral substrates. (B) The majority of reported drug resistance mutations cluster around the protease active site with the catalytic triad depicted in yellow. Table 1. Drug resistance mutations reported in replicon studies and clinical trials* Residue Mutation Drug V36 A, M, L, G Boceprevir, telaprevir Q41 R Boceprevir, ITMN-191 F43 S, C, V, I Boceprevir, telaprevir, ITMN-191, TMC435 V55 A Boceprevir T54 A, S Boceprevir, telaprevir Q80 K, R, H, G, L TMC435 S138 T ITMN-191, TMC435† R155 K, T, I, M, G, L, S, Q Boceprevir, telaprevir, ITMN-191, BILN-2061, TMC435 A156 V, T, S, I, G Boceprevir, telaprevir, ITMN-191, BILN-2061, TMC435 V158 I Boceprevir D168 A, V, E, G, N, T, Y, H, I ITMN-191, BILN-2061, TMC435 V170 A Boceprevir, telaprevir M175 L Boceprevir *References (18, 25, 26, 28, 30–37). †TMC435 displays reduced activity against S138T, but the mutation was not observed in selection experiments. Romano et al. PNAS ∣ December 7, 2010 ∣ vol. 107 ∣ no. 49 ∣ 20987 BIOPHYSICS AND COMPUTATIONAL BIOLOGY CHEMISTRY and burying 500–600 Å2 of solvent accessible surface area as cal- culated by PISA (40). The peptide backbone torsions are very similar, being most conserved at position P1 and deviating slightly toward position P4. Eight hydrogen bonds between backbone amide and carbonyl groups are completely conserved, involving protease residues S159 (C159 in product 3-4A), A157, R155, S139A, S138, and G137. S159 (C159 in product 3-4A), and A157 each contribute two hydrogen bonds with the P5 and P3 peptide residues, respectively. All P1 terminal carboxyl groups sit in the oxyanion hole, hydrogen bonding with the Nϵ atom of H57 and the amide nitrogens of residues 137–139. Although only product 4B5A contains a proline at P2, the other substrate sequences still adopt constrained P2 φ torsion angles. Thus products bind similarly despite their high sequence diversity. The P1 and P6 residues are most conserved among the substrate sequences, as are most of their interactions with the protease. The P1 side chains interact with the aromatic ring of F154. In all structures but product complex 4B5A, K165 forms salt-bridges with the P6 acids, while residues R123, D168, R155, and the catalytic D81 form an ionic network along one surface of the bound products (Fig. 2). In complex 4B5A, R123 interacts directly with the P6 acid, while D168 reorients and no longer contacts R155. Other molecular interactions in the product complexes are more diverse. Notably, K136 interacts differently with the cleavage products, forming: (i) a hydrogen bond with the P2 backbone carbonyl oxygen of 3-4A; (ii) a salt-bridge with the P3 glutamate of product 4A4B; and (iii) non- specific van der Waals interactions with the P2 and P3 side chains of products 4B5A and 5A5B. Also, in product complex 4A4B, an intramolecular hydrogen bond forms between the P3 and P5 glu- tamate residues, while the unique P4 acid of product 3-4A forms salt-bridges with the guanidinium groups of R123 and R155. Thus distinct patterns of side chain interactions underlie the set of con- served features involved in NS3/4A cleavage product binding. The Substrate Envelope. To further analyze the structural similari- ties of the four NS3/4A product complexes, the active sites were superposed on the Cα atoms of residues 137–139 and 154–160, revealing that both the active site residues and substrate products spanning P6-P1 align closely with an average Cα rmsd of 0.24 Å and 0.35 Å, respectively. The consensus van der Waals volume shared by any three of the four cleavage products was then cal- culated to generate the NS3/4A substrate envelope (Fig. 3A). This shape could not be predicted by the primary sequences alone and highlights the conserved mode of viral substrate recognition despite their high sequence diversity. Analysis of Inhibitor Complexes. ITMN-191, TMC435, and bocepre- vir are all peptidomimetic NS3/4A protease inhibitors. Active site superpositions of these drug complexes reveal that the inhibitors interact with many of the same protease residues as the cleavage products. Despite the P3-P1 cyclization of ITMN-191 and TMC435, the functional groups are positioned similarly in all three inhibitor complexes. The P1 cysteine surrogates interact with the aromatic ring of F154, while the P2 and P3 moieties over- lap closely. Although TMC435 does not contain a P4 substituent, the P4 tert-butyl groups of ITMN-191 and boceprevir also align closely. In addition, the P1 and P3 backbone atoms of all inhibi- tors hydrogen bond with the carbonyl oxygens of R155 and A157, respectively. These observations verify the peptidomimetic nat- ure of these drugs and support their observed mechanism as competitive active site inhibitors. The largest variation between these three protease inhibitors occurs at P2 where the aromatic rings of ITMN-191 and TMC435 stack against the guanidinium group of R155 (Fig. 3). This molecular interaction alters the electrostatic network involving R123, D168, R155, and D81. R155 rotates nearly 180° around Cδ relative to its conformation observed in product complexes, losing its hydrogen bond with D81 but maintaining interaction with D168. Mutations at R155 or D168 would disrupt the elec- trostatic network and destabilize this packing thereby lowering the affinity of these macrocyclic drugs. This observation provides a structural rationale for the drug resistance mutations R155K, as previously proposed (19), and D168A/V, which both confer a selective advantage in vitro in the presence of ITMN-191 or TMC435 (26, 30). In addition, the TMC435 complex reveals that R155 is stabilized by a hydrogen bond with Q80, which also mutates to confer resistance to TMC435 (30). Thus many of the primary drug resistance mutations can be explained by the disrup- tion of atomic interactions involving the P2 functional groups of the drugs. Insights into Drug Resistance. To determine the locations where the inhibitors protrude from the substrate envelope, the inhibitor and product complexes were also superposed using residues 137–139 and 154–160. The van der Waals volumes of inhibitor protrusion from the substrate envelope (V out) (41, 42) were calculated for each drug and compared with published EC50 fold-change data for drug resistance variants (30). The magnitudes of the EC50 fold-change data determined for each NS3/4A mutant generally trend with the V out values for the three drugs. The P2 moieties of boceprevir, ITMN-191 and TMC435 protrude most extensively from the substrate envelope with V out values of 105, 294, and Fig. 2. Stereo view of N-terminal cleavage product binding to NS3/4A pro- tease. N-terminal protease cleavage products (A) 3-4A, (B) 4A4B, (C) 4B5A, and (D) 5A5B are depicted as they bind to the protease active site. All conserved interactions are indicated by black dashes, while red lines depict interactions that are not present in all structures. The electrostatic network involving residues R123, D168, R155, and D81 is indicated by blue dashes. 20988 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1006370107 Romano et al. 496 Å3, respectively (Table 2). However, the precise level of drug resistance observed is also determined by the particular change in molecular interaction occurring for a given mutation. For ex- ample, A156 and R155 pack with the P2 moieties of these three inhibitors where they protrude beyond the substrate envelope. Mutations of A156 to bulkier side chains would result in a steric clash with the P2 drug moieties. Indeed, the rigid dimethylcyclo- propane group of boceprevir protrudes from the substrate envel- ope at the P2 subsite, and A156V or A156T confer 65 and 75 fold-changes in EC50, respectively (Table 2). Similarly, molecular changes at R155 and D168 would result in a substantial loss of interactions with P2. The most extensive protrusions of ITMN- 191 and TMC435 at P2 trend with their greatest fold-change in potency of nearly 450 and 600, respectively, from mutations in this subsite. Thus the extent by which an inhibitor protrudes from the substrate envelope in a given subsite is indicative of its vulnerability to resistance. Further structural analyses with the substrate envelope provide insights into other NS3/4A drug resistance mutations. The P1′ sulfonamide groups of ITMN-191 and TMC435, as well as the P1′ ketoamide of boceprevir, protrude from the substrate envel- ope near residues Q41 and F43, which both mutate to confer low-level resistance to these drugs (25, 30, 43). The keto group of boceprevir also projects outside the substrate envelope near T54 and V55. T54A/S confers low-level resistance to boceprevir, while V55A was recently identified in patient isolates after treat- ment with boceprevir (44). The analogous carbonyl groups of ITMN-191 and TMC435, however, are orientated in the opposite direction and protrude toward S138. In fact, in vitro studies reveal reduced activity for ITMN-191 and TMC435 against S138T variants, while boceprevir remains fully active (30, 43). The bulky P4 tert-butyl group of boceprevir extends outside the substrate envelope contacting V158; the V158I variant has lower affinity for this drug, likely due to a steric clash (45). This variant may also impact the affinity of ITMN-191, as its P4 tert-butyl also pro- trudes at the same location. These findings demonstrate that in regions outside the P2 subsite, positions where ITMN-191, TMC435, and boceprevir protrude from the substrate envelope also correlate with many other known sites of drug resistance mutations. Discussion The emergence of drug resistance is a major obstacle in modern medicine that limits the long-term usefulness of the most promis- ing therapeutics. By considering how HCV NS3/4A protease in- hibitors bind relative to natural viral substrates, we discovered that primary sites of resistance occur in regions of the protease where drugs protrude from the substrate envelope. In particular, R155 and A156, which mutate to confer severe resistance against ITMN-191, TMC435, and boceprevir, interact closely with the P2 drug moieties where they protrude most extensively from the substrate envelope. Molecular changes at these residues confer resistance by selectively weakening inhibitor binding without compromising the binding of viral substrates. We further specu- late that these mutations will not considerably affect the binding of the host cellular substrates TRIF and MAVS, which likely fit well within the substrate envelope as they share many features with the viral substrates. However, TRIF contains a track of eight proline residues instead of an acidic residue at position P6, which may modulate its binding. Further structural studies are warranted to better ascertain the molecular details of how these cellular substrates are recognized by the NS3/4A protease. Although this study focuses on ITMN-191, TMC435, and boceprevir, other NS3/4A protease inhibitors in clinical trials, in- cluding telaprevir, narlaprevir, and vaniprevir (Fig. 1A), contain similar functional groups that likely protrude from the substrate envelope. Most notably, all these drug candidates contain bulky P2 moieties and are therefore susceptible to cross-resistance against mutations at R155 and A156. R155 and A156 mutations have been shown to confer telaprevir resistance in treated patients (46). Cross-resistance studies have also shown that nar- Table 2. EC50 fold-change (FC) data * for several NS3/4A drug resistant variants tabulated with Vout, the van der Waals volume of protrusion from the substrate envelope, at each subsite of the enzyme Boceprevir ITMN-191 TMC435 Subsite Resistance mutation Vout (Å3) EC50 FC Vout (Å3) EC50 FC Vout (Å3) EC50 FC Total 292 500 649 P1 76 67 64 P2 105 294 496 Q80R 0.5 3.5 6.9 Q80K 0.8 2.3 7.7 R155K 4.7 447 30 A156V 75 63 177 A156T 65 41 44 D168A 0.7 153 594 D168E 0.8 75 40 P3 34 70 67 P4 76 69 0 V158I 3.3† ND ND *Antiviral activity was reported previously by Lenz et al., 2010 (30). †Fold-change in EC50 reported in replicon assay by Qiu et al., 2009 (45). Fig. 3. Stereo view of the NS3/4A substrate envelope and protease inhibitors. (A) After active site superpositions, the overlapping van der Waals volume shared by any three of the four cleavage products defines the substrate envelope, depicted in blue. NS3/4A protease residues which mutate to confer drug resistance are shown in brown. (B) ITMN-191, (C) boceprevir and (D) TMC435 protrude from the substrate envelope at several locations, which correlate with known sites of drug-resistant mutations to each inhibi- tor, shown in red. Romano et al. PNAS ∣ December 7, 2010 ∣ vol. 107 ∣ no. 49 ∣ 20989 BIOPHYSICS AND COMPUTATIONAL BIOLOGY CHEMISTRY laprevir displays similar fold losses in activity against most of the known drug resistance mutations for telaprevir and boceprevir (47). Ultimately, to slow the emergence of multidrug resistant viral strains, inhibitors should be confined within the substrate envelope, particularly at the P2 position. To compensate for the loss of binding affinity that will likely accompany these changes, additional interactions could potentially be optimized spanning the S4-S6 subsites of the protease and the catalytic triad. Our findings further suggest a general model for using the sub- strate envelope to predict patterns of drug resistance in other quickly evolving diseases. For drug resistance to occur, mutations must selectively weaken a target’s affinity for an inhibitor without significantly altering its natural biological function. Mutations occurring outside the substrate envelope are better able to achieve this effect, as these molecular changes can selectively alter inhibitor binding without compromising the binding of natural substrates. Whenever the interaction of a drug target with its biological substrates can be structurally characterized, we pre- dict that drugs designed to fit within the substrate envelope will be less susceptible to resistance. Structure-based design strategies can utilize this model as an added constraint to develop inhibitors that fit within the substrate envelope. In fact, previous work in our laboratory provides proof-of-concept for the successful incor- poration of the substrate envelope in the design of unique HIV protease inhibitors, which maintain high affinities against a panel of multidrug resistant variants of HIV-1 protease (42, 48–53). As a general paradigm, design efforts incorporating the substrate envelope would facilitate a more rationale evaluation of drug candidates and lead to the development of more robust inhibitors that are less susceptible to resistance. Materials and Methods Protein Crystallization. The NS3/4A protease construct was expressed and purified as reported previously (33, 54), detailed in SI Text. Purified protein was concentrated to ∼3 mg∕mL and loaded on a HiLoad Superdex75 16/60 column equilibrated with 25 mM 2-(N-morpholino)ethanesulfonic acid (MES) at pH 6.5, 500 mM NaCl, 10% glycerol, 30 μM zinc chloride, and 2 mM DTT. The protease fractions were pooled and concentrated to 20–25 mg∕mL with an Amicon Ultra-15 device (10 kD; Millipore). The concentrated samples were then incubated at 4 °C for 1 h with 2–20 M excess of viral substrate 4A4B, peptide products 4B5A or 5A5B, or ITMN-191. Information about the synth- esis of viral peptides and ITMN-191 is provided in SI Text. Diffraction-quality crystals were obtained overnight for all ligands by mixing equal volume of concentrated protein solution with precipitant solution (20–26% PEG-3350, 0.1 M sodium MES buffer at pH 6.5, and 4% ammonium sulfate) in 24-well VDX hanging drop trays. Data Collection and Structure Solution. Crystals were flash-frozen in liquid nitrogen and mounted under constant cryostream. X-ray diffraction data were collected at Advanced Photon Source BioCARS 14-IDB, 14-BMC, and LS-CAT 21-ID-F. Diffraction intensities of the complexes were indexed, integrated, and scaled using the programs HKL2000 (55) and XDS (56). 5% of the data was used to calculate R-free (57). All structure solutions were generated using isomorphous molecular replacement with PHASER (58) or AMORE. The NS3/4A protease domain (PDB code 2A4G) (59) was used for molecular replacement in solving the product 4A4B structure, and this struc- ture was subsequently used for solving the other complexes. In all cases, initial refinement was carried out in the absence of modeled ligand, which was subsequently built in. Phases were improved using ARP/wARP (60). Itera- tive rounds of translation, libration, and screw (TLS) and restrained refine- ment with CCP4 (61) and graphical model building with COOT (62) until convergence was achieved. The final structures were evaluated with MolProbity (63) prior to deposition in the protein data bank. Structural Analysis. Double-difference plots (64) were used to determine the structurally invariant regions of the protease, consisting of residues 32–36, 42–47, 50–54, 84–86, and 140–143. Structures were superposed with PyMOL (65) using the Cα atoms of these residues for all protease molecules from the solved structures (nine total). The B chain of product complex 4A4B was used as the reference structure in all alignments. Fit of individual inhibitors into the substrate envelope was quantified by mapping the substrate envelope and the van der Waals volume of each inhibitor on a three-dimensional grid with spacing of 0.5 Å. Vout for each drug moiety was computed by counting the grid cells, which were occupied by any inhi- bitor atom of that site but not the substrate envelope, and multiplied by the grid cell size, 0.125 Å3 (41, 42). Substrate Envelope and Inhibitor Analyses. NS3/4A substrate envelope was computed using product complexes 4A4B (B chain), 4B5A (D chain), and 5A5B (A chain). In structures with multiple protease molecules in the asym- metric unit, the one containing the most ordered peptide product was used for the alignment. The protease domain of the full-length NS3/4A structure (A chain; PDB code 1CU1) (39), including the C-terminal six amino acids, was included as a product complex 3-4A. All active site alignments were performed in PyMOL using Cα atoms of protease residues 137–139 and 154–160. After superposition, Gaussian object maps were generated in Py- MOL for each cleavage product. Four consensus Gaussian maps were then calculated, representing the intersecting volume of a group of three object maps. Finally, the summation of these four consensus maps was generated to construct the substrate envelope, depicting the van der Waals volume shared by any three of the four products. The previously determined boceprevir complex (PDB code 2OC8) (24) and TMC435 complex (PDB code 3KEE) (23) were used in this study (66). ACKNOWLEDGMENTS. We thank H. Klei for helpful discussions. We also thank Z. Wawrzak, M. Bolbat, and K. Brister of the LS-CAT beamline at Argonne National Laboratory for data collection of the ITMN-191 complex; M. Nalam and R. Bandaranayake for assistance with structural refinement; A. Ozen for providing V out calculations; and S. Shandilya and Y. Cai for computational support. The National Institute of Health (NIH) Grants R01-GM65347 and R01-AI085051 supported this work. Use of Advanced Photon Source (APS) was supported by the Department of Energy (DOE), Basic Energy Sciences, Office of Science, under Contract No. DE-AC02-06CH11357. Use of the Bio- CARS Sector 14 was supported by NIH-NCRR RR007707. Use of the LS-CAT Sector 21 was supported by the Michigan Economic Development Corpora- tion and the Michigan Technology TriCorridor under Grant 085P1000817. 1. World Health Organization. Barnes E, ed. (2010) Vaccine research: hepatitis C. Hepa- titis C Virus: Disease Burden. Available at http://www.who.int/vaccine_research/ diseases/viral_cancers/en/index2.html. March 15, 2010. 2. Major ME, Feinstone SM (1997) The molecular virology of hepatitis C. Hepatology 25:1527–1538. 3. Qureshi SA (2007) Hepatitis C virus—biology, host evasion strategies, and promising new therapies on the horizon. Med Res Rev 27:353–373. 4. Martell M, et al. (1992) Hepatitis C virus (HCV) circulates as a population of different but closely related genomes: quasispecies nature of HCV genome distribution. J Virol 66:3225–3229. 5. Paolucci S, et al. (2001) Analysis of HIV drug-resistant quasispecies in plasma, peripheral blood mononuclear cells and viral isolates from treatment-naive and HAART patients. J Med Virol 65:207–217. 6. Chen Z, et al. (2007) GB virus B disrupts RIG-I signaling by NS3/4A-mediated cleavage of the adaptor protein MAVS. J Virol 81:964–976. 7. Li XD, Sun L, Seth RB, Pineda G, Chen ZJ (2005) Hepatitis C virus protease NS3/4A cleaves mitochondrial antiviral signaling protein off the mitochondria to evade innate immunity. Proc Natl Acad Sci USA 102:17717–17722. 8. Li K, et al. (2005) Immune evasion by hepatitis C virus NS3/4A protease-mediated cleavage of the Toll-like receptor 3 adaptor protein TRIF. Proc Natl Acad Sci USA 102:2992–2997. 9. Llinas-Brunet M, et al. (1998) Peptide-based inhibitors of the hepatitis C virus serine protease. Bioorg Med Chem Lett 8:1713–1718. 10. Steinkuhler C, et al. (1998) Product inhibition of the hepatitis C virus NS3 protease. Biochemistry 37:8899–8905. 11. Arasappan A, et al. (2006) P2-P4 macrocyclic inhibitors of hepatitis C virus NS3-4A serine protease. Bioorg Med Chem Lett 16:3960–3965. 12. Bogen S, et al. (2008) Hepatitis C virus NS3-4A serine protease inhibitors: SAR of new P1 derivatives of SCH 503034. Bioorg Med Chem Lett 18:4219–4223. 13. Malancona S, et al. (2004) SAR and pharmacokinetic studies on phenethylamide inhibitors of the hepatitis C virus NS3/NS4A serine protease. Bioorg Med Chem Lett 14:4575–4579. 14. Nilsson M, et al. (2010) Synthesis and SAR of potent inhibitors of the hepatitis C virus NS3/4A protease: exploration of P2 quinazoline substituents. Bioorg Med Chem Lett 20:4004–4011. 15. Venkatraman S, et al. (2009) Discovery and structure-activity relationship of P1-P3 ketoamide derived macrocyclic inhibitors of hepatitis C virus NS3 protease. J Med Chem 52:336–346. 16. Raboisson P, et al. (2008) Structure-activity relationship study on a novel series of cyclopentane-containing macrocyclic inhibitors of the hepatitis C virus NS3/4A pro- tease leading to the discovery of TMC435350. Bioorg Med Chem Lett 18:4853–4858. 17. Perni RB, et al. (2004) Inhibitors of hepatitis C virus NS3.4A protease 2. Warhead SAR and optimization. Bioorg Med Chem Lett 14:1441–1446. 20990 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1006370107 Romano et al. 18. Zhou Y, et al. (2008) Phenotypic characterization of resistant Val36 variants of hepatitis C virus NS3-4A serine protease. Antimicrob Agents Chemother 52:110–120. 19. Zhou Y, et al. (2007) Phenotypic and structural analyses of hepatitis C virus NS3 protease Arg155 variants: sensitivity to telaprevir (VX-950) and interferon alpha. J Biol Chem 282:22619–22628. 20. Lamarre D, et al. (2003) An NS3 protease inhibitor with antiviral effects in humans infected with hepatitis C virus. Nature 426:186–189. 21. Hinrichsen H, et al. (2004) Short-term antiviral efficacy of BILN 2061, a hepatitis C virus serine protease inhibitor, in hepatitis C genotype 1 patients. Gastroenterology 127:1347–1355. 22. Vanwolleghem T, et al. (2007) Ultra-rapid cardiotoxicity of the hepatitis C virus protease inhibitor BILN 2061 in the urokinase-type plasminogen activator mouse. Gastroenterology 133:1144–1155. 23. Cummings MD, et al. (2010) Induced-fit binding of the macrocyclic noncovalent inhi- bitor TMC435 to its HCV NS3/NS4A protease target. Angew Chem Int Ed Engl 49:1652–1655. 24. Prongay AJ, et al. (2007) Discovery of the HCV NS3/4A protease inhibitor (1R,5S)-N- [3-amino-1-(cyclobutylmethyl)-2,3-dioxopropyl]-3- [2(S)-[[[(1,1-dimethylethyl)amino] carbonyl]amino]-3,3-dimethyl-1-oxobutyl]–6,6-dimethyl-3-azabicyclo[3.1.0]hexan-2(S)- carboxamide (Sch 503034) II. Key steps in structure-based optimization. J Med Chem 50:2310–2318. 25. Tong X, et al. (2008) Characterization of resistance mutations against HCV ketoamide protease inhibitors. Antiviral Res 77:177–185. 26. He Y, et al. (2008) Relative replication capacity and selective advantage profiles of protease inhibitor-resistant hepatitis C virus (HCV) NS3 protease mutants in the HCV genotype 1b replicon system. Antimicrob Agents Chemother 52:1101–1110. 27. Sarrazin C, et al. (2007) SCH 503034, a novel hepatitis C virus protease inhibitor, plus pegylated interferon alpha-2b for genotype 1 nonresponders. Gastroenterology 132:1270–1278. 28. Kieffer TL, et al. (2007) Telaprevir and pegylated interferon-alpha-2a inhibit wild-type and resistant genotype 1 hepatitis C virus replication in patients. Hepatology 46:631–639. 29. Yi M, Ma Y, Yates J, Lemon SM (2009) Transcomplementation of an NS2 defect in a late step in hepatitis C virus (HCV) particle assembly and maturation. PLoS Pathog 5:e1000403. 30. Lenz O, et al. (2010) In vitro resistance profile of the HCV NS3/4A protease inhibitor TMC435. Antimicrob Agents Chemother 54:1878–1887. 31. Arasappan A, Njoroge FG, Girijavallabhan VM (2005) PCT Int.Appl. WO 2005/113581 Patent Application. 32. Wang XA, et al. (2006) US Patent 6995174. 33. Wittekind M, Weinheirner S, Zhang Y, Goldfarb V (2002) US Patent 6333186. 34. Taremi SS, et al. (1998) Construction, expression, and characterization of a novel fully activated recombinant single-chain hepatitis C virus protease. Protein Sci 7:2143–2149. 35. Carter P, Wells JA (1988) Dissecting the catalytic triad of a serine protease. Nature 332:564–568. 36. Krishnan R, Sadler JE, Tulinsky A (2000) Structure of the Ser195Ala mutant of human alpha—thrombin complexed with fibrinopeptide A(7—16): evidence for residual catalytic activity. Acta Crystallogr D 56:406–410. 37. Kim JL, et al. (1996) Crystal structure of the hepatitis C virus NS3 protease domain complexed with a synthetic NS4A cofactor peptide. Cell 87:343–355. 38. Bode W, Huber R (1978) Crystal structure analysis and refinement of two variants of trigonal trypsinogen: trigonal trypsin and PEG (polyethylene glycol) trypsinogen and their comparison with orthorhombic trypsin and trigonal trypsinogen. FEBS Lett 90:265–269. 39. Yao N, Reichert P, Taremi SS, Prosise WW, Weber PC (1999) Molecular views of viral polyprotein processing revealed by the crystal structure of the hepatitis C virus bifunctional protease-helicase. Structure 7:1353–1363. 40. Krissinel E, Henrick K (2007) Inference of macromolecular assemblies from crystalline state. J Mol Biol 372:774–797. 41. Nalam MN, et al. (2010) Evaluating the substrate-envelope hypothesis: structural analysis of novel HIV-1 protease inhibitors designed to be robust against drug resis- tance. J Virol 84:5368–5378. 42. Chellappan S, Kairys V, Fernandes MX, Schiffer C, Gilson MK (2007) Evaluation of the substrate envelope hypothesis for inhibitors of HIV-1 protease. Proteins 68:561–567. 43. Kuntzen T, et al. (2008) Naturally occurring dominant resistance mutations to hepatitis C virus protease and polymerase inhibitors in treatment-naive patients. Hepatology 48:1769–1778. 44. Susser S, et al. (2009) Characterization of resistance to the protease inhibitor bocepre- vir in hepatitis C virus-infected patients. Hepatology 50:1709–1718. 45. Qiu P, et al. (2009) Identification of HCV protease inhibitor resistance mutations by selection pressure-based method. Nucleic Acids Res 37:e74. 46. Sarrazin C, et al. (2007) Dynamic hepatitis C virus genotypic and phenotypic changes in patients treated with the protease inhibitor telaprevir. Gastroenterology 132:1767–1777. 47. Tong X, et al. (2010) Preclinical characterization of the antiviral activity of SCH 900518 (Narlaprevir), a novel mechanism-based inhibitor of hepatitis C virus NS3 protease. Antimicrob Agents Chemother 54:2365–2370. 48. Chellappan S, et al. (2007) Design of mutation-resistant HIV protease inhibitors with the substrate envelope hypothesis. Chem Biol Drug Des 69:298–313. 49. Ali A, et al. (2009) Substrate envelope based design of new HIV-1 protease inhibitors active against drug-resistant HIV-1. 238th ACS National Meeting, Washington, DC, United States (American Chemical Society, Washington, DC), MEDI-102. 50. King NM, Prabu-Jeyabalan M, Nalivaika EA, Schiffer CA (2004) Combating susceptibil- ity to drug resistance: lessons from HIV-1 protease. Chem Biol 11:1333–1338. 51. Prabu-Jeyabalan M, et al. (2006) Substrate envelope and drug resistance: crystal struc- ture of RO1 in complex with wild-type human immunodeficiency virus type 1 protease. Antimicrob Agents Chemother 50:1518–1521. 52. Altman MD, Nalivaika EA, Prabu-Jeyabalan M, Schiffer CA, Tidor B (2008) Computa- tional design and experimental study of tighter binding peptides to an inactivated mutant of HIV-1 protease. Proteins 70:678–694. 53. Altman MD, et al. (2008) HIV-1 protease inhibitors from inverse design in the substrate envelope exhibit subnanomolar binding to drug-resistant variants. J Am Chem Soc 130:6099–6113. 54. Gallinari P, et al. (1998) Multiple enzymatic activities associated with recombinant NS3 protein of hepatitis C virus. J Virol 72:6758–6769. 55. Otwinowski Z, Minor W (1997) Processing of X-ray diffraction data collected in oscillation mode. Methods in Enzymology, 276 pp:307–326. 56. Kabsch W (1993) Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Crystallogr 26:795–800. 57. Brunger AT (1992) Free R value: a novel statistical quantity for assessing the accuracy of crystal structures. Nature 355:472–475. 58. McCoy AJ, et al. (2007) Phaser crystallographic software. J Appl Crystallogr 40:658–674. 59. Arasappan A, et al. (2005) Hepatitis C virus NS3-4A serine protease inhibitors: SAR of P′ 2 moiety with improved potency. Bioorg Med Chem Lett 15:4180–4184. 60. Morris RJ, Perrakis A, Lamzin VS (2002) ARP/wARP’s model-building algorithms. I. The main chain. Acta Crystallogr D 58:968–975. 61. Collaborative Computational Project Number 4 (1994) The CCP4 Suite: programs for protein crystallography. Acta Cryst, D50 pp:760–763. 62. Emsley P, Cowtan K (2004) COOT: model-building tools for molecular graphics. Acta Crystallogr D 60:2126–2132. 63. Davis IW, et al. (2007) MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res 35(Web Server issue):W375–383. 64. Prabu-Jeyabalan M, Nalivaika EA, Romano K, Schiffer CA (2006) Mechanism of sub- strate recognition by drug-resistant human immunodeficiency virus type 1 protease variants revealed by a novel structural intermediate. J Virol 80:3607–3616. 65. DeLano WL (2008) The PyMOL Molecular Graphics System (DeLano Scientific LLC, San Carlos, CA). 66. Berman HM, et al. (2000) The protein data bank. Nucleic Acids Res 28:235–242. Romano et al. PNAS ∣ December 7, 2010 ∣ vol. 107 ∣ no. 49 ∣ 20991 BIOPHYSICS AND COMPUTATIONAL BIOLOGY CHEMISTRY
3M5O
Crystal structure of HCV NS3/4A protease in complex with N-terminal product 5A5B
Drug resistance against HCV NS3/4A inhibitors is defined by the balance of substrate recognition versus inhibitor binding Keith P. Romano1, Akbar Ali1, William E. Royer, and Celia A. Schiffer2 Department of Biochemistry and Molecular Pharmacology, University of Massachusetts Medical School, Worcester, MA 01605 Edited by John M. Coffin, Tufts University School of Medicine, Boston, MA, and approved September 14, 2010 (received for review May 13, 2010) Hepatitis C virus infects an estimated 180 million people world- wide, prompting enormous efforts to develop inhibitors targeting the essential NS3/4A protease. Resistance against the most promis- ing protease inhibitors, telaprevir, boceprevir, and ITMN-191, has emerged in clinical trials. In this study, crystal structures of the NS3/4A protease domain reveal that viral substrates bind to the protease active site in a conserved manner defining a consensus volume, or substrate envelope. Mutations that confer the most severe resistance in the clinic occur where the inhibitors protrude from the substrate envelope, as these changes selectively weaken inhibitor binding without compromising the binding of substrates. These findings suggest a general model for predicting the suscept- ibility of protease inhibitors to resistance: drugs designed to fit within the substrate envelope will be less susceptible to resistance, as mutations affecting inhibitor binding would simultaneously interfere with the recognition of viral substrates. drug design ∣hepatitis C ∣substrate envelope D rug resistance is a major obstacle in the treatment of quickly evolving diseases. Hepatitis C virus (HCV) is a genetically diverse Hepacivirus of the Flaviviridae family infecting an esti- mated 180 million people worldwide (1). The viral RNA genome is translated as a single polyprotein and subsequently processed by host-cell and viral proteases into structural (C, E1, E2, and p7) and nonstructural (NS2, NS3, NS4A, NS4B, NS5A, and NS5B) proteins (2). The viral RNA-dependent RNA polymerase, NS5B, is inherently inaccurate and misincorporation of bases accounts for a very high mutation rate (3). While some mutations are neu- tral, others will alter the viability of the virus and propagate with varying efficiencies in each patient. Thus HCV infected indivi- duals will develop a heterogeneous population of virus variants known as quasispecies (4). As patients begin treatment, the selec- tive pressures of antiviral drugs will favor drug resistant variants (5). Therefore, an inhibitor must not only recognize one protein variant, but an ensemble of related enzymes. A detailed under- standing of the atomic mechanisms of resistance is essential to effectively combat drug resistance against HCV antivirals. The essential HCV NS3/4A protease is an attractive therapeu- tic target responsible for cleaving at least four sites along the viral polyprotein. These sites share little sequence homology except for an acid at position P6, Cys or Thr at P1, and Ser or Ala at P1′ (Table S1). The first cleavage event at the 3-4A junction occurs in cis as a unimolecular process, while the remaining sub- strates are processed bimolecularly in trans. The NS3/4A protease also cleaves the human cellular targets TRIF and MAVS, which confounds the innate immune response to viral infection (6–8). Early drug design efforts were hampered by the relatively shallow, featureless architecture of the protease active site. The eventual observation of N-terminal product inhibition served as a stepping stone for the discovery of more potent peptidomimetic inhibitors (9, 10). Over the past decade, pharmaceutical companies have further developed these lead compounds. Many structure-activ- ity-relationship (SAR) studies have been performed to evaluate the effect of different functional moieties on protease inhibition at positions P4-P1′ (11–17). Crystal structures have been deter- mined of the NS3/4A protease domain bound to a variety of inhibitors as well as of several drug resistant protease variants, such as R155K and V36M (18, 19). These data elucidate the mo- lecular interactions of NS3/4A with inhibitors and the effect of specific drug resistance mutations on binding. These efforts, con- ducted in parallel by several pharmaceutical companies, led to the discovery of many protease inhibitors. Proof-of-concept for the successful clinical activity of this drug class was first demon- strated by the macrocyclic inhibitor BILN-2061 (Boehringer Ingelheim) (20, 21), which was later dropped from clinical trials in 2006 due to cardiotoxicity (22). Many other NS3/4A protease inhibitors are currently in development, and telaprevir (Vertex), boceprevir (Schering-Plough), and ITMN-191 (Intermune) lead the way in advanced phases of human clinical trials (Fig. 1A). Despite these successes, the rapid acquisition of drug resis- tance has limited the efficacy of the most potent NS3/4A protease inhibitors in both replicon studies and human clinical trials (Fig. 1B and Table 1). In this study, we show that mutations con- ferring the most severe resistance occur where the protease extensively contacts the inhibitors but not the natural viral sub- strates. Four crystal structures of the NS3/4A protease domain in complex with the N-terminal products of viral substrates reveal a conserved mode of substrate binding, with the consensus volume defining the substrate envelope. The protease inhibitors ITMN- 191 (3M5L), TMC435 (3KEE) (23), and boceprevir (2OC8) (24) protrude extensively from the substrate envelope in regions that correlate with known sites of resistance mutations. Most notably, the P2 moieties of all three drugs protrude to contact A156 and R155, which mutate to confer high-level resistance against nearly all drugs reported in the literature (25–30). These findings sug- gest that drug resistance results from a change in molecular recognition and imply that drugs designed to fit within the sub- strate envelope will be less susceptible to resistance, as mutations altering inhibitor binding will simultaneously interfere with the binding of substrates. Results Synthesis of ITMN-191. We synthesized the macrocyclic inhibitor ITMN-191 using a convergent reaction sequence described in SI Text. Briefly, the P2 and P1-P1′ fragments were preassembled Author contributions: K.P.R., A.A., and C.A.S. designed research; K.P.R. and A.A. performed research; A.A. and W.E.R. contributed new reagents/analytic tools; K.P.R., W.E.R., and C.A.S. analyzed data; and K.P.R., A.A., W.E.R., and C.A.S. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Data deposition: The atomic coordinates and structure factors have been deposited in the Protein Data Bank, www.pdb.org (PDB ID codes 3M5L, 3M5M, 3M5N, and 3M5O). 1K.P.R. and A.A. contributed equally to this work. 2To whom correspondence should be addressed. E-mail: Celia.Schiffer@umassmed.edu. This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1006370107/-/DCSupplemental. 20986–20991 ∣PNAS ∣December 7, 2010 ∣vol. 107 ∣no. 49 www.pnas.org/cgi/doi/10.1073/pnas.1006370107 and the macrocyclic drug compound was generated by a four- step reaction sequence, including P2-P3 amide coupling, ester hydrolysis, coupling with the P1-P1′ fragment, and ring-closing metathesis. The P2-P3 fragment was assembled by coupling the commercially available Boc-protected amino acid (S)-2-(tert- butoxycarbonylamino)non-8-enoic acid (Acme Biosciences, Inc) with the preassembled P2 fragment, (3R, 5S)-5-(methoxy- carbonyl)pyrrolidin-3-yl 4-fluoroisoindoline-2-carboxylate (31), using O-(7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexa- fluorophosphate (HATU)/diisopropylethylamine (DIPEA). Hy- drolysis of the P2-P3 methyl ester with LiOH:H2O in a mixture of THF-MeOH-H2O followed by coupling of the resulting acid under HATU/DIPEA conditions with the preassembled P1-P1′ fragment, (1R, 2S)-1-amino-N-(cyclopropylsulfonyl)-2-vinylcyclo- propanecarboxamide (32), provided the bis-olefin precursor for ring-closing metathesis. Cyclization of the bis-olefin intermediate was accomplished using a highly efficient ring-closing metathesis catalyst Zhan 1B and provided the protease inhibitor ITMN-191. Structure Determination of Inhibitor and Substrate Complexes. Although NS3/4A cleaves the viral polyprotein of over 3,000 residues at four specific sites in vivo, we focused on the local interactions of the protease domain with short peptide sequences corresponding to the immediate cleavage sites. All structural studies were carried out with the highly soluble, single-chain con- struct of the NS3/4A protease domain described previously (33), which contains a fragment of the essential cofactor NS4A cova- lently linked at the N terminus by a flexible linker. A similar pro- tease construct was shown to retain comparable catalytic activity to the authentic protein complex (34). Crystallization trials were initially carried out using the inactive (S139A) protease variant in complex with substrate peptides spanning P7-P5′. The 4A4B substrate complex revealed cleavage of the scissile bond and no ordered regions for the C-terminal fragment of the substrate. Similar observations were previously described for two other serine proteases where catalytic activity was observed, presum- ably facilitated by water, despite Ala substitutions of the catalytic Ser (35, 36). Thus all subsequent crystallization trials with the NS3/4A protease were performed using N-terminal cleavage products of the viral substrates spanning P7-P1. NS3/4A crystal structures in complex with ITMN-191 and peptide products 4A4B, 4B5A, and 5A5B were determined and refined at 1.25 Å, 1.70 Å, 1.90 Å, and 1.60 Å resolution, respec- tively (Table S2). The complexes crystallized in the space groups P212121 and P21 with one, two, or four molecules in the asym- metric unit. The average B factors range from 16.8–29.7 Å2 and there are no outliers in the Ramachandran plots. These structures represent the highest resolution crystal structures of NS3/4A protease reported to date. Overall Structure Analysis. The NS3/4A protease domain adopts a tertiary fold characteristic of serine proteases of the chymotrypsin family (37, 38). A total of nine protease molecules were modeled in the four crystal structures solved in this study with an overall rms deviation (rmsd) of 0.28 Å. The rmsds reveal the five most variable regions of the protease to be (Fig. S1): (i) the linker con- necting cofactor 4A at the N terminus, (ii) the loop containing residues 65–70, (iii) the zinc-binding site containing residues 95–105, (iv) the 310 helix region spanning residues 128–136, and (v) the active site antiparallel β-sheet containing residues 156–168. These structural differences likely indicate inherent flexibility in the protease and do not appear to correlate with ligand type or active site occupancy. Analysis of Product Complexes. Product complexes 4A4B, 4B5A, and 5A5B were further analyzed with the C terminus of the full-length NS3/4A structure (1CU1), which contains the N-term- inal cleavage product of viral substrate 3-4A (39). All four products bind to the protease active site in a conserved manner (Fig. 2), forming an antiparallel β-sheet with residues 154–160 Fig. 1. NS3/4A protease inhibitors and reported sites of drug resistance. (A) The leading protease inhibitors in development mimic the N-terminal side of the viral substrates. (B) The majority of reported drug resistance mutations cluster around the protease active site with the catalytic triad depicted in yellow. Table 1. Drug resistance mutations reported in replicon studies and clinical trials* Residue Mutation Drug V36 A, M, L, G Boceprevir, telaprevir Q41 R Boceprevir, ITMN-191 F43 S, C, V, I Boceprevir, telaprevir, ITMN-191, TMC435 V55 A Boceprevir T54 A, S Boceprevir, telaprevir Q80 K, R, H, G, L TMC435 S138 T ITMN-191, TMC435† R155 K, T, I, M, G, L, S, Q Boceprevir, telaprevir, ITMN-191, BILN-2061, TMC435 A156 V, T, S, I, G Boceprevir, telaprevir, ITMN-191, BILN-2061, TMC435 V158 I Boceprevir D168 A, V, E, G, N, T, Y, H, I ITMN-191, BILN-2061, TMC435 V170 A Boceprevir, telaprevir M175 L Boceprevir *References (18, 25, 26, 28, 30–37). †TMC435 displays reduced activity against S138T, but the mutation was not observed in selection experiments. Romano et al. PNAS ∣ December 7, 2010 ∣ vol. 107 ∣ no. 49 ∣ 20987 BIOPHYSICS AND COMPUTATIONAL BIOLOGY CHEMISTRY and burying 500–600 Å2 of solvent accessible surface area as cal- culated by PISA (40). The peptide backbone torsions are very similar, being most conserved at position P1 and deviating slightly toward position P4. Eight hydrogen bonds between backbone amide and carbonyl groups are completely conserved, involving protease residues S159 (C159 in product 3-4A), A157, R155, S139A, S138, and G137. S159 (C159 in product 3-4A), and A157 each contribute two hydrogen bonds with the P5 and P3 peptide residues, respectively. All P1 terminal carboxyl groups sit in the oxyanion hole, hydrogen bonding with the Nϵ atom of H57 and the amide nitrogens of residues 137–139. Although only product 4B5A contains a proline at P2, the other substrate sequences still adopt constrained P2 φ torsion angles. Thus products bind similarly despite their high sequence diversity. The P1 and P6 residues are most conserved among the substrate sequences, as are most of their interactions with the protease. The P1 side chains interact with the aromatic ring of F154. In all structures but product complex 4B5A, K165 forms salt-bridges with the P6 acids, while residues R123, D168, R155, and the catalytic D81 form an ionic network along one surface of the bound products (Fig. 2). In complex 4B5A, R123 interacts directly with the P6 acid, while D168 reorients and no longer contacts R155. Other molecular interactions in the product complexes are more diverse. Notably, K136 interacts differently with the cleavage products, forming: (i) a hydrogen bond with the P2 backbone carbonyl oxygen of 3-4A; (ii) a salt-bridge with the P3 glutamate of product 4A4B; and (iii) non- specific van der Waals interactions with the P2 and P3 side chains of products 4B5A and 5A5B. Also, in product complex 4A4B, an intramolecular hydrogen bond forms between the P3 and P5 glu- tamate residues, while the unique P4 acid of product 3-4A forms salt-bridges with the guanidinium groups of R123 and R155. Thus distinct patterns of side chain interactions underlie the set of con- served features involved in NS3/4A cleavage product binding. The Substrate Envelope. To further analyze the structural similari- ties of the four NS3/4A product complexes, the active sites were superposed on the Cα atoms of residues 137–139 and 154–160, revealing that both the active site residues and substrate products spanning P6-P1 align closely with an average Cα rmsd of 0.24 Å and 0.35 Å, respectively. The consensus van der Waals volume shared by any three of the four cleavage products was then cal- culated to generate the NS3/4A substrate envelope (Fig. 3A). This shape could not be predicted by the primary sequences alone and highlights the conserved mode of viral substrate recognition despite their high sequence diversity. Analysis of Inhibitor Complexes. ITMN-191, TMC435, and bocepre- vir are all peptidomimetic NS3/4A protease inhibitors. Active site superpositions of these drug complexes reveal that the inhibitors interact with many of the same protease residues as the cleavage products. Despite the P3-P1 cyclization of ITMN-191 and TMC435, the functional groups are positioned similarly in all three inhibitor complexes. The P1 cysteine surrogates interact with the aromatic ring of F154, while the P2 and P3 moieties over- lap closely. Although TMC435 does not contain a P4 substituent, the P4 tert-butyl groups of ITMN-191 and boceprevir also align closely. In addition, the P1 and P3 backbone atoms of all inhibi- tors hydrogen bond with the carbonyl oxygens of R155 and A157, respectively. These observations verify the peptidomimetic nat- ure of these drugs and support their observed mechanism as competitive active site inhibitors. The largest variation between these three protease inhibitors occurs at P2 where the aromatic rings of ITMN-191 and TMC435 stack against the guanidinium group of R155 (Fig. 3). This molecular interaction alters the electrostatic network involving R123, D168, R155, and D81. R155 rotates nearly 180° around Cδ relative to its conformation observed in product complexes, losing its hydrogen bond with D81 but maintaining interaction with D168. Mutations at R155 or D168 would disrupt the elec- trostatic network and destabilize this packing thereby lowering the affinity of these macrocyclic drugs. This observation provides a structural rationale for the drug resistance mutations R155K, as previously proposed (19), and D168A/V, which both confer a selective advantage in vitro in the presence of ITMN-191 or TMC435 (26, 30). In addition, the TMC435 complex reveals that R155 is stabilized by a hydrogen bond with Q80, which also mutates to confer resistance to TMC435 (30). Thus many of the primary drug resistance mutations can be explained by the disrup- tion of atomic interactions involving the P2 functional groups of the drugs. Insights into Drug Resistance. To determine the locations where the inhibitors protrude from the substrate envelope, the inhibitor and product complexes were also superposed using residues 137–139 and 154–160. The van der Waals volumes of inhibitor protrusion from the substrate envelope (V out) (41, 42) were calculated for each drug and compared with published EC50 fold-change data for drug resistance variants (30). The magnitudes of the EC50 fold-change data determined for each NS3/4A mutant generally trend with the V out values for the three drugs. The P2 moieties of boceprevir, ITMN-191 and TMC435 protrude most extensively from the substrate envelope with V out values of 105, 294, and Fig. 2. Stereo view of N-terminal cleavage product binding to NS3/4A pro- tease. N-terminal protease cleavage products (A) 3-4A, (B) 4A4B, (C) 4B5A, and (D) 5A5B are depicted as they bind to the protease active site. All conserved interactions are indicated by black dashes, while red lines depict interactions that are not present in all structures. The electrostatic network involving residues R123, D168, R155, and D81 is indicated by blue dashes. 20988 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1006370107 Romano et al. 496 Å3, respectively (Table 2). However, the precise level of drug resistance observed is also determined by the particular change in molecular interaction occurring for a given mutation. For ex- ample, A156 and R155 pack with the P2 moieties of these three inhibitors where they protrude beyond the substrate envelope. Mutations of A156 to bulkier side chains would result in a steric clash with the P2 drug moieties. Indeed, the rigid dimethylcyclo- propane group of boceprevir protrudes from the substrate envel- ope at the P2 subsite, and A156V or A156T confer 65 and 75 fold-changes in EC50, respectively (Table 2). Similarly, molecular changes at R155 and D168 would result in a substantial loss of interactions with P2. The most extensive protrusions of ITMN- 191 and TMC435 at P2 trend with their greatest fold-change in potency of nearly 450 and 600, respectively, from mutations in this subsite. Thus the extent by which an inhibitor protrudes from the substrate envelope in a given subsite is indicative of its vulnerability to resistance. Further structural analyses with the substrate envelope provide insights into other NS3/4A drug resistance mutations. The P1′ sulfonamide groups of ITMN-191 and TMC435, as well as the P1′ ketoamide of boceprevir, protrude from the substrate envel- ope near residues Q41 and F43, which both mutate to confer low-level resistance to these drugs (25, 30, 43). The keto group of boceprevir also projects outside the substrate envelope near T54 and V55. T54A/S confers low-level resistance to boceprevir, while V55A was recently identified in patient isolates after treat- ment with boceprevir (44). The analogous carbonyl groups of ITMN-191 and TMC435, however, are orientated in the opposite direction and protrude toward S138. In fact, in vitro studies reveal reduced activity for ITMN-191 and TMC435 against S138T variants, while boceprevir remains fully active (30, 43). The bulky P4 tert-butyl group of boceprevir extends outside the substrate envelope contacting V158; the V158I variant has lower affinity for this drug, likely due to a steric clash (45). This variant may also impact the affinity of ITMN-191, as its P4 tert-butyl also pro- trudes at the same location. These findings demonstrate that in regions outside the P2 subsite, positions where ITMN-191, TMC435, and boceprevir protrude from the substrate envelope also correlate with many other known sites of drug resistance mutations. Discussion The emergence of drug resistance is a major obstacle in modern medicine that limits the long-term usefulness of the most promis- ing therapeutics. By considering how HCV NS3/4A protease in- hibitors bind relative to natural viral substrates, we discovered that primary sites of resistance occur in regions of the protease where drugs protrude from the substrate envelope. In particular, R155 and A156, which mutate to confer severe resistance against ITMN-191, TMC435, and boceprevir, interact closely with the P2 drug moieties where they protrude most extensively from the substrate envelope. Molecular changes at these residues confer resistance by selectively weakening inhibitor binding without compromising the binding of viral substrates. We further specu- late that these mutations will not considerably affect the binding of the host cellular substrates TRIF and MAVS, which likely fit well within the substrate envelope as they share many features with the viral substrates. However, TRIF contains a track of eight proline residues instead of an acidic residue at position P6, which may modulate its binding. Further structural studies are warranted to better ascertain the molecular details of how these cellular substrates are recognized by the NS3/4A protease. Although this study focuses on ITMN-191, TMC435, and boceprevir, other NS3/4A protease inhibitors in clinical trials, in- cluding telaprevir, narlaprevir, and vaniprevir (Fig. 1A), contain similar functional groups that likely protrude from the substrate envelope. Most notably, all these drug candidates contain bulky P2 moieties and are therefore susceptible to cross-resistance against mutations at R155 and A156. R155 and A156 mutations have been shown to confer telaprevir resistance in treated patients (46). Cross-resistance studies have also shown that nar- Table 2. EC50 fold-change (FC) data * for several NS3/4A drug resistant variants tabulated with Vout, the van der Waals volume of protrusion from the substrate envelope, at each subsite of the enzyme Boceprevir ITMN-191 TMC435 Subsite Resistance mutation Vout (Å3) EC50 FC Vout (Å3) EC50 FC Vout (Å3) EC50 FC Total 292 500 649 P1 76 67 64 P2 105 294 496 Q80R 0.5 3.5 6.9 Q80K 0.8 2.3 7.7 R155K 4.7 447 30 A156V 75 63 177 A156T 65 41 44 D168A 0.7 153 594 D168E 0.8 75 40 P3 34 70 67 P4 76 69 0 V158I 3.3† ND ND *Antiviral activity was reported previously by Lenz et al., 2010 (30). †Fold-change in EC50 reported in replicon assay by Qiu et al., 2009 (45). Fig. 3. Stereo view of the NS3/4A substrate envelope and protease inhibitors. (A) After active site superpositions, the overlapping van der Waals volume shared by any three of the four cleavage products defines the substrate envelope, depicted in blue. NS3/4A protease residues which mutate to confer drug resistance are shown in brown. (B) ITMN-191, (C) boceprevir and (D) TMC435 protrude from the substrate envelope at several locations, which correlate with known sites of drug-resistant mutations to each inhibi- tor, shown in red. Romano et al. PNAS ∣ December 7, 2010 ∣ vol. 107 ∣ no. 49 ∣ 20989 BIOPHYSICS AND COMPUTATIONAL BIOLOGY CHEMISTRY laprevir displays similar fold losses in activity against most of the known drug resistance mutations for telaprevir and boceprevir (47). Ultimately, to slow the emergence of multidrug resistant viral strains, inhibitors should be confined within the substrate envelope, particularly at the P2 position. To compensate for the loss of binding affinity that will likely accompany these changes, additional interactions could potentially be optimized spanning the S4-S6 subsites of the protease and the catalytic triad. Our findings further suggest a general model for using the sub- strate envelope to predict patterns of drug resistance in other quickly evolving diseases. For drug resistance to occur, mutations must selectively weaken a target’s affinity for an inhibitor without significantly altering its natural biological function. Mutations occurring outside the substrate envelope are better able to achieve this effect, as these molecular changes can selectively alter inhibitor binding without compromising the binding of natural substrates. Whenever the interaction of a drug target with its biological substrates can be structurally characterized, we pre- dict that drugs designed to fit within the substrate envelope will be less susceptible to resistance. Structure-based design strategies can utilize this model as an added constraint to develop inhibitors that fit within the substrate envelope. In fact, previous work in our laboratory provides proof-of-concept for the successful incor- poration of the substrate envelope in the design of unique HIV protease inhibitors, which maintain high affinities against a panel of multidrug resistant variants of HIV-1 protease (42, 48–53). As a general paradigm, design efforts incorporating the substrate envelope would facilitate a more rationale evaluation of drug candidates and lead to the development of more robust inhibitors that are less susceptible to resistance. Materials and Methods Protein Crystallization. The NS3/4A protease construct was expressed and purified as reported previously (33, 54), detailed in SI Text. Purified protein was concentrated to ∼3 mg∕mL and loaded on a HiLoad Superdex75 16/60 column equilibrated with 25 mM 2-(N-morpholino)ethanesulfonic acid (MES) at pH 6.5, 500 mM NaCl, 10% glycerol, 30 μM zinc chloride, and 2 mM DTT. The protease fractions were pooled and concentrated to 20–25 mg∕mL with an Amicon Ultra-15 device (10 kD; Millipore). The concentrated samples were then incubated at 4 °C for 1 h with 2–20 M excess of viral substrate 4A4B, peptide products 4B5A or 5A5B, or ITMN-191. Information about the synth- esis of viral peptides and ITMN-191 is provided in SI Text. Diffraction-quality crystals were obtained overnight for all ligands by mixing equal volume of concentrated protein solution with precipitant solution (20–26% PEG-3350, 0.1 M sodium MES buffer at pH 6.5, and 4% ammonium sulfate) in 24-well VDX hanging drop trays. Data Collection and Structure Solution. Crystals were flash-frozen in liquid nitrogen and mounted under constant cryostream. X-ray diffraction data were collected at Advanced Photon Source BioCARS 14-IDB, 14-BMC, and LS-CAT 21-ID-F. Diffraction intensities of the complexes were indexed, integrated, and scaled using the programs HKL2000 (55) and XDS (56). 5% of the data was used to calculate R-free (57). All structure solutions were generated using isomorphous molecular replacement with PHASER (58) or AMORE. The NS3/4A protease domain (PDB code 2A4G) (59) was used for molecular replacement in solving the product 4A4B structure, and this struc- ture was subsequently used for solving the other complexes. In all cases, initial refinement was carried out in the absence of modeled ligand, which was subsequently built in. Phases were improved using ARP/wARP (60). Itera- tive rounds of translation, libration, and screw (TLS) and restrained refine- ment with CCP4 (61) and graphical model building with COOT (62) until convergence was achieved. The final structures were evaluated with MolProbity (63) prior to deposition in the protein data bank. Structural Analysis. Double-difference plots (64) were used to determine the structurally invariant regions of the protease, consisting of residues 32–36, 42–47, 50–54, 84–86, and 140–143. Structures were superposed with PyMOL (65) using the Cα atoms of these residues for all protease molecules from the solved structures (nine total). The B chain of product complex 4A4B was used as the reference structure in all alignments. Fit of individual inhibitors into the substrate envelope was quantified by mapping the substrate envelope and the van der Waals volume of each inhibitor on a three-dimensional grid with spacing of 0.5 Å. Vout for each drug moiety was computed by counting the grid cells, which were occupied by any inhi- bitor atom of that site but not the substrate envelope, and multiplied by the grid cell size, 0.125 Å3 (41, 42). Substrate Envelope and Inhibitor Analyses. NS3/4A substrate envelope was computed using product complexes 4A4B (B chain), 4B5A (D chain), and 5A5B (A chain). In structures with multiple protease molecules in the asym- metric unit, the one containing the most ordered peptide product was used for the alignment. The protease domain of the full-length NS3/4A structure (A chain; PDB code 1CU1) (39), including the C-terminal six amino acids, was included as a product complex 3-4A. All active site alignments were performed in PyMOL using Cα atoms of protease residues 137–139 and 154–160. After superposition, Gaussian object maps were generated in Py- MOL for each cleavage product. Four consensus Gaussian maps were then calculated, representing the intersecting volume of a group of three object maps. Finally, the summation of these four consensus maps was generated to construct the substrate envelope, depicting the van der Waals volume shared by any three of the four products. The previously determined boceprevir complex (PDB code 2OC8) (24) and TMC435 complex (PDB code 3KEE) (23) were used in this study (66). ACKNOWLEDGMENTS. We thank H. Klei for helpful discussions. We also thank Z. Wawrzak, M. Bolbat, and K. Brister of the LS-CAT beamline at Argonne National Laboratory for data collection of the ITMN-191 complex; M. Nalam and R. Bandaranayake for assistance with structural refinement; A. Ozen for providing V out calculations; and S. Shandilya and Y. Cai for computational support. The National Institute of Health (NIH) Grants R01-GM65347 and R01-AI085051 supported this work. Use of Advanced Photon Source (APS) was supported by the Department of Energy (DOE), Basic Energy Sciences, Office of Science, under Contract No. DE-AC02-06CH11357. Use of the Bio- CARS Sector 14 was supported by NIH-NCRR RR007707. Use of the LS-CAT Sector 21 was supported by the Michigan Economic Development Corpora- tion and the Michigan Technology TriCorridor under Grant 085P1000817. 1. World Health Organization. Barnes E, ed. (2010) Vaccine research: hepatitis C. Hepa- titis C Virus: Disease Burden. Available at http://www.who.int/vaccine_research/ diseases/viral_cancers/en/index2.html. March 15, 2010. 2. Major ME, Feinstone SM (1997) The molecular virology of hepatitis C. Hepatology 25:1527–1538. 3. Qureshi SA (2007) Hepatitis C virus—biology, host evasion strategies, and promising new therapies on the horizon. Med Res Rev 27:353–373. 4. Martell M, et al. (1992) Hepatitis C virus (HCV) circulates as a population of different but closely related genomes: quasispecies nature of HCV genome distribution. J Virol 66:3225–3229. 5. Paolucci S, et al. (2001) Analysis of HIV drug-resistant quasispecies in plasma, peripheral blood mononuclear cells and viral isolates from treatment-naive and HAART patients. J Med Virol 65:207–217. 6. Chen Z, et al. (2007) GB virus B disrupts RIG-I signaling by NS3/4A-mediated cleavage of the adaptor protein MAVS. J Virol 81:964–976. 7. Li XD, Sun L, Seth RB, Pineda G, Chen ZJ (2005) Hepatitis C virus protease NS3/4A cleaves mitochondrial antiviral signaling protein off the mitochondria to evade innate immunity. Proc Natl Acad Sci USA 102:17717–17722. 8. Li K, et al. (2005) Immune evasion by hepatitis C virus NS3/4A protease-mediated cleavage of the Toll-like receptor 3 adaptor protein TRIF. Proc Natl Acad Sci USA 102:2992–2997. 9. Llinas-Brunet M, et al. (1998) Peptide-based inhibitors of the hepatitis C virus serine protease. Bioorg Med Chem Lett 8:1713–1718. 10. Steinkuhler C, et al. (1998) Product inhibition of the hepatitis C virus NS3 protease. Biochemistry 37:8899–8905. 11. Arasappan A, et al. (2006) P2-P4 macrocyclic inhibitors of hepatitis C virus NS3-4A serine protease. Bioorg Med Chem Lett 16:3960–3965. 12. Bogen S, et al. (2008) Hepatitis C virus NS3-4A serine protease inhibitors: SAR of new P1 derivatives of SCH 503034. Bioorg Med Chem Lett 18:4219–4223. 13. Malancona S, et al. (2004) SAR and pharmacokinetic studies on phenethylamide inhibitors of the hepatitis C virus NS3/NS4A serine protease. Bioorg Med Chem Lett 14:4575–4579. 14. Nilsson M, et al. (2010) Synthesis and SAR of potent inhibitors of the hepatitis C virus NS3/4A protease: exploration of P2 quinazoline substituents. Bioorg Med Chem Lett 20:4004–4011. 15. Venkatraman S, et al. (2009) Discovery and structure-activity relationship of P1-P3 ketoamide derived macrocyclic inhibitors of hepatitis C virus NS3 protease. J Med Chem 52:336–346. 16. Raboisson P, et al. (2008) Structure-activity relationship study on a novel series of cyclopentane-containing macrocyclic inhibitors of the hepatitis C virus NS3/4A pro- tease leading to the discovery of TMC435350. Bioorg Med Chem Lett 18:4853–4858. 17. Perni RB, et al. (2004) Inhibitors of hepatitis C virus NS3.4A protease 2. Warhead SAR and optimization. Bioorg Med Chem Lett 14:1441–1446. 20990 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1006370107 Romano et al. 18. Zhou Y, et al. (2008) Phenotypic characterization of resistant Val36 variants of hepatitis C virus NS3-4A serine protease. Antimicrob Agents Chemother 52:110–120. 19. Zhou Y, et al. (2007) Phenotypic and structural analyses of hepatitis C virus NS3 protease Arg155 variants: sensitivity to telaprevir (VX-950) and interferon alpha. J Biol Chem 282:22619–22628. 20. Lamarre D, et al. (2003) An NS3 protease inhibitor with antiviral effects in humans infected with hepatitis C virus. Nature 426:186–189. 21. Hinrichsen H, et al. (2004) Short-term antiviral efficacy of BILN 2061, a hepatitis C virus serine protease inhibitor, in hepatitis C genotype 1 patients. Gastroenterology 127:1347–1355. 22. Vanwolleghem T, et al. (2007) Ultra-rapid cardiotoxicity of the hepatitis C virus protease inhibitor BILN 2061 in the urokinase-type plasminogen activator mouse. Gastroenterology 133:1144–1155. 23. Cummings MD, et al. (2010) Induced-fit binding of the macrocyclic noncovalent inhi- bitor TMC435 to its HCV NS3/NS4A protease target. Angew Chem Int Ed Engl 49:1652–1655. 24. Prongay AJ, et al. (2007) Discovery of the HCV NS3/4A protease inhibitor (1R,5S)-N- [3-amino-1-(cyclobutylmethyl)-2,3-dioxopropyl]-3- [2(S)-[[[(1,1-dimethylethyl)amino] carbonyl]amino]-3,3-dimethyl-1-oxobutyl]–6,6-dimethyl-3-azabicyclo[3.1.0]hexan-2(S)- carboxamide (Sch 503034) II. Key steps in structure-based optimization. J Med Chem 50:2310–2318. 25. Tong X, et al. (2008) Characterization of resistance mutations against HCV ketoamide protease inhibitors. Antiviral Res 77:177–185. 26. He Y, et al. (2008) Relative replication capacity and selective advantage profiles of protease inhibitor-resistant hepatitis C virus (HCV) NS3 protease mutants in the HCV genotype 1b replicon system. Antimicrob Agents Chemother 52:1101–1110. 27. Sarrazin C, et al. (2007) SCH 503034, a novel hepatitis C virus protease inhibitor, plus pegylated interferon alpha-2b for genotype 1 nonresponders. Gastroenterology 132:1270–1278. 28. Kieffer TL, et al. (2007) Telaprevir and pegylated interferon-alpha-2a inhibit wild-type and resistant genotype 1 hepatitis C virus replication in patients. Hepatology 46:631–639. 29. Yi M, Ma Y, Yates J, Lemon SM (2009) Transcomplementation of an NS2 defect in a late step in hepatitis C virus (HCV) particle assembly and maturation. PLoS Pathog 5:e1000403. 30. Lenz O, et al. (2010) In vitro resistance profile of the HCV NS3/4A protease inhibitor TMC435. Antimicrob Agents Chemother 54:1878–1887. 31. Arasappan A, Njoroge FG, Girijavallabhan VM (2005) PCT Int.Appl. WO 2005/113581 Patent Application. 32. Wang XA, et al. (2006) US Patent 6995174. 33. Wittekind M, Weinheirner S, Zhang Y, Goldfarb V (2002) US Patent 6333186. 34. Taremi SS, et al. (1998) Construction, expression, and characterization of a novel fully activated recombinant single-chain hepatitis C virus protease. Protein Sci 7:2143–2149. 35. Carter P, Wells JA (1988) Dissecting the catalytic triad of a serine protease. Nature 332:564–568. 36. Krishnan R, Sadler JE, Tulinsky A (2000) Structure of the Ser195Ala mutant of human alpha—thrombin complexed with fibrinopeptide A(7—16): evidence for residual catalytic activity. Acta Crystallogr D 56:406–410. 37. Kim JL, et al. (1996) Crystal structure of the hepatitis C virus NS3 protease domain complexed with a synthetic NS4A cofactor peptide. Cell 87:343–355. 38. Bode W, Huber R (1978) Crystal structure analysis and refinement of two variants of trigonal trypsinogen: trigonal trypsin and PEG (polyethylene glycol) trypsinogen and their comparison with orthorhombic trypsin and trigonal trypsinogen. FEBS Lett 90:265–269. 39. Yao N, Reichert P, Taremi SS, Prosise WW, Weber PC (1999) Molecular views of viral polyprotein processing revealed by the crystal structure of the hepatitis C virus bifunctional protease-helicase. Structure 7:1353–1363. 40. Krissinel E, Henrick K (2007) Inference of macromolecular assemblies from crystalline state. J Mol Biol 372:774–797. 41. Nalam MN, et al. (2010) Evaluating the substrate-envelope hypothesis: structural analysis of novel HIV-1 protease inhibitors designed to be robust against drug resis- tance. J Virol 84:5368–5378. 42. Chellappan S, Kairys V, Fernandes MX, Schiffer C, Gilson MK (2007) Evaluation of the substrate envelope hypothesis for inhibitors of HIV-1 protease. Proteins 68:561–567. 43. Kuntzen T, et al. (2008) Naturally occurring dominant resistance mutations to hepatitis C virus protease and polymerase inhibitors in treatment-naive patients. Hepatology 48:1769–1778. 44. Susser S, et al. (2009) Characterization of resistance to the protease inhibitor bocepre- vir in hepatitis C virus-infected patients. Hepatology 50:1709–1718. 45. Qiu P, et al. (2009) Identification of HCV protease inhibitor resistance mutations by selection pressure-based method. Nucleic Acids Res 37:e74. 46. Sarrazin C, et al. (2007) Dynamic hepatitis C virus genotypic and phenotypic changes in patients treated with the protease inhibitor telaprevir. Gastroenterology 132:1767–1777. 47. Tong X, et al. (2010) Preclinical characterization of the antiviral activity of SCH 900518 (Narlaprevir), a novel mechanism-based inhibitor of hepatitis C virus NS3 protease. Antimicrob Agents Chemother 54:2365–2370. 48. Chellappan S, et al. (2007) Design of mutation-resistant HIV protease inhibitors with the substrate envelope hypothesis. Chem Biol Drug Des 69:298–313. 49. Ali A, et al. (2009) Substrate envelope based design of new HIV-1 protease inhibitors active against drug-resistant HIV-1. 238th ACS National Meeting, Washington, DC, United States (American Chemical Society, Washington, DC), MEDI-102. 50. King NM, Prabu-Jeyabalan M, Nalivaika EA, Schiffer CA (2004) Combating susceptibil- ity to drug resistance: lessons from HIV-1 protease. Chem Biol 11:1333–1338. 51. Prabu-Jeyabalan M, et al. (2006) Substrate envelope and drug resistance: crystal struc- ture of RO1 in complex with wild-type human immunodeficiency virus type 1 protease. Antimicrob Agents Chemother 50:1518–1521. 52. Altman MD, Nalivaika EA, Prabu-Jeyabalan M, Schiffer CA, Tidor B (2008) Computa- tional design and experimental study of tighter binding peptides to an inactivated mutant of HIV-1 protease. Proteins 70:678–694. 53. Altman MD, et al. (2008) HIV-1 protease inhibitors from inverse design in the substrate envelope exhibit subnanomolar binding to drug-resistant variants. J Am Chem Soc 130:6099–6113. 54. Gallinari P, et al. (1998) Multiple enzymatic activities associated with recombinant NS3 protein of hepatitis C virus. J Virol 72:6758–6769. 55. Otwinowski Z, Minor W (1997) Processing of X-ray diffraction data collected in oscillation mode. Methods in Enzymology, 276 pp:307–326. 56. Kabsch W (1993) Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Crystallogr 26:795–800. 57. Brunger AT (1992) Free R value: a novel statistical quantity for assessing the accuracy of crystal structures. Nature 355:472–475. 58. McCoy AJ, et al. (2007) Phaser crystallographic software. J Appl Crystallogr 40:658–674. 59. Arasappan A, et al. (2005) Hepatitis C virus NS3-4A serine protease inhibitors: SAR of P′ 2 moiety with improved potency. Bioorg Med Chem Lett 15:4180–4184. 60. Morris RJ, Perrakis A, Lamzin VS (2002) ARP/wARP’s model-building algorithms. I. The main chain. Acta Crystallogr D 58:968–975. 61. Collaborative Computational Project Number 4 (1994) The CCP4 Suite: programs for protein crystallography. Acta Cryst, D50 pp:760–763. 62. Emsley P, Cowtan K (2004) COOT: model-building tools for molecular graphics. Acta Crystallogr D 60:2126–2132. 63. Davis IW, et al. (2007) MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res 35(Web Server issue):W375–383. 64. Prabu-Jeyabalan M, Nalivaika EA, Romano K, Schiffer CA (2006) Mechanism of sub- strate recognition by drug-resistant human immunodeficiency virus type 1 protease variants revealed by a novel structural intermediate. J Virol 80:3607–3616. 65. DeLano WL (2008) The PyMOL Molecular Graphics System (DeLano Scientific LLC, San Carlos, CA). 66. Berman HM, et al. (2000) The protein data bank. Nucleic Acids Res 28:235–242. Romano et al. PNAS ∣ December 7, 2010 ∣ vol. 107 ∣ no. 49 ∣ 20991 BIOPHYSICS AND COMPUTATIONAL BIOLOGY CHEMISTRY
3M5Q
0.93 A Structure of Manganese-Bound Manganese Peroxidase
Ultrahigh (0.93 Å) Resolution Structure of Manganese Peroxidase from Phanerochaete chrysosporium: Implications for the Catalytic Mechanism†,£ Munirathinam Sundaramoorthy‡,*, Michael H. Gold∥, and Thomas L. Poulos⊥ ‡ Department of Biochemistry, Vanderbilt University Medical Center, Nashville, TN 37232 ∥ Department of Biochemistry and Molecular Biology, OGI School of Science and Engineering, Oregon Health and Science University, Portland, OR 97291-1000 ⊥ Departments of Molecular Biology & Biochemistry, Chemistry, and Pharmaceutical Sciences University of California, Irvine, CA 92697-3900 Abstract Manganese peroxidase (MnP) is an extracellular heme enzyme produced by the lignin-degrading white-rot fungus Phanerochaete chrysosporium. MnP catalyzes the peroxide-dependent oxidation of MnII to MnIII. The MnIII is released from the enzyme in complex with oxalate, enabling the oxalate- MnIII complex to serve as a diffusible redox mediator capable of oxidizing lignin, especially under the mediation of unsaturated fatty acids. One heme propionate and the side chains of Glu35, Glu39 and Asp179 have been identified as MnII ligands in our previous crystal structures of native MnP. In our current work, new 0.93 Å and 1.05 Å crystal structures of MnP with and without bound MnII, respectively, have been solved. This represents only the fourth structure of a protein of this size at 0.93 Å resolution. In addition, this is the first structure of a heme peroxidase from a eukaryotic organism at sub-Ångstrom resolution. These new structures reveal an ordering/disordering of the C- terminal loop, which is likely required for Mn binding and release. In addition, the catalytic Arg42 residue at the active site, normally thought to function only in the peroxide activation process, also undergoes ordering/disordering that is coupled to a transient H-bond with the Mn ligand, Glu39. Finally, these high-resolution structures also reveal the exact H atoms in several parts of the structure that are relevant to the catalytic mechanism. Keywords Manganese; peroxidase; crystallography; atomic resolution; refinement £Atomic coordinates have been submitted to the Protein Data Bank as entries XXX and YYY. © 2010 Elsevier Inc. All rights reserved. *To whom correspondence should be addressed: Department of Biochemistry, Vanderbilt University School of Medicine, 23rd @ Pierce, 626 RRB, Nashville, TN 37232. m.sundaramoorthy@vanderbilt.edu. Phone: (615) 343-1373. Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. NIH Public Access Author Manuscript J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. Published in final edited form as: J Inorg Biochem. 2010 June ; 104(6): 683–690. doi:10.1016/j.jinorgbio.2010.02.011. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript INTRODUCTION White-rot basidiomycetous fungi are the only organisms capable of degrading the phenylpropanoid, plant cell wall polymer, lignin [1-4]. The lignin-degrading system of these fungi also can oxidize a variety of economically and environmentally important aromatic pollutants [5-9] and will probably play an important role in processes for the conversion of lignocellulosics materials to ethanol. Under ligninolytic conditions, the best-studied lignin- degrading fungus, Phanerochaete chrysosporium, secretes two families of extracellular heme peroxidases, lignin peroxidase (LiP) and manganese peroxidase (MnP) [2-4,10] and a hydrogen peroxide generating system [4,6,11,12]. MnP from P. chrysosporium has been studied by a variety of biochemical and biophysical methods [2,13-16]. The crystal structure of MnP illustrates that the heme environment of this enzyme is similar to that of other plant and fungal peroxidases [17,18]. However, MnP is the only heme peroxidase capable of the one-electron oxidation of MnII in a typical peroxidase reaction cycle: (1) (2) (3) where MnPI and MnPII are the oxidized intermediates MnP compounds I and II, respectively. Our earlier crystal structures of MnP show that the substrate, MnII, binds to one heme propionate and the side chains of three amino acids, Glu35, Glu39, and Asp179, as well as two solvent ligands [17,18]. This site was proven by kinetic and biophysical studies of wild-type MnP and of proteins containing point mutations in the putative binding site. Alteration of the proposed amino acid ligands in the Mn-binding site significantly affects Mn-binding and oxidation [19-24] and crystals of both the single variant, D179N, and the double variant, E35Q- D179N, lack electron density at the proposed Mn-binding site [25] suggesting that MnII is not bound. Furthermore, competitive inhibitors such as CdII, bind at the identical site, although with alternative geometry [18,24]. MnP is unique among enzymes, using manganese as a redox cofactor. Rather than permanently sequestering MnII in an interior binding site, MnP selectively binds MnII near the surface of the protein, oxidizes it and then releases the MnIII product in complex with organic acids such as oxalate. The relatively stable MnIII-oxalate complex acts as a diffusible mediator to oxidize the terminal substrate lignin, usually in the presence of a radical mediator such as an unsaturated fatty acid [26,27]. Whether a MnII-chelator complex binds to the enzyme to form a ternary complex or the chelator simply facilitates release of MnIII via ligand displacement has now been resolved. Both NMR and crystal structure studies of the MnP-MnII complex in the absence of chelators indicate that the later alternative occurs [16,18,28-30]. Crystals of MnP are very stable and are good candidates for high resolution structure determination [18]. Although a good deal is known about the catalytic mechanism of MnP and of other plant and fungal peroxidases, atomic level structures could help reveal details of heme geometry and H-bonding critical to the catalytic cycle of this enzyme. Atomic-detail structures Sundaramoorthy et al. Page 2 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript can also show the exact stereochemical properties of various prosthetic groups bound to proteins. Protein crystallographers normally rely on stereochemical parameters from crystal structures of small molecules to constrain prosthetic groups during refinement and hence, the final results are prejudiced toward the small molecule model. This is an important problem since it is a common axiom in structural biology that the protein modulates the reactivity of the prosthetic group by altering both electronic and geometric properties. Heme proteins provide a classic example of how the same heme, iron protoporphyrin IX, is used in a vast majority of heme proteins including the globins, cytochromes, and many heme enzymes, yet these heme proteins exhibit widely different functions. A detailed comparison of known heme protein structures illustrates that the heme group deviates significantly from planarity and that heme proteins, which are functionally related, exhibit similar types of non-planarity [31]. Thus, the ability to detail the geometry of such prosthetic groups is another benefit of ultrahigh resolution protein structures. The use of a new generation of synchrotron radiation sources coupled with cryogenic techniques has enabled the solution of an increasing number of protein structures at true atomic resolution. This has provided an unprecedented level of detail in analyzing functionally important features of protein structures, including the identification of individual H atoms. The precise location of H atoms is important in understanding enzyme reactions, where H-bonds and acid-base catalysis are often utilized. Recent examples include the serine protease structures from Bacillus lentus [32] and Titiachium album limber [33] solved at 0.78Å and 0.98Å, respectively. For Bacillus lentus subtilisin an H atom was found between the essential catalytic His and Asp which forms part of the well-documented serine protease Ser-His-Asp catalytic triad. In addition, the His-Asp H-bond distance was found to be relatively short, 2.62Å [32]. Such results have important implications for the details of enzyme catalyzed reactions and, in the subtilisin example, suggest the role that “low-barrier” H-bonds [34] play in the reaction. In this study, the structure of MnP containing 357 amino acid residues was refined at 0.93 Å resolution. Of the total of about 37,000 protein structures determined by x-ray crystallography and deposited in the PDB, only five unique structures (a bacterial catalase, 1GWE [35]; PfluDING, a DING protein from Pseudomonas fluorescens, 3G63 [36]; cholesterol oxidase from Streptomyces sp., 1N4W [37]; xylose isomerase, 1MUW [38]; and pentaerythritol tetranitrate reducatase, 1VYR [39] are comparable to MnP in both chain length (> 350 residues) and resolution (≥ 0.93 Å). Thus MnP is the first eukaryotic heme peroxidase to be analyzed at sub-Angstrom resolution. MATERIAL AND METHODS Protein Purification and Crystallization Wild-type MnP was purified from shaking cultures of Phanerochaete chrysosporium grown on high carbon, low nitrogen medium, as previously described [13,14,24]. MnII free MnP was prepared using a metal Chelax-100 column as described [24]. Crystals of MnII bound MnP (Mn-MnP) were grown at room temperature using the hanging drop vapor diffusion method with an excess of 5 mM MnCl2 as described [18,40]. The reservoir contained 30% (w/v) of polyethylene glycol 8,000, 0.2 M ammonium sulfate, and 0.1 M sodium cacodylate buffer, pH 6.5. The crystallization drops were composed of 5 μl of the protein solution (10-15 mg/ml) mixed with an equal volume of reservoir solution. The crystallization was initiated by a seeding procedure, using serially diluted seed stocks prepared from old native MnP crystals. Crystals of MnII free MnP were grown, using similar reservoir conditions except that the drops contained 4 mM EDTA instead of MnCl2 and the crystallization temperature was 4 °C. Sundaramoorthy et al. Page 3 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Data Collection and Processing Crystals were harvested in synthetic mother liquor identical to the reservoir solution and transferred to the cryo-solution containing 10% (v/v) glycerol in the mother liquor. The cryo- soaked crystals were flash-frozen in an N2 cryostream and all data sets were collected at -160 °C. The MnII free MnP data set was collected at the Stanford Synchrotron Radiation Laboratory (SSRL) beam line 7-1 at 1.08 Å wave length. The diffraction images were recorded in two steps using a single crystal on a MAR imaging plate at different detector settings and exposure times. The detector distance was 80 mm for the MAR300 plate to record high resolution reflections in dose mode with 3600 unit (~45 s) per frame and moved to 120 mm for the MAR180 plate to collect low resolution data with a 20 s exposure per frame. A total of 180 frames of 1° oscillation per frame were collected for each set. The images of each set were processed separately to extract raw intensities using MOSFLM [41] and were scaled together using SCALA to obtain a single data set at 1.05 Å (Table 1). The data set for the MnII bound MnP (Mn-MnP) was collected at SSRL Beam Line 9-1 at 0.78 Å wave length, using a single crystal. High resolution frames were collected using the MAR345 imaging plate at a 110 mm detector distance in dose mode with 6,000 units per frame. A total of 180 frames of 0.75° oscillation angle were collected to record high resolution data. Low resolution frames were collected using the MAR240 at 120 mm distance with 10 s exposure per frame. A total of 180 frames of 1° per frame were collected for each set. The intensities of each set were integrated separately using DENZO [42] and merged using SCALEPACK to obtain a single data set 0.93 Å (Table 1). Refinement Coordinates of the native MnP structure previously refined at 1.45 Å resolution (PDB Accession Code: 1YYD) were used as the starting model for the refinement with the MnII free MnP data set which was collected first. The reflections used for Rfree calculation in the 1.45 Å data set were flagged in the present data set and extended to full resolution at 1.05 Å using XPLOR [43]. However, in order to keep the unused reflections to a minimum, only 2% of the reflections were flagged for the test set (Table 2). The first round of refinement was carried out using data in the 8.0-1.45 Å resolution range by conjugate gradient least squares (CGLS) protocol in SHELXL [44]. The model was examined from amino- to carboxy-terminus guided by σA-weighted 2Fo-Fc and Fo-Fc maps. The electron density for the previously disordered C- terminal loop was more ordered for this data set and residues 342-352 were rebuilt. The resolution was extended gradually in a few rounds and the automatic water picking option (PLAN 50 2.4) was used to add new waters to the model. The refined model and new water positions were examined and manually corrected at each stage. The iterative cycles of manual model adjustment and refinement coupled with resolution extension was continued until the highest resolution of 1.05 Å was reached. Water molecules refined with a B-factor > 50 Å2 (with occupancy set to unity) were removed from the refinement throughout this exercise. At the end of 10 rounds of refinement the Rcryst (Fo > 4 σ) was 0.159 and the corresponding Rfree was 0.177. At this stage, anisotropic B-factor or atomic displacement parameter (ADP) refinement was turned on using ANIS keyword for all atoms and the model was refined in 20 CGLS cycles. The Rcryst and Rfree dropped to 0.122 and 0.154, respectively. Following this step, a few disordered sides chains were rebuilt in multiple conformations and the model was further refined. When the refinement converged, 20 cycles of CGLS refinement was run with riding hydrogen atoms for the polypeptide, sugar residues, and heme group. Both Rcryst and Rfree dropped to 0.110 and 0.123, respectively. Finally, 10 cycles of CGLS refinement was carried out with all reflections (i.e. working set + test set) to the final R-factor of 0.111. Sundaramoorthy et al. Page 4 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript For the 0.93 Å Mn-MnP data set, test reflections were flagged using the 1.05 Å data set as the reference and extended to full resolution. The refined model of the MnII free structure was used as the starting model for the refinement. However, the refinement was started at 1.5 Å with isotropic B-factors. The substrate MnII ion and its amino acid ligands were modeled using σA-weighted 2Fo-Fc and Fo-Fc maps. The resolution was increased to 0.93 Å in five rounds (1.2 Å, 1.05 Å, 1.00 Å, 0.95 Å and 0.93 Å) of CGLS refinement and the STIR command was used to extend the resolution in finer steps in several cycles (20-35) in each round. The maps were examined at the end of each refinement and minor changes were made to the model. New waters were added in each round but those refined with high B-factors (> 50 Å2) were removed in the subsequent refinement. The refinement converged with an Rcryst of 0.167 and an Rfree of 0.179 with isotropic B-factors. Subsequently anisotropic B-factor refinement was turned on and after extensive refinement the Rcryst and Rfree values converged to 0.12 and 0.143, respectively. At this stage, a few disordered side chains were identified and modeled in multiple conformations and refined further. When the refinement converged, 20 cycles of CGLS refinement was run with riding H atoms for the polypeptide, sugar residues and heme group (Rcryst = 0.109 and Rfree = 0.128). However, riding H atoms were not included for the proximal heme ligand, His173, the residue H-bonded to this histidine, Asp242, the distal base catalyst His46, and the residue H-bonded to this histidine, Asn80. Finally, 10 cycles of CGLS refinement was carried out with all reflections (i.e. working set + test set) to the final R-factor of 0.109. The refinement statistics are shown in Table 2. The metal-ligand distances for the heme FeIII, the substrate MnII, and two structural CaII ions, were unrestrained throughout the refinement. After the anisotropic B-factor refinement began, side chain disorders were examined and several side chains were modeled in multiple conformations with appropriate partial occupancies and total occupancy was restrained to unity using the FVAR instruction. A few other residues were found to be highly disordered but their side chains were not truncated. For the calculation of estimated standard deviations (ESDs) for the atomic positions, one cycle of full-matrix least-squares refinement was performed (BLOC 1). RESULTS AND DISCUSSION Quality of the Maps and Models Crystals of MnP from P. chrysosporium are very robust with low mosaicity ( < 0.2°) and can tolerate prolonged exposure as well as freeze-thaw cycles as observed in our previous study [18]. These well ordered crystals can diffract to 1.0 Å or better at a synchrotron source. The maps calculated using ultrahigh resolution data show high clarity that is typical of such data sets (Figure 1). The previously reported 1.45 Å structure showed an O-glycosylation site at Ser336 with a β-mannose residue and extra density at the N-glycosylation site at Asn131 that could be due to a potential β-mannose residue [18]. These features are observed in the ultrahigh resolution maps as well, but the maps do not show additional sugar residues at these sites nor do they reveal new glycosylation sites. The electron density for most of the bound waters is very clear and all of them were refined with unit occupancy. There remain small pieces of uninterpretable density in several places in the solvent regions. The final model of the substrate-bound MnP (Mn-MnP) consists of 357 amino acid residues, three sugar residues (GlcNac,GlcNac at Asn131 and a single mannose at Ser336) , a heme prosthetic group, two structural calcium ions, a substrate MnII ion, and 478 solvent molecules, including two glycerol molecules. The substrate-free MnP model differs only in lacking the MnII ion in the Mn binding site and in the number of solvent molecules, 549, which includes two glycerol molecules. The two models superimpose with an overall RMS deviation of 0.253 Å for all 357 Cα atoms and of 0.146 Å for all residues minus the C-terminal loop (residues 344-350). There is no significant difference between these structures and the previously Sundaramoorthy et al. Page 5 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript reported 1.45 Å cryo structure or the 2.06 Å room temperature structure in overall topology (Figure 2). About 91% of the non-glycine residues are in the most favored regions of the Ramachandran plot and the remaining are in either allowed or generously allowed regions. Disorder The N-terminal disorder persists for the first 10 residues in the maps of both the 1.05 Å and the 0.93 Å data sets. However, there is an improvement in the quality of density near the C- terminal loop encompassing Gly344-Gly350 in the 1.05 Å map of the substrate-free MnP data set. Interestingly, this region is still disordered in the 0.93 Å map of the Mn-MnP data set (Figure 3). This difference in the quality of electron density between the substrate-free and substrate-bound forms is probably not due to the difference in their resolution. Instead, this dynamic disorder is likely due to the binding of the substrate with implications for the catalytic mechanism of the enzyme. This C-terminal loop traverses near the Mn binding pocket and must be sufficiently flexible to allow binding of the substrate MnII, the entry of dicarboxylic acid chelators such as oxalate to bind the oxidized MnIII product [16,29], and the release of the MnIII-chelator complex. When we revisited the maps of the previously reported data sets of native MnP (1.45 Å), CdII-MnP complex (1.6 Å), SmIII-MnP complex (1.6 Å) and oxalate soaked SmIII-MnP complex (1.4 Å) similar differences are observed [18]. Whereas all three metal bound structures show disorder, the map of oxalate soaked SmIII-MnP complex, which lacks a metal ion, shows a more ordered C-terminal loop (data not shown). Besides the disorder at the N- and C-termini, there are small sections of surface loops and the side chains of a few surface residues that are disordered. Some of the disordered side chains could be modeled in two conformations with partial occupancies in each. Notable among them are a few cysteine residues, (Cys3, Cys14 and Cys253) which are found to be partially reduced due to radiation, and the active site residue Arg42. In addition, two of the MnII ligands, Glu35 and Glu39, are disordered and exhibit multiple conformations in the MnII free structure. Other disordered residues show a complete lack of density which could not be modeled with confidence. Nevertheless, the side chains were not truncated and were modeled according to the known sequence information [45,46]. These residues are characterized by high isotropic B-factors and exhibit unrealistic atomic displacement parameters (ADPs) as analyzed by PARVATI [47]. Mn Binding Site The manganese binding site in both the substrate free MnP and Mn-MnP structures is shown in Figure 4 and Table 3 provides ligand coordination distances. As described in our earlier study, treating MnP with EDTA did not completely remove the metal from the substrate binding site but did lead to a substantial decrease in electron density [25]. In our current work, however, we were able to crystallize metal free MnP protein and the structure confirms the dynamic nature of the active site described in our previous studies [18,25]. Two metal ligands, Glu35 and Glu39, move from their original MnII binding conformations and this provides insights into the mechanism of MnP. Once MnP oxidizes MnII to MnIII, the MnIII acts as a diffusible oxidant of lignin and other oxidizable substrates only when complexed with a suitable chelator [16,29]. The latter is required to stabilize the highly reactive MnIII. The question is how such a complex forms. In the structures of the single mutant (D179N) and the double mutant (E35Q, D179N) MnPs, which are essentially inactive and, which do not bind MnII, the side chain of Glu39 moves away (“open” conformation) from the metal binding site [25]. Similarly, the metal free structure of oxalate soaked SmIII-MnP complex structure showed Glu39 in the “open” conformation [18]. Another Mn ligand, Glu35, undergoes a less dramatic conformational change in these structures as its movement is restricted by a salt bridge with the side chain of Arg177 [21]. Sundaramoorthy et al. Page 6 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Whereas the mobility of Glu39 is effected by a rotation of the Cα–Cβ bond resulting in a major movement, Glu35 side chain is rotated about the Cβ–Cγ bond resulting in variable, but minor changes in the wild-type and mutant structures depending upon the electronic environment of the site. Other ligands—the heme propionate and Asp179—do not move from their original positions whether a metal ion is bound or not, strongly suggesting that precise geometry is required for efficient MnII-binding and oxidation. This is confirmed by steady-state and transient-state kinetic analyses of a MnP E39D single mutant and an E35D-E39D-D179E triple mutant [23]. The single and triple mutant variants exhibit 20- and 40-fold increase in Km, and 103 and 104 decrease in catalytic efficiency, respectively. Although the overall charge is retained, a decrease in the chain length of one ligand could not be compensated for by an increase in the chain length of another ligand for either MnII binding or oxidation. Our previous study using the non-reducible/oxidizable trivalent cation SmIII showed that it could bind at the Mn binding site of MnP [18]. In addition, the release of SmIII by oxalate from the pregrown SmIII-MnP crystals provides insights into how the Mn binding site can bind both divalent and trivalent cations and how the trivalent cation can be released from the resting enzyme by organic acids [18]. Taken together, our past studies and the present high resolution structures imply that MnII binding is precise and the site is relatively rigid, except for the ability of Glu35 and Glu39 to adopt two conformations—“closed” conformations in the metal bound state and “open” conformations in the metal free state, possibly acting as a “gate”, enabling a small carboxylic acid like oxalate or malonate to remove MnIII from the binding site. Without some flexibility in the Mn ligands, it is difficult to envisage how oxalate could remove MnIII from its coordination shell unless MnIII is first freely released. Another residue, Arg42, is also disordered, and this has significance for peroxidase function. This arginine residue is conserved in all peroxidases and is implicated in stabilizing the compound I and II intermediates by forming a hydrogen bond with the oxyferryl oxygen [48, 49]. This active site arginine residue appears disordered or in multiple conformations in this high resolution structures of MnP and in cytochrome c peroxidase (CcP) [48] but is in a well defined single conformation in the compound I structure of CcP [48]. Arg42 is disordered in both substrate-bound and -free structures of MnP and is modeled in two conformations—“in” and “out”, in which Arg42 moves closer to and away from the oxyferryl center. However, the main difference occurs in its interaction with Glu39. Besides providing a ligand to MnII, the carboxylate group of Glu39 also forms a salt bridge with the guanidium group of Arg42 in the “out” conformation in the substrate-bound structure. This is similar to the salt bridge between Glu35 and Arg177 described earlier [17,18,21,50]. However, in the substrate-free structure, movement of Glu39 to the “open” conformation breaks this salt bridge and a water molecule occupies the position of the carboxylate oxygen of Glu39. Thus, Arg42 may also stabilize MnII binding in the resting state through its interaction with Glu39, but its movement to the “in” conformation to stabilize the oxyferryl group in compounds I and II may destabilize Glu39 enabling it to move out for the oxidized MnIII to be chelated by oxalate. In contrast, the other two amino acid ligands for MnII, Glu35 and Asp179, are held in place by the rigid guanidium group of the Arg177 side chain and backbone amide group of residue Ala187, respectively. In support of this idea, disruption of the salt bridge between Glu35 and Arg177, through mutation of the arginine residue, has been shown to significantly lower the efficiency of MnII binding and oxidation in our previous work [21,50]. This new role for Arg42 is revealed for the first time in this high resolution structure of MnP. Heme Stereochemistry The overall structure of Mn-MnP (Figure 2) at 0.93Å appears essentially the same as that in our earlier lower resolution work [17,18]. Given the favorable ratio of data to parameters to be refined, the least squares matrix could be inverted which provides an accurate estimate of bond distances and angles. The electron density for the heme and proximal heme ligand together, Sundaramoorthy et al. Page 7 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript shown in Figure 5 and Table 4, provides specific heme parameters. As expected from spectroscopic studies [51], the heme Fe is pentacoordinate and high-spin. A functionally important parameter of heme geometry is the displacement of the iron from the porphyrin core. In MnP the Fe is 0.26Å out of plane compared to 0.48Å for typical model heme complexes [52]. The ability of the Fe atom to move in and out of the porphyrin core as a function of ligation, redox state, and spin state is a central feature of the hemoglobin allosteric mechanism proposed by Perutz [53]. Compared to the globins, the Fe atom in MnP and CcP is closer to the porphyrin core. We have attributed this difference [54] to the surrounding protein environment which in the peroxidases places the proximal helix containing the heme His ligand much closer to the protein. This prevents the spring-like up/down motion of the Fe-His-helix as seen in hemoglobin. The functional relevance of this difference is that in peroxidases, the iron is oxidized from FeIII to FeIV=O. The higher oxidation state, lower-spin state, and strong FeIV=O bond all favor the FeIV closer to the heme plane. Hence, the resting FeIII state with only partial displacement of the Fe from the porphyrin plane is poised for oxidation to FeIV, which is reminiscent of the entatic state [55], where the kinetic and/or thermodynamic barriers required for changing redox and spin-state are lowered owing to protein-metal interactions. Mechanism and H atoms As with many enzyme systems, H-bonds and proton transfer play a critical role in peroxidase catalysis. The original stereochemical mechanism proposed for heme peroxidases [56] indicates that the distal histidine residue (His 46 in MnP, Figure 2) acts as an acid-base catalyst by removing a proton from the iron-linked peroxide O atom and delivering it to the leaving OH moiety to produce water. As shown in Figure 5A, which of the N atoms of the imidazole moiety in His46 carries the proton is clearly visible in Fo-Fc electron density maps. The Nδ1atom is protonated and donates an H-bond to Asn80. This His-Asn H-bonding arrangement is conserved in all peroxidases and is thought to play an important role in orientating the distal His for catalysis as well as ensuring that the Nε2 is free to accept a proton from hydrogen peroxide. Another level of detail at 0.93 Å is the unambiguous orientation of amide side chains such as that for Asn80. At lower resolution, it is not possible to differentiate between the Nδ2 and Oδ1 side chain atoms. However, as shown in Figure 5A the amide side chain Oδ1 atom of Asn80 exhibits smeared electron density with the Cγ side chain atom while the Cδ-Nε2 density is much shaper. This indicates a Cγ-Oδ1 double bond and an H bond between the Nε1 of His 46 and the Oδ1 of Asn80. The role of H-bonds involving the proximal His ligand is not well documented. In all known heme peroxidase structures, the His ligand is within H-bonding distance of a buried Asp, Asp242 in MnP (Figure 2). A number of important functional properties have been attributed to this H-bond, one of which involves redox potential. A strong His-Asp H-bond favors a lower redox potential by stabilizing the additional positive charge on FeIV compared to FeIII. In the globins the histidine ligand forms an H-bond with a peptide carbonyl O atom, which presumably is weaker than the His-Asp H-bond in peroxidases. This difference helps to explain why the peroxidases exhibit lower redox potentials than globins. This view is supported by both model heme systems [57-60], protein NMR [61], and site directed mutagenesis [62,63]. The present structure provides a somewhat complex picture of the His-Asp interaction. As shown in Figure 5B there is a strong lobe of electron density in both Fo-Fc and 2Fo-Fc maps at about 1.34Å from Oδ2 of Asp242. Initially we considered that this peak might be an H atom which would mean that Asp242 carries the proton and not His173. However, the O-H distance, 1.34Å, is too long, compared to other H atoms in the structure, and the electron density peak too large for an H atom. Two additional data sets to 1.15Å and 1.05Å have been collected and maps generated from these data sets also exhibit additional electron density between Asp242 and His173 (data not shown). The prospects of a low-barrier H-bond seems unlikely since the Sundaramoorthy et al. Page 8 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript O-N hydrogen bonding distance is quite normal at 2.83Å while low-barrier H-bonds are much shorter.. The exact explanation for this extra electron density remains unclear. Acknowledgments This research was supported by a grant GM42614 (to T.L.P.) from the National Institutes of Health and grants MCB-9808420 from the National Science Foundation and DE-03-96ER20235 from the Division of Energy Biosciences, U.S. Department of Energy (to M.H.G). Portions of this research were carried out at the Stanford Synchrotron Radiation Laboratory, a national user facility operated by Stanford University on behalf of the U.S. Department of Energy, Office of Basic Energy Sciences. The SSRL Structural Molecular Biology Program is supported by the Department of Energy, Office of Biological and Environmental Research, and by the National Institutes of Health, National Center for Research Resources, Biomedical Technology Program, and the National Institute of General Medical Sciences. ABBREVIATIONS MnP manganese peroxidase Mn-MnP manganese-enzyme complex L.S. least squares CGLS conjugate gradient least squares ESD estimated standard deviation ADP atomic displacement parameter REFERENCES 1. Buswell JA, Odier E. CRC Crit. Rev. Biotechnol 1987;6:1–60. 2. Gold MH, Alic M. Microbiol. Rev 1993;57:605–622. [PubMed: 8246842] 3. Gold MH, Wariishi H, Valli K. ACS Symp. Ser 1989;389:127–140. 4. Kirk TK, Farrell RL. Annu. Rev. Microbiol 1987;41:465–505. [PubMed: 3318677] 5. Bumpus JA, Aust SD. BioEssays 1989;6:166–170. 6. Hammel KE. Enzyme Mirob. Technol 1989;11:776–777. 7. Reddy GV, Gelpke MD, Gold MH. J Bacteriol 1998;180:5159–5164. [PubMed: 9748450] 8. Valli K, Brock BJ, Joshi DK, Gold MH. Appl Environ Microbiol 1992;58:221–228. [PubMed: 1539977] 9. Valli K, Wariishi H, Gold MH. J Bacteriol 1992;174:2131–2137. [PubMed: 1551837] 10. Kuwahara M, Glenn JK, Morgan MA, Gold MH. FEBS Lett 1984;169:247–250. 11. Kersten PJ, Kirk TK. J Bacteriol 1987;169:2195–2201. [PubMed: 3553159] 12. Perie FH, Sheng D, Gold MH. Biochim. Biophys. Acta 1996;1297:139–148. [PubMed: 8917615] 13. Glenn JK, Akileswaran L, Gold MH. Arch. Biochem. Biophys 1986;251:688–696. [PubMed: 3800395] 14. Glenn JK, Gold MH. Arch. Biochem. Biophys 1985;242:329–341. [PubMed: 4062285] 15. Gold, MH.; Youngs, HL.; Sollewijn Gelpke, MD. Manganese and Its Role in Biological Processes. Sigel, A.; Sigel, H., editors. Macel Dekker; New York: 2000. p. 559-586. 16. Wariishi H, Valli K, Gold MH. J. Biol. CHem 1992;267:23688–23695. [PubMed: 1429709] 17. Sundaramoorthy M, Kishi K, Gold MH, Poulos TL. J Biol Chem 1994;269:32759–32767. [PubMed: 7806497] 18. Sundaramoorthy M, Youngs HL, Gold MH, Poulos TL. Biochemistry 2005;44:6463–6470. [PubMed: 15850380] 19. Kishi K, Kusters-van Someren M, Mayfield MB, Sun J, Loehr TM, Gold MH. Biochemistry 1996;35:8986–8994. [PubMed: 8688436] Sundaramoorthy et al. Page 9 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 20. Kusters-van Someren M, Kishi K, Lundell T, Gold MH. Biochemistry 1995;34:10620–10627. [PubMed: 7654716] 21. Sollewijn Gelpke MD, Moenne-Loccoz P, Gold MH. Biochemistry 1999;38:11482–11489. [PubMed: 10471300] 22. Whitwam RE, Brown KR, Musick M, Natan MJ, Tien M. Biochemistry 1997;27:5365–5370. 23. Youngs HL, Sollewijn Gelpke MD, Li D, Sundaramoorthy M, Gold MH. Biochemistry 2001;40:2243–2250. [PubMed: 11329293] 24. Youngs HL, Sundaramoorthy M, Gold MH. Eur J Biochem 2000;267:1761–1769. [PubMed: 10712608] 25. Sundaramoorthy M, Kishi K, Gold MH, Poulos TL. J Biol Chem 1997;272:17574–17580. [PubMed: 9211904] 26. Bao W, Fukushima Y, Jensen KA Jr. Moen MA, Hammel KE. FEBS Lett 1994;354:297–300. [PubMed: 7957943] 27. Reddy GV, Sridhar M, Gold MH. Eur J Biochem 2003;270:284–292. [PubMed: 12605679] 28. Banci L, Bertini I, Dat Pozzo L, Del Conte R, Tien M. Biochemistry 1998;37:9009–9015. [PubMed: 9636044] 29. Kishi K, Wariishi H, Marquez L, Dunford HB, Gold MH. Biochemistry 1994;33:8694–8701. [PubMed: 8038159] 30. Kuan IC, Tien M. Proc. Natl. Acad. Sci. USA 1993;90:1242–1246. [PubMed: 8433984] 31. Jentzen W, Ma JG, Shelnutt JA. Biophys J 1998;74:753–763. [PubMed: 9533688] 32. Kuhn P, Knapp M, Soltis SM, Ganshaw G, Thoene M, Bott R. Biochemistry 1998;37:13446–13452. [PubMed: 9753430] 33. Betzel C, Gourinath S, Kumar P, Kaur P, Perbandt M, Eschenburg S, Singh TP. Biochemistry 2001;40:3080–3088. [PubMed: 11258922] 34. Frey PA, Whitt SA, Tobin JB. Science 1994;264:1927–1930. [PubMed: 7661899] 35. Murshudov GN, Grebenko AA, Brannigan JA, Anston AA, Barynin VV, Dodson GG, Dauter Z, Wilson KS, Melik-Adamyan WR. Acta Crystallogr D Biol Crystallogr 2002;58:1972–1982. [PubMed: 12454454] 36. Liebschner D, Elias M, Moniot S, Fournier B, Scott K, Jelsch C, Guillot B, Lecomte C, Chabriere E. Journal of American Chemical Society 2009;131:7879–7886. 37. Lyubimov AY, Lario PI, Moustafa I, Vrielink A. Nature Chemical Biology 2006;2:259–264. 38. Fenn TD, Ringe D, Petsko GA. Biochemistry 2004;43:6464–6474. [PubMed: 15157080] 39. Khan H, Barna T, Harris RJ, Bruce NC, Barsukov I, Munro AW, Moody PC, Scrutton NS. J Biol Chem 2004;279:30563–30572. [PubMed: 15128738] 40. Sundaramoorthy M, Kishi K, Gold MH, Poulos TL. J Mol Biol 1994;238:845–848. [PubMed: 8182752] 41. Leslie AGW. Joint CCP4 + ESF-EAMCB Newsletter on Protein Crystallography 1992;26 42. Otwinowski Z, Minor W. Methods in Enzymol 1997;276:307–326. 43. Brunger, AT. X-PLOR Manual Version 3.1: A system for X-ray Crystallography and NMR. Yale University; New Haven, CT: 1992. 44. Sheldrick GM, Scheider TR. Methods in Enzymol 1997;277:319–343. [PubMed: 18488315] 45. Godfrey BJ, Mayfield MB, Brown JA, Gold MH. Gene 1990;93:119–124. [PubMed: 2227420] 46. Pribnow D, Mayfield MB, Nipper VJ, Brown JA, Gold MH. J Biol Chem 1989;264:5036–5040. [PubMed: 2925681] 47. Merritt EA. Acta Crystallogr D Biol Crystallogr 1999;55:1109–1117. [PubMed: 10329772] 48. Bonagura CA, Bhaskar B, Shimizu H, Li H, Sundaramoorthy M, McRee DE, Goodin DB, Poulos TL. Biochemistry 2003;42:5600–5608. [PubMed: 12741816] 49. Edwards SL, Nguyen HX, Hamlin RC, Kraut J. Biochemistry 1987;26:1503–1511. [PubMed: 3036202] 50. Gelpke MD, Youngs HL, Gold MH. Eur J Biochem 2000;267:7038–7045. [PubMed: 11106414] 51. Mino Y, Wariishi H, Blackburn NJ, Loehr TM, Gold MH. J Biol Chem 1988;263:7029–7036. [PubMed: 2835361] Sundaramoorthy et al. Page 10 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 52. Hoard JL. Science 1971;174:1295–1302. [PubMed: 4332625] 53. Perutz MF. Q Rev Biophys 1989;22:139–237. [PubMed: 2675171] 54. Poulos TL. Nat Struct Biol 1996;3:401–403. [PubMed: 8612066] 55. Williams RJ. Eur J Biochem 1995;234:363–381. [PubMed: 8536678] 56. Poulos TL, Kraut J. J Biol Chem 1980;255:8199–8205. [PubMed: 6251047] 57. Chang YT, Stiffelman OB, Loew GH. Biochimie 1996;78:771–779. [PubMed: 9010606] 58. Doef MM, Sweigart DA, O'Brien P. Inorg. Chem 1983;22:851–852. 59. Nappa M, Valentine JS, Snyder PA. Journal of American Chemical Society 1977;99:5799–5800. 60. Valentine JS, Sheridan RP, Allen LC, Kahn PC. Proceedings of National Academy of Sciences U S A 1979;76:1009–1013. 61. Banci L, Bertini I, Pease EA, Tien M, Turano P. Biochemistry 1992;31:10009–10017. [PubMed: 1327129] 62. Choudhury K, Sundaramoorthy M, Hickman A, Yonetani T, Woehl E, Dunn MF, Poulos TL. J Biol Chem 1994;269:20239–20249. [PubMed: 8051115] 63. Goodin DB, McRee DE. Biochemistry 1993;32:3313–3324. [PubMed: 8384877] Sundaramoorthy et al. Page 11 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. (A) 2Fo-Fc map density for the refined heme group, calculated using the 0.93 Å native Mn- MnP data set. The contours are drawn at 1.0 σ (green) and 4.0 σ (magenta). (B) Thermal ellipsoid diagram for the heme group drawn using ORTEP. The structure, including anisotropic B-factors, was refined using the SHELXL program [44]. Sundaramoorthy et al. Page 12 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. (A) The overall structure of MnP. The red spheres are structural CaII ions conserved in extracellular heme peroxidases. The location of the substrate, MnII, near the heme, is indicated. (B) The active site structure of MnP. This architecture is highly conserved in heme peroxidase. The main variations are the Phe residues which are Trp in the intercellular peroxidases, cytochrome c and ascorbate peroxidase. The Asp242-His173 pair is conserved. Sundaramoorthy et al. Page 13 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. (A). C-terminal region of the MnP structure. The loop encompassing residues 342-352 traverses close to the MnII binding site. This loop is disordered when MnII is bound in the 0.93 Å structure of native Mn-MnP but is ordered in the substrate free MnP structure at 1.05 Å. 2Fo-Fc electron density for the ordered (B) and disordered (C) C-terminal loop is shown contoured at 1.0 σ. Sundaramoorthy et al. Page 14 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. Stereo views of the electron density maps obtained from the 0.93 Å Mn-MnP structure contoured at 1.6σ (top) and 1.05Å Mn free MnP structure obtained from EDTA-treated crystal contoured at 1.0σbottom). MnII (cyan) and all its ligands including two water molecules are labeled for the MnII-bound structure. Glu39 is in a single conformation in this structure where it is in two conformations in the MnII-free structure, as shown in the figure. Sundaramoorthy et al. Page 15 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. 2Fo – Fc (green, 1 σ and red, 3 σ and Fo – Fc (blue, 3 σ) electron density maps in the distal (A) and proximal (B) regions. The peaks (blue) in Fo – Fc map indicate potential hydrogen peaks near the proximal Asp242 and distal His46. Sundaramoorthy et al. Page 16 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Sundaramoorthy et al. Page 17 Table 1 Data collection statistics Dataset MnP (-Mn) Mn-MnP Resolution range (Å) ∞ - 1.05 ∞ - 0.93 No. of observations 653,391 1,246,031 Unique reflections 150,666 264,958 Completeness (%) 91.2 (79.3) 97.5 (93.0) Redundancy 4.1 (2.7) 4.9 (2.0) <I>/<σ(I)> 9.0 (3.4) 26.3 (1.9) Rsym 0.05 (0.21) 0.066 (0.512) J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Sundaramoorthy et al. Page 18 Table 2 Refinement statistics MnP (-Mn) Mn-MnP Resolution range (Å) 8.0 – 1.05 8.0 – 0.93 No of atoms Amino acid residues 2,622 2,622 Sugar residues 39 39 Heme 43 43 CaII 2 2 MnII - 1 Glycerol 24 12 Water 549 477 Rwork [working set, Fo > 4σ(Fo)] 11.1 (144,730) 10.7 (191,300) Rwork (working set, all reflections) 11.7 (156,881) 12.4 (243,170) Rfree [test set, Fo > 4σ(Fo)] 13.6 (3,181) 12.4 (3,967) Rfree (test set, all reflections) 13. 9 (3,352) 13.4 (4,736) Rcryst [working + test set, Fo > 4σ(Fo)] 11.1 (147,911) 10.8 (195,267) Rcryst (working + test set, all reflections) 11.6 (160,232) 12.4 (247,906) No. of Parameters 30,135 29,226 No. of observations/No. of parameters 5.3 8.5 No. of restraints 36,426 35,548 Mean isotropic B-factor (Å2) All atoms 12.973 13.670 Main chain atoms 9.279 10.098 Side chain atoms 11.776 13.276 Sugar residues 17.171 16.416 Heme and metal ions 8.394 7.150 Solvent 24.909 25.687 RMS deviation from ideal geometry Bond length (1-2) (Å) 0.017 0.018 Angle distance (1-3) (Å) 0.038 0.033 Chiral volume (Å3) 0.129 0.096 Non-zero chiral volume (Å3) 0.219 0.105 Deviation from planes (Å) 0.029 0.028 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Sundaramoorthy et al. Page 19 Table 3 Metal-ligand distances Metal – ligand MnP (-Mn) (Å) Mn-MnP (Å) H2O* MnII Heme propionate 2.25 2.12 Glu35 OE1 2.36 2.14 Glu39 OE1 -- 2.27 Asp179 OD1 -- 2.28 Water 1 (1040) 2.56 2.25 Water 2 (1108) 2.64 2.26 Proximal CaII Carbonyl O-Ser174 2.36 2.37 Side chain OG1-Ser174 2.47 2.47 Side chain carboxyl Asp191 2.42 2.44 Carbonyl O-Thr193 2.37 2.38 Side chain OG1-Thr193 2.50 2.49 Carbonyl O-Thr196 2.49 2.51 Side chain carboxyl Asp198 2.46 2.48 Distal CaII Carbonyl O-Asp47 2.44 2.44 Side chain carboxyl Asp47 2.31 2.31 Carbonyl O-Gly62 2.44 2.46 Side chain carboxyl Asp64 2.39 2.42 Side chain O-Ser66 2.49 2.49 Water 1 (1085) 2.39 2.41 Water 2 (1027) 2.35 2.33 *A water or unidentified monovalent cation occupies the Mn-binding site in the substrate free structure. J Inorg Biochem. Author manuscript; available in PMC 2011 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Sundaramoorthy et al. Page 20 Table 4 Heme parameters Distance (Å) MnP (-Mn) (Å) Mn-MnP (Å) CcP (Å) Fe—His N 2.10 2.07 2.07 Fe—pyrrole N 2.02 2.04 2.05 Fe—pyrrole N plane 0.28 J Inorg Biochem. Author manuscript; available in PMC 2011 June 1.
3M5X
Crystal structure of the mutant V182A,I199A of orotidine 5'-monophosphate decarboxylase from Methanobacterium thermoautotrophicum
Conformational Changes in Orotidine 5′-Monophosphate Decarboxylase: “Remote” Residues that Stabilize the Active Conformation† B. McKay Wood‡, Tina L. Amyes§, Alexander A. Fedorov∥, Elena V. Fedorov∥, Andrew Shabila‡, Steven C. Almo∥, John P. Richard§, and John A. Gerlt*,‡ ‡Departments of Biochemistry and Chemistry, University of Illinois at Urbana-Champaign §Department of Chemistry, University at Buffalo, Buffalo, NY 14260 ∥Department of Biochemistry, Albert Einstein College of Medicine, Bronx, NY 10461 Abstract The structural factors responsible for the extraordinary rate enhancement (~1017) of the reaction catalyzed by orotidine 5′-monophosphate decarboxylase (OMPDC) have not been defined. Catalysis requires a conformational change that closes an active site loop and “clamps” the orotate base proximal to hydrogen-bonded networks that destabilize the substrate and stabilize the intermediate. In the OMPDC from Methanobacter thermoautotrophicus, a “remote” structurally conserved cluster of hydrophobic residues that includes Val 182 in the active site loop is assembled in the closed, catalytically active conformation. Substitution of these residues with Ala decreases kcat/Km with a minimal effect on kcat, providing evidence that the cluster stabilizes the closed conformation. The intrinsic binding energies of the 5′-phosphate group of OMP for the mutant enzymes are similar to that for the wild type, supporting this conclusion. Orotidine 5′-monophosphate decarboxylase (OMPDC1) is one of Nature’s best catalysts: the reaction occurs with a rate enhancement of ~1017 and a proficiency of ~1023 M (2). The reaction coordinate includes a vinyl carbanion intermediate (Scheme 1) (3, 4). Structures of OMPDCs complexed with UMP or 6-hydroxyUMP, a n intermediate analog (5-9), reveal hydrogen-bonded networks proximal to 1) C6 of the pyrimidine (Asp 70-Lys 72-Asp 75 in the OMPDC from Methanobacter thermoautotrophicus; MtOMPDC); and 2) O2, N3, and O4 of the base (Ser 127-Gln 185). These destabilize the substrate (9) and stabilize the intermediate, although the structural strategy for the latter is uncertain. The 5′-phosphate group binds in a conserved motif at the ends of the seventh and eighth β- strands of the (β/α)8-barrel structure, with interactions to backbone NH groups as well as the guanidinium group of a conserved Arg (Arg 203 in MtOMPDC). These interactions are important for catalysis: 1) kcat/Km for OMP exceeds that for orotidine by a factor of ~1011 for the OMPDC from Saccharomyces cerevisiae (ScOMPDC) (10); 2) kcat/Km for OMP †This work was supported by NIH Grants GM039754 to J.P.R. and GM065155 to J.A.G. *To whom correspondence should be addressed: J.A.G.: telephone, (217) 244-7414; fax: (217) 244-6538; j-gerlt@uiuc.edu.. SUPPORTING INFORMATION AVAILABLE Descriptions of the experimental procedures. This material is available free of charge via the Internet at http://pubs.acs.org. 1Abbreviations: OMP, orotidine 5′-monophosphate; OMPDC, OMP decarboxylase; MtOMPDC, OMPDC from Methanobacter thermoautotrophicus; ScOMPDC, OMPDC from Saccharomyces cerevisiae; EO, 1-(β-D-erythrofuranosyl)-orotic acid; HPi, phosphate dianion; IBE, intrinsic binding energy. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 December 26. Published in final edited form as: Biochemistry. 2010 May 4; 49(17): 3514–3516. doi:10.1021/bi100443a. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript exceeds that for 1-(β-D-erythrofuranosyl)orotic acid (EO; 5′-truncated OMP analog) by factors of 5.2 × 108 and 3.6 × 108 for ScOMPDC and MtOMPDC, respectively (11, 12); and 3 ) phosphite dianion ( H Pi) activates decarboxylation of EO by factors of 5.6 × 105 M−1 and 2.9 × 105 M−1 for ScOMPDC and MtOMPDC, respectively (11, 12). The “intrinsic binding energy” [IBE; (13)] of the 5′-phosphate/HPi 1) increases the affinity for the substrate (e.g., OMP vs. orotidine/EO); and 2) enables decarboxylation by juxtaposition of the substrate with the active site hydrogen-bonded networks (substrate destabilization and intermediate stabilization). How the IBE promotes catalysis is unknown but required to understand the structural basis for the rate enhancement. A loop located at the end of the seventh β-strand closes over the active site when OMP binds (Figure 1). Although the active site loops differ in both length and sequence in divergent OMPDCs (12), each includes a spatially conserved Gln (Gln 185 in MtOMPDC) hydrogen- bonded to both the 5′-phosphate and O2 of the pyrimidine as well as a conserved Ser at the end of the fifth β-strand that also is hydrogen-bonded to N3 of the pyrimidine (Ser 127). We characterized the importance of these “clamp” residues in ScOMPDC (Gln 215-Ser 154) using EO/HPi and confirmed that the 5′-phosphate/HPi “switch” is required to activate the enzyme (14). The structures of divergent OMPDCs reveal that OMP binding is always accompanied by a conformational change (Figure 1). The most obvious component is closure of the active site loop. However, the (β/α)8-barrel structure can be divided into two domains, one formed from the second, third, fourth, and fifth β-strands (where the hydrogen-bonded Asp 70-Lys 72-Asp 75 motif and Ser 127 are located) and the second from the sixth, seventh, eighth, and first β-strands (where the phosphate binding motif and the active site loop, including Gln 185, are located) (15). OMP binding reorients the domains, with the latter domain moving toward the former, enforcing the orotate carboxylate group to be juxtaposed vis a vis Asp 70- Lys 72-Asp 75 and, also, allowing formation of the Ser 127-Gln 185 “clamp”. Thus, the transition between the open and closed conformations is more complicated than “simple” hinge motion of the loop on the rigid framework of the (β/α)8-barrel structure. In this Report we identify “remote” residues involved in this conformational change and quantitate their importance in promoting and stabilizing the catalytically competent form of the enzyme. The active site loop of MtOMPDC, P180-G181-V182-G183-A184-Q185-G186-G187-D188, is disordered in the absence of substrate but ordered and closed in its presence (Figure 1). In the liganded structure two residues in the loop make contacts with the (β/α)8-barrel scaffold: 1) Gln 185 is hydrogen-bonded to Ser 127 (vide infra); and 2) Val 182 is embedded in a hydrophobic cluster also formed by Ile 199, Val 201, and Ile 218. The conservation of this hydrophobic cluster in all OMPDCs suggests its importance in a common catalytic strategy. We probed this strategy by mutagenesis of these hydrophobic residues. Ala substitutions for residues in the hydrophobic cluster cause substantial decreases in kcat/ Km for decarboxylation of OMP but have little effect on kcat (Table 1). We determined high resolution X-ray structures (≤ 1.4 Å) for each mutant in the presence of 6-hydroxyUMP (Figure 2). The liganded structures superimpose well with that of wild type, with only small differences observed at the sites of the substitutions (panel A). The active sites are identical to that of wild type (panel B), explaining the minimal impact on kcat. The mutated residues are remote from the active site (Figures 1 and 2). Thus, the effects of the substitutions on kcat/Km cannot be explained by altered direct interactions with the substrate. Instead, the effects can be explained by decreased stabilities of the closed conformation in which the substrate is destabilized (9) and the anionic intermediate is stabilized. A consistent model (Scheme 2) is that an equilibrium of open (Eo) and closed Wood et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (Ec) conformations exists in the absence or presence of substrate (KC’ or KC; KC’ << 1). The substitutions destabilize Ec (decrease KC and KC ’), so kcat/Km is decreased (the energy difference between Eo + S and [Ec•S]‡ is increased). However, the invariance of kcat establishes that the substitutions do not alter the reactivitiy of Ec•S (the energy difference between Ec•S and [Ec•S]‡). The energy to form Ec•S from Eo + S is achieved from 1) interactions of the 5′-phosphate group with its binding motif (IBE; vide infra), and 2) an increased concentration of OMP to form Ec•S (KC/KS or, equivalently, KC’/KS’, although the former is expected to be the relevant pathway). We also used the two part EO•HPi substrate. The values of kcat/Km for EO are decreased relative to that for wild type (Table 1); these can be explained by decreased populations of Ec (Scheme 2), assuming that EO, without a 5′-phosphate group, is unable to promote the transition from Eo to Ec. The values of kcat/Km for OMP and kcat/Km for EO allow calculation of the IBE for the 5′-phosphate group of OMP (Table 1). HPi activates the mutants as judged by the values of the third-order rate constant, (kcat/ Km)EO•HPi/Kd. The equivalent changes (ΔΔG‡) in kcat/Km for both OMP and EO and the third-order rate constant indicate that all three measure the effects of the substitutions on the values of KC. The values of the IBEs for the 5′-phosphate group for the mutants are the same as that for the wild type, establishing that decarboxylation in the Ec•S complex occurs with equivalent amounts of ground state destabilization (9) and transition state stabilization (as also reflected by the invariant values of kcat). The IBEs provide further support for the role of the “remote” hydrophobic cluster in stabilizing Ec relative to Eo but do not directly participating in catalysis. Our experiments implicate a structurally conserved hydrophobic cluster, Val 182, Ile 199, Val 201, and Ile 218 in MtOMPDC, in stabilizing the closed conformation required for catalysis. Its identification provides evidence that structural elements distal from the active site, in addition to the proximal active site loop that closes to “clamp” the substrate, are required for OMPDC’s extraordinary catalytic efficiency and proficiency. Supplementary Material Refer to Web version on PubMed Central for supplementary material. REFERENCES (1). Wood BM, Chan KK, Amyes TL, Richard JP, Gerlt JA. Biochemistry. 2009; 48:5510–5517. [PubMed: 19435313] (2). Radzicka A, Wolfenden R. Science. 1995; 267:90–93. [PubMed: 7809611] (3). Toth K, Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2007; 129:12946–12947. [PubMed: 17918849] (4). Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2008; 130:1574–1575. [PubMed: 18186641] (5). Miller BG, Hassell AM, Wolfenden R, Milburn MV, Short SA. Proc. Natl. Acad. Sci. U A. 2000; 97:2011–2016. (6). Appleby TC, Kinsland C, Begley TP, Ealick SE. Proc atl Acad Sci U S A. 2000; 97:2005–2010. (7). Harris P, Navarro Poulsen JC, Jensen KF, Larsen S. Biochemistry. 2000; 39:4217–4224. [PubMed: 10757968] (8). Wu N, Mo Y, Gao J, Pai EF. Proc. Natl. Acad. Sci. U S A. 2000; 97:2017–2022. [PubMed: 10681441] Wood et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (9). Chan KK, Wood BM, Fedorov AA, Fedorov EV, Imker HJ, Amyes TL, Richard JP, Almo SC, Gerlt JA. Biochemistry. 2009; 48:5518–5531. [PubMed: 19435314] (10). Sievers A, Wolfenden R. Bioorg. Chem. 2005; 33:45–52. [PubMed: 15668182] (11). Amyes TL, Richard JP, Tait JJ. J. Am. Chem. Soc. 2005; 127:15708–15709. [PubMed: 16277505] (12). Toth K, Amyes TL, Wood BM, Chan KK, Gerlt JA, Richard JP. Biochemistry. 2009; 48:8006– 8013. [PubMed: 19618917] (13). Jencks WP. Adv. Enzymol. Rel. Areas. Mol. Biol. 1975; 43:219–410. (14). Barnett SA, Amyes TL, Wood BM, Gerlt JA, Richard JP. Biochemistry. 2008; 47:7785–4487. [PubMed: 18598058] (15). Harris P, Poulsen JC, Jensen KF, Larsen S. J. Mol. Biol. 2002; 318:1019–1029. [PubMed: 12054799] (16). Go MK, Amyes TL, Richard JP. Biochemistry. 2009; 48:5769–5778. [PubMed: 19425580] Wood et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Superposition of the active site of wild type MtOMPDC in the absence (cyan; disordered loop depicted with the dotted line ) and presence ( grey ) of 6-hydroxyUMP. The carbons of 6-hydroxyUMP are highlighted in yellow; the carbons of Val 182 and Arg 203 in the liganded structure are highlighted in orange. Wood et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Superpositions of the 6-hydroxyUMP-liganded structures of wild type MtOMPDC and the single mutants in the hydrophobic cluster. Panel A, hydrophobic cluster; panel B, active sites. Wood et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. Wood et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 2. Wood et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Wood et al. Page 9 Table 1 Kinetic Constants for OMP, EO and EO/HPi and Intrinsic Binding Energies of the 5′-Phosphate Group of OMP at pH 7.1 and 25 °C. MtOMPDC kcat OMP s−1 kcat/Km OMP M−1 s−1 ΔΔG‡ kcal/mola kcat/Km EO M−1 s−1 ΔΔG‡ kcal/mola (kcat/Km)/K D EO•HPib M−2 s−1 ΔΔG‡ kcal/mola 5′-Phosphate IBEc kcal/mol Wild type 4.6 2.9 × 106 8.7 × 10−3 2500 11.6 d V182A 3.4 1.4 × 105 1.8 1.3 × 10−3 1.1 190 1.5 10.9 I199A 3.9 9.1 × 105 0.7 1.9 × 10−3 0.9 980 0.6 11.8 V201A 4.0 9.5 × 105 0.7 3.1 × 10−3 0.6 690 0.8 11.5 I218A 3.3 2.8 × 105 1.4 2.3 × 10−3 0.8 340 1.2 11.0 V182A/I199A 3.1 4.9 × 104 2.4 3.9 × 10−4 1.8 81 2.0 11.0 V182A/V201A 2.5 4.9 × 104 2.4 5.0 × 10−4 1.7 30 2.6 10.9 aCalculated from the ratio of the second-order or third-order rate constants for the wild type and mutant enzyme. bThird-order rate constant for reaction of EO/HPi. cTransition state stabilization by the 5′-phosphate group of OMP, calculated from the ratio of the values of kcat/Km for OMP and EO. IBEs for phosphite dianion can be calculated from the ratio of (kcat/ Km)/KD for EOHPi and kcat/Km for EO. Biochemistry. Author manuscript; available in PMC 2011 December 26.
3M5Y
Crystal structure of the mutant V182A,V201A of orotidine 5'-monophosphate decarboxylase from Methanobacterium thermoautotrophicum
Conformational Changes in Orotidine 5′-Monophosphate Decarboxylase: “Remote” Residues that Stabilize the Active Conformation† B. McKay Wood‡, Tina L. Amyes§, Alexander A. Fedorov∥, Elena V. Fedorov∥, Andrew Shabila‡, Steven C. Almo∥, John P. Richard§, and John A. Gerlt*,‡ ‡Departments of Biochemistry and Chemistry, University of Illinois at Urbana-Champaign §Department of Chemistry, University at Buffalo, Buffalo, NY 14260 ∥Department of Biochemistry, Albert Einstein College of Medicine, Bronx, NY 10461 Abstract The structural factors responsible for the extraordinary rate enhancement (~1017) of the reaction catalyzed by orotidine 5′-monophosphate decarboxylase (OMPDC) have not been defined. Catalysis requires a conformational change that closes an active site loop and “clamps” the orotate base proximal to hydrogen-bonded networks that destabilize the substrate and stabilize the intermediate. In the OMPDC from Methanobacter thermoautotrophicus, a “remote” structurally conserved cluster of hydrophobic residues that includes Val 182 in the active site loop is assembled in the closed, catalytically active conformation. Substitution of these residues with Ala decreases kcat/Km with a minimal effect on kcat, providing evidence that the cluster stabilizes the closed conformation. The intrinsic binding energies of the 5′-phosphate group of OMP for the mutant enzymes are similar to that for the wild type, supporting this conclusion. Orotidine 5′-monophosphate decarboxylase (OMPDC1) is one of Nature’s best catalysts: the reaction occurs with a rate enhancement of ~1017 and a proficiency of ~1023 M (2). The reaction coordinate includes a vinyl carbanion intermediate (Scheme 1) (3, 4). Structures of OMPDCs complexed with UMP or 6-hydroxyUMP, a n intermediate analog (5-9), reveal hydrogen-bonded networks proximal to 1) C6 of the pyrimidine (Asp 70-Lys 72-Asp 75 in the OMPDC from Methanobacter thermoautotrophicus; MtOMPDC); and 2) O2, N3, and O4 of the base (Ser 127-Gln 185). These destabilize the substrate (9) and stabilize the intermediate, although the structural strategy for the latter is uncertain. The 5′-phosphate group binds in a conserved motif at the ends of the seventh and eighth β- strands of the (β/α)8-barrel structure, with interactions to backbone NH groups as well as the guanidinium group of a conserved Arg (Arg 203 in MtOMPDC). These interactions are important for catalysis: 1) kcat/Km for OMP exceeds that for orotidine by a factor of ~1011 for the OMPDC from Saccharomyces cerevisiae (ScOMPDC) (10); 2) kcat/Km for OMP †This work was supported by NIH Grants GM039754 to J.P.R. and GM065155 to J.A.G. *To whom correspondence should be addressed: J.A.G.: telephone, (217) 244-7414; fax: (217) 244-6538; j-gerlt@uiuc.edu.. SUPPORTING INFORMATION AVAILABLE Descriptions of the experimental procedures. This material is available free of charge via the Internet at http://pubs.acs.org. 1Abbreviations: OMP, orotidine 5′-monophosphate; OMPDC, OMP decarboxylase; MtOMPDC, OMPDC from Methanobacter thermoautotrophicus; ScOMPDC, OMPDC from Saccharomyces cerevisiae; EO, 1-(β-D-erythrofuranosyl)-orotic acid; HPi, phosphate dianion; IBE, intrinsic binding energy. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 December 26. Published in final edited form as: Biochemistry. 2010 May 4; 49(17): 3514–3516. doi:10.1021/bi100443a. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript exceeds that for 1-(β-D-erythrofuranosyl)orotic acid (EO; 5′-truncated OMP analog) by factors of 5.2 × 108 and 3.6 × 108 for ScOMPDC and MtOMPDC, respectively (11, 12); and 3 ) phosphite dianion ( H Pi) activates decarboxylation of EO by factors of 5.6 × 105 M−1 and 2.9 × 105 M−1 for ScOMPDC and MtOMPDC, respectively (11, 12). The “intrinsic binding energy” [IBE; (13)] of the 5′-phosphate/HPi 1) increases the affinity for the substrate (e.g., OMP vs. orotidine/EO); and 2) enables decarboxylation by juxtaposition of the substrate with the active site hydrogen-bonded networks (substrate destabilization and intermediate stabilization). How the IBE promotes catalysis is unknown but required to understand the structural basis for the rate enhancement. A loop located at the end of the seventh β-strand closes over the active site when OMP binds (Figure 1). Although the active site loops differ in both length and sequence in divergent OMPDCs (12), each includes a spatially conserved Gln (Gln 185 in MtOMPDC) hydrogen- bonded to both the 5′-phosphate and O2 of the pyrimidine as well as a conserved Ser at the end of the fifth β-strand that also is hydrogen-bonded to N3 of the pyrimidine (Ser 127). We characterized the importance of these “clamp” residues in ScOMPDC (Gln 215-Ser 154) using EO/HPi and confirmed that the 5′-phosphate/HPi “switch” is required to activate the enzyme (14). The structures of divergent OMPDCs reveal that OMP binding is always accompanied by a conformational change (Figure 1). The most obvious component is closure of the active site loop. However, the (β/α)8-barrel structure can be divided into two domains, one formed from the second, third, fourth, and fifth β-strands (where the hydrogen-bonded Asp 70-Lys 72-Asp 75 motif and Ser 127 are located) and the second from the sixth, seventh, eighth, and first β-strands (where the phosphate binding motif and the active site loop, including Gln 185, are located) (15). OMP binding reorients the domains, with the latter domain moving toward the former, enforcing the orotate carboxylate group to be juxtaposed vis a vis Asp 70- Lys 72-Asp 75 and, also, allowing formation of the Ser 127-Gln 185 “clamp”. Thus, the transition between the open and closed conformations is more complicated than “simple” hinge motion of the loop on the rigid framework of the (β/α)8-barrel structure. In this Report we identify “remote” residues involved in this conformational change and quantitate their importance in promoting and stabilizing the catalytically competent form of the enzyme. The active site loop of MtOMPDC, P180-G181-V182-G183-A184-Q185-G186-G187-D188, is disordered in the absence of substrate but ordered and closed in its presence (Figure 1). In the liganded structure two residues in the loop make contacts with the (β/α)8-barrel scaffold: 1) Gln 185 is hydrogen-bonded to Ser 127 (vide infra); and 2) Val 182 is embedded in a hydrophobic cluster also formed by Ile 199, Val 201, and Ile 218. The conservation of this hydrophobic cluster in all OMPDCs suggests its importance in a common catalytic strategy. We probed this strategy by mutagenesis of these hydrophobic residues. Ala substitutions for residues in the hydrophobic cluster cause substantial decreases in kcat/ Km for decarboxylation of OMP but have little effect on kcat (Table 1). We determined high resolution X-ray structures (≤ 1.4 Å) for each mutant in the presence of 6-hydroxyUMP (Figure 2). The liganded structures superimpose well with that of wild type, with only small differences observed at the sites of the substitutions (panel A). The active sites are identical to that of wild type (panel B), explaining the minimal impact on kcat. The mutated residues are remote from the active site (Figures 1 and 2). Thus, the effects of the substitutions on kcat/Km cannot be explained by altered direct interactions with the substrate. Instead, the effects can be explained by decreased stabilities of the closed conformation in which the substrate is destabilized (9) and the anionic intermediate is stabilized. A consistent model (Scheme 2) is that an equilibrium of open (Eo) and closed Wood et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (Ec) conformations exists in the absence or presence of substrate (KC’ or KC; KC’ << 1). The substitutions destabilize Ec (decrease KC and KC ’), so kcat/Km is decreased (the energy difference between Eo + S and [Ec•S]‡ is increased). However, the invariance of kcat establishes that the substitutions do not alter the reactivitiy of Ec•S (the energy difference between Ec•S and [Ec•S]‡). The energy to form Ec•S from Eo + S is achieved from 1) interactions of the 5′-phosphate group with its binding motif (IBE; vide infra), and 2) an increased concentration of OMP to form Ec•S (KC/KS or, equivalently, KC’/KS’, although the former is expected to be the relevant pathway). We also used the two part EO•HPi substrate. The values of kcat/Km for EO are decreased relative to that for wild type (Table 1); these can be explained by decreased populations of Ec (Scheme 2), assuming that EO, without a 5′-phosphate group, is unable to promote the transition from Eo to Ec. The values of kcat/Km for OMP and kcat/Km for EO allow calculation of the IBE for the 5′-phosphate group of OMP (Table 1). HPi activates the mutants as judged by the values of the third-order rate constant, (kcat/ Km)EO•HPi/Kd. The equivalent changes (ΔΔG‡) in kcat/Km for both OMP and EO and the third-order rate constant indicate that all three measure the effects of the substitutions on the values of KC. The values of the IBEs for the 5′-phosphate group for the mutants are the same as that for the wild type, establishing that decarboxylation in the Ec•S complex occurs with equivalent amounts of ground state destabilization (9) and transition state stabilization (as also reflected by the invariant values of kcat). The IBEs provide further support for the role of the “remote” hydrophobic cluster in stabilizing Ec relative to Eo but do not directly participating in catalysis. Our experiments implicate a structurally conserved hydrophobic cluster, Val 182, Ile 199, Val 201, and Ile 218 in MtOMPDC, in stabilizing the closed conformation required for catalysis. Its identification provides evidence that structural elements distal from the active site, in addition to the proximal active site loop that closes to “clamp” the substrate, are required for OMPDC’s extraordinary catalytic efficiency and proficiency. Supplementary Material Refer to Web version on PubMed Central for supplementary material. REFERENCES (1). Wood BM, Chan KK, Amyes TL, Richard JP, Gerlt JA. Biochemistry. 2009; 48:5510–5517. [PubMed: 19435313] (2). Radzicka A, Wolfenden R. Science. 1995; 267:90–93. [PubMed: 7809611] (3). Toth K, Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2007; 129:12946–12947. [PubMed: 17918849] (4). Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2008; 130:1574–1575. [PubMed: 18186641] (5). Miller BG, Hassell AM, Wolfenden R, Milburn MV, Short SA. Proc. Natl. Acad. Sci. U A. 2000; 97:2011–2016. (6). Appleby TC, Kinsland C, Begley TP, Ealick SE. Proc atl Acad Sci U S A. 2000; 97:2005–2010. (7). Harris P, Navarro Poulsen JC, Jensen KF, Larsen S. Biochemistry. 2000; 39:4217–4224. [PubMed: 10757968] (8). Wu N, Mo Y, Gao J, Pai EF. Proc. Natl. Acad. Sci. U S A. 2000; 97:2017–2022. [PubMed: 10681441] Wood et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (9). Chan KK, Wood BM, Fedorov AA, Fedorov EV, Imker HJ, Amyes TL, Richard JP, Almo SC, Gerlt JA. Biochemistry. 2009; 48:5518–5531. [PubMed: 19435314] (10). Sievers A, Wolfenden R. Bioorg. Chem. 2005; 33:45–52. [PubMed: 15668182] (11). Amyes TL, Richard JP, Tait JJ. J. Am. Chem. Soc. 2005; 127:15708–15709. [PubMed: 16277505] (12). Toth K, Amyes TL, Wood BM, Chan KK, Gerlt JA, Richard JP. Biochemistry. 2009; 48:8006– 8013. [PubMed: 19618917] (13). Jencks WP. Adv. Enzymol. Rel. Areas. Mol. Biol. 1975; 43:219–410. (14). Barnett SA, Amyes TL, Wood BM, Gerlt JA, Richard JP. Biochemistry. 2008; 47:7785–4487. [PubMed: 18598058] (15). Harris P, Poulsen JC, Jensen KF, Larsen S. J. Mol. Biol. 2002; 318:1019–1029. [PubMed: 12054799] (16). Go MK, Amyes TL, Richard JP. Biochemistry. 2009; 48:5769–5778. [PubMed: 19425580] Wood et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Superposition of the active site of wild type MtOMPDC in the absence (cyan; disordered loop depicted with the dotted line ) and presence ( grey ) of 6-hydroxyUMP. The carbons of 6-hydroxyUMP are highlighted in yellow; the carbons of Val 182 and Arg 203 in the liganded structure are highlighted in orange. Wood et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Superpositions of the 6-hydroxyUMP-liganded structures of wild type MtOMPDC and the single mutants in the hydrophobic cluster. Panel A, hydrophobic cluster; panel B, active sites. Wood et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. Wood et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 2. Wood et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Wood et al. Page 9 Table 1 Kinetic Constants for OMP, EO and EO/HPi and Intrinsic Binding Energies of the 5′-Phosphate Group of OMP at pH 7.1 and 25 °C. MtOMPDC kcat OMP s−1 kcat/Km OMP M−1 s−1 ΔΔG‡ kcal/mola kcat/Km EO M−1 s−1 ΔΔG‡ kcal/mola (kcat/Km)/K D EO•HPib M−2 s−1 ΔΔG‡ kcal/mola 5′-Phosphate IBEc kcal/mol Wild type 4.6 2.9 × 106 8.7 × 10−3 2500 11.6 d V182A 3.4 1.4 × 105 1.8 1.3 × 10−3 1.1 190 1.5 10.9 I199A 3.9 9.1 × 105 0.7 1.9 × 10−3 0.9 980 0.6 11.8 V201A 4.0 9.5 × 105 0.7 3.1 × 10−3 0.6 690 0.8 11.5 I218A 3.3 2.8 × 105 1.4 2.3 × 10−3 0.8 340 1.2 11.0 V182A/I199A 3.1 4.9 × 104 2.4 3.9 × 10−4 1.8 81 2.0 11.0 V182A/V201A 2.5 4.9 × 104 2.4 5.0 × 10−4 1.7 30 2.6 10.9 aCalculated from the ratio of the second-order or third-order rate constants for the wild type and mutant enzyme. bThird-order rate constant for reaction of EO/HPi. cTransition state stabilization by the 5′-phosphate group of OMP, calculated from the ratio of the values of kcat/Km for OMP and EO. IBEs for phosphite dianion can be calculated from the ratio of (kcat/ Km)/KD for EOHPi and kcat/Km for EO. Biochemistry. Author manuscript; available in PMC 2011 December 26.
3M5Z
Crystal structure of the mutant V182A,I218A of orotidine 5'-monophosphate decarboxylase from Methanobacterium thermoautotrophicum
Conformational Changes in Orotidine 5′-Monophosphate Decarboxylase: “Remote” Residues that Stabilize the Active Conformation† B. McKay Wood‡, Tina L. Amyes§, Alexander A. Fedorov∥, Elena V. Fedorov∥, Andrew Shabila‡, Steven C. Almo∥, John P. Richard§, and John A. Gerlt*,‡ ‡Departments of Biochemistry and Chemistry, University of Illinois at Urbana-Champaign §Department of Chemistry, University at Buffalo, Buffalo, NY 14260 ∥Department of Biochemistry, Albert Einstein College of Medicine, Bronx, NY 10461 Abstract The structural factors responsible for the extraordinary rate enhancement (~1017) of the reaction catalyzed by orotidine 5′-monophosphate decarboxylase (OMPDC) have not been defined. Catalysis requires a conformational change that closes an active site loop and “clamps” the orotate base proximal to hydrogen-bonded networks that destabilize the substrate and stabilize the intermediate. In the OMPDC from Methanobacter thermoautotrophicus, a “remote” structurally conserved cluster of hydrophobic residues that includes Val 182 in the active site loop is assembled in the closed, catalytically active conformation. Substitution of these residues with Ala decreases kcat/Km with a minimal effect on kcat, providing evidence that the cluster stabilizes the closed conformation. The intrinsic binding energies of the 5′-phosphate group of OMP for the mutant enzymes are similar to that for the wild type, supporting this conclusion. Orotidine 5′-monophosphate decarboxylase (OMPDC1) is one of Nature’s best catalysts: the reaction occurs with a rate enhancement of ~1017 and a proficiency of ~1023 M (2). The reaction coordinate includes a vinyl carbanion intermediate (Scheme 1) (3, 4). Structures of OMPDCs complexed with UMP or 6-hydroxyUMP, a n intermediate analog (5-9), reveal hydrogen-bonded networks proximal to 1) C6 of the pyrimidine (Asp 70-Lys 72-Asp 75 in the OMPDC from Methanobacter thermoautotrophicus; MtOMPDC); and 2) O2, N3, and O4 of the base (Ser 127-Gln 185). These destabilize the substrate (9) and stabilize the intermediate, although the structural strategy for the latter is uncertain. The 5′-phosphate group binds in a conserved motif at the ends of the seventh and eighth β- strands of the (β/α)8-barrel structure, with interactions to backbone NH groups as well as the guanidinium group of a conserved Arg (Arg 203 in MtOMPDC). These interactions are important for catalysis: 1) kcat/Km for OMP exceeds that for orotidine by a factor of ~1011 for the OMPDC from Saccharomyces cerevisiae (ScOMPDC) (10); 2) kcat/Km for OMP †This work was supported by NIH Grants GM039754 to J.P.R. and GM065155 to J.A.G. *To whom correspondence should be addressed: J.A.G.: telephone, (217) 244-7414; fax: (217) 244-6538; j-gerlt@uiuc.edu.. SUPPORTING INFORMATION AVAILABLE Descriptions of the experimental procedures. This material is available free of charge via the Internet at http://pubs.acs.org. 1Abbreviations: OMP, orotidine 5′-monophosphate; OMPDC, OMP decarboxylase; MtOMPDC, OMPDC from Methanobacter thermoautotrophicus; ScOMPDC, OMPDC from Saccharomyces cerevisiae; EO, 1-(β-D-erythrofuranosyl)-orotic acid; HPi, phosphate dianion; IBE, intrinsic binding energy. NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 December 26. Published in final edited form as: Biochemistry. 2010 May 4; 49(17): 3514–3516. doi:10.1021/bi100443a. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript exceeds that for 1-(β-D-erythrofuranosyl)orotic acid (EO; 5′-truncated OMP analog) by factors of 5.2 × 108 and 3.6 × 108 for ScOMPDC and MtOMPDC, respectively (11, 12); and 3 ) phosphite dianion ( H Pi) activates decarboxylation of EO by factors of 5.6 × 105 M−1 and 2.9 × 105 M−1 for ScOMPDC and MtOMPDC, respectively (11, 12). The “intrinsic binding energy” [IBE; (13)] of the 5′-phosphate/HPi 1) increases the affinity for the substrate (e.g., OMP vs. orotidine/EO); and 2) enables decarboxylation by juxtaposition of the substrate with the active site hydrogen-bonded networks (substrate destabilization and intermediate stabilization). How the IBE promotes catalysis is unknown but required to understand the structural basis for the rate enhancement. A loop located at the end of the seventh β-strand closes over the active site when OMP binds (Figure 1). Although the active site loops differ in both length and sequence in divergent OMPDCs (12), each includes a spatially conserved Gln (Gln 185 in MtOMPDC) hydrogen- bonded to both the 5′-phosphate and O2 of the pyrimidine as well as a conserved Ser at the end of the fifth β-strand that also is hydrogen-bonded to N3 of the pyrimidine (Ser 127). We characterized the importance of these “clamp” residues in ScOMPDC (Gln 215-Ser 154) using EO/HPi and confirmed that the 5′-phosphate/HPi “switch” is required to activate the enzyme (14). The structures of divergent OMPDCs reveal that OMP binding is always accompanied by a conformational change (Figure 1). The most obvious component is closure of the active site loop. However, the (β/α)8-barrel structure can be divided into two domains, one formed from the second, third, fourth, and fifth β-strands (where the hydrogen-bonded Asp 70-Lys 72-Asp 75 motif and Ser 127 are located) and the second from the sixth, seventh, eighth, and first β-strands (where the phosphate binding motif and the active site loop, including Gln 185, are located) (15). OMP binding reorients the domains, with the latter domain moving toward the former, enforcing the orotate carboxylate group to be juxtaposed vis a vis Asp 70- Lys 72-Asp 75 and, also, allowing formation of the Ser 127-Gln 185 “clamp”. Thus, the transition between the open and closed conformations is more complicated than “simple” hinge motion of the loop on the rigid framework of the (β/α)8-barrel structure. In this Report we identify “remote” residues involved in this conformational change and quantitate their importance in promoting and stabilizing the catalytically competent form of the enzyme. The active site loop of MtOMPDC, P180-G181-V182-G183-A184-Q185-G186-G187-D188, is disordered in the absence of substrate but ordered and closed in its presence (Figure 1). In the liganded structure two residues in the loop make contacts with the (β/α)8-barrel scaffold: 1) Gln 185 is hydrogen-bonded to Ser 127 (vide infra); and 2) Val 182 is embedded in a hydrophobic cluster also formed by Ile 199, Val 201, and Ile 218. The conservation of this hydrophobic cluster in all OMPDCs suggests its importance in a common catalytic strategy. We probed this strategy by mutagenesis of these hydrophobic residues. Ala substitutions for residues in the hydrophobic cluster cause substantial decreases in kcat/ Km for decarboxylation of OMP but have little effect on kcat (Table 1). We determined high resolution X-ray structures (≤ 1.4 Å) for each mutant in the presence of 6-hydroxyUMP (Figure 2). The liganded structures superimpose well with that of wild type, with only small differences observed at the sites of the substitutions (panel A). The active sites are identical to that of wild type (panel B), explaining the minimal impact on kcat. The mutated residues are remote from the active site (Figures 1 and 2). Thus, the effects of the substitutions on kcat/Km cannot be explained by altered direct interactions with the substrate. Instead, the effects can be explained by decreased stabilities of the closed conformation in which the substrate is destabilized (9) and the anionic intermediate is stabilized. A consistent model (Scheme 2) is that an equilibrium of open (Eo) and closed Wood et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (Ec) conformations exists in the absence or presence of substrate (KC’ or KC; KC’ << 1). The substitutions destabilize Ec (decrease KC and KC ’), so kcat/Km is decreased (the energy difference between Eo + S and [Ec•S]‡ is increased). However, the invariance of kcat establishes that the substitutions do not alter the reactivitiy of Ec•S (the energy difference between Ec•S and [Ec•S]‡). The energy to form Ec•S from Eo + S is achieved from 1) interactions of the 5′-phosphate group with its binding motif (IBE; vide infra), and 2) an increased concentration of OMP to form Ec•S (KC/KS or, equivalently, KC’/KS’, although the former is expected to be the relevant pathway). We also used the two part EO•HPi substrate. The values of kcat/Km for EO are decreased relative to that for wild type (Table 1); these can be explained by decreased populations of Ec (Scheme 2), assuming that EO, without a 5′-phosphate group, is unable to promote the transition from Eo to Ec. The values of kcat/Km for OMP and kcat/Km for EO allow calculation of the IBE for the 5′-phosphate group of OMP (Table 1). HPi activates the mutants as judged by the values of the third-order rate constant, (kcat/ Km)EO•HPi/Kd. The equivalent changes (ΔΔG‡) in kcat/Km for both OMP and EO and the third-order rate constant indicate that all three measure the effects of the substitutions on the values of KC. The values of the IBEs for the 5′-phosphate group for the mutants are the same as that for the wild type, establishing that decarboxylation in the Ec•S complex occurs with equivalent amounts of ground state destabilization (9) and transition state stabilization (as also reflected by the invariant values of kcat). The IBEs provide further support for the role of the “remote” hydrophobic cluster in stabilizing Ec relative to Eo but do not directly participating in catalysis. Our experiments implicate a structurally conserved hydrophobic cluster, Val 182, Ile 199, Val 201, and Ile 218 in MtOMPDC, in stabilizing the closed conformation required for catalysis. Its identification provides evidence that structural elements distal from the active site, in addition to the proximal active site loop that closes to “clamp” the substrate, are required for OMPDC’s extraordinary catalytic efficiency and proficiency. Supplementary Material Refer to Web version on PubMed Central for supplementary material. REFERENCES (1). Wood BM, Chan KK, Amyes TL, Richard JP, Gerlt JA. Biochemistry. 2009; 48:5510–5517. [PubMed: 19435313] (2). Radzicka A, Wolfenden R. Science. 1995; 267:90–93. [PubMed: 7809611] (3). Toth K, Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2007; 129:12946–12947. [PubMed: 17918849] (4). Amyes TL, Wood BM, Chan K, Gerlt JA, Richard JP. J. Am. Chem. Soc. 2008; 130:1574–1575. [PubMed: 18186641] (5). Miller BG, Hassell AM, Wolfenden R, Milburn MV, Short SA. Proc. Natl. Acad. Sci. U A. 2000; 97:2011–2016. (6). Appleby TC, Kinsland C, Begley TP, Ealick SE. Proc atl Acad Sci U S A. 2000; 97:2005–2010. (7). Harris P, Navarro Poulsen JC, Jensen KF, Larsen S. Biochemistry. 2000; 39:4217–4224. [PubMed: 10757968] (8). Wu N, Mo Y, Gao J, Pai EF. Proc. Natl. Acad. Sci. U S A. 2000; 97:2017–2022. [PubMed: 10681441] Wood et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript (9). Chan KK, Wood BM, Fedorov AA, Fedorov EV, Imker HJ, Amyes TL, Richard JP, Almo SC, Gerlt JA. Biochemistry. 2009; 48:5518–5531. [PubMed: 19435314] (10). Sievers A, Wolfenden R. Bioorg. Chem. 2005; 33:45–52. [PubMed: 15668182] (11). Amyes TL, Richard JP, Tait JJ. J. Am. Chem. Soc. 2005; 127:15708–15709. [PubMed: 16277505] (12). Toth K, Amyes TL, Wood BM, Chan KK, Gerlt JA, Richard JP. Biochemistry. 2009; 48:8006– 8013. [PubMed: 19618917] (13). Jencks WP. Adv. Enzymol. Rel. Areas. Mol. Biol. 1975; 43:219–410. (14). Barnett SA, Amyes TL, Wood BM, Gerlt JA, Richard JP. Biochemistry. 2008; 47:7785–4487. [PubMed: 18598058] (15). Harris P, Poulsen JC, Jensen KF, Larsen S. J. Mol. Biol. 2002; 318:1019–1029. [PubMed: 12054799] (16). Go MK, Amyes TL, Richard JP. Biochemistry. 2009; 48:5769–5778. [PubMed: 19425580] Wood et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Superposition of the active site of wild type MtOMPDC in the absence (cyan; disordered loop depicted with the dotted line ) and presence ( grey ) of 6-hydroxyUMP. The carbons of 6-hydroxyUMP are highlighted in yellow; the carbons of Val 182 and Arg 203 in the liganded structure are highlighted in orange. Wood et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Superpositions of the 6-hydroxyUMP-liganded structures of wild type MtOMPDC and the single mutants in the hydrophobic cluster. Panel A, hydrophobic cluster; panel B, active sites. Wood et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. Wood et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 2. Wood et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 December 26. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Wood et al. Page 9 Table 1 Kinetic Constants for OMP, EO and EO/HPi and Intrinsic Binding Energies of the 5′-Phosphate Group of OMP at pH 7.1 and 25 °C. MtOMPDC kcat OMP s−1 kcat/Km OMP M−1 s−1 ΔΔG‡ kcal/mola kcat/Km EO M−1 s−1 ΔΔG‡ kcal/mola (kcat/Km)/K D EO•HPib M−2 s−1 ΔΔG‡ kcal/mola 5′-Phosphate IBEc kcal/mol Wild type 4.6 2.9 × 106 8.7 × 10−3 2500 11.6 d V182A 3.4 1.4 × 105 1.8 1.3 × 10−3 1.1 190 1.5 10.9 I199A 3.9 9.1 × 105 0.7 1.9 × 10−3 0.9 980 0.6 11.8 V201A 4.0 9.5 × 105 0.7 3.1 × 10−3 0.6 690 0.8 11.5 I218A 3.3 2.8 × 105 1.4 2.3 × 10−3 0.8 340 1.2 11.0 V182A/I199A 3.1 4.9 × 104 2.4 3.9 × 10−4 1.8 81 2.0 11.0 V182A/V201A 2.5 4.9 × 104 2.4 5.0 × 10−4 1.7 30 2.6 10.9 aCalculated from the ratio of the second-order or third-order rate constants for the wild type and mutant enzyme. bThird-order rate constant for reaction of EO/HPi. cTransition state stabilization by the 5′-phosphate group of OMP, calculated from the ratio of the values of kcat/Km for OMP and EO. IBEs for phosphite dianion can be calculated from the ratio of (kcat/ Km)/KD for EOHPi and kcat/Km for EO. Biochemistry. Author manuscript; available in PMC 2011 December 26.
3M62
Crystal structure of Ufd2 in complex with the ubiquitin-like (UBL) domain of Rad23
The Yeast E4 Ubiquitin Ligase Ufd2 Interacts with the Ubiquitin-like Domains of Rad23 and Dsk2 via a Novel and Distinct Ubiquitin-like Binding Domain*□ S Received for publication,February 12, 2010, and in revised form, March 22, 2010 Published, JBC Papers in Press,April 28, 2010, DOI 10.1074/jbc.M110.112532 Petra Ha¨nzelmann‡1, Julian Stingele§1, Kay Hofmann¶, Hermann Schindelin‡2, and Shahri Raasi§3 From the ‡Rudolf Virchow Center for Experimental Biomedicine, University of Wu¨rzburg, Josef-Schneider-Strasse 2, 97080 Wu¨rzburg, the §Laboratory of Cellular Biochemistry, Department of Biology, University of Konstanz, 78457 Konstanz, and the ¶Bioinformatics Group, Miltenyi Biotec GmbH, Friedrich-Ebert-Strasse 68, 51429 Bergisch-Gladbach, Germany Proteins containing ubiquitin-like (UBL) and ubiquitin-asso- ciated (UBA) domains interact with various binding partners and function as hubs during ubiquitin-mediated protein degra- dation. A common interaction of the budding yeast UBL-UBA proteins Rad23 and Dsk2 with the E4 ubiquitin ligase Ufd2 has been described in endoplasmic reticulum-associated degrada- tion among other pathways. The UBL domains of Rad23 and Dsk2 play a prominent role in this process by interacting with Ufd2 and different subunits of the 26 S proteasome. Here, we report crystal structures of Ufd2 in complex with the UBL domains of Rad23 and Dsk2. The N-terminal UBL-interacting region of Ufd2 exhibits a unique sequence pattern, which is dis- tinct from any known ubiquitin- or UBL-binding domain iden- tified so far. Residue-specific differences exist in the interac- tions of these UBL domains with Ufd2, which are coupled to subtle differences in their binding affinities. The molecular details of their differential interactions point to a role for adap- tive evolution in shaping these interfaces. The ubiquitin proteasome system regulates diverse cellular processes including cell cycle progression, immune response, neurodegenerative diseases, and protein quality control (1–4). Ubiquitin-like (UBL)4 domains and ubiquitin- or UBL-binding domains (UBD) (5) are small and highly diversified domains that occur as integral parts of larger proteins (6–9). Integral UBLs display a similar fold as ubiquitin (Ub) and like Ub are described as protein-protein interaction modules without the modifier function of Ub (5, 10). So far more than 20 different classes of UBDs have been reported with a wide range of Ub binding specificities (11, 12). The ubiquitin-associated (UBA) domain was the first identified UBD, which exhibits the highest representation of all UBDs in the eukaryotic genome (13) with diverse Ub and Ub chain binding properties (14, 15). Although the source of this binding diversity in vivo remained elusive so far, remarkable structural studies have recently unraveled the unique poly-Ub binding mode of a few other UBDs (16–20) and contributed further to the understanding of how UBDs might have acquired their respective ligand specificity. UBL-UBA proteins contain both a UBL domain and at least one UBA domain. Via these domains they interact simulta- neously with ubiquitylated substrates and 26 S proteasome, thereby delivering substrates to the proteasome for degradation (21). Interestingly, UBL-UBA proteins are also binding partners of other proteins (22–25). For instance, the budding yeast UBL- UBA proteins Rad23 and Dsk2 can interact with the E4 ligase Ufd2 via their UBL domains (22, 26, 27). A common involve- ment of Ufd2, Rad23, and Dsk2 has been described in the endo- plasmic reticulum-associated degradation, ubiquitin fusion degradation, and OLE-1 gene induction pathway (22, 28–30), where the UBL-Ufd2 interaction is indispensable. The associa- tion of UBL-UBA proteins with Ub ligases, their reported sub- strate specificity (31, 32), and the inhibitory effect of UBL-UBA proteins on Ub chain disassembly (33, 34) support the idea that UBL-UBA proteins might function as important regulatory and specificity factors in Ub-mediated cellular proteolysis (21). Therefore, understanding the binding behavior of the UBL domains of UBL-UBA proteins with their various interacting proteins will shed light on the regulatory role of these proteins. Despite the identification of a large number of UBDs, structural details of integral UBL-binding domains are limited. In some cases, the intra- and intermolecular interactions between these UBLs with known UBDs such as UBA or the ubiquitin-interact- ing motif (UIM) have been demonstrated by solution NMR (35–38). Here, we are reporting crystal structures of budding yeast Ufd2 in complex with the UBL domains of Rad23 and Dsk2 and the molecular details of their interaction interfaces. We identify a novel sequence pattern in the N-terminal UBL-binding region of budding yeast Ufd2, which is conserved in lower eukaryotes and is distinct from any known UBD identified so far. More- over, despite engaging the same binding region, residue-spe- * This work was supported by Deutsche Forschungsgemeinschaft Grant RA1643/2-1 (to S. R.) and Rudolf Virchow Center for Experimental Biomed- icine Grant FZ 82 (to H. S.). The atomic coordinates and structure factors (codes 3M62 and 3M63) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Experimental Procedures, Tables S1 and S2, and Figs. S1–S5. 1 Both authors contributed equally to this work. 2 To whom correspondence may be addressed. E-mail: hermann.schindelin@ virchow.uni-wuerzburg.de. 3 To whom correspondence may be addressed. E-mail: shahri.raasi@ uni-konstanz.de. 4 The abbreviations used are: UBL, ubiquitin-like; UBA, ubiquitin-associated; UIM, ubiquitin-interacting motif; UBD, ubiquitin-binding or ubiquitin-like binding domain; UFD, ubiquitin fusion degradation; Ub, ubiquitin; GST, glutathione S-transferase; ITC, isothermal titration calorimetry; SPR, sur- face plasmon resonance; r.m.s., root mean square; WT, wild type; PDB, Pro- tein Data Bank; h, human; Sc, S. cerevisiae; Sp, S. pombe. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 26, pp. 20390–20398, June 25, 2010 © 2010 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. 20390 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 26•JUNE 25, 2010 cific differences exist in the interactions of the UBL domains of Rad23 and Dsk2 with Ufd2, which are coupled to subtle differ- ences in their overall binding affinities. Mutational analyses of the binding surface of the UBL domains and a closer inspection of the thermodynamic contributions of those residues point to adaptive evolution as a factor shaping these interfaces. EXPERIMENTAL PROCEDURES Cloning, site-directed mutagenesis, protein expression, and purification are described in the supplemental Experi- mental Procedures. Crystallization of Ufd2Rad23-UBL and Ufd2Dsk2-UBL— For crystallization of the Ufd2Rad23-UBL and Ufd2Dsk2-UBL complexes, Ufd2 was incubated with Rad23-UBL or Dsk2-UBL at a molar ratio of 1:1.5 (77 M Ufd2 and 115.5 M UBL) for 1 h at 4 °C in the presence of 2 mM dithiothreitol. Crystals were grown by vapor diffusion in hanging drops containing equal volumes of protein in 50 mM HEPES, pH 7.4, 150 mM NaCl, and 2 mM dithiothreitol and a reservoir solution consisting of 16–18% (w/v) polyethylene glycol 3500 and 200 mM K3-citrate, pH 8.3, equilibrated against the reservoir solution. Crystals were cryo-protected by soaking in mother liquor containing 15–20% (v/v) glycerol. They belong to space group P212121 with approximate cell dimensions of a  65 Å, b  126 Å, and c  181 Å with one complex per asymmetric unit. Data Collection and Structure Determination—Crystals were flash-cooled in liquid nitrogen, and data collection was per- formed at 100 K. Data were collected at beamlines ID14–4 (European Synchrotron Radiation Facility (ESRF), Grenoble, France) and BL 14.1 (Berliner Elektronenspeicherring-Gesell- schaft fu¨r Synchrotronstrahlung (BESSY), Berlin, Germany) and processed using Mosflm and Scala (39, 40). Data collection statistics are summarized in supplemental Table S1. For subse- quent calculations, the CCP4 suite was utilized (41) with excep- tions as indicated. The Ufd2 structure was solved by molecular replacement using Phaser (42) with Protein Data Bank (PDB) entry 2QIZ as search model. Because Phaser could not find a solution for the UBL domain with different search models, this domain was fitted manually into the electron density using human ubiquilin 3 (PDB entry 1YQB) for the Ufd2Rad23-UBL complex and the Dsk2-UBL domain (PDB entry 2BWF) for the Ufd2Dsk2-UBL complex as a model. The structures were refined with Phenix (43) and REFMAC5 incorporating transla- tion, libration, screw-rotation (TLS) refinement in all cycles (44, 45). Solvent molecules were automatically added with Coot (46). The figures were produced with PyMOL (65). In Vitro Binding Assays—For pulldown assays, GST-tagged Ufd2 and variants were immobilized on glutathione (GSH) beads. In all experiments, 20 l of GSH beads were incubated with 0.95 M purified Ufd2 in 400 l of phosphate-buffered saline buffer with 1 mM dithiothreitol and 0.1% (v/v) Triton X-100 at 4 °C for 1 h. WT-Ufd2 and GST alone were included as controls. After centrifugation (1250  g, 30 s), beads were washed five times with 400 l of binding buffer. Purified UBL proteins (0.95 M) in a total volume of 400 l of binding buffer were added to immobilized Ufd2 and treated in the same way as in the first step. Immobilized proteins were analyzed by 17% (v/v) SDS-PAGE or by immunoblotting with an anti-His antibody. Isothermal Titration Calorimetry (ITC)—Proteins were extensively dialyzed against phosphate-buffered saline buffer (pH 7.4, 1 mM -mercaptoethanol) followed by degassing. In all experiments, 75–150 M Rad23- and Dsk2-UBL proteins were titrated as the ligand into the sample cell containing 5–10 M Ufd2. A volume of 10 l of ligand was added at a time with a total number of 30 injections, resulting in a final molar ratio of ligand-to-protein varying between 3:1 and 4:1. All experiments were performed using a VP-ITC instrument (MicroCal, GE Healthcare) at 25 °C. Buffer-to-buffer, buffer-to-Ufd2, as well as Rad23-UBL/Dsk2-UBL-to-buffer titrations were performed as described above. Corrected data were analyzed with a single- site binding model using software supplied by the ITC manu- facturer and non-linear least squares fitting to calculate the dissociation constant (Kd). Surface Plasmon Resonance (SPR) Measurements—SPR binding assays were performed alternatively on BIAcore X or BIAcore T100 instruments (GE Healthcare) at 25 °C in 10 mM HEPES, pH 7.4, 150 mM NaCl, 50 M EDTA, 1 mM -mercap- toethanol, and 0.005% (v/v) Surfactant P20. 100 response units of His-tagged Rad23- or Dsk2-UBL were captured on a nickel- nitrilotriacetic acid (Ni-NTA) sensor chip. GST-tagged Ufd2 for comparative binding assays and untagged Ufd2 for affin- ity analysis were applied to the UBL surfaces in random duplicates at a flow rate of 50 l/min. After each cycle, the surface was regenerated using 350 mM EDTA in running buffer to remove bound Ni2 and captured proteins. The BIAcore T100 evaluation software was used to calculate the steady state affinity constants. Data were plotted using GraphPad Prism. For comparative assays, the relative bind- ing responses of the mutants to WT proteins were deter- mined by obtaining the maximum response for each interac- tion at the end of each injection. RESULTS Ufd2 Binds the UBL Domains of Rad23 and Dsk2 with High Affinity—Although Rad23 and Dsk2 interact with Ufd2 via their UBL domains (22, 26), yeast two hybrid assays could only identify the isolated N-terminal fragment (residues 1–380) of Ufd2 as its UBL-interacting region (26). Additional details regarding the Ufd2-UBL interactions have not been unraveled so far. To further characterize the interactions of Ufd2 with the UBLs of Rad23 and Dsk2, we performed GST pulldown assays with GST-tagged full-length Ufd2 and C-terminally His-tagged UBLs (Fig. 1A). Both UBLs were readily captured using immo- bilized GST-Ufd2. In contrast, the UBL domain of Ddi1, a third UBL-UBA protein, does not interact with Ufd2 (Fig. 1A) (22). The differential binding of the Rad23- and Dsk2-UBLs to the proteasomal subunits Rpn1 and Rpn10 has been described (25, 47, 48). Hence, we used SPR interaction analysis to search for quantitative differences in their interactions. Steady state affin- ity analysis of Ufd2 on both Rad23-UBL (Fig. 1B, left panel, and 1C) and Dsk2-UBL surfaces (Fig. 1B, right panel, and 1C) pro- vided a Kd of 55  3 nM for the interaction of Rad23-UBL and a lower affinity for Dsk2-UBL with a Kd of 418  56 nM. UBL-binding Domain of Ufd2 JUNE 25, 2010•VOLUME 285•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 20391 The binding of the UBLs of Rad23 and Dsk2 to Ufd2 was also analyzed by ITC to allow for a complete thermodynamic char- acterization (Fig. 1D). These studies resulted in a Kd of 70  6 nM for the interaction of Rad23-UBL with Ufd2 and for the binding of Dsk2-UBL to Ufd2 in a 2–3-fold higher Kd of 175  19 nM. Although there is an excellent agreement between SPR and ITC for the Rad23-Ufd2 interaction, the two methods show an 2-fold difference for the Dsk2-Ufd2 interaction. More importantly, the enthalpic and entropic components to the free energy are highly different between the two UBLs. The interac- tion of Rad23-UBL and Ufd2 is more exothermic (H  17.3 kcal/mol) when compared with Dsk2-UBL (H  10.1 kcal/ mol). However, this is offset by a substantial decrease in entropy for Rad23-UBL (TS  7.4 kcal/mol), whereas the entropic contribution is minimal for the Dsk2-UBL interaction (TS  0.8 kcal/mol). Crystal Structures of Ufd2 in Complex with Rad23- and Dsk2-UBL—We solved the structures of Ufd2 in complex with Rad23-UBL carrying either an N-terminal or a C-terminal His tag, which showed no significant structural differences. Due to better data quality, the structure of Ufd2 with a C-terminal His-tagged UBL is presented here. The Ufd2Rad23-UBL com- plex was refined at 2.4 Å resolution to a crystallographic R-fac- tor of 20.3% and a free R-factor of 25.7% (Table 1). As described previously (49), Ufd2 is composed of an N-terminal variable domain, a core domain, and a C-terminal U-box domain with a fold similar to that of RING (really interesting new gene) domains, which are present in certain Ub ligases (Fig. 2A). Despite some conformational variability of the U-box domain, our Ufd2 structure in the complex is quite similar (1.5 Å root mean square (r.m.s.) deviation for 954 C atoms) to the pub- lished Ufd2 structure (49). The N-terminal variable region of Ufd2 that binds to the UBL domain consists of eight -helices. Helices 1 to 4 are arranged in a four-helix bundle, whereas helices 5 and 6 interact with 3 and 4 through hydrophobic contacts that are partly mediated by their connecting loops (Fig. 2B). The struc- ture of Rad23-UBL is comprised of a five-stranded -sheet, one -helix, and one 310-helix (Fig. 2B). It displays a high degree of similarity with Ub (PDB entry 1UBQ, 1.1 Å r.m.s. deviation for 72 C atoms, z-score 14, 25% sequence identity) and the UBL domain of hHR23A (PDB entry 1P98, 1.6 Å r.m.s. deviation for FIGURE 1. Interactions of Ufd2 with the UBL domains of Rad23 and Dsk2. A, GST-Ufd2 immobilized on GSH-beads was tested for binding to C-terminally His-tagged UBLs of Rad23, Dsk2, and Ddi1. Captured UBLs were visualized by immunoblotting (WB) with an anti-His antibody. 2% of the input and GST beads incubatedwithUBLswereloadedascontrols.B,aseriesof2-foldUfd2dilutions(233–3.6nM)wasappliedonaRad23-orDsk2-UBLsurfacefor120s(leftandright panel,respectively).RU,responseunits.C,SPRbindingisothermsofWT-Rad23-andWT-Dsk2-UBLandthequintupleandseptupleDsk2-UBLvariantswithUfd2. conc., concentration. D, ITC analysis of Ufd2Rad23-UBL (closed circles) and Ufd2Dsk2-UBL (open circles) complexes. TABLE 1 Refinement statistics Ufd2Rad23-UBL Ufd2Dsk2-UBL Resolution limit (Å) 45.2-2.4 73.5-2.4 No. of reflections 56,268 55,087 No. of protein/ligand/solvent atoms 8303/17/298 8288/17/182 Rcryst (Rfree)a,b 0.203 (0.257) 0.210 (0.270) r.m.s. deviations in: Bond lengths (Å) 0.016 0.015 Bond angles (°) 1.711 1.610 Estimated coordinate error (Å) 0.25 0.26 Overall average B-factor (Å2) 25.7 42.9 Ramachandran statistics (%)c 93.1/97.9/2.1 93.8/98.4/1.6 aRcryst  hklFo  Fc/hklFo where Fo and Fc are the observed and calculated structure factor amplitudes. bRfree, same as Rcryst for 5% of the data randomly omitted from the refinement. The estimated coordinate error is based on Rfree. c Ramachandran statistics indicate the fraction of residues in the favored (98%), allowed ( 99.8%), and disallowed regions of the Ramachandran diagram, as defined by MolProbity (64). UBL-binding Domain of Ufd2 20392 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 26•JUNE 25, 2010 72 C atoms, z-score 11.3, 26% sequence identity), one of the two human homologs of budding yeast Rad23. Subsequently, we solved the structure of Ufd2 with Dsk2- UBL by molecular replacement. The UBL domain in the Ufd2Dsk2-UBL complex exhibits increased flexibility; in par- ticular, with a C-terminally tagged UBL domain, the first 30 amino acids of this domain were largely disordered (data not shown). With an N-terminally tagged protein, the Ufd2Dsk2- UBL structure was refined at 2.4 Å resolution to a crystallo- graphic R-factor of 21.0% and a free R-factor of 27.0% (Table 1). Both Rad23-UBL and Dsk2-UBL structures can be superim- posed with an r.m.s. deviation of 1.1 Å for 71 aligned residues (z-score 13.6, 30% sequence identity). Analysis of the Ufd2Rad23-UBL Interface—The Ufd2UBL interface in the structure of the complex buries a total molecu- lar surface of about 1260 Å2, which is comprised to 590 Å2 of the molecular surface of Ufd2 (1.3% of the total surface area) and 670 Å2 from UBL (14.6% of the total surface area). This interface is composed of almost equal parts of non-polar resi- dues (38%), polar residues (33%), and charged residues (29%); however, there are only one salt bridge (UBL-Lys-10 N–Ufd2- Glu-49 O1 with a distance of 2.6 Å) and two direct hydrogen bonds (UBL-Ser-47 O–Ufd2-Arg-92 N2, UBL-Gln-52 N2– Ufd2-Glu-141 O at distances of 2.3 and 3.2 Å, respectively) present (Fig. 3A). Three UBL segments are contacting Ufd2 (Fig. 3A). Segment I is located in the loop connecting -strands one and two, seg- ment II involves -strands three and four, and segment III is located in -strand five. Ufd2 residues from helix 2 and 4 as well as the loop connecting 4 with 5 contribute to the Ufd2UBL interface. These residues contact the hydrophobic surface of the UBL -sheet in the region of -strands 3, 4, and 5. Participating residues from Ufd2 include Leu-44, Tyr-97, Val- 100, and Trp-107, which are located in the hydrophobic UBL pocket formed by residues Phe-9, Ile-45, Val-50, Val-69, and Met-71 of Rad23 (Fig. 3, A and B). For comparison, the principal recognition determinants in Ub are: 1) the hydrophobic pocket formed by the side chains of Leu-8 (Phe-9 in Rad23), Ile-44 (Ile-45 in Rad23), His-68 (Val-69 in Rad23), and Val-70 (Met-71 in Rad23) and 2) the main chain amide group of Gly-47 (Gly-48 in Rad23), which is involved in hydrogen bonding (50). Although the hydrophobic patch of Rad23-UBL is also crucial for its interaction with Ufd2, the main chain of Gly-48 does not form a hydrogen bond. Instead, the -turn (Ser47–Gly48) connecting -strands 3 and 4 is stabi- lized by the aforementioned strong hydrogen bond between Ufd2-Arg-92 and UBL-Ser-47, whereas Ufd2-Gly-96 and Ufd2- Tyr-97 contact UBL-Gly-48 (Fig. 3A). The aromatic ring of Ufd2-Tyr-97 is involved in a stacking interaction with the pep- tide bond between UBL residues 47 and 48 in this -turn. Probing the Ufd2Rad23-UBL Interface—The importance of interface residues was analyzed by mutagenesis experiments. Eleven residues from Ufd2 and nine from Rad23-UBL were each replaced with Ala. With the exception of the Rad23-UBL- G48A variant that showed a reduced expression, all Ufd2 and Rad23-UBL variants behaved like the WT protein during and after purification, indicating that they were correctly folded (data not shown). Initially, the contribution of these residues was studied by GST pulldown and comparative SPR binding assays (Table 2, supplemental Figs. S1 and S2A). In SPR studies, the relative binding responses of mutants to WT proteins were determined and compared. The majority of Rad23-UBL single mutants revealed reduced binding to Ufd2 with Rad23-UBL- I45A displaying the most prominent binding defect. The con- tribution of the remaining residues to the interaction is aug- mented in double mutants (supplemental Fig. S1C). Analysis of the Ufd2 variants by SPR showed a largely reduced binding of the residues located in the hydrophobic region of the UBL- binding pocket (Leu-44, Tyr-97, Val-100, and Phe-107) and Asp-40 (Table 2 and supplemental Fig. S2A). ITC studies confirmed these results and allowed for a quantitative analysis (Table 2, supplemental Fig. S3 and supplemental Table S2). The most significant effect for Ufd2 was observed for all residues located in the hydrophobic UBL pocket. Mutation of Val-100 and Phe-107 to Ala completely abolished binding, the Y97A variant strongly reduced binding (1900-fold), and the I104A and L44A variants showed signifi- cantly decreased affinities (20- and 120-fold, respectively). Although not directly involved in complex formation (Fig. 3A), the Ufd2-D40A variant showed a 110-fold reduced affinity (Table 2), which probably is the result of the missing intramo- lecular hydrogen bond between Ufd2-Asp-40 and Ufd2-Tyr-97 (O2–OH 2.5 Å). This hydrogen bond seems to be crucial for proper positioning of the aromatic side chain of Tyr-97 in the interface region and might be important to align helices 2 and 4 for interaction with the Rad23-UBL. FIGURE 2. Structure of Ufd2 in complex with the UBL domain of Rad23. A, ribbon representation of the overall structure of the Ufd2Rad23-UBL com- plex. The Rad23-UBL domain is shown in green, the N-terminal Ufd2 region is in orange, the Ufd2 core domain is in gray, and the Ufd2 U-box domain is in red. B, close-up view of the N-terminal Ufd2 domain in complex with Rad23- UBL with secondary structural elements labeled and color-coded as in A. UBL-binding Domain of Ufd2 JUNE 25, 2010•VOLUME 285•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 20393 In Rad23-UBL, Ile-45 was shown to be integral for binding to Ufd2 by the detrimental effect (130-fold decrease) after exchange to Ala (Table 2). Mutation of Phe-9, Val-50, and Val- 69, residues adjacent to Ile-45 in the hydrophobic patch, also decreased the affinity of Rad23-UBL 5–7-fold. Ser-47, which is in hydrogen-bonding distance to Ufd2-Arg-92 and next to UBL-Gly-48, showed a 9-fold reduced affinity. In Ub and in the human Rad23 homolog hHR23A, Ser-47 is replaced by Ala. Charged residues found in the interface (Ufd2, Glu-26, Glu-49, and Arg-92; UBL, Lys-10) do not contribute significantly to the interaction. In summary, our data indicate that the most prom- inent contact between Ufd2 and Rad23-UBL is the strong hydrophobic interaction between UBL-Ile-45 and Ufd2-Val- 100 as well as Ufd2-Phe-107, which defines the core of the UBL- interacting region of Ufd2. Molecular Discrimination between Rad23 and Dsk2—De- spite a similar fold, the UBL domains of Rad23 and Dsk2 display only 30% sequence identity, which could give rise to differences in their interactions. A superposition of the bound Rad23-UBL and Dsk2-UBL in the two complex structures showed signifi- cant changes (Fig. 3C). Of the three UBL segments involved in the Ufd2 interaction (Fig. 3A), segment II including Ile-45 (Ile-44 in Ub) is highly conserved, and there are no conforma- tional changes in both UBL structures, whereas segments I and III are not conserved and display structural changes (Fig. 3C). The loop, connecting -strands one and two, adopts different conformations, and -strand five shows a displacement that might affect binding (Fig. 3C). Segment I includes Phe-9 in Rad23-UBL, corresponding to Leu-8 in Ub, where this residue is also involved in Ub recogni- tion by UBDs (50, 51). Phe-9 is replaced by Gly-10 in Dsk2-UBL, and there is no corresponding hydrophobic interacting residue (supplemental Fig. S4A). Dsk2 residues Gly-10 and Gln-11 adopt different conformations when compared with Leu-8/ Thr-9 of Ub and Phe-9/Lys-10 of Rad23-UBL. In the Ufd2Dsk2-UBL structure, the Ufd2Rad23-UBL salt bridge (Lys-10/Glu-49) is missing due to the Lys-10 to Gln-11 exchange, with the latter side chain no longer being located in the protein interface (supplemental Fig. S4A). The missing interaction from segment I in Dsk2 might be compensated by the displacement of -strand five toward Ufd2 and a replace- ment of Val-69 to His-69 found in segment III resulting in a more pronounced interaction in this region when compared with Rad23-UBL (supplemental Fig. S4A). The presence of the salt bridge seems to be the reason for the more exothermic character of the Ufd2Rad23-UBL interaction, a view that is also supported by the corresponding Ufd2-E49A and Rad23-K10A variants, which both display binding enthalpies similar to the FIGURE 3. The Ufd2Rad23-UBL interface. A, residues involved in binding are shown in stick representation. Carbon atoms of Ufd2 residues are colored in orange and in green for Rad23-UBL. Dashed lines indicate H-bonds. B, structure-based sequence alignment of Rad23-UBL, Dsk2-UBL, hHR23A-UBL, and Ub. Secondary structure elements of Rad23-UBL were assigned using DSSP (61) and are labeled above the sequences. The alignment was performed using DaliLite (62), and the figure was prepared with ESPript (63). Strictly conserved amino acids are highlighted with a red background, and similar amino acids are shown as redletters.ThethreeUfd2-bindingsegmentsareindicated.ResiduesinvolvedinUfd2Rad23-UBLinteractionarelabeledwithgreenstars.C,superpositionofthe Ufd2Rad23-UBL/Dsk2-UBL complex structures with the N-terminal binding domain of Ufd2 in orange (Rad23 complex) and gray (Dsk2 complex), with Rad23- UBL in green and Dsk2-UbL in yellow. UBL-binding Domain of Ufd2 20394 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 26•JUNE 25, 2010 Ufd2Dsk2-UBL complex (supplemental Table S2). At the same time, the absence of the salt bridge in both mutants is accom- panied by a more favorable entropic contribution, which is on a level similar to the Ufd2Dsk2-UBL complex. To identify residues important for the subtle molecular dis- crimination between the UBL domains of Rad23 and Dsk2, the interaction of Dsk2-UBL with Ufd2 mutants was analyzed by GST pulldown assays (data not shown), SPR, and ITC (Table 2, supplemental Figs. S2B and S3C). Quantitative ITC analysis showed reduced binding of Dsk2 to Ufd2 mutants Y97A (470- fold), V100A (22-fold), I104A (6-fold), and F107A (20-fold) (Table 2). However, binding of the V100A and F107A variants is not completely abolished, and when compared with Rad23- UBL, the binding affinities are less affected by a factor of about 3–7 in most of the mutants analyzed. In addition, the L44A mutant, which has a 120-fold reduced affinity with Rad23-UBL, is only three times reduced in the case of Dsk2-UBL. In general agreement with the ITC affinity data, the compar- ative SPR binding assay revealed significant differences in the association of Ufd2 variants Y97A, V100A, I104A, and F107A with Rad23- and Dsk2-UBL surfaces (supplemental Fig. S4B). The observed SPR decrease for the binding of the T48A and E49A variants of Ufd2 to Dsk2-UBL seems to be compensated by slower dissociations, thus explaining why these mutants show no significant defect in the ITC analysis. To further analyze the contribution of segments I and III to complex formation, a G10F/Q11K/S67Q/H69V/V71M quintu- ple Dsk2-UBL mutant was generated, where key residues in binding segments I and III were replaced with the correspond- ing residues from Rad23-UBL. Comparative binding as well as steady state affinity analysis by SPR revealed only a small increase (Kd  348 nM) in binding affinity for Ufd2 when com- pared with WT-Dsk2-UBL (Kd  418 nM) (data not shown and Fig. 1C). In addition, neither a crystal structure of the quintuple Ufd2Dsk2-UBL complex (data not shown) nor the KD of 240 nM deduced by ITC revealed significant differences from WT-Dsk2-UBL (Kd  175 nM). The ITC analysis did, however, reveal that the binding is now driven by an increase in entropy (TS  6.5 kcal/mol versus 0.8 and 7.4 kcal/mol for WT-Dsk2-UBL and -Rad23-UBL, respectively), whereas the binding enthalpy is reduced to only 2.5 kcal/mol when com- pared with 10.1 and 17.3 kcal/mol (supplemental Table S2). Interestingly, SPR and ITC analysis of a G10F/Q11K/I50V/ K52Q/S67Q/H69V/V71M septuple Dsk2-UBL mutant, which has the additional I50V and K52Q substitutions in segment II, showed an even lower affinity (SPR, Kd  648 nM; ITC, Kd  875 nM) to Ufd2 when compared with WT-Dsk2-UBL (Fig. 1C). The N Terminus of Ufd2 Represents a Unique and Conserved UBL-binding Domain—A multiple sequence alignment of Ufd2 from different yeast species displays a distinct pattern of con- served residues involved in UBL binding (Fig. 4A). Among the available yeast genomes, the Schizosaccharomyces pombe sequence is most similar to those from higher eukaryotes; thus we isolated cDNA fragments for the coding region of the UBL domains of Rad23 and Dsk2 and full-length Ufd2 from this organism and examined their interactions by GST pulldown assays (Fig. 4B) as well as SPR (data not shown). We could show that SpUfd2 interacts strongly with the UBL domains of SpRad23 and SpDsk2 as well as with the UBL domains of ScRad23 and ScDsk2 and vice versa. This cross species interac- tion, despite the diversified UBL and Ufd2 amino acid sequences, indicates that the identified sequence pattern repre- sents a real UBL-interacting domain. A surface representation of this motif is shown in Fig. 4C. The N terminus of budding yeast Ufd2 displays only limited sequence homology with the human Ufd2s, E4A and E4B (supplemental Fig. S5) and other Ufd2s from higher eukaryotes. In agreement with this finding, there are no reports that hHR23A/B interacts with either of the human homologs of Ufd2. Interestingly, our SPR studies showed that the UBL domain of hHR23A interacts with ScUfd2, albeit with lower affinity (data not shown). Apparently, the high affinity interac- tion of the UBL domains of Rad23 and Dsk2 has been lost dur- ing the evolution of this domain. The absence of conservation of the Ufd2-UBL interface could potentially be used for thera- peutic interventions against pathogenic yeasts such as Candida albicans by designing low molecular weight compounds that disrupt this interface. However, further functional studies in pathogenic yeasts are required to examine the suitability of this surface as a drug target. TABLE 2 ITC and SPR parameters of Ufd2, Rad23-UBL, Dsk2-UBL, and variants  indicates no change; ND indicates not detected (corresponding to at least a 104-fold decrease in binding affinity). Ufd2 WT-UBL ITC SPRa (% of relative response) Kd Fold decrease nM WT Rad23 70 100 Dsk2 175 100 E26A Rad23 284 4 91 Dsk2 521 3 83 D40A Rad23 7900 110 20 Dsk2 7600 40 0 L44A Rad23 8300 120 31 Dsk2 463 3 52 T48A Rad23 72  70 Dsk2 296 2 29 E49A Rad23 413 6 69 Dsk2 314 2 44 R92A Rad23 265 4 76 Dsk2 128  59 G96A Rad23 592 8 51 Dsk2 216  60 Y97A Rad23 134,000 1900 3 Dsk2 83,000 470 0 V100A Rad23 ND 10,000 9 Dsk2 3900 22 1 I104A Rad23 1600 20 43 Dsk2 1100 6 12 F107A Rad23 ND 10,000 11 Dsk2 3600 20 0 Ufd2 Rad23-UBL ITC SPRa (% of relative response) Kd Fold decrease nM WT F9A 376 5 80 WT K10A 162 2 96 WT I45A 9100 130 17 WT S47A 606 9 62 WT V50A 441 6 88 WT Q52A 415 6 79 WT Q67A 113 2 92 WT V69A 478 7 70 WT M71A 221 3 88 a For comparative SPR assays, the relative binding responses of the mutants to wt proteins were determined by obtaining the maximum response for each interac- tion at the end of injection. UBL-binding Domain of Ufd2 JUNE 25, 2010•VOLUME 285•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 20395 DISCUSSION Families and superfamilies of protein domains or folds have evolved through a process of homologous recombination and gene duplication (52) followed by sequence divergence. Mem- bers of different classes of UBDs such as UBA or UBL domains result from these processes. For instance, the UBL domains of Rad23 and Dsk2 display only 30% sequence identity but adopt the same fold and utilize the same binding surface to recognize a common UBL-binding domain of Ufd2 to form complexes that display similarly high affinity. Nevertheless, not all inter- acting residues are conserved; in particular, there is sequence diversity in binding segments I and III of UBLs. Our attempts to interconvert the UBL domains by altering non-conserved inter- facial residues were not successful, thus suggesting that addi- tional elements exist and play a role in the respective Ufd2-UBL interaction. Interestingly, these results resemble earlier studies on WW domains (53, 54), where a statistical analysis of multiple sequence alignments was utilized to identify co-evolving resi- dues. The authors demonstrated that not only interfacial resi- dues but also buried residues distal to the interface co-evolved with interfacial residues and contribute significantly to the interactions. They concluded that certain sequence patterns in interacting domains are due to adaptive evolution. In agree- ment with these findings, our data prove that substitution of key interfacial residues of Dsk2-UBL has no significant effect on its overall binding affinity to Ufd2. In case of the septuple mutant, we even observed a decrease in binding affinity, which could be due to the imposed disorder into the evolutionary inter-residue relations within the UBL fold. This is supported by the fact that when compared with Dsk2-UBL and in partic- ular Rad23-UBL, the binding of the quintuple Dsk2-UBL mutant is driven strongly by entropy. These findings indicate that binding interfaces can be modulated by changes in residues that affect either the binding enthalpy or the entropy, thus pro- viding additional freedom to maintain an interaction during the course of evolution, an effect that has been described previously as entropy/enthalpy compensation (55, 56). Our studies suggest that UBL domains have co-evolved with Ufd2 to reach optimal binding affinities by altering specific res- idue-to-residue interactions (co-evolution at the residue level) (57), while at the same time, all functional aspects of Rad23 or Dsk2 are preserved. Therefore, the primary sequence degener- FIGURE 4. The N terminus of Ufd2 represents a conserved UBL-interacting domain in lower eukaryotes. A, alignment of the N-terminal sequences of fungal Ufd2s. Invariant or conserved residues with surface access are colored in dark blue, buried ones are in light blue. Residues labeled with red stars represent the core region of the binding domain, which is essential for UBL interaction, whereas residues labeled with yellow stars contribute moderately to the interaction. K. lactis, Kluyveromyces lactis; C. glabrata, Candida glabrata; Z. rouxii, Zygosaccharomyces rouxii; L. thermotolerans, Lachancea thermotolerans; C. tropicalis, Candida tropicalis; C. dubliniensis, Candida dubliniensis; P. guilliermondii, Pichia guilliermondii; D. hansenii, Debaryomyces hansenii. B, GST pulldown assay demonstrates the cross interactions of S. pombe and S. cerevisiae proteins. 5% of inputs and GST beads incubated with UBLs were loaded as controls. WB, Western blot. C, surface representation of the N-terminal UBL-binding domain of Ufd2, color-coded as in A. UBL-binding Domain of Ufd2 20396 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 26•JUNE 25, 2010 acy of protein domains such as UBAs or UBLs has been toler- ated and established in favor of the cooperative nature of the interactions and their functionality within their respective pro- tein complexes. This further suggests that differential binding properties observed for the interactions of Ufd2 with UBLs (this study) or for the interactions of UBAs with Ub and Ub chains (14) can arise not necessarily due to their interaction with dif- ferent ligands but can also result from the adaptive co-evolution of these domains with the same interacting partners. Seem- ingly, these interfacial domains have evolved to hold protein- protein interactions in a suitable form within multicomponent complexes until they are challenged by downstream events. Numerous structures of Ub receptors in complex with their respective Ub/UBL-binding domains have been reported. The so far characterized Ub receptors of the 26 S proteasome in budding yeast encompass the two proteasomal subunits Rpn10 (S5a in humans) and Rpn13 and the three UBL-UBA proteins Rad23, Dsk2, and Ddi1, which associate with the proteasome and function as shuttle factors (21). Experimental evidence for the existence of additional candidates exist (21, 58). Rad23 and Dsk2 interact with the proteasomal subunit Rpn1 via their UBL domains (21, 47). Aside from their known interactions with Ub, Rpn13 and Rpn10/S5a alternatively interact with UBL-UBA proteins (21, 35, 37, 38, 48, 51, 59). For instance, the preferential association of Rpn1 with Rad23 and Rpn10 with Dsk2 has been reported (25, 38, 47, 48). Based on the binding of hpLIC2 (Dsk2 homolog) with Rpn13, an interac- tion of Dsk2 with Rpn13 has been proposed (51, 59). Although the aforementioned examples engage essentially the same surface of Ub/UBL, they diverge in both structure and pat- terns of Ub/UBL recognition (Fig. 5). For instance, hRpn10/S5a recog- nizes the UBL domain of hHR23A, one of the two human homologs of Rad23, via a Ub-interacting motif, which consists of a single -helix (35, 37). Rpn13 binds Ub via a pleck- strin homology domain, which is a seven-stranded -sandwich capped by an -helix (51). The Ub-binding surface of Rpn13 is formed by three loops that bridge -strands. Another Ub-binding element is the UBA domain found for example in Dsk2 (60). The UBA domain is com- posed of a three-helix bundle. With the exception of Rpn13, which exclusively binds via loops, it seems that the majority of Ub/UBL-bind- ing domains fold into -helical structures including the known UBDs, UIM, and UBA, and the UBL-binding domain of Ufd2 iden- tified in this study. Despite the pre- dominant interaction involving -helices as Ub/UBL-binding elements, the three-dimensional structure of the UBL-binding domain of Ufd2 differs from other known examples, hence providing the first structural descrip- tion for how Ufd2 acts as a UBL receptor while at the same time further enhancing the diversity of UBDs in general. Acknowledgments—We thank Martin Scheffner and Keith Wilkinson for critical reading of the manuscript. We thank Stefan Jentsch for providing the original plasmids for the expression of Rad23, Dsk2, and Ufd2 and for Ufd2-specific antibodies used in the initial phase of this study. We also thank David Fischer and Rodrigo Villasen˜or for the contribution to this study and Sven Eiselein for providing us with C-terminal GST-tagging plasmid. REFERENCES 1. Hershko, A., and Ciechanover, A. (1998) Annu. Rev. Biochem. 67, 425–479 2. Ross, C. A., and Pickart, C. M. (2004) Trends Cell Biol. 14, 703–711 3. Haglund, K., and Dikic, I. (2005) EMBO J. 24, 3353–3359 4. Mukhopadhyay, D., and Riezman, H. (2007) Science 315, 201–205 5. Grabbe, C., and Dikic, I. (2009) Chem. Rev. 109, 1481–1494 6. Buchberger, A. (2002) Trends Cell Biol. 12, 216–221 7. Hicke, L., Schubert, H. L., and Hill, C. P. (2005) Nat. Rev. Mol. Cell Biol. 6, 610–621 8. Hurley, J. H., Lee, S., and Prag, G. (2006) Biochem. J. 399, 361–372 9. Harper, J. W., and Schulman, B. A. (2006) Cell 124, 1133–1136 10. Hochstrasser, M. (2009) Nature 458, 422–429 FIGURE 5. Mode of Ub/UBL recognition by different Ub/UBL-binding domains. In each panel, ribbon rep- resentationstogetherwiththemolecularsurfacesofbothbindingpartnersareshownwithUb/UBLinthesame orientation. A, Ufd2Rad23-UBL. B, Dsk2-UBADsk2-UBL (PDB entry 2BWE). C, Rpn13Ub (PDB entry 2Z59). D, S5a/Rpn10hHR23A-UBL (PDB entry 1P9D). UBL-binding Domain of Ufd2 JUNE 25, 2010•VOLUME 285•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 20397 11. Dikic, I., Wakatsuki, S., and Walters, K. J. (2009) Nat. Rev. Mol. Cell Biol. 10, 659–671 12. Hofmann, K. (2009) DNA Repair 8, 544–556 13. Hofmann, K., and Bucher, P. (1996) Trends Biochem. Sci. 21, 172–173 14. Raasi, S., Varadan, R., Fushman, D., and Pickart, C. M. (2005) Nat. Struct. Mol. Biol. 12, 708–714 15. Varadan, R., Assfalg, M., Raasi, S., Pickart, C., and Fushman, D. (2005) Mol. Cell 18, 687–698 16. Rahighi, S., Ikeda, F., Kawasaki, M., Akutsu, M., Suzuki, N., Kato, R., Ken- sche, T., Uejima, T., Bloor, S., Komander, D., Randow, F., Wakatsuki, S., and Dikic, I. (2009) Cell 136, 1098–1109 17. Lo, Y. C., Lin, S. C., Rospigliosi, C. C., Conze, D. B., Wu, C. J., Ashwell, J. D., Eliezer, D., and Wu, H. (2009) Mol. Cell 33, 602–615 18. Sato, Y., Yoshikawa, A., Mimura, H., Yamashita, M., Yamagata, A., and Fukai, S. (2009) EMBO J. 28, 2461–2468 19. Sato, Y., Yoshikawa, A., Yamashita, M., Yamagata, A., and Fukai, S. (2009) EMBO J. 28, 3903–3909 20. Kulathu, Y., Akutsu, M., Bremm, A., Hofmann, K., and Komander, D. (2009) Nat. Struct. Mol. Biol. 16, 1328–1330 21. Finley, D. (2009) Annu. Rev. Biochem. 78, 477–513 22. Kim, I., Mi, K., and Rao, H. (2004) Mol. Biol. Cell 15, 3357–3365 23. Hara, T., Kamura, T., Kotoshiba, S., Takahashi, H., Fujiwara, K., Onoyama, I., Shirakawa, M., Mizushima, N., and Nakayama, K. I. (2005) Mol. Cell. Biol. 25, 9292–9303 24. Ivantsiv, Y., Kaplun, L., Tzirkin-Goldin, R., Shabek, N., and Raveh, D. (2006) Mol. Cell. Biol. 26, 1579–1588 25. Ishii, T., Funakoshi, M., and Kobayashi, H. (2006) EMBO J. 25, 5492–5503 26. Richly, H., Rape, M., Braun, S., Rumpf, S., Hoege, C., and Jentsch, S. (2005) Cell 120, 73–84 27. Hoppe, T. (2005) Trends Biochem. Sci. 30, 183–187 28. Rape, M., and Jentsch, S. (2004) Biochim. Biophys. Acta 1695, 209–213 29. Medicherla, B., Kostova, Z., Schaefer, A., and Wolf, D. H. (2004) EMBO Rep. 5, 692–697 30. Raasi, S., and Wolf, D. H. (2007) Semin. Cell Dev. Biol. 18, 780–791 31. Verma, R., Oania, R., Graumann, J., and Deshaies, R. J. (2004) Cell 118, 99–110 32. Liu, C., van Dyk, D., Li, Y., Andrews, B., and Rao, H. (2009) BMC Biol. 7, 75 33. Raasi, S., and Pickart, C. M. (2003) J. Biol. Chem. 278, 8951–8959 34. Hartmann-Petersen, R., Hendil, K. B., and Gordon, C. (2003) FEBS Lett. 535, 77–81 35. Walters, K. J., Kleijnen, M. F., Goh, A. M., Wagner, G., and Howley, P. M. (2002) Biochemistry 41, 1767–1777 36. Walters, K. J., Lech, P. J., Goh, A. M., Wang, Q., and Howley, P. M. (2003) Proc. Natl. Acad. Sci. U.S.A. 100, 12694–12699 37. Mueller, T. D., and Feigon, J. (2003) EMBO J. 22, 4634–4645 38. Zhang, D., Chen, T., Ziv, I., Rosenzweig, R., Matiuhin, Y., Bronner, V., Glickman, M. H., and Fushman, D. (2009) Mol Cell. 36, 1018–1033 39. Leslie, A. G. W. (1992) Joint CCP4ESF-EAMCB Newsletter on Protein Crystallography, Vol. 26, Daresbury Laboratory, Warrington, UK 40. Evans, P. (2006) Acta Crystallogr. D Biol. Crystallogr. 62, 72–82 41. Bailey, S. (1994) Acta Crystallogr. D Biol. Crystallogr. 50, 760–763 42. McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Sto- roni, L. C., and Read, R. J. (2007) J. Appl. Crystallogr. 40, 658–674 43. Adams, P. D., Grosse-Kunstleve, R. W., Hung, L. W., Ioerger, T. R., Mc- Coy, A. J., Moriarty, N. W., Read, R. J., Sacchettini, J. C., Sauter, N. K., and Terwilliger, T. C. (2002) Acta Crystallogr. D Biol. Crystallogr. 58, 1948–1954 44. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Acta Crystallogr. D Biol. Crystallogr. 53, 240–255 45. Winn, M. D., Isupov, M. N., and Murshudov, G. N. (2001) Acta Crystal- logr. D Biol. Crystallogr. 57, 122–133 46. Emsley, P., and Cowtan, K. (2004) Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 47. Elsasser, S., Gali, R. R., Schwickart, M., Larsen, C. N., Leggett, D. S., Mu¨ller, B., Feng, M. T., Tu¨bing, F., Dittmar, G. A., and Finley, D. (2002) Nat. Cell Biol. 4, 725–730 48. Matiuhin, Y., Kirkpatrick, D. S., Ziv, I., Kim, W., Dakshinamurthy, A., Kleifeld, O., Gygi, S. P., Reis, N., and Glickman, M. H. (2008) Mol. Cell. 32, 415–425 49. Tu, D., Li, W., Ye, Y., and Brunger, A. T. (2007) Proc. Natl. Acad. Sci. U.S.A. 104, 15599–15606 50. Ohno, A., Jee, J., Fujiwara, K., Tenno, T., Goda, N., Tochio, H., Kobayashi, H., Hiroaki, H., and Shirakawa, M. (2005) Structure 13, 521–532 51. Schreiner, P., Chen, X., Husnjak, K., Randles, L., Zhang, N., Elsasser, S., Finley, D., Dikic, I., Walters, K. J., and Groll, M. (2008) Nature 453, 548–552 52. Te Velthuis, A. J., and Bagowski, C. P. (2008) Curr. Genomics. 9, 88–96 53. Russ, W. P., Lowery, D. M., Mishra, P., Yaffe, M. B., and Ranganathan, R. (2005) Nature 437, 579–583 54. Socolich, M., Lockless, S. W., Russ, W. P., Lee, H., Gardner, K. H., and Ranganathan, R. (2005) Nature 437, 512–518 55. Reyes-Turcu, F. E., Shanks, J. R., Komander, D., and Wilkinson, K. D. (2008) J. Biol. Chem. 283, 19581–19592 56. Hunter, C. A., and Tomas, S. (2003) Chem. Biol. 10, 1023–1032 57. Pazos, F., and Valencia, A. (2008) EMBO J. 27, 2648–2655 58. Lam, Y. A., Lawson, T. G., Velayutham, M., Zweier, J. L., and Pickart, C. M. (2002) Nature 416, 763–767 59. Husnjak, K., Elsasser, S., Zhang, N., Chen, X., Randles, L., Shi, Y., Hof- mann, K., Walters, K. J., Finley, D., and Dikic, I. (2008) Nature 453, 481–488 60. Lowe, E. D., Hasan, N., Trempe, J. F., Fonso, L., Noble, M. E., Endicott, J. A., Johnson, L. N., and Brown, N. R. (2006) Acta Crystallogr. D Biol. Crystallogr. 62, 177–188 61. Kabsch, W., and Sander, C. (1983) Biopolymers 22, 2577–2637 62. Holm, L., Ka¨a¨ria¨inen, S., Rosenstro¨m, P., and Schenkel, A. (2008) Bioin- formatics 24, 2780–2781 63. Gouet, P., Courcelle, E., Stuart, D. I., and Me´toz, F. (1999) Bioinformatics 15, 305–308 64. Davis, I. W., Leaver-Fay, A., Chen, V. B., Block, J. N., Kapral, G. J., Wang, X., Murray, L. W., Arendall, W. B., 3rd, Snoeyink, J., Richardson, J. S., and Richardson, D. C. (2007) Nucleic Acids Res. 35, W375–W383 65. DeLano, W. L. (2002) The PyMOL Molecular Graphics System, DeLano Scientific LLC, San Carlos, CA UBL-binding Domain of Ufd2 20398 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 26•JUNE 25, 2010
3M63
Crystal structure of Ufd2 in complex with the ubiquitin-like (UBL) domain of Dsk2
The Yeast E4 Ubiquitin Ligase Ufd2 Interacts with the Ubiquitin-like Domains of Rad23 and Dsk2 via a Novel and Distinct Ubiquitin-like Binding Domain*□ S Received for publication,February 12, 2010, and in revised form, March 22, 2010 Published, JBC Papers in Press,April 28, 2010, DOI 10.1074/jbc.M110.112532 Petra Ha¨nzelmann‡1, Julian Stingele§1, Kay Hofmann¶, Hermann Schindelin‡2, and Shahri Raasi§3 From the ‡Rudolf Virchow Center for Experimental Biomedicine, University of Wu¨rzburg, Josef-Schneider-Strasse 2, 97080 Wu¨rzburg, the §Laboratory of Cellular Biochemistry, Department of Biology, University of Konstanz, 78457 Konstanz, and the ¶Bioinformatics Group, Miltenyi Biotec GmbH, Friedrich-Ebert-Strasse 68, 51429 Bergisch-Gladbach, Germany Proteins containing ubiquitin-like (UBL) and ubiquitin-asso- ciated (UBA) domains interact with various binding partners and function as hubs during ubiquitin-mediated protein degra- dation. A common interaction of the budding yeast UBL-UBA proteins Rad23 and Dsk2 with the E4 ubiquitin ligase Ufd2 has been described in endoplasmic reticulum-associated degrada- tion among other pathways. The UBL domains of Rad23 and Dsk2 play a prominent role in this process by interacting with Ufd2 and different subunits of the 26 S proteasome. Here, we report crystal structures of Ufd2 in complex with the UBL domains of Rad23 and Dsk2. The N-terminal UBL-interacting region of Ufd2 exhibits a unique sequence pattern, which is dis- tinct from any known ubiquitin- or UBL-binding domain iden- tified so far. Residue-specific differences exist in the interac- tions of these UBL domains with Ufd2, which are coupled to subtle differences in their binding affinities. The molecular details of their differential interactions point to a role for adap- tive evolution in shaping these interfaces. The ubiquitin proteasome system regulates diverse cellular processes including cell cycle progression, immune response, neurodegenerative diseases, and protein quality control (1–4). Ubiquitin-like (UBL)4 domains and ubiquitin- or UBL-binding domains (UBD) (5) are small and highly diversified domains that occur as integral parts of larger proteins (6–9). Integral UBLs display a similar fold as ubiquitin (Ub) and like Ub are described as protein-protein interaction modules without the modifier function of Ub (5, 10). So far more than 20 different classes of UBDs have been reported with a wide range of Ub binding specificities (11, 12). The ubiquitin-associated (UBA) domain was the first identified UBD, which exhibits the highest representation of all UBDs in the eukaryotic genome (13) with diverse Ub and Ub chain binding properties (14, 15). Although the source of this binding diversity in vivo remained elusive so far, remarkable structural studies have recently unraveled the unique poly-Ub binding mode of a few other UBDs (16–20) and contributed further to the understanding of how UBDs might have acquired their respective ligand specificity. UBL-UBA proteins contain both a UBL domain and at least one UBA domain. Via these domains they interact simulta- neously with ubiquitylated substrates and 26 S proteasome, thereby delivering substrates to the proteasome for degradation (21). Interestingly, UBL-UBA proteins are also binding partners of other proteins (22–25). For instance, the budding yeast UBL- UBA proteins Rad23 and Dsk2 can interact with the E4 ligase Ufd2 via their UBL domains (22, 26, 27). A common involve- ment of Ufd2, Rad23, and Dsk2 has been described in the endo- plasmic reticulum-associated degradation, ubiquitin fusion degradation, and OLE-1 gene induction pathway (22, 28–30), where the UBL-Ufd2 interaction is indispensable. The associa- tion of UBL-UBA proteins with Ub ligases, their reported sub- strate specificity (31, 32), and the inhibitory effect of UBL-UBA proteins on Ub chain disassembly (33, 34) support the idea that UBL-UBA proteins might function as important regulatory and specificity factors in Ub-mediated cellular proteolysis (21). Therefore, understanding the binding behavior of the UBL domains of UBL-UBA proteins with their various interacting proteins will shed light on the regulatory role of these proteins. Despite the identification of a large number of UBDs, structural details of integral UBL-binding domains are limited. In some cases, the intra- and intermolecular interactions between these UBLs with known UBDs such as UBA or the ubiquitin-interact- ing motif (UIM) have been demonstrated by solution NMR (35–38). Here, we are reporting crystal structures of budding yeast Ufd2 in complex with the UBL domains of Rad23 and Dsk2 and the molecular details of their interaction interfaces. We identify a novel sequence pattern in the N-terminal UBL-binding region of budding yeast Ufd2, which is conserved in lower eukaryotes and is distinct from any known UBD identified so far. More- over, despite engaging the same binding region, residue-spe- * This work was supported by Deutsche Forschungsgemeinschaft Grant RA1643/2-1 (to S. R.) and Rudolf Virchow Center for Experimental Biomed- icine Grant FZ 82 (to H. S.). The atomic coordinates and structure factors (codes 3M62 and 3M63) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Experimental Procedures, Tables S1 and S2, and Figs. S1–S5. 1 Both authors contributed equally to this work. 2 To whom correspondence may be addressed. E-mail: hermann.schindelin@ virchow.uni-wuerzburg.de. 3 To whom correspondence may be addressed. E-mail: shahri.raasi@ uni-konstanz.de. 4 The abbreviations used are: UBL, ubiquitin-like; UBA, ubiquitin-associated; UIM, ubiquitin-interacting motif; UBD, ubiquitin-binding or ubiquitin-like binding domain; UFD, ubiquitin fusion degradation; Ub, ubiquitin; GST, glutathione S-transferase; ITC, isothermal titration calorimetry; SPR, sur- face plasmon resonance; r.m.s., root mean square; WT, wild type; PDB, Pro- tein Data Bank; h, human; Sc, S. cerevisiae; Sp, S. pombe. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 26, pp. 20390–20398, June 25, 2010 © 2010 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. 20390 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 26•JUNE 25, 2010 cific differences exist in the interactions of the UBL domains of Rad23 and Dsk2 with Ufd2, which are coupled to subtle differ- ences in their overall binding affinities. Mutational analyses of the binding surface of the UBL domains and a closer inspection of the thermodynamic contributions of those residues point to adaptive evolution as a factor shaping these interfaces. EXPERIMENTAL PROCEDURES Cloning, site-directed mutagenesis, protein expression, and purification are described in the supplemental Experi- mental Procedures. Crystallization of Ufd2Rad23-UBL and Ufd2Dsk2-UBL— For crystallization of the Ufd2Rad23-UBL and Ufd2Dsk2-UBL complexes, Ufd2 was incubated with Rad23-UBL or Dsk2-UBL at a molar ratio of 1:1.5 (77 M Ufd2 and 115.5 M UBL) for 1 h at 4 °C in the presence of 2 mM dithiothreitol. Crystals were grown by vapor diffusion in hanging drops containing equal volumes of protein in 50 mM HEPES, pH 7.4, 150 mM NaCl, and 2 mM dithiothreitol and a reservoir solution consisting of 16–18% (w/v) polyethylene glycol 3500 and 200 mM K3-citrate, pH 8.3, equilibrated against the reservoir solution. Crystals were cryo-protected by soaking in mother liquor containing 15–20% (v/v) glycerol. They belong to space group P212121 with approximate cell dimensions of a  65 Å, b  126 Å, and c  181 Å with one complex per asymmetric unit. Data Collection and Structure Determination—Crystals were flash-cooled in liquid nitrogen, and data collection was per- formed at 100 K. Data were collected at beamlines ID14–4 (European Synchrotron Radiation Facility (ESRF), Grenoble, France) and BL 14.1 (Berliner Elektronenspeicherring-Gesell- schaft fu¨r Synchrotronstrahlung (BESSY), Berlin, Germany) and processed using Mosflm and Scala (39, 40). Data collection statistics are summarized in supplemental Table S1. For subse- quent calculations, the CCP4 suite was utilized (41) with excep- tions as indicated. The Ufd2 structure was solved by molecular replacement using Phaser (42) with Protein Data Bank (PDB) entry 2QIZ as search model. Because Phaser could not find a solution for the UBL domain with different search models, this domain was fitted manually into the electron density using human ubiquilin 3 (PDB entry 1YQB) for the Ufd2Rad23-UBL complex and the Dsk2-UBL domain (PDB entry 2BWF) for the Ufd2Dsk2-UBL complex as a model. The structures were refined with Phenix (43) and REFMAC5 incorporating transla- tion, libration, screw-rotation (TLS) refinement in all cycles (44, 45). Solvent molecules were automatically added with Coot (46). The figures were produced with PyMOL (65). In Vitro Binding Assays—For pulldown assays, GST-tagged Ufd2 and variants were immobilized on glutathione (GSH) beads. In all experiments, 20 l of GSH beads were incubated with 0.95 M purified Ufd2 in 400 l of phosphate-buffered saline buffer with 1 mM dithiothreitol and 0.1% (v/v) Triton X-100 at 4 °C for 1 h. WT-Ufd2 and GST alone were included as controls. After centrifugation (1250  g, 30 s), beads were washed five times with 400 l of binding buffer. Purified UBL proteins (0.95 M) in a total volume of 400 l of binding buffer were added to immobilized Ufd2 and treated in the same way as in the first step. Immobilized proteins were analyzed by 17% (v/v) SDS-PAGE or by immunoblotting with an anti-His antibody. Isothermal Titration Calorimetry (ITC)—Proteins were extensively dialyzed against phosphate-buffered saline buffer (pH 7.4, 1 mM -mercaptoethanol) followed by degassing. In all experiments, 75–150 M Rad23- and Dsk2-UBL proteins were titrated as the ligand into the sample cell containing 5–10 M Ufd2. A volume of 10 l of ligand was added at a time with a total number of 30 injections, resulting in a final molar ratio of ligand-to-protein varying between 3:1 and 4:1. All experiments were performed using a VP-ITC instrument (MicroCal, GE Healthcare) at 25 °C. Buffer-to-buffer, buffer-to-Ufd2, as well as Rad23-UBL/Dsk2-UBL-to-buffer titrations were performed as described above. Corrected data were analyzed with a single- site binding model using software supplied by the ITC manu- facturer and non-linear least squares fitting to calculate the dissociation constant (Kd). Surface Plasmon Resonance (SPR) Measurements—SPR binding assays were performed alternatively on BIAcore X or BIAcore T100 instruments (GE Healthcare) at 25 °C in 10 mM HEPES, pH 7.4, 150 mM NaCl, 50 M EDTA, 1 mM -mercap- toethanol, and 0.005% (v/v) Surfactant P20. 100 response units of His-tagged Rad23- or Dsk2-UBL were captured on a nickel- nitrilotriacetic acid (Ni-NTA) sensor chip. GST-tagged Ufd2 for comparative binding assays and untagged Ufd2 for affin- ity analysis were applied to the UBL surfaces in random duplicates at a flow rate of 50 l/min. After each cycle, the surface was regenerated using 350 mM EDTA in running buffer to remove bound Ni2 and captured proteins. The BIAcore T100 evaluation software was used to calculate the steady state affinity constants. Data were plotted using GraphPad Prism. For comparative assays, the relative bind- ing responses of the mutants to WT proteins were deter- mined by obtaining the maximum response for each interac- tion at the end of each injection. RESULTS Ufd2 Binds the UBL Domains of Rad23 and Dsk2 with High Affinity—Although Rad23 and Dsk2 interact with Ufd2 via their UBL domains (22, 26), yeast two hybrid assays could only identify the isolated N-terminal fragment (residues 1–380) of Ufd2 as its UBL-interacting region (26). Additional details regarding the Ufd2-UBL interactions have not been unraveled so far. To further characterize the interactions of Ufd2 with the UBLs of Rad23 and Dsk2, we performed GST pulldown assays with GST-tagged full-length Ufd2 and C-terminally His-tagged UBLs (Fig. 1A). Both UBLs were readily captured using immo- bilized GST-Ufd2. In contrast, the UBL domain of Ddi1, a third UBL-UBA protein, does not interact with Ufd2 (Fig. 1A) (22). The differential binding of the Rad23- and Dsk2-UBLs to the proteasomal subunits Rpn1 and Rpn10 has been described (25, 47, 48). Hence, we used SPR interaction analysis to search for quantitative differences in their interactions. Steady state affin- ity analysis of Ufd2 on both Rad23-UBL (Fig. 1B, left panel, and 1C) and Dsk2-UBL surfaces (Fig. 1B, right panel, and 1C) pro- vided a Kd of 55  3 nM for the interaction of Rad23-UBL and a lower affinity for Dsk2-UBL with a Kd of 418  56 nM. UBL-binding Domain of Ufd2 JUNE 25, 2010•VOLUME 285•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 20391 The binding of the UBLs of Rad23 and Dsk2 to Ufd2 was also analyzed by ITC to allow for a complete thermodynamic char- acterization (Fig. 1D). These studies resulted in a Kd of 70  6 nM for the interaction of Rad23-UBL with Ufd2 and for the binding of Dsk2-UBL to Ufd2 in a 2–3-fold higher Kd of 175  19 nM. Although there is an excellent agreement between SPR and ITC for the Rad23-Ufd2 interaction, the two methods show an 2-fold difference for the Dsk2-Ufd2 interaction. More importantly, the enthalpic and entropic components to the free energy are highly different between the two UBLs. The interac- tion of Rad23-UBL and Ufd2 is more exothermic (H  17.3 kcal/mol) when compared with Dsk2-UBL (H  10.1 kcal/ mol). However, this is offset by a substantial decrease in entropy for Rad23-UBL (TS  7.4 kcal/mol), whereas the entropic contribution is minimal for the Dsk2-UBL interaction (TS  0.8 kcal/mol). Crystal Structures of Ufd2 in Complex with Rad23- and Dsk2-UBL—We solved the structures of Ufd2 in complex with Rad23-UBL carrying either an N-terminal or a C-terminal His tag, which showed no significant structural differences. Due to better data quality, the structure of Ufd2 with a C-terminal His-tagged UBL is presented here. The Ufd2Rad23-UBL com- plex was refined at 2.4 Å resolution to a crystallographic R-fac- tor of 20.3% and a free R-factor of 25.7% (Table 1). As described previously (49), Ufd2 is composed of an N-terminal variable domain, a core domain, and a C-terminal U-box domain with a fold similar to that of RING (really interesting new gene) domains, which are present in certain Ub ligases (Fig. 2A). Despite some conformational variability of the U-box domain, our Ufd2 structure in the complex is quite similar (1.5 Å root mean square (r.m.s.) deviation for 954 C atoms) to the pub- lished Ufd2 structure (49). The N-terminal variable region of Ufd2 that binds to the UBL domain consists of eight -helices. Helices 1 to 4 are arranged in a four-helix bundle, whereas helices 5 and 6 interact with 3 and 4 through hydrophobic contacts that are partly mediated by their connecting loops (Fig. 2B). The struc- ture of Rad23-UBL is comprised of a five-stranded -sheet, one -helix, and one 310-helix (Fig. 2B). It displays a high degree of similarity with Ub (PDB entry 1UBQ, 1.1 Å r.m.s. deviation for 72 C atoms, z-score 14, 25% sequence identity) and the UBL domain of hHR23A (PDB entry 1P98, 1.6 Å r.m.s. deviation for FIGURE 1. Interactions of Ufd2 with the UBL domains of Rad23 and Dsk2. A, GST-Ufd2 immobilized on GSH-beads was tested for binding to C-terminally His-tagged UBLs of Rad23, Dsk2, and Ddi1. Captured UBLs were visualized by immunoblotting (WB) with an anti-His antibody. 2% of the input and GST beads incubatedwithUBLswereloadedascontrols.B,aseriesof2-foldUfd2dilutions(233–3.6nM)wasappliedonaRad23-orDsk2-UBLsurfacefor120s(leftandright panel,respectively).RU,responseunits.C,SPRbindingisothermsofWT-Rad23-andWT-Dsk2-UBLandthequintupleandseptupleDsk2-UBLvariantswithUfd2. conc., concentration. D, ITC analysis of Ufd2Rad23-UBL (closed circles) and Ufd2Dsk2-UBL (open circles) complexes. TABLE 1 Refinement statistics Ufd2Rad23-UBL Ufd2Dsk2-UBL Resolution limit (Å) 45.2-2.4 73.5-2.4 No. of reflections 56,268 55,087 No. of protein/ligand/solvent atoms 8303/17/298 8288/17/182 Rcryst (Rfree)a,b 0.203 (0.257) 0.210 (0.270) r.m.s. deviations in: Bond lengths (Å) 0.016 0.015 Bond angles (°) 1.711 1.610 Estimated coordinate error (Å) 0.25 0.26 Overall average B-factor (Å2) 25.7 42.9 Ramachandran statistics (%)c 93.1/97.9/2.1 93.8/98.4/1.6 aRcryst  hklFo  Fc/hklFo where Fo and Fc are the observed and calculated structure factor amplitudes. bRfree, same as Rcryst for 5% of the data randomly omitted from the refinement. The estimated coordinate error is based on Rfree. c Ramachandran statistics indicate the fraction of residues in the favored (98%), allowed ( 99.8%), and disallowed regions of the Ramachandran diagram, as defined by MolProbity (64). UBL-binding Domain of Ufd2 20392 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 26•JUNE 25, 2010 72 C atoms, z-score 11.3, 26% sequence identity), one of the two human homologs of budding yeast Rad23. Subsequently, we solved the structure of Ufd2 with Dsk2- UBL by molecular replacement. The UBL domain in the Ufd2Dsk2-UBL complex exhibits increased flexibility; in par- ticular, with a C-terminally tagged UBL domain, the first 30 amino acids of this domain were largely disordered (data not shown). With an N-terminally tagged protein, the Ufd2Dsk2- UBL structure was refined at 2.4 Å resolution to a crystallo- graphic R-factor of 21.0% and a free R-factor of 27.0% (Table 1). Both Rad23-UBL and Dsk2-UBL structures can be superim- posed with an r.m.s. deviation of 1.1 Å for 71 aligned residues (z-score 13.6, 30% sequence identity). Analysis of the Ufd2Rad23-UBL Interface—The Ufd2UBL interface in the structure of the complex buries a total molecu- lar surface of about 1260 Å2, which is comprised to 590 Å2 of the molecular surface of Ufd2 (1.3% of the total surface area) and 670 Å2 from UBL (14.6% of the total surface area). This interface is composed of almost equal parts of non-polar resi- dues (38%), polar residues (33%), and charged residues (29%); however, there are only one salt bridge (UBL-Lys-10 N–Ufd2- Glu-49 O1 with a distance of 2.6 Å) and two direct hydrogen bonds (UBL-Ser-47 O–Ufd2-Arg-92 N2, UBL-Gln-52 N2– Ufd2-Glu-141 O at distances of 2.3 and 3.2 Å, respectively) present (Fig. 3A). Three UBL segments are contacting Ufd2 (Fig. 3A). Segment I is located in the loop connecting -strands one and two, seg- ment II involves -strands three and four, and segment III is located in -strand five. Ufd2 residues from helix 2 and 4 as well as the loop connecting 4 with 5 contribute to the Ufd2UBL interface. These residues contact the hydrophobic surface of the UBL -sheet in the region of -strands 3, 4, and 5. Participating residues from Ufd2 include Leu-44, Tyr-97, Val- 100, and Trp-107, which are located in the hydrophobic UBL pocket formed by residues Phe-9, Ile-45, Val-50, Val-69, and Met-71 of Rad23 (Fig. 3, A and B). For comparison, the principal recognition determinants in Ub are: 1) the hydrophobic pocket formed by the side chains of Leu-8 (Phe-9 in Rad23), Ile-44 (Ile-45 in Rad23), His-68 (Val-69 in Rad23), and Val-70 (Met-71 in Rad23) and 2) the main chain amide group of Gly-47 (Gly-48 in Rad23), which is involved in hydrogen bonding (50). Although the hydrophobic patch of Rad23-UBL is also crucial for its interaction with Ufd2, the main chain of Gly-48 does not form a hydrogen bond. Instead, the -turn (Ser47–Gly48) connecting -strands 3 and 4 is stabi- lized by the aforementioned strong hydrogen bond between Ufd2-Arg-92 and UBL-Ser-47, whereas Ufd2-Gly-96 and Ufd2- Tyr-97 contact UBL-Gly-48 (Fig. 3A). The aromatic ring of Ufd2-Tyr-97 is involved in a stacking interaction with the pep- tide bond between UBL residues 47 and 48 in this -turn. Probing the Ufd2Rad23-UBL Interface—The importance of interface residues was analyzed by mutagenesis experiments. Eleven residues from Ufd2 and nine from Rad23-UBL were each replaced with Ala. With the exception of the Rad23-UBL- G48A variant that showed a reduced expression, all Ufd2 and Rad23-UBL variants behaved like the WT protein during and after purification, indicating that they were correctly folded (data not shown). Initially, the contribution of these residues was studied by GST pulldown and comparative SPR binding assays (Table 2, supplemental Figs. S1 and S2A). In SPR studies, the relative binding responses of mutants to WT proteins were determined and compared. The majority of Rad23-UBL single mutants revealed reduced binding to Ufd2 with Rad23-UBL- I45A displaying the most prominent binding defect. The con- tribution of the remaining residues to the interaction is aug- mented in double mutants (supplemental Fig. S1C). Analysis of the Ufd2 variants by SPR showed a largely reduced binding of the residues located in the hydrophobic region of the UBL- binding pocket (Leu-44, Tyr-97, Val-100, and Phe-107) and Asp-40 (Table 2 and supplemental Fig. S2A). ITC studies confirmed these results and allowed for a quantitative analysis (Table 2, supplemental Fig. S3 and supplemental Table S2). The most significant effect for Ufd2 was observed for all residues located in the hydrophobic UBL pocket. Mutation of Val-100 and Phe-107 to Ala completely abolished binding, the Y97A variant strongly reduced binding (1900-fold), and the I104A and L44A variants showed signifi- cantly decreased affinities (20- and 120-fold, respectively). Although not directly involved in complex formation (Fig. 3A), the Ufd2-D40A variant showed a 110-fold reduced affinity (Table 2), which probably is the result of the missing intramo- lecular hydrogen bond between Ufd2-Asp-40 and Ufd2-Tyr-97 (O2–OH 2.5 Å). This hydrogen bond seems to be crucial for proper positioning of the aromatic side chain of Tyr-97 in the interface region and might be important to align helices 2 and 4 for interaction with the Rad23-UBL. FIGURE 2. Structure of Ufd2 in complex with the UBL domain of Rad23. A, ribbon representation of the overall structure of the Ufd2Rad23-UBL com- plex. The Rad23-UBL domain is shown in green, the N-terminal Ufd2 region is in orange, the Ufd2 core domain is in gray, and the Ufd2 U-box domain is in red. B, close-up view of the N-terminal Ufd2 domain in complex with Rad23- UBL with secondary structural elements labeled and color-coded as in A. UBL-binding Domain of Ufd2 JUNE 25, 2010•VOLUME 285•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 20393 In Rad23-UBL, Ile-45 was shown to be integral for binding to Ufd2 by the detrimental effect (130-fold decrease) after exchange to Ala (Table 2). Mutation of Phe-9, Val-50, and Val- 69, residues adjacent to Ile-45 in the hydrophobic patch, also decreased the affinity of Rad23-UBL 5–7-fold. Ser-47, which is in hydrogen-bonding distance to Ufd2-Arg-92 and next to UBL-Gly-48, showed a 9-fold reduced affinity. In Ub and in the human Rad23 homolog hHR23A, Ser-47 is replaced by Ala. Charged residues found in the interface (Ufd2, Glu-26, Glu-49, and Arg-92; UBL, Lys-10) do not contribute significantly to the interaction. In summary, our data indicate that the most prom- inent contact between Ufd2 and Rad23-UBL is the strong hydrophobic interaction between UBL-Ile-45 and Ufd2-Val- 100 as well as Ufd2-Phe-107, which defines the core of the UBL- interacting region of Ufd2. Molecular Discrimination between Rad23 and Dsk2—De- spite a similar fold, the UBL domains of Rad23 and Dsk2 display only 30% sequence identity, which could give rise to differences in their interactions. A superposition of the bound Rad23-UBL and Dsk2-UBL in the two complex structures showed signifi- cant changes (Fig. 3C). Of the three UBL segments involved in the Ufd2 interaction (Fig. 3A), segment II including Ile-45 (Ile-44 in Ub) is highly conserved, and there are no conforma- tional changes in both UBL structures, whereas segments I and III are not conserved and display structural changes (Fig. 3C). The loop, connecting -strands one and two, adopts different conformations, and -strand five shows a displacement that might affect binding (Fig. 3C). Segment I includes Phe-9 in Rad23-UBL, corresponding to Leu-8 in Ub, where this residue is also involved in Ub recogni- tion by UBDs (50, 51). Phe-9 is replaced by Gly-10 in Dsk2-UBL, and there is no corresponding hydrophobic interacting residue (supplemental Fig. S4A). Dsk2 residues Gly-10 and Gln-11 adopt different conformations when compared with Leu-8/ Thr-9 of Ub and Phe-9/Lys-10 of Rad23-UBL. In the Ufd2Dsk2-UBL structure, the Ufd2Rad23-UBL salt bridge (Lys-10/Glu-49) is missing due to the Lys-10 to Gln-11 exchange, with the latter side chain no longer being located in the protein interface (supplemental Fig. S4A). The missing interaction from segment I in Dsk2 might be compensated by the displacement of -strand five toward Ufd2 and a replace- ment of Val-69 to His-69 found in segment III resulting in a more pronounced interaction in this region when compared with Rad23-UBL (supplemental Fig. S4A). The presence of the salt bridge seems to be the reason for the more exothermic character of the Ufd2Rad23-UBL interaction, a view that is also supported by the corresponding Ufd2-E49A and Rad23-K10A variants, which both display binding enthalpies similar to the FIGURE 3. The Ufd2Rad23-UBL interface. A, residues involved in binding are shown in stick representation. Carbon atoms of Ufd2 residues are colored in orange and in green for Rad23-UBL. Dashed lines indicate H-bonds. B, structure-based sequence alignment of Rad23-UBL, Dsk2-UBL, hHR23A-UBL, and Ub. Secondary structure elements of Rad23-UBL were assigned using DSSP (61) and are labeled above the sequences. The alignment was performed using DaliLite (62), and the figure was prepared with ESPript (63). Strictly conserved amino acids are highlighted with a red background, and similar amino acids are shown as redletters.ThethreeUfd2-bindingsegmentsareindicated.ResiduesinvolvedinUfd2Rad23-UBLinteractionarelabeledwithgreenstars.C,superpositionofthe Ufd2Rad23-UBL/Dsk2-UBL complex structures with the N-terminal binding domain of Ufd2 in orange (Rad23 complex) and gray (Dsk2 complex), with Rad23- UBL in green and Dsk2-UbL in yellow. UBL-binding Domain of Ufd2 20394 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 26•JUNE 25, 2010 Ufd2Dsk2-UBL complex (supplemental Table S2). At the same time, the absence of the salt bridge in both mutants is accom- panied by a more favorable entropic contribution, which is on a level similar to the Ufd2Dsk2-UBL complex. To identify residues important for the subtle molecular dis- crimination between the UBL domains of Rad23 and Dsk2, the interaction of Dsk2-UBL with Ufd2 mutants was analyzed by GST pulldown assays (data not shown), SPR, and ITC (Table 2, supplemental Figs. S2B and S3C). Quantitative ITC analysis showed reduced binding of Dsk2 to Ufd2 mutants Y97A (470- fold), V100A (22-fold), I104A (6-fold), and F107A (20-fold) (Table 2). However, binding of the V100A and F107A variants is not completely abolished, and when compared with Rad23- UBL, the binding affinities are less affected by a factor of about 3–7 in most of the mutants analyzed. In addition, the L44A mutant, which has a 120-fold reduced affinity with Rad23-UBL, is only three times reduced in the case of Dsk2-UBL. In general agreement with the ITC affinity data, the compar- ative SPR binding assay revealed significant differences in the association of Ufd2 variants Y97A, V100A, I104A, and F107A with Rad23- and Dsk2-UBL surfaces (supplemental Fig. S4B). The observed SPR decrease for the binding of the T48A and E49A variants of Ufd2 to Dsk2-UBL seems to be compensated by slower dissociations, thus explaining why these mutants show no significant defect in the ITC analysis. To further analyze the contribution of segments I and III to complex formation, a G10F/Q11K/S67Q/H69V/V71M quintu- ple Dsk2-UBL mutant was generated, where key residues in binding segments I and III were replaced with the correspond- ing residues from Rad23-UBL. Comparative binding as well as steady state affinity analysis by SPR revealed only a small increase (Kd  348 nM) in binding affinity for Ufd2 when com- pared with WT-Dsk2-UBL (Kd  418 nM) (data not shown and Fig. 1C). In addition, neither a crystal structure of the quintuple Ufd2Dsk2-UBL complex (data not shown) nor the KD of 240 nM deduced by ITC revealed significant differences from WT-Dsk2-UBL (Kd  175 nM). The ITC analysis did, however, reveal that the binding is now driven by an increase in entropy (TS  6.5 kcal/mol versus 0.8 and 7.4 kcal/mol for WT-Dsk2-UBL and -Rad23-UBL, respectively), whereas the binding enthalpy is reduced to only 2.5 kcal/mol when com- pared with 10.1 and 17.3 kcal/mol (supplemental Table S2). Interestingly, SPR and ITC analysis of a G10F/Q11K/I50V/ K52Q/S67Q/H69V/V71M septuple Dsk2-UBL mutant, which has the additional I50V and K52Q substitutions in segment II, showed an even lower affinity (SPR, Kd  648 nM; ITC, Kd  875 nM) to Ufd2 when compared with WT-Dsk2-UBL (Fig. 1C). The N Terminus of Ufd2 Represents a Unique and Conserved UBL-binding Domain—A multiple sequence alignment of Ufd2 from different yeast species displays a distinct pattern of con- served residues involved in UBL binding (Fig. 4A). Among the available yeast genomes, the Schizosaccharomyces pombe sequence is most similar to those from higher eukaryotes; thus we isolated cDNA fragments for the coding region of the UBL domains of Rad23 and Dsk2 and full-length Ufd2 from this organism and examined their interactions by GST pulldown assays (Fig. 4B) as well as SPR (data not shown). We could show that SpUfd2 interacts strongly with the UBL domains of SpRad23 and SpDsk2 as well as with the UBL domains of ScRad23 and ScDsk2 and vice versa. This cross species interac- tion, despite the diversified UBL and Ufd2 amino acid sequences, indicates that the identified sequence pattern repre- sents a real UBL-interacting domain. A surface representation of this motif is shown in Fig. 4C. The N terminus of budding yeast Ufd2 displays only limited sequence homology with the human Ufd2s, E4A and E4B (supplemental Fig. S5) and other Ufd2s from higher eukaryotes. In agreement with this finding, there are no reports that hHR23A/B interacts with either of the human homologs of Ufd2. Interestingly, our SPR studies showed that the UBL domain of hHR23A interacts with ScUfd2, albeit with lower affinity (data not shown). Apparently, the high affinity interac- tion of the UBL domains of Rad23 and Dsk2 has been lost dur- ing the evolution of this domain. The absence of conservation of the Ufd2-UBL interface could potentially be used for thera- peutic interventions against pathogenic yeasts such as Candida albicans by designing low molecular weight compounds that disrupt this interface. However, further functional studies in pathogenic yeasts are required to examine the suitability of this surface as a drug target. TABLE 2 ITC and SPR parameters of Ufd2, Rad23-UBL, Dsk2-UBL, and variants  indicates no change; ND indicates not detected (corresponding to at least a 104-fold decrease in binding affinity). Ufd2 WT-UBL ITC SPRa (% of relative response) Kd Fold decrease nM WT Rad23 70 100 Dsk2 175 100 E26A Rad23 284 4 91 Dsk2 521 3 83 D40A Rad23 7900 110 20 Dsk2 7600 40 0 L44A Rad23 8300 120 31 Dsk2 463 3 52 T48A Rad23 72  70 Dsk2 296 2 29 E49A Rad23 413 6 69 Dsk2 314 2 44 R92A Rad23 265 4 76 Dsk2 128  59 G96A Rad23 592 8 51 Dsk2 216  60 Y97A Rad23 134,000 1900 3 Dsk2 83,000 470 0 V100A Rad23 ND 10,000 9 Dsk2 3900 22 1 I104A Rad23 1600 20 43 Dsk2 1100 6 12 F107A Rad23 ND 10,000 11 Dsk2 3600 20 0 Ufd2 Rad23-UBL ITC SPRa (% of relative response) Kd Fold decrease nM WT F9A 376 5 80 WT K10A 162 2 96 WT I45A 9100 130 17 WT S47A 606 9 62 WT V50A 441 6 88 WT Q52A 415 6 79 WT Q67A 113 2 92 WT V69A 478 7 70 WT M71A 221 3 88 a For comparative SPR assays, the relative binding responses of the mutants to wt proteins were determined by obtaining the maximum response for each interac- tion at the end of injection. UBL-binding Domain of Ufd2 JUNE 25, 2010•VOLUME 285•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 20395 DISCUSSION Families and superfamilies of protein domains or folds have evolved through a process of homologous recombination and gene duplication (52) followed by sequence divergence. Mem- bers of different classes of UBDs such as UBA or UBL domains result from these processes. For instance, the UBL domains of Rad23 and Dsk2 display only 30% sequence identity but adopt the same fold and utilize the same binding surface to recognize a common UBL-binding domain of Ufd2 to form complexes that display similarly high affinity. Nevertheless, not all inter- acting residues are conserved; in particular, there is sequence diversity in binding segments I and III of UBLs. Our attempts to interconvert the UBL domains by altering non-conserved inter- facial residues were not successful, thus suggesting that addi- tional elements exist and play a role in the respective Ufd2-UBL interaction. Interestingly, these results resemble earlier studies on WW domains (53, 54), where a statistical analysis of multiple sequence alignments was utilized to identify co-evolving resi- dues. The authors demonstrated that not only interfacial resi- dues but also buried residues distal to the interface co-evolved with interfacial residues and contribute significantly to the interactions. They concluded that certain sequence patterns in interacting domains are due to adaptive evolution. In agree- ment with these findings, our data prove that substitution of key interfacial residues of Dsk2-UBL has no significant effect on its overall binding affinity to Ufd2. In case of the septuple mutant, we even observed a decrease in binding affinity, which could be due to the imposed disorder into the evolutionary inter-residue relations within the UBL fold. This is supported by the fact that when compared with Dsk2-UBL and in partic- ular Rad23-UBL, the binding of the quintuple Dsk2-UBL mutant is driven strongly by entropy. These findings indicate that binding interfaces can be modulated by changes in residues that affect either the binding enthalpy or the entropy, thus pro- viding additional freedom to maintain an interaction during the course of evolution, an effect that has been described previously as entropy/enthalpy compensation (55, 56). Our studies suggest that UBL domains have co-evolved with Ufd2 to reach optimal binding affinities by altering specific res- idue-to-residue interactions (co-evolution at the residue level) (57), while at the same time, all functional aspects of Rad23 or Dsk2 are preserved. Therefore, the primary sequence degener- FIGURE 4. The N terminus of Ufd2 represents a conserved UBL-interacting domain in lower eukaryotes. A, alignment of the N-terminal sequences of fungal Ufd2s. Invariant or conserved residues with surface access are colored in dark blue, buried ones are in light blue. Residues labeled with red stars represent the core region of the binding domain, which is essential for UBL interaction, whereas residues labeled with yellow stars contribute moderately to the interaction. K. lactis, Kluyveromyces lactis; C. glabrata, Candida glabrata; Z. rouxii, Zygosaccharomyces rouxii; L. thermotolerans, Lachancea thermotolerans; C. tropicalis, Candida tropicalis; C. dubliniensis, Candida dubliniensis; P. guilliermondii, Pichia guilliermondii; D. hansenii, Debaryomyces hansenii. B, GST pulldown assay demonstrates the cross interactions of S. pombe and S. cerevisiae proteins. 5% of inputs and GST beads incubated with UBLs were loaded as controls. WB, Western blot. C, surface representation of the N-terminal UBL-binding domain of Ufd2, color-coded as in A. UBL-binding Domain of Ufd2 20396 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 26•JUNE 25, 2010 acy of protein domains such as UBAs or UBLs has been toler- ated and established in favor of the cooperative nature of the interactions and their functionality within their respective pro- tein complexes. This further suggests that differential binding properties observed for the interactions of Ufd2 with UBLs (this study) or for the interactions of UBAs with Ub and Ub chains (14) can arise not necessarily due to their interaction with dif- ferent ligands but can also result from the adaptive co-evolution of these domains with the same interacting partners. Seem- ingly, these interfacial domains have evolved to hold protein- protein interactions in a suitable form within multicomponent complexes until they are challenged by downstream events. Numerous structures of Ub receptors in complex with their respective Ub/UBL-binding domains have been reported. The so far characterized Ub receptors of the 26 S proteasome in budding yeast encompass the two proteasomal subunits Rpn10 (S5a in humans) and Rpn13 and the three UBL-UBA proteins Rad23, Dsk2, and Ddi1, which associate with the proteasome and function as shuttle factors (21). Experimental evidence for the existence of additional candidates exist (21, 58). Rad23 and Dsk2 interact with the proteasomal subunit Rpn1 via their UBL domains (21, 47). Aside from their known interactions with Ub, Rpn13 and Rpn10/S5a alternatively interact with UBL-UBA proteins (21, 35, 37, 38, 48, 51, 59). For instance, the preferential association of Rpn1 with Rad23 and Rpn10 with Dsk2 has been reported (25, 38, 47, 48). Based on the binding of hpLIC2 (Dsk2 homolog) with Rpn13, an interac- tion of Dsk2 with Rpn13 has been proposed (51, 59). Although the aforementioned examples engage essentially the same surface of Ub/UBL, they diverge in both structure and pat- terns of Ub/UBL recognition (Fig. 5). For instance, hRpn10/S5a recog- nizes the UBL domain of hHR23A, one of the two human homologs of Rad23, via a Ub-interacting motif, which consists of a single -helix (35, 37). Rpn13 binds Ub via a pleck- strin homology domain, which is a seven-stranded -sandwich capped by an -helix (51). The Ub-binding surface of Rpn13 is formed by three loops that bridge -strands. Another Ub-binding element is the UBA domain found for example in Dsk2 (60). The UBA domain is com- posed of a three-helix bundle. With the exception of Rpn13, which exclusively binds via loops, it seems that the majority of Ub/UBL-bind- ing domains fold into -helical structures including the known UBDs, UIM, and UBA, and the UBL-binding domain of Ufd2 iden- tified in this study. Despite the pre- dominant interaction involving -helices as Ub/UBL-binding elements, the three-dimensional structure of the UBL-binding domain of Ufd2 differs from other known examples, hence providing the first structural descrip- tion for how Ufd2 acts as a UBL receptor while at the same time further enhancing the diversity of UBDs in general. Acknowledgments—We thank Martin Scheffner and Keith Wilkinson for critical reading of the manuscript. We thank Stefan Jentsch for providing the original plasmids for the expression of Rad23, Dsk2, and Ufd2 and for Ufd2-specific antibodies used in the initial phase of this study. We also thank David Fischer and Rodrigo Villasen˜or for the contribution to this study and Sven Eiselein for providing us with C-terminal GST-tagging plasmid. REFERENCES 1. Hershko, A., and Ciechanover, A. (1998) Annu. Rev. Biochem. 67, 425–479 2. Ross, C. A., and Pickart, C. M. (2004) Trends Cell Biol. 14, 703–711 3. Haglund, K., and Dikic, I. (2005) EMBO J. 24, 3353–3359 4. Mukhopadhyay, D., and Riezman, H. (2007) Science 315, 201–205 5. Grabbe, C., and Dikic, I. (2009) Chem. Rev. 109, 1481–1494 6. Buchberger, A. (2002) Trends Cell Biol. 12, 216–221 7. Hicke, L., Schubert, H. L., and Hill, C. P. (2005) Nat. Rev. Mol. Cell Biol. 6, 610–621 8. Hurley, J. H., Lee, S., and Prag, G. (2006) Biochem. J. 399, 361–372 9. Harper, J. W., and Schulman, B. A. (2006) Cell 124, 1133–1136 10. Hochstrasser, M. (2009) Nature 458, 422–429 FIGURE 5. Mode of Ub/UBL recognition by different Ub/UBL-binding domains. In each panel, ribbon rep- resentationstogetherwiththemolecularsurfacesofbothbindingpartnersareshownwithUb/UBLinthesame orientation. A, Ufd2Rad23-UBL. B, Dsk2-UBADsk2-UBL (PDB entry 2BWE). C, Rpn13Ub (PDB entry 2Z59). D, S5a/Rpn10hHR23A-UBL (PDB entry 1P9D). UBL-binding Domain of Ufd2 JUNE 25, 2010•VOLUME 285•NUMBER 26 JOURNAL OF BIOLOGICAL CHEMISTRY 20397 11. Dikic, I., Wakatsuki, S., and Walters, K. J. (2009) Nat. Rev. Mol. Cell Biol. 10, 659–671 12. Hofmann, K. (2009) DNA Repair 8, 544–556 13. Hofmann, K., and Bucher, P. (1996) Trends Biochem. Sci. 21, 172–173 14. Raasi, S., Varadan, R., Fushman, D., and Pickart, C. M. (2005) Nat. Struct. Mol. Biol. 12, 708–714 15. Varadan, R., Assfalg, M., Raasi, S., Pickart, C., and Fushman, D. (2005) Mol. Cell 18, 687–698 16. Rahighi, S., Ikeda, F., Kawasaki, M., Akutsu, M., Suzuki, N., Kato, R., Ken- sche, T., Uejima, T., Bloor, S., Komander, D., Randow, F., Wakatsuki, S., and Dikic, I. (2009) Cell 136, 1098–1109 17. Lo, Y. C., Lin, S. C., Rospigliosi, C. C., Conze, D. B., Wu, C. J., Ashwell, J. D., Eliezer, D., and Wu, H. (2009) Mol. Cell 33, 602–615 18. Sato, Y., Yoshikawa, A., Mimura, H., Yamashita, M., Yamagata, A., and Fukai, S. (2009) EMBO J. 28, 2461–2468 19. Sato, Y., Yoshikawa, A., Yamashita, M., Yamagata, A., and Fukai, S. (2009) EMBO J. 28, 3903–3909 20. Kulathu, Y., Akutsu, M., Bremm, A., Hofmann, K., and Komander, D. (2009) Nat. Struct. Mol. Biol. 16, 1328–1330 21. Finley, D. (2009) Annu. Rev. Biochem. 78, 477–513 22. Kim, I., Mi, K., and Rao, H. (2004) Mol. Biol. Cell 15, 3357–3365 23. Hara, T., Kamura, T., Kotoshiba, S., Takahashi, H., Fujiwara, K., Onoyama, I., Shirakawa, M., Mizushima, N., and Nakayama, K. I. (2005) Mol. Cell. Biol. 25, 9292–9303 24. Ivantsiv, Y., Kaplun, L., Tzirkin-Goldin, R., Shabek, N., and Raveh, D. (2006) Mol. Cell. Biol. 26, 1579–1588 25. Ishii, T., Funakoshi, M., and Kobayashi, H. (2006) EMBO J. 25, 5492–5503 26. Richly, H., Rape, M., Braun, S., Rumpf, S., Hoege, C., and Jentsch, S. (2005) Cell 120, 73–84 27. Hoppe, T. (2005) Trends Biochem. Sci. 30, 183–187 28. Rape, M., and Jentsch, S. (2004) Biochim. Biophys. Acta 1695, 209–213 29. Medicherla, B., Kostova, Z., Schaefer, A., and Wolf, D. H. (2004) EMBO Rep. 5, 692–697 30. Raasi, S., and Wolf, D. H. (2007) Semin. Cell Dev. Biol. 18, 780–791 31. Verma, R., Oania, R., Graumann, J., and Deshaies, R. J. (2004) Cell 118, 99–110 32. Liu, C., van Dyk, D., Li, Y., Andrews, B., and Rao, H. (2009) BMC Biol. 7, 75 33. Raasi, S., and Pickart, C. M. (2003) J. Biol. Chem. 278, 8951–8959 34. Hartmann-Petersen, R., Hendil, K. B., and Gordon, C. (2003) FEBS Lett. 535, 77–81 35. Walters, K. J., Kleijnen, M. F., Goh, A. M., Wagner, G., and Howley, P. M. (2002) Biochemistry 41, 1767–1777 36. Walters, K. J., Lech, P. J., Goh, A. M., Wang, Q., and Howley, P. M. (2003) Proc. Natl. Acad. Sci. U.S.A. 100, 12694–12699 37. Mueller, T. D., and Feigon, J. (2003) EMBO J. 22, 4634–4645 38. Zhang, D., Chen, T., Ziv, I., Rosenzweig, R., Matiuhin, Y., Bronner, V., Glickman, M. H., and Fushman, D. (2009) Mol Cell. 36, 1018–1033 39. Leslie, A. G. W. (1992) Joint CCP4ESF-EAMCB Newsletter on Protein Crystallography, Vol. 26, Daresbury Laboratory, Warrington, UK 40. Evans, P. (2006) Acta Crystallogr. D Biol. Crystallogr. 62, 72–82 41. Bailey, S. (1994) Acta Crystallogr. D Biol. Crystallogr. 50, 760–763 42. McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Sto- roni, L. C., and Read, R. J. (2007) J. Appl. Crystallogr. 40, 658–674 43. Adams, P. D., Grosse-Kunstleve, R. W., Hung, L. W., Ioerger, T. R., Mc- Coy, A. J., Moriarty, N. W., Read, R. J., Sacchettini, J. C., Sauter, N. K., and Terwilliger, T. C. (2002) Acta Crystallogr. D Biol. Crystallogr. 58, 1948–1954 44. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Acta Crystallogr. D Biol. Crystallogr. 53, 240–255 45. Winn, M. D., Isupov, M. N., and Murshudov, G. N. (2001) Acta Crystal- logr. D Biol. Crystallogr. 57, 122–133 46. Emsley, P., and Cowtan, K. (2004) Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 47. Elsasser, S., Gali, R. R., Schwickart, M., Larsen, C. N., Leggett, D. S., Mu¨ller, B., Feng, M. T., Tu¨bing, F., Dittmar, G. A., and Finley, D. (2002) Nat. Cell Biol. 4, 725–730 48. Matiuhin, Y., Kirkpatrick, D. S., Ziv, I., Kim, W., Dakshinamurthy, A., Kleifeld, O., Gygi, S. P., Reis, N., and Glickman, M. H. (2008) Mol. Cell. 32, 415–425 49. Tu, D., Li, W., Ye, Y., and Brunger, A. T. (2007) Proc. Natl. Acad. Sci. U.S.A. 104, 15599–15606 50. Ohno, A., Jee, J., Fujiwara, K., Tenno, T., Goda, N., Tochio, H., Kobayashi, H., Hiroaki, H., and Shirakawa, M. (2005) Structure 13, 521–532 51. Schreiner, P., Chen, X., Husnjak, K., Randles, L., Zhang, N., Elsasser, S., Finley, D., Dikic, I., Walters, K. J., and Groll, M. (2008) Nature 453, 548–552 52. Te Velthuis, A. J., and Bagowski, C. P. (2008) Curr. Genomics. 9, 88–96 53. Russ, W. P., Lowery, D. M., Mishra, P., Yaffe, M. B., and Ranganathan, R. (2005) Nature 437, 579–583 54. Socolich, M., Lockless, S. W., Russ, W. P., Lee, H., Gardner, K. H., and Ranganathan, R. (2005) Nature 437, 512–518 55. Reyes-Turcu, F. E., Shanks, J. R., Komander, D., and Wilkinson, K. D. (2008) J. Biol. Chem. 283, 19581–19592 56. Hunter, C. A., and Tomas, S. (2003) Chem. Biol. 10, 1023–1032 57. Pazos, F., and Valencia, A. (2008) EMBO J. 27, 2648–2655 58. Lam, Y. A., Lawson, T. G., Velayutham, M., Zweier, J. L., and Pickart, C. M. (2002) Nature 416, 763–767 59. Husnjak, K., Elsasser, S., Zhang, N., Chen, X., Randles, L., Shi, Y., Hof- mann, K., Walters, K. J., Finley, D., and Dikic, I. (2008) Nature 453, 481–488 60. Lowe, E. D., Hasan, N., Trempe, J. F., Fonso, L., Noble, M. E., Endicott, J. A., Johnson, L. N., and Brown, N. R. (2006) Acta Crystallogr. D Biol. Crystallogr. 62, 177–188 61. Kabsch, W., and Sander, C. (1983) Biopolymers 22, 2577–2637 62. Holm, L., Ka¨a¨ria¨inen, S., Rosenstro¨m, P., and Schenkel, A. (2008) Bioin- formatics 24, 2780–2781 63. Gouet, P., Courcelle, E., Stuart, D. I., and Me´toz, F. (1999) Bioinformatics 15, 305–308 64. Davis, I. W., Leaver-Fay, A., Chen, V. B., Block, J. N., Kapral, G. J., Wang, X., Murray, L. W., Arendall, W. B., 3rd, Snoeyink, J., Richardson, J. S., and Richardson, D. C. (2007) Nucleic Acids Res. 35, W375–W383 65. DeLano, W. L. (2002) The PyMOL Molecular Graphics System, DeLano Scientific LLC, San Carlos, CA UBL-binding Domain of Ufd2 20398 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 26•JUNE 25, 2010
3M6B
Crystal Structure of the Ertapenem Pre-isomerized Covalent Adduct with TB B-lactamase
Biochemical and Structural Characterization of Mycobacterium tuberculosis β-Lactamase (BlaC) with the Carbapenems Ertapenem and Doripenem Lee W. Tremblay#, Fan Fan#, and John S Blanchard‡,* Department of Biochemistry, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461 Abstract Despite the enormous success of β-lactams as broad-spectrum antibacterials, they have never been widely used for the treatment of TB due to intrinsic resistance that is caused by the presence of a chromosomally-encoded gene (blaC) in Mycobacterium tuberculosis. Our previous studies of TB BlaC revealed that this enzyme is an extremely broad-spectrum β-lactamase hydrolyzing all β- lactam classes. Carbapenems are slow substrates that acylate the enzyme but are only slowly deacylated and can therefore act also as potent inhibitors of BlaC. We carried out the in vitro characterization of doripenem and ertapenem with BlaC. A steady-state kinetic burst was observed with both compounds with magnitudes proportional to the concentration of BlaC used. The results show apparent Km and kcat values of 0.18 µM and 0.016 min−1 for doripenem and 0.18 µM and 0.017 min−1 for ertapenem. FTICR mass spectrometry demonstrated that the doripenem and ertapenem acyl-enzyme complexes remain stable over a time period of 90 min. The BlaC- doripenem covalent complex obtained after 90 minutes of soaking was solved to 2.2 Å, while the BlaC-ertapenem complex obtained after a 90 minute soak was solved to 2.0 Å. The 1.3 Å diffraction data from a 10 minute ertapenem-soaked crystal revealed an isomerization occurring in the BlaC-ertapenem adduct in which the original Δ2 pyrroline ring was tautomerized to generate the Δ1 pyrroline ring. The isomerization leads to the flipping of the carbapenem-hydroxyethyl group to hydrogen bond to the carboxyl O2 of Glu166. The hydroxyethyl flip results in both decreased basicity of Glu166 and in a significant increase in the distance between the carboxyl O2 of Glu166 and the catalytic water molecule, slowing hydrolysis. Tuberculosis (TB), caused by Mycobacterium tuberculosis, continues to be a worldwide health concern (1). There were an estimated 9.3 million new cases of TB in 2007 and approximately 1.3 million HIV-negative patient fatalities as well as nearly half a million deaths amongst HIV-positive populations (2). Even fifty years after the introduction of powerful antibiotics to treat TB, it has been estimated that one person is infected in the world every few seconds (3). The failure to control TB is due to the emergence of M. tuberculosis strains that are multiply drug resistant towards the front line antimycobacterial drugs such as isoniazid and rifampicin. Phone: (718) 430-3096; Fax: (718) 430-8565. *AUTHOR EMAIL ADDRESS: blanchar@aecom.yu.edu #These authors contributed equally to this work. Supporting Information Available One figure showing the dependence of kburst on the [ertapenem] and [doripenem]. This material is available free of charge via the Internet at http://pubs.acs.org NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 May 4. Published in final edited form as: Biochemistry. 2010 May 4; 49(17): 3766–3773. doi:10.1021/bi100232q. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript As one of the most important antibiotic families, β-lactams include a broad range of molecules including penicillin derivatives, cephalosporins, monobactams, carbapenems, and β-lactamase inhibitors. The carbapenems exhibit the broadest spectrum of activity among the β-lactam antimicrobials, providing safe and efficacious therapies in the treatment of serious infections caused by Gram-positive, Gram-negative, and anaerobic bacterial pathogens (4,5). Carbapenem antibiotics were originally developed from thienamycin, a natural product identified in culture filtrates of Streptomyces cattleya (6). There are four carbapenems approved thus far for human use: imipenem, meropenem, ertapenem, and doripenem (5). Imipenem was the first carbapenem approved by the US Food and Drug Administration (FDA) in 1985, and is by far the most widely used carbapenem. The use of meropenem was approved in 1995, followed by ertapenem and doripenem in 2001 and 2007, respectively. Except for imipenem, all carbapenems are stable against the mammalian kidney dehydropeptidase (7). In clinical usage, imipenem and meropenem have to be given frequently to maintain high circulating levels. Also, weight-dosage adjustment of imipenem is required to minimize the chance of seizures (8). Ertapenem and doripenem can be given once per day due to their high target affinity and circulating stability (5,9). The lower effective doses of these latter drugs reduces potential side effects, as well as the development of resistance (10). Currently, ertapenem and doripenem are used for complicated intra-abdominal, and urinary tract infections (11,12). Despite the general success of β-lactam antibiotics, they have not been widely used for the treatment of TB due to intrinsic resistance that is caused by the presence of a chromosomally-encoded gene (blaC) in M. tuberculosis for a Class A Ambler β-lactamase (BlaC). Like other Class A β-lactamases, BlaC catalyzes the opening of the β- lactam ring via nucleophilic attack by an active site serine residue to generate the acylenzyme, followed by the hydrolysis of the ester bond to generate the ring-opened, inactive product. Our previous studies of TB BlaC revealed that this enzyme is an extremely broad-spectrum β- lactamase hydrolyzing all β-lactam classes, including the carbapenems meropenem and imipenem (13). Being slow substrates that exhibit rapid acylation followed by a slow deacylation step, meropenem and imipenem also act as potent inhibitors of BlaC (14). FTICR mass spectrometry demonstrated that the acylated intermediate remains stable for many minutes (14). Such slow turnover rates allowed the determination of three- dimensional structure of BlaC in complex with meropenem at a resolution of 1.8 Å. In vivo studies showed that meropenem in combination with the β-lactamase inhibitor, clavulante, is bactericidal against clinical TB strains that are phenotypically exensively drug resistant (XDR-TB) (14). As an extension of our prior work, we carried out an in vitro characterization of doripenem and ertapenem with BlaC. Materials and Methods All chromatographic materials were purchased from Pharmacia. Meropenem and faropenem were from IKT Laboratories. Doripenem (as Doribax) was from Ortho-McNeil Pharmaceutical Inc (Raritan, NJ). Ertapenem (as Invanz) was from Merck & Co. Inc. The potassium salt of clavulanic acid was from Sigma Aldrich. All other chemicals were purchased from Sigma or Aldrich. Nitrocefin was purchased from Beckton Dickinson. Purification of BlaC Recombinant and truncated BlaC from M. tuberculosis expressed from plasmid pET28a(+) and purified to homogeneity as described by Hugonnet and Blanchard (13). Tremblay et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Kinetics The steady state rate of hydrolysis of β-lactam ring was monitored as a decrease in the absorbance in the UV region, as described previously (13). Assays using doripenem, ertapenem, faropenem and meropenem were performed at 296 nm (ε = 7,540 M−1 cm−1), 295 nm (ε = 9,970 M−1 cm−1), 306 nm (ε = 3,445 M−1 cm−1), and 297 nm (ε = 6,152 M−1 cm−1), respectively. Assays using the chromogenic substrate nitrocefin were performed at 486 nm (ε = 20,500 M−1 cm−1). Assays were performed in 100 mM MES (pH 6.5). Reactions were initiated by the addition of enzyme at concentrations between 0.1–25 µM using 100 µM of the carbapenem substrate. Inhibition Studies Carbapenems at concentrations ranging from 0.1–10 µM were tested as inhibitors of 1.5 nM BlaC using 60 µM nitrocefin as substrate. Time courses were followed for 15 min. For slow onset inhibition, reaction velocities as a function of time were fitted to eq 1: (1) where [P] is the concentration of the product, vi and vs are the initial and final reaction velocities respectively for the reaction in the presence of inhibitor and kiso is the apparent first order rate constant for the inter-conversion between vi and vs, and t is time. The general mechanism can be modeled as: (2) where k1 and k−1 represent the reversible binding to and dissociation from the carbapenem to BlaC, k2 represents the irreversible cleavage of the carbapenem β-lactam ring and k3 represents the hydrolysis of the BlaC-carbapenem adduct. For this model, the rate constant that describes kiso is given by eq 3, where Kd equals k−1/k1. (3) In eq 4, the Km value can be expressed as: (4) In addition, from the determined k2 and k3 values, kcat is calculated from eq 7, assuming k2,k3≪k1,k−1. (5) Tremblay et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Mass Spectrometry All mass spectra were acquired on a 9.6 T Fourier Transform Ion Cyclotron Resonance (FTICR) mass spectrometer (Ionspec, Lake Forest, CA). To avoid salt interference, BlaC was dialyzed against 20 mM ammonium bicarbonate, pH 6.5. The molecular mass of each protein sample was determined for the 25+ charge state using the equation m = (m/z x 25)−25 on the isotopic centroid. To monitor the intermediate of steady state turnover or small molecular mass spectrometry, 51 µM of enzyme was incubated with 25 µM carbapenem in a total volume of 20 µL. An aliquot of 1 µL was withdrawn at desired time (0, 30, 60, and 90 min) and mixed with 9 µL of mixing solution (containing 50% acetonitrile and 0.1% formic acid). The resulting mixture was injected into the FTICR mass spectrometer. Crystallization BlaC was crystallized in the hanging drop vapor diffusion configuration over well conditions of 0.1 M HEPES, pH 7.5 and 2 M NH4H2PO4. The final pH of the well solution was 4.1. Protein at a concentration of 10 mg/ml was mixed 1:1 with the well solution and incubated at 18 °C. Initial crystals grew within a week but were small, sparse and amorphous. New wells were sealed and allowed to equilibrate overnight. Equilibrated drops were micro-seeded, which resulted in efficient crystal growth as well as improved morphology. Iterative seeding resulted in diffraction quality crystals of active enzyme. Data collection and refinement Crystals were soaked with either ~ 50 mM ertapenem or doripenem in mother liquor plus 20% glycerol as a cryo-protectant. Data were collected after 10 and 90 minute soaks with ertapenem and a 90 minute soak with doripenem at Brookhaven National Laboratory on beamlines X12C and X29, in which various resolutions of diffraction were obtained dependent on the soaking times and beamline. The data were processed using either HKL2000 (15) or Mosflm (16). Our previous structure of clavulanate bound M. tuberculosis β-lactamase (17) (PDB entry 3CG5) was used to phase all the data, using the CCP4 software suite (18). Iterative rounds of structural refinement and model building were performed in Refmac5 (19,20) and Coot (21). Table 1 lists the data collection statistics for the structures as well as the final refinement statistics. RESULTS and DISCUSSION Kinetics The accurate determination of the kinetic parameters for doripenem and ertapenem was severely hampered by apparent very low Km values, very low kcat values and the modest extinction coefficients accompanying hydrolysis. At the [BlaC] required to see any significant rate of reaction (~2 µM), variation of the [doripenem] or [ertapenem] at concentrations from 2–20 µM showed almost no difference in rate, suggesting their Km values were less than 2 µM. The steady-state kinetic parameters determined for faropenem, a structurally distinct penem, were Km = 55 ± 11 µM, and kcat = 0.65 ± 0.04 min−1 (data not shown). This Km value is ~17 times larger and the kcat value 8 times faster than those of meropenem (14). Detailed investigations of the kinetics of carbapenem hydrolysis under near stoichiometric enzyme concentrations were carried out over 30 minute time periods. As shown in Figure 1, a steady-state kinetic burst was observed with both compounds where the magnitudes of the burst are proportional to the concentration of BlaC used. Extrapolation of the rates of hydrolysis to the y-axis demonstrates that the acylation is stoichiometric with the concentration of enzyme. Tremblay et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Due to the extremely feeble turnover rate, we further tested these carbapenems as inhibitors of the reaction of nitrocefin with BlaC. Nitrocefin is an extremely good substrate for BlaC and its β-lactam ring-opened form is extremely chromogenic. As shown in Figure 2, doripenem and ertapenem act as slow-onset, tight binding inhibitors of BlaC when the hydrolysis of nitrocefin was monitored. In contrast, faropenem exhibited standard, competitive inhibition with no time dependent component (data not shown). This type of time-dependent inhibition for a dead-end inhibitor is modeled as being due to the reversible formation of a non-covalent complex (E–I), followed by the reversible conversion to an isomerized complex (E–I*). However, in the case of a slow substrate for BlaC, the initially formed Michaelis complex reacts with the enzyme in an irreversible step to generate the BlaC-carbapenem covalent intermediate. This is then hydrolyzed slowly to regenerate the free enzyme that can react with nitrocefin. While the same equation is used to fit the two models, the kinetic constants that contribute to kiso (Figure S1) and Ki (or Kd) are different. Using the fits of the slow-onset data and eq 3 to calculate Kd, k2 and k3, we can then used eqs 4 and 5 to calculate the apparent Km and kcat values for doripenem (0.18 µM and 0.016 min−1, respectively), and for ertapenem (0.18 µM and 0.017 min−1, respectively). We have not corrected for the concentration of nitrocefin used in these experiments because of the large standard errors (>40%) associated with these kinetic parameters (the reported Km values are apparent values). However, the extremely tight binding and extremely low turnover of these carbapenems is evident from these rather imprecise kinetic data. Mass Spectrometry The rapid acylation and slow deacylation of BlaC by the carbapenems allows the observation of the covalently bound, acyl-enzyme intermediate by Fourier transform ion cyclotron resonance. A freshly prepared solution containing excess BlaC and doripenem displayed three peaks: the first peak corresponds to free BlaC with mass/charge ratio (m/z) = 28,785.0, a second peak corresponding to the covalently acylated BlaC-doripenem complex with mass/charge ratio (m/z) = 29, 204.1 and a third peak whose mass corresponds to the mass of the covalently acylated BlaC-doripenem complex minus 44 mass unit (m/z = 29, 161.0), as shown in Figure 3. With ertapenem, the two covalent acylated BlaC complex peaks observed had molecular masses of 29,260.0 and 29,217.1, corresponding to acylated BlaC-ertapenem complex and acylated BlaC-doripenem complete minus 44 mass units, respectively. This data demonstrates that both doripenem and ertapenem undergo the same chemical breakdown in the active site as meropenem (14). Once the acyl-enzyme forms, the carbapenems partition between hydrolysis and enzyme-catalyzed decomposition of the C6 hydroxyethyl substituent, via a retro-Aldol decomposition, which yields acetaldehyde (14). Intriguingly, the intensities of the acyl-enzyme complexes remain stable over the time period of 90 min for doripenem and ertapenem. This is in contrast with previous observations with meropenem, where the acylated forms of the enzyme started to diminish after several minutes. These data suggest that doripenem and ertapenem form more stable complexes with BlaC than meropenem, reinforcing the kinetic data. X-ray Crystallography The 2.2 Å data from a 90 min doripenem-soaked crystal were refined to an Rwork of 0.161 and an Rfree of 0.205. The 1.3 Å diffraction data from a 10 minute ertapenem-soaked crystal refined to an Rwork of 0.147 and an Rfree of 0.176. The 2.0 Å diffraction data from a 90 min ertapenem-soaked crystal were refined to an Rwork of 0.175 and an Rfree of 0.222. In these three structures, the active site Ambler residue Ser70 has been covalently linked with the ring open form of these β-lactams in accordance with the acylation chemistry of the first half of the enzymatic reaction (Scheme 1). The quality of the electron density is displayed in Figure 4 and Figure 5 under a Fo-Fc omit calculated map contoured at 2.0 σ. Tremblay et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The C3 atom of the pyrroline rings of doripenem and ertapenem covalent adducts are sp3 hybridized in the 90 min soaks. These results require an isomerization event occurring in the BlaC-carbapenem adducts in which the original Δ2 pyrroline ring was tautomerized to generate the Δ1 pyrroline ring, evidenced by the collinear positioning of the C5, N1, C2, and C3 atoms. In addition, the BlaC-doripenem and ertapenem covalent adduct densities allow for the positioning of the thioether sulfur atom in the unambiguous assignment of the S configuration at C3, requiring protonation at the re face of the C2–C3 double bond. This is similar to our earlier findings with BlaC crystals soaked with meropenem on similar time scales (14), and represents the thermodynamically preferred product with the trans orientation of the C4 methyl and thioether substituents. Interestingly, in the recently reported structure of the Class D β-lactamase, OXA-1 covalent adduct with doripenem, revealed that while an identical isomerization had taken place, that the final Δ1 pyrroline ring product was of the opposite, R, stereochemistry (22). In the structure determined after the shorter 10-minute soak, the BlaC-ertapenem adduct was covalently bound in the active site, but in a different geometry. On this shorter time scale, the BlaC-ertapenem adduct C3 atom is found in its original sp2 hybridization with the definitive collinear positioning of the thioether sulfer atom in line with the N1, C2, C3, and C4 bonds indicating the presence of the Δ2 pyrroline ring. This result requires that β-lactam ring cleavage and isomerization of the methyl pyrroline ring not be concerted. The active site interactions vary in some subtle ways between the pre and post-isomerization ertapenem complexes, yet a number of common interactions are observed in all complexes. Both the BlaC-ertapenem and -doripenem adducts bind as covalent adducts with the active site Ser70 and position their lactam ring-opened ester carbonyl oxygen atom within the oxyanion hole formed from hydrogen bonding interactions with the Ser70 and Thr253 amide nitrogen atoms. All structures contain a hydrophobic interaction between the methyl group of the pyrroline ring and the sidechain of Ile117 and different forms of hydrogen bonding interactions between the C6 hydroxyethyl substituent of the carbapenem and Glu166. All three structures also show a conserved interaction between the sidechain hydroxyl of Ser130, which consistently hydrogen bonds the pyrroline ring nitrogen atom at a distance of 2.7–2.8 Å. The pyrroline C2 carboxylate group forms hydrogen bonds with the Thr251 hydroxyethyl side chain and an active site water molecule. In structures of other β- lactamase-carbapenem adducts, this carboxylate electrostatically interacts with a conserved arginine residue (R244 in TEM-1) (23), but this is not the case for BlaC. In the pre- isomerized ertapenem structure, there is an additional hydrogen bond between the C2 carboxylate of the pyrroline ring and Thr253, which is broken upon the repositioning of the meta-amino-benzoate ‘arm’ observed in the post-isomerized ertapenem structure. The isomerization and stereospecific protonation leads to a reorientation of the terminal portion of the molecule within the active site, allowing for the formation of the hydrogen bond between the terminal carboxylate group of the meta-amino-benzoic acid moiety and Ser118. A final difference between the initially formed Δ2-pyrroline isomer and the final Δ1- pyrroline form is the orientation of the C6 hydroxyethyl substituent. In the pre-isomerized complex it is oriented away from Lys73 and hydrogen bonds to the carboxyl O1 of Glu166 as well as the Asn186 nitrogen, but rotates upon isomerization to hydrogen bond with the carboxyl O2 of Glu166 and the ε-amino group of Lys73 in a manner similar to that observed for the doripenem complex. The structures of the pre and post-isomerization states reveal the mechanistic basis for the relative stability of the carbapenems within the active site of BlaC and their ability to resist hydrolysis by the enzyme. As seen in Scheme1, deacylation-hydrolysis from the enzyme requires the activation of the conserved active site water by the side chain carboxyl O2 of Glu166. The probability of water activation by Glu166 decreases with the increased distance Tremblay et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript between the two. The distance between the carboxyl O2 of Glu166 and the active site water is significantly increased after isomerization from 2.4 Å to 2.7 Å, making water activation less probable. The isomerization event does not directly cause these changes but rather alters the positioning of the adduct such that the adduct-hydroxyethyl flips in the active site, breaking the hydrogen bond formed between the hydroxyl group with the side chain amide nitrogen of Asn186 and the adjacent carboxyl O1 of Glu166 oxygen. The hydroxyethyl substituent then rotates to generate a hydrogen bond network with Lys73 and the carboxyl O2 of Glu166, effectively ‘pulling’ this essential base away from the conserved active site water. These residues (Lys 73 and Glu166) are involved as general bases in the acylation and deacylation reactions, respectively. In addition, by hydrogen bonding carboxyl O2 of Glu166, the reoriented hydroxyethyl substituent reduced the basicity of Glu166. These two factors, introduced in the reorientation of the hydroxyethyl substituent in the post- isomerization complex, reduce water activation and thereby stabilize the acyl-intermediate in the active site. The studies reported here allow us to directly visualize the changes that occur between the pre- and post-isomerization adduct structures and are atomic level observations relevant to the biphasic kinetics previously reported for the reactions between carbapenems and the RTEM β-lactamase (24). The Δ2 to Δ1-pyrroline isomerized forms of carbapenems have been known to form within the active sites of various β-lactamase enzymes (25). In confirmation of this, crystal structures of carbapenems bound within the active sites of the Class A β-lactamases TEM-1 (PDB entry 1BT5) (26) and SHV-1 (PDB entry 2ZD8) (27) as well as AmpC (PDB entry 1LL5) (28) a Class C β-lactamase all revealed the Δ2 form of the carbapenem bound in the active sites, while the Class D OXA-1 (PDB entry 3ISG) (22) and the Class A BlaC (PDB entry 3DWZ) (14) were both bound with carbapenems in the Δ1 isomerized forms with respective R and S-stereochemistries. Our findings are the first to show the structures of both the Δ2 and Δ1 forms of a carbapenem bound to a single β-lactamase. Interestingly, several of the structures of carbapenems bound as the Δ2 isomers show evidence for alternate conformations for the carbapenem-carbonyl oxygen position. This oxygen is found buried within the oxyanion-hole as well as bound in a position rotated by 180 degrees, usually facing an opposing serine residue (Ser130). In these instances it has been proposed that the flipping of the carbonyl oxygen from the oxyanion-hole blocks formation of the deacylation tetrahedral intermediate to inhibit the enzyme. In the cases of OXA-1 and BlaC, the carbapenem-carbonyl oxygen is only found bound tightly within the oxyanion-hole and no evidence of alternate conformers has been observed. In these cases inhibition by the carbapenem is likely due to disruption of water activation. A second possible reason for the observed alternate conformers at the carbapenem-carbonyl is likely due to the position of the carbapenem-carboxylate moiety within those active sites. To date, those β-lactamases with alternate conformations for the carbapenem-carbonyl, have a highly conserved Arg244 reside which electrostatically interacts with the carbapenem- carboxylate moiety. The OXA-1 and BlaC active sites lack this arginine interaction and instead use a combination of threonine and/or serine residues coordinated with waters to bind the carboxylate moiety. These residues are located closer to the oxyanion hole and act to ‘clamp’ the carboxylate into a proximal position, as opposed to the Arg244 mechanism of carboxylate binding, where distance introduces flexibility, allowing for the alternate positioning of the pyrroline ring. This bonding pattern to the carbapenem allows for alternate “in/out” conformations of the carbapenem carbonyl in the oxyanion hole. In contrast, the carbapenem carbonyl is tightly bound in the oxyanion hole in BlaC in both the Δ2 and Δ1 forms reported here. Tremblay et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments This work was supported in part by a grant from the National Institute of Health (AI33696 to J. S. B.) and in part by the Charles Revson Foundation (to L.W.T.) ABBREVIATIONS BlaC Mycobacterium tuberculosis beta-lactamase TB Tuberculosis XDR-TB extensively drug resistant References 1. Dye, C.; Floyd, K.; Uplekar, M.; Bierrenbach, A.; Bergstrom, K.; Blanc, L.; Grezmska, M.; Gunneberg, C.; Lonnroth, K.; Nunn, P.; Pantoja, A.; Raviglione, S.; Weyer, K. WHO report 2008. World Health Organization; 2008. Global tuberculosis control : surveillance, planning, financing. 2. Bauquerez, R.; Blanc, L.; Bierrenbach, A.; Brands, A.; Ciceri, K.; Falzon, D.; Floyd, K.; Glaziou, P.; Gunneberg, C.; Hiatt, T.; Hosseini, M.; Pantoja, A.; Uplekar, M.; Watt, C.; Wright, A. WHO Report 2009. World Health Organization; 2009. Global Tuberculosis Control: Epidemiology, Strategy, Financing. 3. Netto EM, Dye C, Raviglione MC. Progress in global tuberculosis control 1995–1996, with emphasis on 22 high-incidence countries. Global Monitoring and Surveillance Project. Int J Tuberc Lung Dis 1999;3:310–320. [PubMed: 10206501] 4. Mandell L. Doripenem: a new carbapenem in the treatment of nosocomial infection. Clin Infect Dis 2009;49(Suppl 1):S1–S3. [PubMed: 19619016] 5. Baughman RP. The use of carbapenems in the treatment of serious infections. J Intensive Care Med 2009;24:230–241. [PubMed: 19617229] 6. Birnbaum J, Kahan FM, Kropp H, MacDonald JS. Carbapenems, a new class of beta-lactam antibiotics. Discovery and development of imipenem/cilastatin. Am J Med 1985;78:3–21. [PubMed: 3859213] 7. Livermore DM. Of Pseudomonas, porins, pumps and carbapenems. J Antimicrob Chemother 2001;47:247–250. [PubMed: 11222556] 8. Calandra G, Lydick E, Carrigan J, Weiss L, Guess H. Factors predisposing to seizures in seriously ill infected patients receiving antibiotics: experience with imipenem/cilastatin. Am J Med 1988;84:911–918. [PubMed: 3284342] 9. Bhavnani SM, Hammel JP, Cirincione BB, Wikler MA, Ambrose PG. Use of pharmacokinetic- pharmacodynamic target attainment analyses to support phase 2 and 3 dosing strategies for doripenem. Antimicrob Agents Chemother 2005;49:3944–3947. [PubMed: 16127078] 10. Lynch MJ, Drusano GL, Mobley HL. Emergence of resistance to imipenem in Pseudomonas aeruginosa. Antimicrob Agents Chemother 1987;31:1892–1896. [PubMed: 3125787] 11. Behera B, Mathur P, Das A, Kapil A. Ertapenem susceptibility of extended spectrum beta- lactamase-producing Enterobacteriaceae at a tertiary care centre in India. Singapore Med J 2009;50:628–632. [PubMed: 19551319] 12. Paterson DL, Depestel DD. Doripenem. Clin Infect Dis 2009;49:291–298. [PubMed: 19527173] 13. Hugonnet JE, Blanchard JS. Irreversible inhibition of the Mycobacterium tuberculosis beta- lactamase by clavulanate. Biochemistry 2007;46:11998–12004. [PubMed: 17915954] 14. Hugonnet JE, Tremblay LW, Boshoff HI, Barry CE 3rd, Blanchard JS. Meropenem-clavulanate is effective against extensively drug-resistant Mycobacterium tuberculosis. Science 2009;323:1215– 1218. [PubMed: 19251630] Tremblay et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 15. Otwinowski Z, a MW. Processing of X-ray Diffraction Data Collected in Oscillation Mode. Methods in Enzymology 1997;276:307–326. 16. Leslie AGW. Mosflm. Joint CCP4 + ESF-EAMCB Newsletter on Protein Crystallography No. 26. 1992 17. Tremblay LW, Hugonnet JE, Blanchard JS. Structure of the covalent adduct formed between Mycobacterium tuberculosis beta-lactamase and clavulanate. Biochemistry 2008;47:5312–5316. [PubMed: 18422342] 18. Potterton EBP, Turkenburg M, Dodson E. A graphical user interface to the CCP4 program suite. Acta. Cryst 2003;D59:1131–1137. 19. Murshudov GN, Vagin AA, Dodson EJ. Refinement of Macromolecular Structures by the Maximum-Likelihood Method. Acta Cryst 1997;D53:240–255. 20. Pannu NS, Murshudov GN, Dodson EJ, Read RJ. Incorporation of prior phase information strengthens maximum-likelihood structure refinement. Acta Crystallogr D Biol Crystallogr 1998;54:1285–1294. [PubMed: 10089505] 21. Emsley PCK. Coot: model-building tools for molecular graphics. Acta Crystallogr 2004;D60:2126–2132. 22. Schneider KD, Karpen ME, Bonomo RA, Leonard DA, Powers RA. The 1.4 A crystal structure of the class D beta-lactamase OXA-1 complexed with doripenem. Biochemistry 2009;48:11840– 11847. [PubMed: 19919101] 23. Zafaralla G, Manavathu EK, Lerner SA, Mobashery S. Elucidation of the role of arginine-244 in the turnover processes of class A beta-lactamases. Biochemistry 1992;31:3847–3852. [PubMed: 1567841] 24. Easton CJ, Knowles J. Inhibition of the RTEM beta-lactamase from Escherichia coli. Interaction of the enzyme with derivatives of olivanic acid. Biochemistry 1982;21(12):2857–2862. [PubMed: 7049231] 25. Kalp M, Carey PR. Carbapenems and SHV-1 beta-lactamase form different acyl-enzyme populations in crystals and solution. Biochemistry 2008;47:11830–11837. [PubMed: 18922024] 26. Maveyraud L, Mourey L, Kotra LP, Pedelac J, Guillet V, Mobashery S, Samama J. Structural Basis for Clinical Longevity of Carbapenem Antibiotics in the Face of Challenge by the Common Class A β-Lactamases from the Antibiotic-Resistant Bacteria. Journal of the American Chemical Society 1998;120:9748–9752. 27. Nukaga M, Bethel CR, Thomson JM, Hujer AM, Distler A, Anderson VE, Knox JR, Bonomo RA. Inhibition of class A beta-lactamases by carbapenems: crystallographic observation of two conformations of meropenem in SHV-1. J Am Chem Soc 2008;130:12656–12662. [PubMed: 18761444] 28. Beadle BM, Shoichet BK. Structural basis for imipenem inhibition of class C beta-lactamases. Antimicrob Agents Chemother 2002;46:3978–3980. [PubMed: 12435704] Tremblay et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Time courses of doripenem (A) and ertapenem (B) hydrolysis with various concentrations of BlaC. Tremblay et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Time courses of nitrocefin hydrolysis by BlaC in the presence of doripenem (upper) and ertapenem (lower). Tremblay et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Mass spectra of enzyme-carbapenem species. The 25+ charge state ions are shown. Tremblay et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. (A) Overall structure of BlaC displayed in rainbow from N term (blue) to the C term (red), with the doripenem adduct displayed in red surface mesh. (B) Fo-Fc omit density (green) contoured at 2.0 σ surrounds the covalent doripenem adduct formed at the Ambler active- site residue serine 70. All structure figures were produced using Pymol. Tremblay et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. (A) Fo-Fc omit density (green) contoured at 2.0 σ surrounds the covalent ertapenem adduct formed at the Ambler active-site residue serine 70 in the pre-isomerization state. (B) Fo-Fc omit density (green) contoured at 2.0 σ surrounds the covalent ertapenem adduct formed at the Ambler active-site residue serine 70 in the post-isomerization state. The resolution of the densities unambiguously demonstrates the shift in stereochemistry with the change from sp2 to sp3 hybridization of the C3 carbapenem carbon atom with the change in the position of the density associated with the meta-amino-benzoate and the hydoxyethyl ertapenem moieties. Tremblay et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. (A) the structures of doripenem and ertapenem. (B) The chemical mechanism of hydrolysis of ertapenem by the Mycobacterium tuberculosis BlaC. Tremblay et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Tremblay et al. Page 16 Table 1 Data Collection and Refinement Statistics Data Collection Doripenem Δ1-isomerζ Ertapenem Δ2-isomer Ertapenem Δ1-isomer Resolution (Å) 50.0-2.2 (2.32-2.20) 50.0-1.30 (1.33-1.30) 50.0-2.0 (2.07-2.00) Completeness 100% (100%) 100.0% (100%) 99.5 (99.9) Redundancy 7.6 (7.4) 7.5 (5.7) 4.4 (4.4) I/sigma(I) 3.8 (1.6) 21.4 (1.8) 9.8 (4.0) Rmerge 0.077 (0.47) 0.057 (0.757) 0.158 (0.373) Space Group P212121 P212121 P212121 Unit cell (Å) a =49.989 b =68.068 c =75.792 α = β = γ = 90.0° a = 49.66 b = 67.92 c = 75.55 α = β = γ = 90.0° a =49.934 b =67.830 c =75.201 α = β = γ = 90.0° Reflections 13,695 (1,943) 60,263 (4,388) 17,920 (1,762) Refinement Statistics Rwork 0.161 (0.176) 0.147 (0.265) 0.175 (0.191) Rfree 0.205 (0.237) 0.176 (0.278) 0.222 (0.281) Average B-factors (Å2) Protein 6.97 10.49 6.09 Adduct 27.32 18.64 15.50 Solvent 17.51 32.64 14.36 PO4 12.89 14.40 10.53 RMS deviations bonds (Å) 0.010 0.010 0.012 angles (°) 1.204 1.428 1.386 Ramachandra Favored= 97.7% outliers= 0.0% Favored= 97.7% outliers= 0.0% Favored= 98.1% outliers= 0.0% PDB accession code 3IQA 3M6B 3M6H Values in parentheses are for the highest resolution bin. ζThis data processed using Mosflm Biochemistry. Author manuscript; available in PMC 2011 May 4.
3M6C
Crystal structure of Mycobacterium tuberculosis GroEL1 apical domain
Structural and Functional Conservation of Mycobacterium tuberculosis GroEL Paralogs Suggests that GroEL1 is a Chaperonin Bernhard Sielaff*, Ki Seog Lee*,2, and Francis T.F. Tsai1 Verna and Marrs McLean Department of Biochemistry and Molecular Biology, and Department of Molecular and Cellular Biology, Baylor College of Medicine, One Baylor Plaza, Houston, Texas 77030, USA Abstract GroEL is a group I chaperonin that facilitates protein folding and prevents protein aggregation in the bacterial cytosol. Mycobacteria are unusual in encoding two or more copies of GroEL in their genome. While GroEL2 is essential for viability and likely functions as the general housekeeping chaperonin, GroEL1 is dispensable but its structure and function remain unclear. Here we present the 2.2 Å resolution crystal structure of a 23 kDa fragment of Mycobacterium tuberculosis GroEL1 consisting of an extended apical domain. Our X-ray structure of the GroEL1 apical domain closely resembles those of Escherichia coli GroEL and M. tuberculosis GroEL2; thus, highlighting the remarkable structural conservation of bacterial chaperonins. Notably, in our structure, the proposed substrate-binding site of GroEL1 interacts with the N-terminal region of a symmetry related, neighboring GroEL1 molecule. The latter is consistent with the known GroEL apical domain function in substrate binding, and is supported by results obtained from using peptide array technology. Taken together, we show that the apical domains of M. tuberculosis GroEL paralogs are conserved in three-dimensional structure, suggesting that GroEL1, like GroEL2, is a chaperonin. Keywords Molecular chaperones; Hsp60; apical domain; protein folding; KasA Introduction Molecular chaperones assist protein folding by facilitating the productive folding of newly synthesized polypeptides and by preventing protein aggregation in the crowded environment © 2010 Elsevier Ltd. All rights reserved. 1Corresponding author: Department of Biochemistry, Baylor College of Medicine, One Baylor Plaza, MS: BCM125, Houston, Texas 77030. ftsai@bcm.edu; Phone: +1-713-798-8668; Fax: +1-713-796-9436. *These authors contributed equally to this study. 2Present address: Department of Clinical Laboratory Science, College of Health Science, Catholic University of Pusan, Pusan 609-757, Republic of Korea. Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Author contributions B.S. designed and performed research, analyzed data, and wrote the paper. K.S.L. performed research, analyzed data, and wrote the paper. F.T.FT. designed research, analyzed data, and wrote the paper. NIH Public Access Author Manuscript J Mol Biol. Author manuscript; available in PMC 2012 January 21. Published in final edited form as: J Mol Biol. 2011 January 21; 405(3): 831–839. doi:10.1016/j.jmb.2010.11.021. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript of the cell.1 Escherichia coli GroEL is a group I chaperonin that assembles into an 800 kDa homo-tetradecamer composed of two heptameric rings that are stacked back-to-back.2; 3 Each GroEL subunit has a molecular weight of 57 kDa and consists of an equatorial, an intermediate, and an apical domain.3 The equatorial domain contains the ATP-binding site and mediates contacts between subunits in the cis and trans rings. The intermediate domain functions as a hinge that connects the equatorial domain to the apical domain. The latter forms the entrance to the GroEL cavity and is involved in GroES binding4 as well as polypeptide recognition.5; 6 It has been suggested from X-ray crystallographic studies that helices H and I of the E. coli GroEL apical domain form the substrate binding site.7; 8; 9 Interestingly, many mycobacteria contain genes encoding two or more GroEL paralogs.10 GroEL1 and GroEL2 from the human pathogen Mycobacterium tuberculosis11; 12 are of particular interest as both proteins are involved in the host immune response to M. tuberculosis infection.13; 14 While GroEL2 is essential and likely functions as the principal housekeeping chaperonin,10; 11 GroEL1 is non-essential and is dispensable for viability. It has been proposed that M. tuberculosis GroEL1 is a nucleoid-associated protein,15 and that the closely related GroEL1 ortholog from M. smegmatis plays a role in biofilm formation by modulating mycolic acid biosynthesis through direct interaction with the β-ketoacyl ACP synthase KasA.10 Like other bacterial chaperonins, M. tuberculosis GroEL1 and GroEL2 are up-regulated upon heat shock16 as well as in response to oxidative stress,17 indicating that both copies may have chaperone activity inside cells. In contrast, recombinant GroEL1 and GroEL2 overexpressed in E. coli exist as dimers, and exhibit low ATPase and no folding activities.18 Since native GroEL1 forms higher-order oligomers in M. tuberculosis cells,19 lack of chaperone activity might be attributed to the inability of the recombinant proteins to self- assemble. Consistent with its essential cellular role, the X-ray structure of a M. tuberculosis GroEL2 dimer20 showed that the GroEL2 monomer has the same fold as E. coli GroEL,20 supporting the notion that GroEL2 is a chaperonin. However, at present, no high-resolution structural information is available for M. tuberculosis GroEL1, and its structure-function relationship remains unclear. Here we present the 2.2 Å resolution crystal structure of a 23 kDa M. tuberculosis GroEL1 fragment consisting of the GroEL1 apical domain flanked by flexible segments that are part of the intermediate domain. This structure is hereafter referred to as the GroEL1 apical domain. We found that the atomic structure of the GroEL1 apical domain is very similar to those of M. tuberculosis GroEL220 and E. coli GroEL.7; 8 Fortuitously, in our crystal structure, the N-terminus of one molecule interacts with the putative GroEL substrate- binding site of a symmetry related molecule. This interaction is reminiscent of the X-ray structures of E. coli chaperonin-substrate peptide complexes.7; 8; 9 Moreover, we found using peptide array technology that both full-length M. tuberculosis GroEL1 and the isolated GroEL1 apical domain recognize the same peptide motifs present in the M. tuberculosis KasA sequence, which resemble binding motifs reported for E. coli GroEL.21 Thus, our combined structural and functional data suggest that M. tuberculosis GroEL1, like GroEL2, is a chaperonin and support the notion that the apical domain is sufficient for substrate interaction. Sielaff et al. Page 2 J Mol Biol. Author manuscript; available in PMC 2012 January 21. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Results and Discussion Crystal Structure of the M. tuberculosis GroEL1 Apical Domain Crystals of the GroEL1 apical domain (residues 184–377) diffracted to 2.2 Å resolution on a home X-ray source, and belonged to the orthorhombic space group P21212 with one molecule in the asymmetric unit. After rigid body refinement of the initial model in CNS22 the resulting electron density map was readily interpretable. Several rounds of model building, simulated annealing, least-squares positional, and individual B-factor refinement were performed resulting in a final model with a Rfactor and Rfree 23 of 21.0% and 23.2%, respectively. The final model consists of 194 residues and 65 water molecules. The Ramachandran plot showed that 91.1% of all residues were in the most favored regions and 8.9% in additional allowed regions. No residue was in a generously allowed or disallowed region. The refinement statistics are summarized in Table 1. The structure of the M. tuberculosis GroEL1 apical domain consists of a β-sandwich scaffold flanked by several α-helices and loops (Fig. 1a and b). Structural comparison of the M. tuberculosis GroEL1 apical domain with those of M. tuberculosis GroEL2 (PDB ID: 1SJP-A)20 and E. coli GroEL (PDB ID: 1KID and 1DKD-A)7; 8 showed that they are very similar (Fig. 1c). The Cα atoms of the refined M. tuberculosis GroEL1 apical domain structure (residues 191–372) superimposed pairwise with those of M. tuberculosis GroEL2 residues 190–371 (PDB ID: 1SJP-A) 20 with a root mean square deviation (RMSD) of 0.82 Å, and with those of E. coli GroEL residues 193–375 (PDB ID: 1KID)7 and residues 193– 336 (PDB ID: 1DKD-A)8 with a RMSD of 0.63 Å and 0.75 Å, respectively. Protein-protein interactions between symmetry related GroEL1 molecules Our crystal structure showed two types of protein interfaces with crystallographic symmetry related GroEL1 apical domain molecules in the crystal lattice. While the observed contacts may hint at the existence of higher oligomers, none of the observed interactions are consistent with the known interfaces in the E. coli GroEL tetradecamer structure.2; 3 It has been proposed that both recombinant, full-length M. tuberculosis GroEL1 and GroEL2 form dimers18 through inter-subunit contacts between apical domains.20 While our biochemical results support the notion that recombinantly expressed, full-length GroEL1 is a dimer, we found that the isolated GroEL1 apical domain is a monomer in solution (Fig. 1d). In our structure, the N-terminal residues 184 to 187 with sequence Glu-Leu-Glu-Phe interact with helices H and I of a symmetry-related neighboring molecule (Fig. 2a and b). Notably, the Leu185 side-chain is in van der Waals contact with the side-chains of Leu232, Leu235, Ala239, and Leu269, which form a hydrophobic pocket (Fig. 2b and 3a). In addition, there is a network of hydrogen bond interactions between the polar side chains of Asn263 and Arg266 and the main chain carbonyl oxygens of Glu184, Leu185, and Glu186 (Fig. 2b and 3a). These interactions are reminiscent of the previously observed E. coli GroEL apical domain contacts with different peptides (Fig. 3b–c).7; 8; 9 GroEL1 substrate binding site It has been proposed that the apical domain of E. coli GroEL contains the substrate binding site.5; 6 This is supported by crystal structures of the E. coli GroEL apical domain bound to a strong binding peptide selected by phage-display,8 and to an extended N-terminal segment of a symmetry related molecule,7 respectively. In the latter two structures, protein-protein interactions were mediated by both non-polar and hydrogen bond interactions with helices H and I of the GroEL apical domain.7; 8 Although the sequence of the bound peptides differed in each case, the two previously reported X-ray structures of the E. coli GroEL apical domain7; 8 and our atomic structure of the M. tuberculosis GroEL1 apical domain display Sielaff et al. Page 3 J Mol Biol. Author manuscript; available in PMC 2012 January 21. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript common binding features. First, the hydrophobic pocket is lined by the conserved residues Leu232, Leu235, and Ala239 from helix H and the less conserved Leu269, which correspond to Leu234, Leu237, Ala241, and Val271 in E. coli GroEL (Fig. 3a–c). Second, hydrogen bond interactions mainly involve residues from helix I (Fig. 3a–c). Third, hydrogen bonds are exclusively formed between the side chains of GroEL and the main chain of the bound peptide (Fig. 3a–c). Notably, in our structure, only four residues (Glu- Leu-Glu-Phe) contact the substrate binding site of the apical domain, which may represent the minimal binding region. GroEL1-KasA peptide interaction To gain further insight into substrate recognition by M. tuberculosis GroEL1, we synthesized a miniaturized peptide array of overlapping 12-mer peptides derived from the M. tuberculosis KasA amino acid sequence. This work followed from an earlier report demonstrating a direct interaction between KasA and GroEL1 in M. smegmatis cells,10 which serves as a model organism for different mycobacterial species, including M. tuberculosis. Probing our KasA-derived peptide array with full-length GroEL1 showed binding to 67 out of 192 peptides (Fig. 4a). Eleven of those peptides (C01, C14, C16 – C18, E21 – E24, G13, and G15) also bound to a control protein and, therefore, were excluded from further analysis (data not shown). Remarkably, we found that the isolated GroEL1 apical domain bound to the same peptides as full-length GroEL1, suggesting that the apical domain is sufficient for substrate binding, although, some spots differed in intensity indicating different binding affinities (Fig. 4a and b). Moreover, the same KasA peptides, in addition to other KasA motifs, were also recognized by full-length GroEL2 in a control experiment (Fig. S1). Consecutive binding peptides with at least three members were grouped together to identify consensus motifs within each group (Table 2). A consensus motif was defined as an amino acid sequence that was present in most members of a group, and was at least four amino acid residues long. Only one acidic residue (Asp) was found in all consensus motifs (Table 2; Group B), which is very low compared to the abundance of negatively charged residues (14 %) in the KasA sequence. On the other hand, positively charged (Arg), and hydrophobic residues (Val and Met) were enriched 2 to 2.6-fold in the consensus motifs relative to the full-length sequence. The latter is in good agreement with previous studies of peptide sequences bound by E. coli GroEL, which also revealed a strong preference for hydrophobic and positively charged residues.21; 24; 25; 26 To provide a spatial explanation for the location of our identified peptides, we mapped our consensus motifs onto the available X-ray structure of M. tuberculosis KasA27 (Fig. 4c). We found that six of our consensus motifs (Groups A, D, G, H, I, J) are mostly buried in the native protein structure, as it might be expected for a chaperonin substrate.28 However, other consensus motifs (Groups B, C, E, and F) are solvent-exposed (Fig. 4c), suggesting that GroEL1 may also interact with native, folded KasA. Taken together, our findings suggest that M. tuberculosis GroEL1 and GroEL2 are chaperonins that recognize both distinct and overlapping KasA peptide motifs. How GroEL1 modulates fatty-acid synthesis through interaction with KasA remains unclear and is subject to further investigation. Materials and Methods Protein preparation and analysis Cloning, expression, and purification of full-length GroEL1 from M. tuberculosis H37Rv, as well as the crystallization and preliminary crystallographic analysis of the GroEL1 fragment have been described.29 For peptide array analysis, the apical domain of GroEL1 (Glu188 to Sielaff et al. Page 4 J Mol Biol. Author manuscript; available in PMC 2012 January 21. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Val373) was cloned by PCR into pProEx Htb generating a plasmid that harbors the GroEL1 apical domain with a Tobacco Etch Virus protease-cleavable N-terminal His6-tag. Expression and purification of the GroEL1 apical domain was performed as previously described for full-length GroEL1.29 GroEL2 was cloned, overexpressed, and purified in a similar manner, except for addition of a complete EDTA-free protease inhibitor cocktail (Roche Diagnostics) in the lysis buffer. To determine the oligomeric state, purified full-length GroEL1 and GroEL1 apical domain were analyzed at 4 °C by size-exclusion chromatography on a Superdex 200 HR 10/30 column (GE Healthcare) in 50 mM Tris pH 7.5 and 150 mM NaCl. Molecular weight protein standards (BioRad) were run in the same buffer in order to correlate elution volume with protein size. X-ray crystallographic analysis and structure refinement The structure of the GroEL1 apical domain was determined by the molecular replacement technique using the apical domain (residues 188 to 372) of a M. tuberculosis GroEL2 monomer (PDB ID: 1SJP-A)20 as search model. This model shares 64% sequence identity with the M. tuberculosis GroEL1 apical domain, distributed evenly over the amino acid sequence (Fig. 1a). 5% of the observed data were randomly chosen and excluded from refinement for cross-validation purposes. The model was refined in CNS 1.2.22 Model refinement was interspersed by manual rebuilding of the atomic structure using COOT.30 Water molecules were selected automatically in CNS 1.2,22 and confirmed manually according to the peak height and distance criteria in the calculated Fo-Fc and 2Fo-Fc maps. The stereochemistry of the final model was analyzed using PROCHECK.31 Superpositioning of molecules was done in PyMol32, the SSAP Server33 was used for RMSD calculations, and ESPript34 was used for preparing the secondary structure alignment. Peptide array synthesis A peptide array of 12-mer overlapping peptides derived from the amino acid sequence of M. tuberculosis KasA was prepared by the SPOT synthesis technique using a semi-automated ASP 222 peptide synthesis robot (Intavis) essentially as described.35 The sequence was walked through by advancing two to three amino acids at each position in order to fit two complete sets of peptides onto one membrane. The membrane was cut into halves, blocked for 2 h in 1× Pierce Superblock in TBS-1 (20 mM Tris-HCl pH 7.5, 137 mM NaCl, and 0.1 % Tween-20), and washed for 10 min in blocking buffer consisting of 10% Superblock and 5% sucrose in TBS-2 (20 mM Tris-HCl pH 7.5, 137 mM NaCl, and 0.05 % Tween-20). Next, each membrane half was incubated in blocking buffer for 1 h in the presence of either 500 nM His6-tagged GroEL1 or 500 nM His6-tagged GroEL1 apical domain. After washing the membranes three times in TBS-1, bound protein was electro-transferred to a Hybond- ECL nitrocellulose membrane (GE Healthcare), and probed and detected as previously described.35 In addition, membranes were also probed directly with 50 ng/ml of a commercially available, anti-His6 monoclonal antibody-horse radish peroxidase conjugate (BD Biosciences) for 1 h in blocking buffer without sucrose. Despite slight differences in spot intensities, both methods generated identical results. After detection, bound protein and antibody were stripped off the membrane by washing the membrane three times in 8 M urea, 1% SDS, 0.5% 2-mercaptoethanol for 30 min each, followed by three times in 20% acetic acid and 50% ethanol for 15 min each, and three times in 100% ethanol for 10 min each. A negative control was performed with His6-tagged GrpE which was expressed and purified as described.36 After stripping the membrane, another control was performed with His6-tagged GroEL2. All incubations were carried out under gentle rocking at room temperature. Sielaff et al. Page 5 J Mol Biol. Author manuscript; available in PMC 2012 January 21. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript PDB accession code The atomic coordinates and corresponding structure factors have been deposited in the Protein Data Bank under PDB ID: 3M6C. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We thank Dr. T. Palzkill for advice and suggestions, and Dr. A. Reger for help with X-ray data collection. Work in F.T.F.T.'s laboratory is supported by grants from the National Institutes of Health (R01-AI076239), the Welch Foundation (Q-1530), the Department of Defense, and the American Cancer Society. B.S. was a Welch postdoctoral fellow and a recipient of a training fellowship from the Pharmacoinformatics Training Program of the Keck Center of the Gulf Coast Consortia (NIH Grant No. R90-DK071505). References 1. Houry WA, Frishman D, Eckerskorn C, Lottspeich F, Hartl FU. Identification of in vivo substrates of the chaperonin GroEL. Nature. 1999; 402:147–154. [PubMed: 10647006] 2. Bartolucci C, Lamba D, Grazulis S, Manakova E, Heumann H. Crystal structure of wild-type chaperonin GroEL. J Mol Biol. 2005; 354:940–951. [PubMed: 16288915] 3. Braig K, Otwinowski Z, Hegde R, Boisvert DC, Joachimiak A, Horwich AL, Sigler PB. The crystal structure of the bacterial chaperonin GroEL at 2.8 A. Nature. 1994; 371:578–586. [PubMed: 7935790] 4. Xu Z, Horwich AL, Sigler PB. The crystal structure of the asymmetric GroEL-GroES-(ADP)7 chaperonin complex. Nature. 1997; 388:741–750. [PubMed: 9285585] 5. Chen S, Roseman AM, Hunter AS, Wood SP, Burston SG, Ranson NA, Clarke AR, Saibil HR. Location of a folding protein and shape changes in GroEL-GroES complexes imaged by cryo- electron microscopy. Nature. 1994; 371:261–264. [PubMed: 7915827] 6. Fenton WA, Kashi Y, Furtak K, Horwich AL. Residues in chaperonin GroEL required for polypeptide binding and release. Nature. 1994; 371:614–619. [PubMed: 7935796] 7. Buckle AM, Zahn R, Fersht AR. A structural model for GroEL-polypeptide recognition. Proc Natl Acad Sci U S A. 1997; 94:3571–3575. [PubMed: 9108017] 8. Chen L, Sigler PB. The crystal structure of a GroEL/peptide complex: plasticity as a basis for substrate diversity. Cell. 1999; 99:757–768. [PubMed: 10619429] 9. Wang J, Chen L. Domain motions in GroEL upon binding of an oligopeptide. J Mol Biol. 2003; 334:489–499. [PubMed: 14623189] 10. Ojha A, Anand M, Bhatt A, Kremer L, Jacobs WR Jr, Hatfull GF. GroEL1: a dedicated chaperone involved in mycolic acid biosynthesis during biofilm formation in mycobacteria. Cell. 2005; 123:861–873. [PubMed: 16325580] 11. Hu Y, Henderson B, Lund PA, Tormay P, Ahmed MT, Gurcha SS, Besra GS, Coates AR. A Mycobacterium tuberculosis mutant lacking the groEL homologue cpn60.1 is viable but fails to induce an inflammatory response in animal models of infection. Infect Immun. 2008; 76:1535– 1546. [PubMed: 18227175] 12. Kong TH, Coates AR, Butcher PD, Hickman CJ, Shinnick TM. Mycobacterium tuberculosis expresses two chaperonin-60 homologs. Proc Natl Acad Sci U S A. 1993; 90:2608–2612. [PubMed: 7681982] 13. Lewthwaite JC, Coates AR, Tormay P, Singh M, Mascagni P, Poole S, Roberts M, Sharp L, Henderson B. Mycobacterium tuberculosis chaperonin 60.1 is a more potent cytokine stimulator than chaperonin 60.2 (Hsp 65) and contains a CD14-binding domain. Infect Immun. 2001; 69:7349–7355. [PubMed: 11705907] Sielaff et al. Page 6 J Mol Biol. Author manuscript; available in PMC 2012 January 21. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 14. Orme IM, Roberts AD, Griffin JP, Abrams JS. Cytokine secretion by CD4 T lymphocytes acquired in response to Mycobacterium tuberculosis infection. J Immunol. 1993; 151:518–525. [PubMed: 8100846] 15. Basu D, Khare G, Singh S, Tyagi A, Khosla S, Mande SC. A novel nucleoid-associated protein of Mycobacterium tuberculosis is a sequence homolog of GroEL. Nucleic Acids Res. 2009; 37:4944– 4954. [PubMed: 19528065] 16. Stewart GR, Wernisch L, Stabler R, Mangan JA, Hinds J, Laing KG, Young DB, Butcher PD. Dissection of the heat-shock response in Mycobacterium tuberculosis using mutants and microarrays. Microbiology. 2002; 148:3129–3138. [PubMed: 12368446] 17. Dosanjh NS, Rawat M, Chung JH, Av-Gay Y. Thiol specific oxidative stress response in Mycobacteria. FEMS Microbiol Lett. 2005; 249:87–94. [PubMed: 16006064] 18. Qamra R, Srinivas V, Mande SC. Mycobacterium tuberculosis GroEL homologues unusually exist as lower oligomers and retain the ability to suppress aggregation of substrate proteins. J Mol Biol. 2004; 342:605–617. [PubMed: 15327959] 19. Kumar CM, Khare G, Srikanth CV, Tyagi AK, Sardesai AA, Mande SC. Facilitated oligomerization of mycobacterial GroEL: evidence for phosphorylation-mediated oligomerization. J Bacteriol. 2009; 191:6525–6538. [PubMed: 19717599] 20. Qamra R, Mande SC. Crystal structure of the 65-kilodalton heat shock protein, chaperonin 60.2, of Mycobacterium tuberculosis. J Bacteriol. 2004; 186:8105–8113. [PubMed: 15547284] 21. Coyle JE, Jaeger J, Gross M, Robinson CV, Radford SE. Structural and mechanistic consequences of polypeptide binding by GroEL. Fold Des. 1997; 2:R93–R104. [PubMed: 9427006] 22. Brunger AT, Adams PD, Clore GM, DeLano WL, Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges M, Pannu NS, Read RJ, Rice LM, Simonson T, Warren GL. Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr. 1998; 54:905–921. [PubMed: 9757107] 23. Brunger AT. Free R value: a novel statistical quantity for assessing the accuracy of crystal structures. Nature. 1992; 355:472–475. [PubMed: 18481394] 24. Li Y, Gao X, Chen L. GroEL Recognizes an Amphipathic Helix and Binds to the Hydrophobic Side. J Biol Chem. 2009; 284:4324–4331. [PubMed: 19074438] 25. Preuss M, Hutchinson JP, Miller AD. Secondary structure forming propensity coupled with amphiphilicity is an optimal motif in a peptide or protein for association with chaperonin 60 (GroEL). Biochemistry. 1999; 38:10272–10286. [PubMed: 10441121] 26. Wang Z, Feng H, Landry SJ, Maxwell J, Gierasch LM. Basis of substrate binding by the chaperonin GroEL. Biochemistry. 1999; 38:12537–12546. [PubMed: 10504222] 27. Luckner SR, Machutta CA, Tonge PJ, Kisker C. Crystal structures of Mycobacterium tuberculosis KasA show mode of action within cell wall biosynthesis and its inhibition by thiolactomycin. Structure. 2009; 17:1004–1013. [PubMed: 19604480] 28. Stan G, Brooks BR, Lorimer GH, Thirumalai D. Residues in substrate proteins that interact with GroEL in the capture process are buried in the native state. Proc Natl Acad Sci U S A. 2006; 103:4433–4438. [PubMed: 16537402] 29. Sielaff B, Lee KS, Tsai FT. Crystallization and preliminary X-ray crystallographic analysis of a GroEL1 fragment from Mycobacterium tuberculosis H37Rv. Acta Crystallogr Sect F Struct Biol Cryst Commun. 2010; 66:418–420. 30. Paul E, Kevin C. Coot: model-building tools for molecular graphics. Acta Cryst. 2004; D60:2126– 2132. 31. Laskowski R, MacArthur M, Moss D, Thornton J. Procheck - a program to check the stereochemical quality of protein structures. J Appl Cryst. 1996; 26:283–291. 32. DeLano, W. The PYMOL Molecular Graphics System. San Carlos, CA, USA: DeLano Scientic LLC; 2002. (http://www.pymol.org) 33. Orengo CA, Michie AD, Jones S, Jones DT, Swindells MB, Thornton JM. CATH--a hierarchic classification of protein domain structures. Structure. 1997; 5:1093–1108. [PubMed: 9309224] 34. Gouet P, Courcelle E, Stuart DI, Metoz F. ESPript: analysis of multiple sequence alignments in PostScript. Bioinformatics. 1999; 15:305–308. [PubMed: 10320398] Sielaff et al. Page 7 J Mol Biol. Author manuscript; available in PMC 2012 January 21. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 35. Rees I, Lee S, Kim H, Tsai FT. The E3 ubiquitin ligase CHIP binds the androgen receptor in a phosphorylation-dependent manner. Biochim Biophys Acta. 2006; 1764:1073–1079. [PubMed: 16725394] 36. Sielaff B, Tsai FT. The M-Domain Controls Hsp104 Protein Remodeling Activity in an Hsp70/ Hsp40-Dependent Manner. J Mol Biol. 2010; 402:30–37. [PubMed: 20654624] Sielaff et al. Page 8 J Mol Biol. Author manuscript; available in PMC 2012 January 21. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Fig. 1. Bacterial chaperonins are structurally conserved. (a) Sequence alignment of the apical domains of M. tuberculosis (Mt) GroEL1, Mt GroEL2, and E. coli (Ec) GroEL. Secondary structure elements of M. tuberculosis GroEL1 are represented by helices (α- and η [310]- helices) and arrows (β-strands). Conserved residues are boxed. Nomenclature of α-helices (H–L) is the same as previously described for the E. coli GroEL apical domain.8 (b) Stereo view of the M. tuberculosis GroEL1 apical domain. The crystal structure is shown as ribbon diagram with α-helices colored blue, β-strands red, and the N- and C-terminal extensions and loops green. η denotes a 310 helix. (c) Superposition of the apical domain structures from M. tuberculosis GroEL1 (orange), M. tuberculosis GroEL2 (magenta),20 and E. coli Sielaff et al. Page 9 J Mol Biol. Author manuscript; available in PMC 2012 January 21. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript GroEL residues 191–376 (blue)7 and residues 191–336 (cyan).8 (d) Analysis of full-length GroEL1 (blue) and GroEL1 apical domain (magenta) by size-exclusion chromatography. The chromatogram shows that the full-length protein elutes as a dimer, whereas the apical domain is a monomer. Sielaff et al. Page 10 J Mol Biol. Author manuscript; available in PMC 2012 January 21. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Fig. 2. Structure of the GroEL1 substrate-binding site. (a) Ribbon diagram of the GroEL1 apical domain (orange) and that of a crystallographic symmetry related molecule (green). The putative substrate binding site is indicated by a purple box. (b) Close-up stereo view of the substrate binding site in GroEL1. Interacting residues from helices H and I (orange) and from the N-terminus of a crystallographic symmetry related molecule (green) are depicted as ball-and-stick models. Hydrogen bonds are indicated by dashed lines and are shown together with the corresponding bond lengths. Sielaff et al. Page 11 J Mol Biol. Author manuscript; available in PMC 2012 January 21. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Fig. 3. Schematic representation of molecular interactions at the putative chaperonin-substrate peptide interface in M. tuberculosis GroEL1 and E. coli GroEL. (a) M. tuberculosis GroEL1 apical domain bound to an N-terminal segment of GroEL1, (b) E. coli GroEL apical domain bound to a non-native peptide,7 and (c) E. coli GroEL apical domain bound to a strong binding peptide selected by phage display.8 Gray lines indicate van der Waals contacts, blue arrows represent hydrogen bonds. Residues surrounded by a dotted box are located on a loop immediately downstream of helix I. Sielaff et al. Page 12 J Mol Biol. Author manuscript; available in PMC 2012 January 21. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Fig. 4. GroEL1 apical domain is sufficient for substrate recognition. Binding of (a) full-length M. tuberculosis GroEL1 or (b) the isolated GroEL1 apical domain to a miniaturized peptide array derived from the M. tuberculosis KasA sequence. Peptides belonging to the same group are boxed together. Groups are differentiated by the color of boxes: Group A (red), Group B (blue), Group C (yellow), Group D (green), Group E (orange), Group F (purple), Group G (white), Group H (pink), Group I (brown), and Group J (teal). (c) Surface representation of the KasA dimer structure (PDB ID: 2WGF)37 with the KasA monomers displayed in different hues. Consensus motifs are mapped onto the KasA structure using the same color scheme as in Fig. 4a and b. Sielaff et al. Page 13 J Mol Biol. Author manuscript; available in PMC 2012 January 21. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Sielaff et al. Page 14 Table 1 Summary of Data Collection and Refinement Statistics M. tuberculosis GroEL1 apical domain Diffraction data statistics Resolution (Å) 40.0 – 2.2 (2.28 – 2.2) Space group P 21212 Unit cell parameters a = 75.47 Å, b = 78.65 Å, c = 34.89 Å α = β = γ = 90° No. of unique reflections 10,994 Completeness (%) 98.7 (91.9) Redundancy 4.3 (2.6) I/sigma (I) 12.3 (2.4) Rsym (%)b 10.1 (36.7) Refinement statistics Resolution (Å) 20.0 – 2.2 Rfactor / Rfree (%)c 21.0 / 23.2 Average B-factor (Å2) 30.3 Number of atoms Protein 1461 Water 65 RMSD of ideal bond length/angle Bond length (Å) 0.006 Bond angle (°) 1.2 aValues in parentheses are for the highest resolution shell. bRsym = Σ |I − (I)| / Σ (I), where I is the observed intensity and (I) is the average intensity. cRfactor Σ |Fobs − Fcalc| / Σ |Fobs|, where Fobs are the observed structure factors and Fcalc are the calculated structure factors. The crystallographic Rfactor is based on 95% of the data used in refinement, and the Rfree is based on 5% of the data withheld for cross-validation test. J Mol Biol. Author manuscript; available in PMC 2012 January 21. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Sielaff et al. Page 15 Table 2 KasA peptides bound by GroEL1 Group Spots Sequencea A A05-A08 9GGFPSVVVTAVTATTSIS26 B A23-B01 49EFVTKWDLAVKIGGHL64 C B09-B17 68DSHMDGRLDMRRMSYVQRMGKLLGGQLWE86 D C02-C04 103VDPDRFAVVVGTGLGG118 E C23-D07 145IMPNGAAAVIGLQLGARAGVMTPVSACS172 F D23-E01 205LPIAAFSMMRAMSTRN220 G F01-F04 257RGAKPLARLLGAGITSDA274 H F12-F17 279APAADGVRAGRAMTRSLELAGL300 I G20-G24 349AVGALESVLTVLTLRDGVIP368 J H17-H19 391YGDYRYAVNNSFGFGG406 aConsensus motifs are depicted in bold. Lower case numbers indicate the position in the M. tuberculosis KasA sequence. J Mol Biol. Author manuscript; available in PMC 2012 January 21.
3M6H
Crystal Structure of Post-isomerized Ertapenem Covalent Adduct with TB B-lactamase
Biochemical and Structural Characterization of Mycobacterium tuberculosis β-Lactamase (BlaC) with the Carbapenems Ertapenem and Doripenem Lee W. Tremblay#, Fan Fan#, and John S Blanchard‡,* Department of Biochemistry, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461 Abstract Despite the enormous success of β-lactams as broad-spectrum antibacterials, they have never been widely used for the treatment of TB due to intrinsic resistance that is caused by the presence of a chromosomally-encoded gene (blaC) in Mycobacterium tuberculosis. Our previous studies of TB BlaC revealed that this enzyme is an extremely broad-spectrum β-lactamase hydrolyzing all β- lactam classes. Carbapenems are slow substrates that acylate the enzyme but are only slowly deacylated and can therefore act also as potent inhibitors of BlaC. We carried out the in vitro characterization of doripenem and ertapenem with BlaC. A steady-state kinetic burst was observed with both compounds with magnitudes proportional to the concentration of BlaC used. The results show apparent Km and kcat values of 0.18 µM and 0.016 min−1 for doripenem and 0.18 µM and 0.017 min−1 for ertapenem. FTICR mass spectrometry demonstrated that the doripenem and ertapenem acyl-enzyme complexes remain stable over a time period of 90 min. The BlaC- doripenem covalent complex obtained after 90 minutes of soaking was solved to 2.2 Å, while the BlaC-ertapenem complex obtained after a 90 minute soak was solved to 2.0 Å. The 1.3 Å diffraction data from a 10 minute ertapenem-soaked crystal revealed an isomerization occurring in the BlaC-ertapenem adduct in which the original Δ2 pyrroline ring was tautomerized to generate the Δ1 pyrroline ring. The isomerization leads to the flipping of the carbapenem-hydroxyethyl group to hydrogen bond to the carboxyl O2 of Glu166. The hydroxyethyl flip results in both decreased basicity of Glu166 and in a significant increase in the distance between the carboxyl O2 of Glu166 and the catalytic water molecule, slowing hydrolysis. Tuberculosis (TB), caused by Mycobacterium tuberculosis, continues to be a worldwide health concern (1). There were an estimated 9.3 million new cases of TB in 2007 and approximately 1.3 million HIV-negative patient fatalities as well as nearly half a million deaths amongst HIV-positive populations (2). Even fifty years after the introduction of powerful antibiotics to treat TB, it has been estimated that one person is infected in the world every few seconds (3). The failure to control TB is due to the emergence of M. tuberculosis strains that are multiply drug resistant towards the front line antimycobacterial drugs such as isoniazid and rifampicin. Phone: (718) 430-3096; Fax: (718) 430-8565. *AUTHOR EMAIL ADDRESS: blanchar@aecom.yu.edu #These authors contributed equally to this work. Supporting Information Available One figure showing the dependence of kburst on the [ertapenem] and [doripenem]. This material is available free of charge via the Internet at http://pubs.acs.org NIH Public Access Author Manuscript Biochemistry. Author manuscript; available in PMC 2011 May 4. Published in final edited form as: Biochemistry. 2010 May 4; 49(17): 3766–3773. doi:10.1021/bi100232q. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript As one of the most important antibiotic families, β-lactams include a broad range of molecules including penicillin derivatives, cephalosporins, monobactams, carbapenems, and β-lactamase inhibitors. The carbapenems exhibit the broadest spectrum of activity among the β-lactam antimicrobials, providing safe and efficacious therapies in the treatment of serious infections caused by Gram-positive, Gram-negative, and anaerobic bacterial pathogens (4,5). Carbapenem antibiotics were originally developed from thienamycin, a natural product identified in culture filtrates of Streptomyces cattleya (6). There are four carbapenems approved thus far for human use: imipenem, meropenem, ertapenem, and doripenem (5). Imipenem was the first carbapenem approved by the US Food and Drug Administration (FDA) in 1985, and is by far the most widely used carbapenem. The use of meropenem was approved in 1995, followed by ertapenem and doripenem in 2001 and 2007, respectively. Except for imipenem, all carbapenems are stable against the mammalian kidney dehydropeptidase (7). In clinical usage, imipenem and meropenem have to be given frequently to maintain high circulating levels. Also, weight-dosage adjustment of imipenem is required to minimize the chance of seizures (8). Ertapenem and doripenem can be given once per day due to their high target affinity and circulating stability (5,9). The lower effective doses of these latter drugs reduces potential side effects, as well as the development of resistance (10). Currently, ertapenem and doripenem are used for complicated intra-abdominal, and urinary tract infections (11,12). Despite the general success of β-lactam antibiotics, they have not been widely used for the treatment of TB due to intrinsic resistance that is caused by the presence of a chromosomally-encoded gene (blaC) in M. tuberculosis for a Class A Ambler β-lactamase (BlaC). Like other Class A β-lactamases, BlaC catalyzes the opening of the β- lactam ring via nucleophilic attack by an active site serine residue to generate the acylenzyme, followed by the hydrolysis of the ester bond to generate the ring-opened, inactive product. Our previous studies of TB BlaC revealed that this enzyme is an extremely broad-spectrum β- lactamase hydrolyzing all β-lactam classes, including the carbapenems meropenem and imipenem (13). Being slow substrates that exhibit rapid acylation followed by a slow deacylation step, meropenem and imipenem also act as potent inhibitors of BlaC (14). FTICR mass spectrometry demonstrated that the acylated intermediate remains stable for many minutes (14). Such slow turnover rates allowed the determination of three- dimensional structure of BlaC in complex with meropenem at a resolution of 1.8 Å. In vivo studies showed that meropenem in combination with the β-lactamase inhibitor, clavulante, is bactericidal against clinical TB strains that are phenotypically exensively drug resistant (XDR-TB) (14). As an extension of our prior work, we carried out an in vitro characterization of doripenem and ertapenem with BlaC. Materials and Methods All chromatographic materials were purchased from Pharmacia. Meropenem and faropenem were from IKT Laboratories. Doripenem (as Doribax) was from Ortho-McNeil Pharmaceutical Inc (Raritan, NJ). Ertapenem (as Invanz) was from Merck & Co. Inc. The potassium salt of clavulanic acid was from Sigma Aldrich. All other chemicals were purchased from Sigma or Aldrich. Nitrocefin was purchased from Beckton Dickinson. Purification of BlaC Recombinant and truncated BlaC from M. tuberculosis expressed from plasmid pET28a(+) and purified to homogeneity as described by Hugonnet and Blanchard (13). Tremblay et al. Page 2 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Kinetics The steady state rate of hydrolysis of β-lactam ring was monitored as a decrease in the absorbance in the UV region, as described previously (13). Assays using doripenem, ertapenem, faropenem and meropenem were performed at 296 nm (ε = 7,540 M−1 cm−1), 295 nm (ε = 9,970 M−1 cm−1), 306 nm (ε = 3,445 M−1 cm−1), and 297 nm (ε = 6,152 M−1 cm−1), respectively. Assays using the chromogenic substrate nitrocefin were performed at 486 nm (ε = 20,500 M−1 cm−1). Assays were performed in 100 mM MES (pH 6.5). Reactions were initiated by the addition of enzyme at concentrations between 0.1–25 µM using 100 µM of the carbapenem substrate. Inhibition Studies Carbapenems at concentrations ranging from 0.1–10 µM were tested as inhibitors of 1.5 nM BlaC using 60 µM nitrocefin as substrate. Time courses were followed for 15 min. For slow onset inhibition, reaction velocities as a function of time were fitted to eq 1: (1) where [P] is the concentration of the product, vi and vs are the initial and final reaction velocities respectively for the reaction in the presence of inhibitor and kiso is the apparent first order rate constant for the inter-conversion between vi and vs, and t is time. The general mechanism can be modeled as: (2) where k1 and k−1 represent the reversible binding to and dissociation from the carbapenem to BlaC, k2 represents the irreversible cleavage of the carbapenem β-lactam ring and k3 represents the hydrolysis of the BlaC-carbapenem adduct. For this model, the rate constant that describes kiso is given by eq 3, where Kd equals k−1/k1. (3) In eq 4, the Km value can be expressed as: (4) In addition, from the determined k2 and k3 values, kcat is calculated from eq 7, assuming k2,k3≪k1,k−1. (5) Tremblay et al. Page 3 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Mass Spectrometry All mass spectra were acquired on a 9.6 T Fourier Transform Ion Cyclotron Resonance (FTICR) mass spectrometer (Ionspec, Lake Forest, CA). To avoid salt interference, BlaC was dialyzed against 20 mM ammonium bicarbonate, pH 6.5. The molecular mass of each protein sample was determined for the 25+ charge state using the equation m = (m/z x 25)−25 on the isotopic centroid. To monitor the intermediate of steady state turnover or small molecular mass spectrometry, 51 µM of enzyme was incubated with 25 µM carbapenem in a total volume of 20 µL. An aliquot of 1 µL was withdrawn at desired time (0, 30, 60, and 90 min) and mixed with 9 µL of mixing solution (containing 50% acetonitrile and 0.1% formic acid). The resulting mixture was injected into the FTICR mass spectrometer. Crystallization BlaC was crystallized in the hanging drop vapor diffusion configuration over well conditions of 0.1 M HEPES, pH 7.5 and 2 M NH4H2PO4. The final pH of the well solution was 4.1. Protein at a concentration of 10 mg/ml was mixed 1:1 with the well solution and incubated at 18 °C. Initial crystals grew within a week but were small, sparse and amorphous. New wells were sealed and allowed to equilibrate overnight. Equilibrated drops were micro-seeded, which resulted in efficient crystal growth as well as improved morphology. Iterative seeding resulted in diffraction quality crystals of active enzyme. Data collection and refinement Crystals were soaked with either ~ 50 mM ertapenem or doripenem in mother liquor plus 20% glycerol as a cryo-protectant. Data were collected after 10 and 90 minute soaks with ertapenem and a 90 minute soak with doripenem at Brookhaven National Laboratory on beamlines X12C and X29, in which various resolutions of diffraction were obtained dependent on the soaking times and beamline. The data were processed using either HKL2000 (15) or Mosflm (16). Our previous structure of clavulanate bound M. tuberculosis β-lactamase (17) (PDB entry 3CG5) was used to phase all the data, using the CCP4 software suite (18). Iterative rounds of structural refinement and model building were performed in Refmac5 (19,20) and Coot (21). Table 1 lists the data collection statistics for the structures as well as the final refinement statistics. RESULTS and DISCUSSION Kinetics The accurate determination of the kinetic parameters for doripenem and ertapenem was severely hampered by apparent very low Km values, very low kcat values and the modest extinction coefficients accompanying hydrolysis. At the [BlaC] required to see any significant rate of reaction (~2 µM), variation of the [doripenem] or [ertapenem] at concentrations from 2–20 µM showed almost no difference in rate, suggesting their Km values were less than 2 µM. The steady-state kinetic parameters determined for faropenem, a structurally distinct penem, were Km = 55 ± 11 µM, and kcat = 0.65 ± 0.04 min−1 (data not shown). This Km value is ~17 times larger and the kcat value 8 times faster than those of meropenem (14). Detailed investigations of the kinetics of carbapenem hydrolysis under near stoichiometric enzyme concentrations were carried out over 30 minute time periods. As shown in Figure 1, a steady-state kinetic burst was observed with both compounds where the magnitudes of the burst are proportional to the concentration of BlaC used. Extrapolation of the rates of hydrolysis to the y-axis demonstrates that the acylation is stoichiometric with the concentration of enzyme. Tremblay et al. Page 4 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Due to the extremely feeble turnover rate, we further tested these carbapenems as inhibitors of the reaction of nitrocefin with BlaC. Nitrocefin is an extremely good substrate for BlaC and its β-lactam ring-opened form is extremely chromogenic. As shown in Figure 2, doripenem and ertapenem act as slow-onset, tight binding inhibitors of BlaC when the hydrolysis of nitrocefin was monitored. In contrast, faropenem exhibited standard, competitive inhibition with no time dependent component (data not shown). This type of time-dependent inhibition for a dead-end inhibitor is modeled as being due to the reversible formation of a non-covalent complex (E–I), followed by the reversible conversion to an isomerized complex (E–I*). However, in the case of a slow substrate for BlaC, the initially formed Michaelis complex reacts with the enzyme in an irreversible step to generate the BlaC-carbapenem covalent intermediate. This is then hydrolyzed slowly to regenerate the free enzyme that can react with nitrocefin. While the same equation is used to fit the two models, the kinetic constants that contribute to kiso (Figure S1) and Ki (or Kd) are different. Using the fits of the slow-onset data and eq 3 to calculate Kd, k2 and k3, we can then used eqs 4 and 5 to calculate the apparent Km and kcat values for doripenem (0.18 µM and 0.016 min−1, respectively), and for ertapenem (0.18 µM and 0.017 min−1, respectively). We have not corrected for the concentration of nitrocefin used in these experiments because of the large standard errors (>40%) associated with these kinetic parameters (the reported Km values are apparent values). However, the extremely tight binding and extremely low turnover of these carbapenems is evident from these rather imprecise kinetic data. Mass Spectrometry The rapid acylation and slow deacylation of BlaC by the carbapenems allows the observation of the covalently bound, acyl-enzyme intermediate by Fourier transform ion cyclotron resonance. A freshly prepared solution containing excess BlaC and doripenem displayed three peaks: the first peak corresponds to free BlaC with mass/charge ratio (m/z) = 28,785.0, a second peak corresponding to the covalently acylated BlaC-doripenem complex with mass/charge ratio (m/z) = 29, 204.1 and a third peak whose mass corresponds to the mass of the covalently acylated BlaC-doripenem complex minus 44 mass unit (m/z = 29, 161.0), as shown in Figure 3. With ertapenem, the two covalent acylated BlaC complex peaks observed had molecular masses of 29,260.0 and 29,217.1, corresponding to acylated BlaC-ertapenem complex and acylated BlaC-doripenem complete minus 44 mass units, respectively. This data demonstrates that both doripenem and ertapenem undergo the same chemical breakdown in the active site as meropenem (14). Once the acyl-enzyme forms, the carbapenems partition between hydrolysis and enzyme-catalyzed decomposition of the C6 hydroxyethyl substituent, via a retro-Aldol decomposition, which yields acetaldehyde (14). Intriguingly, the intensities of the acyl-enzyme complexes remain stable over the time period of 90 min for doripenem and ertapenem. This is in contrast with previous observations with meropenem, where the acylated forms of the enzyme started to diminish after several minutes. These data suggest that doripenem and ertapenem form more stable complexes with BlaC than meropenem, reinforcing the kinetic data. X-ray Crystallography The 2.2 Å data from a 90 min doripenem-soaked crystal were refined to an Rwork of 0.161 and an Rfree of 0.205. The 1.3 Å diffraction data from a 10 minute ertapenem-soaked crystal refined to an Rwork of 0.147 and an Rfree of 0.176. The 2.0 Å diffraction data from a 90 min ertapenem-soaked crystal were refined to an Rwork of 0.175 and an Rfree of 0.222. In these three structures, the active site Ambler residue Ser70 has been covalently linked with the ring open form of these β-lactams in accordance with the acylation chemistry of the first half of the enzymatic reaction (Scheme 1). The quality of the electron density is displayed in Figure 4 and Figure 5 under a Fo-Fc omit calculated map contoured at 2.0 σ. Tremblay et al. Page 5 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The C3 atom of the pyrroline rings of doripenem and ertapenem covalent adducts are sp3 hybridized in the 90 min soaks. These results require an isomerization event occurring in the BlaC-carbapenem adducts in which the original Δ2 pyrroline ring was tautomerized to generate the Δ1 pyrroline ring, evidenced by the collinear positioning of the C5, N1, C2, and C3 atoms. In addition, the BlaC-doripenem and ertapenem covalent adduct densities allow for the positioning of the thioether sulfur atom in the unambiguous assignment of the S configuration at C3, requiring protonation at the re face of the C2–C3 double bond. This is similar to our earlier findings with BlaC crystals soaked with meropenem on similar time scales (14), and represents the thermodynamically preferred product with the trans orientation of the C4 methyl and thioether substituents. Interestingly, in the recently reported structure of the Class D β-lactamase, OXA-1 covalent adduct with doripenem, revealed that while an identical isomerization had taken place, that the final Δ1 pyrroline ring product was of the opposite, R, stereochemistry (22). In the structure determined after the shorter 10-minute soak, the BlaC-ertapenem adduct was covalently bound in the active site, but in a different geometry. On this shorter time scale, the BlaC-ertapenem adduct C3 atom is found in its original sp2 hybridization with the definitive collinear positioning of the thioether sulfer atom in line with the N1, C2, C3, and C4 bonds indicating the presence of the Δ2 pyrroline ring. This result requires that β-lactam ring cleavage and isomerization of the methyl pyrroline ring not be concerted. The active site interactions vary in some subtle ways between the pre and post-isomerization ertapenem complexes, yet a number of common interactions are observed in all complexes. Both the BlaC-ertapenem and -doripenem adducts bind as covalent adducts with the active site Ser70 and position their lactam ring-opened ester carbonyl oxygen atom within the oxyanion hole formed from hydrogen bonding interactions with the Ser70 and Thr253 amide nitrogen atoms. All structures contain a hydrophobic interaction between the methyl group of the pyrroline ring and the sidechain of Ile117 and different forms of hydrogen bonding interactions between the C6 hydroxyethyl substituent of the carbapenem and Glu166. All three structures also show a conserved interaction between the sidechain hydroxyl of Ser130, which consistently hydrogen bonds the pyrroline ring nitrogen atom at a distance of 2.7–2.8 Å. The pyrroline C2 carboxylate group forms hydrogen bonds with the Thr251 hydroxyethyl side chain and an active site water molecule. In structures of other β- lactamase-carbapenem adducts, this carboxylate electrostatically interacts with a conserved arginine residue (R244 in TEM-1) (23), but this is not the case for BlaC. In the pre- isomerized ertapenem structure, there is an additional hydrogen bond between the C2 carboxylate of the pyrroline ring and Thr253, which is broken upon the repositioning of the meta-amino-benzoate ‘arm’ observed in the post-isomerized ertapenem structure. The isomerization and stereospecific protonation leads to a reorientation of the terminal portion of the molecule within the active site, allowing for the formation of the hydrogen bond between the terminal carboxylate group of the meta-amino-benzoic acid moiety and Ser118. A final difference between the initially formed Δ2-pyrroline isomer and the final Δ1- pyrroline form is the orientation of the C6 hydroxyethyl substituent. In the pre-isomerized complex it is oriented away from Lys73 and hydrogen bonds to the carboxyl O1 of Glu166 as well as the Asn186 nitrogen, but rotates upon isomerization to hydrogen bond with the carboxyl O2 of Glu166 and the ε-amino group of Lys73 in a manner similar to that observed for the doripenem complex. The structures of the pre and post-isomerization states reveal the mechanistic basis for the relative stability of the carbapenems within the active site of BlaC and their ability to resist hydrolysis by the enzyme. As seen in Scheme1, deacylation-hydrolysis from the enzyme requires the activation of the conserved active site water by the side chain carboxyl O2 of Glu166. The probability of water activation by Glu166 decreases with the increased distance Tremblay et al. Page 6 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript between the two. The distance between the carboxyl O2 of Glu166 and the active site water is significantly increased after isomerization from 2.4 Å to 2.7 Å, making water activation less probable. The isomerization event does not directly cause these changes but rather alters the positioning of the adduct such that the adduct-hydroxyethyl flips in the active site, breaking the hydrogen bond formed between the hydroxyl group with the side chain amide nitrogen of Asn186 and the adjacent carboxyl O1 of Glu166 oxygen. The hydroxyethyl substituent then rotates to generate a hydrogen bond network with Lys73 and the carboxyl O2 of Glu166, effectively ‘pulling’ this essential base away from the conserved active site water. These residues (Lys 73 and Glu166) are involved as general bases in the acylation and deacylation reactions, respectively. In addition, by hydrogen bonding carboxyl O2 of Glu166, the reoriented hydroxyethyl substituent reduced the basicity of Glu166. These two factors, introduced in the reorientation of the hydroxyethyl substituent in the post- isomerization complex, reduce water activation and thereby stabilize the acyl-intermediate in the active site. The studies reported here allow us to directly visualize the changes that occur between the pre- and post-isomerization adduct structures and are atomic level observations relevant to the biphasic kinetics previously reported for the reactions between carbapenems and the RTEM β-lactamase (24). The Δ2 to Δ1-pyrroline isomerized forms of carbapenems have been known to form within the active sites of various β-lactamase enzymes (25). In confirmation of this, crystal structures of carbapenems bound within the active sites of the Class A β-lactamases TEM-1 (PDB entry 1BT5) (26) and SHV-1 (PDB entry 2ZD8) (27) as well as AmpC (PDB entry 1LL5) (28) a Class C β-lactamase all revealed the Δ2 form of the carbapenem bound in the active sites, while the Class D OXA-1 (PDB entry 3ISG) (22) and the Class A BlaC (PDB entry 3DWZ) (14) were both bound with carbapenems in the Δ1 isomerized forms with respective R and S-stereochemistries. Our findings are the first to show the structures of both the Δ2 and Δ1 forms of a carbapenem bound to a single β-lactamase. Interestingly, several of the structures of carbapenems bound as the Δ2 isomers show evidence for alternate conformations for the carbapenem-carbonyl oxygen position. This oxygen is found buried within the oxyanion-hole as well as bound in a position rotated by 180 degrees, usually facing an opposing serine residue (Ser130). In these instances it has been proposed that the flipping of the carbonyl oxygen from the oxyanion-hole blocks formation of the deacylation tetrahedral intermediate to inhibit the enzyme. In the cases of OXA-1 and BlaC, the carbapenem-carbonyl oxygen is only found bound tightly within the oxyanion-hole and no evidence of alternate conformers has been observed. In these cases inhibition by the carbapenem is likely due to disruption of water activation. A second possible reason for the observed alternate conformers at the carbapenem-carbonyl is likely due to the position of the carbapenem-carboxylate moiety within those active sites. To date, those β-lactamases with alternate conformations for the carbapenem-carbonyl, have a highly conserved Arg244 reside which electrostatically interacts with the carbapenem- carboxylate moiety. The OXA-1 and BlaC active sites lack this arginine interaction and instead use a combination of threonine and/or serine residues coordinated with waters to bind the carboxylate moiety. These residues are located closer to the oxyanion hole and act to ‘clamp’ the carboxylate into a proximal position, as opposed to the Arg244 mechanism of carboxylate binding, where distance introduces flexibility, allowing for the alternate positioning of the pyrroline ring. This bonding pattern to the carbapenem allows for alternate “in/out” conformations of the carbapenem carbonyl in the oxyanion hole. In contrast, the carbapenem carbonyl is tightly bound in the oxyanion hole in BlaC in both the Δ2 and Δ1 forms reported here. Tremblay et al. Page 7 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments This work was supported in part by a grant from the National Institute of Health (AI33696 to J. S. B.) and in part by the Charles Revson Foundation (to L.W.T.) ABBREVIATIONS BlaC Mycobacterium tuberculosis beta-lactamase TB Tuberculosis XDR-TB extensively drug resistant References 1. Dye, C.; Floyd, K.; Uplekar, M.; Bierrenbach, A.; Bergstrom, K.; Blanc, L.; Grezmska, M.; Gunneberg, C.; Lonnroth, K.; Nunn, P.; Pantoja, A.; Raviglione, S.; Weyer, K. WHO report 2008. World Health Organization; 2008. Global tuberculosis control : surveillance, planning, financing. 2. Bauquerez, R.; Blanc, L.; Bierrenbach, A.; Brands, A.; Ciceri, K.; Falzon, D.; Floyd, K.; Glaziou, P.; Gunneberg, C.; Hiatt, T.; Hosseini, M.; Pantoja, A.; Uplekar, M.; Watt, C.; Wright, A. WHO Report 2009. World Health Organization; 2009. Global Tuberculosis Control: Epidemiology, Strategy, Financing. 3. Netto EM, Dye C, Raviglione MC. Progress in global tuberculosis control 1995–1996, with emphasis on 22 high-incidence countries. Global Monitoring and Surveillance Project. Int J Tuberc Lung Dis 1999;3:310–320. [PubMed: 10206501] 4. Mandell L. Doripenem: a new carbapenem in the treatment of nosocomial infection. Clin Infect Dis 2009;49(Suppl 1):S1–S3. [PubMed: 19619016] 5. Baughman RP. The use of carbapenems in the treatment of serious infections. J Intensive Care Med 2009;24:230–241. [PubMed: 19617229] 6. Birnbaum J, Kahan FM, Kropp H, MacDonald JS. Carbapenems, a new class of beta-lactam antibiotics. Discovery and development of imipenem/cilastatin. Am J Med 1985;78:3–21. [PubMed: 3859213] 7. Livermore DM. Of Pseudomonas, porins, pumps and carbapenems. J Antimicrob Chemother 2001;47:247–250. [PubMed: 11222556] 8. Calandra G, Lydick E, Carrigan J, Weiss L, Guess H. Factors predisposing to seizures in seriously ill infected patients receiving antibiotics: experience with imipenem/cilastatin. Am J Med 1988;84:911–918. [PubMed: 3284342] 9. Bhavnani SM, Hammel JP, Cirincione BB, Wikler MA, Ambrose PG. Use of pharmacokinetic- pharmacodynamic target attainment analyses to support phase 2 and 3 dosing strategies for doripenem. Antimicrob Agents Chemother 2005;49:3944–3947. [PubMed: 16127078] 10. Lynch MJ, Drusano GL, Mobley HL. Emergence of resistance to imipenem in Pseudomonas aeruginosa. Antimicrob Agents Chemother 1987;31:1892–1896. [PubMed: 3125787] 11. Behera B, Mathur P, Das A, Kapil A. Ertapenem susceptibility of extended spectrum beta- lactamase-producing Enterobacteriaceae at a tertiary care centre in India. Singapore Med J 2009;50:628–632. [PubMed: 19551319] 12. Paterson DL, Depestel DD. Doripenem. Clin Infect Dis 2009;49:291–298. [PubMed: 19527173] 13. Hugonnet JE, Blanchard JS. Irreversible inhibition of the Mycobacterium tuberculosis beta- lactamase by clavulanate. Biochemistry 2007;46:11998–12004. [PubMed: 17915954] 14. Hugonnet JE, Tremblay LW, Boshoff HI, Barry CE 3rd, Blanchard JS. Meropenem-clavulanate is effective against extensively drug-resistant Mycobacterium tuberculosis. Science 2009;323:1215– 1218. [PubMed: 19251630] Tremblay et al. Page 8 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 15. Otwinowski Z, a MW. Processing of X-ray Diffraction Data Collected in Oscillation Mode. Methods in Enzymology 1997;276:307–326. 16. Leslie AGW. Mosflm. Joint CCP4 + ESF-EAMCB Newsletter on Protein Crystallography No. 26. 1992 17. Tremblay LW, Hugonnet JE, Blanchard JS. Structure of the covalent adduct formed between Mycobacterium tuberculosis beta-lactamase and clavulanate. Biochemistry 2008;47:5312–5316. [PubMed: 18422342] 18. Potterton EBP, Turkenburg M, Dodson E. A graphical user interface to the CCP4 program suite. Acta. Cryst 2003;D59:1131–1137. 19. Murshudov GN, Vagin AA, Dodson EJ. Refinement of Macromolecular Structures by the Maximum-Likelihood Method. Acta Cryst 1997;D53:240–255. 20. Pannu NS, Murshudov GN, Dodson EJ, Read RJ. Incorporation of prior phase information strengthens maximum-likelihood structure refinement. Acta Crystallogr D Biol Crystallogr 1998;54:1285–1294. [PubMed: 10089505] 21. Emsley PCK. Coot: model-building tools for molecular graphics. Acta Crystallogr 2004;D60:2126–2132. 22. Schneider KD, Karpen ME, Bonomo RA, Leonard DA, Powers RA. The 1.4 A crystal structure of the class D beta-lactamase OXA-1 complexed with doripenem. Biochemistry 2009;48:11840– 11847. [PubMed: 19919101] 23. Zafaralla G, Manavathu EK, Lerner SA, Mobashery S. Elucidation of the role of arginine-244 in the turnover processes of class A beta-lactamases. Biochemistry 1992;31:3847–3852. [PubMed: 1567841] 24. Easton CJ, Knowles J. Inhibition of the RTEM beta-lactamase from Escherichia coli. Interaction of the enzyme with derivatives of olivanic acid. Biochemistry 1982;21(12):2857–2862. [PubMed: 7049231] 25. Kalp M, Carey PR. Carbapenems and SHV-1 beta-lactamase form different acyl-enzyme populations in crystals and solution. Biochemistry 2008;47:11830–11837. [PubMed: 18922024] 26. Maveyraud L, Mourey L, Kotra LP, Pedelac J, Guillet V, Mobashery S, Samama J. Structural Basis for Clinical Longevity of Carbapenem Antibiotics in the Face of Challenge by the Common Class A β-Lactamases from the Antibiotic-Resistant Bacteria. Journal of the American Chemical Society 1998;120:9748–9752. 27. Nukaga M, Bethel CR, Thomson JM, Hujer AM, Distler A, Anderson VE, Knox JR, Bonomo RA. Inhibition of class A beta-lactamases by carbapenems: crystallographic observation of two conformations of meropenem in SHV-1. J Am Chem Soc 2008;130:12656–12662. [PubMed: 18761444] 28. Beadle BM, Shoichet BK. Structural basis for imipenem inhibition of class C beta-lactamases. Antimicrob Agents Chemother 2002;46:3978–3980. [PubMed: 12435704] Tremblay et al. Page 9 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Time courses of doripenem (A) and ertapenem (B) hydrolysis with various concentrations of BlaC. Tremblay et al. Page 10 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Time courses of nitrocefin hydrolysis by BlaC in the presence of doripenem (upper) and ertapenem (lower). Tremblay et al. Page 11 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Mass spectra of enzyme-carbapenem species. The 25+ charge state ions are shown. Tremblay et al. Page 12 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. (A) Overall structure of BlaC displayed in rainbow from N term (blue) to the C term (red), with the doripenem adduct displayed in red surface mesh. (B) Fo-Fc omit density (green) contoured at 2.0 σ surrounds the covalent doripenem adduct formed at the Ambler active- site residue serine 70. All structure figures were produced using Pymol. Tremblay et al. Page 13 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. (A) Fo-Fc omit density (green) contoured at 2.0 σ surrounds the covalent ertapenem adduct formed at the Ambler active-site residue serine 70 in the pre-isomerization state. (B) Fo-Fc omit density (green) contoured at 2.0 σ surrounds the covalent ertapenem adduct formed at the Ambler active-site residue serine 70 in the post-isomerization state. The resolution of the densities unambiguously demonstrates the shift in stereochemistry with the change from sp2 to sp3 hybridization of the C3 carbapenem carbon atom with the change in the position of the density associated with the meta-amino-benzoate and the hydoxyethyl ertapenem moieties. Tremblay et al. Page 14 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. (A) the structures of doripenem and ertapenem. (B) The chemical mechanism of hydrolysis of ertapenem by the Mycobacterium tuberculosis BlaC. Tremblay et al. Page 15 Biochemistry. Author manuscript; available in PMC 2011 May 4. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Tremblay et al. Page 16 Table 1 Data Collection and Refinement Statistics Data Collection Doripenem Δ1-isomerζ Ertapenem Δ2-isomer Ertapenem Δ1-isomer Resolution (Å) 50.0-2.2 (2.32-2.20) 50.0-1.30 (1.33-1.30) 50.0-2.0 (2.07-2.00) Completeness 100% (100%) 100.0% (100%) 99.5 (99.9) Redundancy 7.6 (7.4) 7.5 (5.7) 4.4 (4.4) I/sigma(I) 3.8 (1.6) 21.4 (1.8) 9.8 (4.0) Rmerge 0.077 (0.47) 0.057 (0.757) 0.158 (0.373) Space Group P212121 P212121 P212121 Unit cell (Å) a =49.989 b =68.068 c =75.792 α = β = γ = 90.0° a = 49.66 b = 67.92 c = 75.55 α = β = γ = 90.0° a =49.934 b =67.830 c =75.201 α = β = γ = 90.0° Reflections 13,695 (1,943) 60,263 (4,388) 17,920 (1,762) Refinement Statistics Rwork 0.161 (0.176) 0.147 (0.265) 0.175 (0.191) Rfree 0.205 (0.237) 0.176 (0.278) 0.222 (0.281) Average B-factors (Å2) Protein 6.97 10.49 6.09 Adduct 27.32 18.64 15.50 Solvent 17.51 32.64 14.36 PO4 12.89 14.40 10.53 RMS deviations bonds (Å) 0.010 0.010 0.012 angles (°) 1.204 1.428 1.386 Ramachandra Favored= 97.7% outliers= 0.0% Favored= 97.7% outliers= 0.0% Favored= 98.1% outliers= 0.0% PDB accession code 3IQA 3M6B 3M6H Values in parentheses are for the highest resolution bin. ζThis data processed using Mosflm Biochemistry. Author manuscript; available in PMC 2011 May 4.
3M6K
Crystal Structure of N-terminal 44 kDa fragment of topoisomerase V in the presence of guanidium hydrochloride
Structures of minimal catalytic fragments of topoisomerase V reveals conformational changes relevant for DNA binding Rakhi Rajan*, Bhupesh Taneja*,†, and Alfonso Mondragón*,‡ * Department of Biochemistry, Molecular Biology and Cell Biology, Northwestern University, 2205 Tech Dr, Evanston, IL 60208 Summary Topoisomerase V is an archaeal type I topoisomerase that is unique among topoisomerases due to presence of both topoisomerase and DNA repair activities in the same protein. It is organized as an N-terminal topoisomerase domain followed by 24 tandem helix hairpin helix (HhH) motifs. Structural studies have shown that the active site is buried by the (HhH) motifs. Here we show that the N-terminal domain can relax DNA in the absence of any HhH motifs and that the HhH motifs are required for stable protein-DNA complex formation. Crystal structures of various topoisomerase V fragments show changes in the relative orientation of the domains mediated by a long bent linker helix, and these movements are essential for the DNA to enter the active site. Phosphate ions bound to the protein near the active site helped model DNA in the topoisomerase domain and shows how topoisomerase V may interact with DNA. Introduction DNA topoisomerases are enzymes found in all forms of life (bacteria, eukarya, and archaea) and they regulate the topological state of DNA inside the cell. They form a transient break in a single or double stranded DNA and allow the passage of another single or double DNA strand through the break, before resealing the break (Champoux, 2001) (Schoeffler and Berger, 2008). As a result of this, topoisomerases can relax supercoiled DNA, help in the segregation of DNA strands following replication, and lead to the formation and resolution of knots and catenates (Gellert, 1981). Topoisomerases participate in many aspects of DNA metabolism, such as replication, recombination, and transcription (Champoux, 2001). In addition, they are targets of various anti-cancerous drugs and anti-bacterial agents (Maxwell, 1999; Pommier, 1998; Rothenberg, 1997; Wang et al., 1997). DNA topoisomerases are broadly classified into two types, type I and type II enzymes. Type I enzymes cleave a single strand of a DNA molecule and pass another single or double stranded DNA through the break before resealing the opening. Type II enzymes cleave both ‡Corresponding author: Phone: 847-491-7726, Fax: 847-467-6489, a-mondragon@northwestern.edu. †Present address: Institute of Genomics and Integrative Biology, CSIR, Delhi, India Protein data bank accession codes The final structure factors and coordinates of Topo-31, Topo-44 Form I, Form II, and Form III have been deposited in the Protein Data Bank with accession codes 3M7G, 3M7D, 3M6K, and 3M6Z respectively. Supplementary data Supplementary data are available at Structure Journal Online. Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. NIH Public Access Author Manuscript Structure. Author manuscript; available in PMC 2011 July 14. Published in final edited form as: Structure. 2010 July 14; 18(7): 829–838. doi:10.1016/j.str.2010.03.006. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript strands of a double stranded DNA in concert and pass another double stranded DNA through the break. Type I enzymes use the torsional energy stored in the supercoiled DNA to drive DNA relaxation and hence they do not require high energy cofactors, such as ATP, for their activity (Baker et al., 2009) Type II enzymes, on the other hand, require ATP and Mg2+ for their activity. Type I topoisomerases are further subdivided into three subtypes: IA, IB, and IC (Forterre et al., 2007). Type IA and IB enzymes have been studied extensively (Baker et al., 2009) and there is ample information available about their general mechanism of DNA relaxation and the mode of DNA binding. Type IC, on the other hand, is a relatively new subtype. Currently topoisomerase V is the only member of this family and it has been identified only in the Methanopyrus genus. Previously, topoisomerase V had been considered as a type IB enzyme based on its biochemical characteristics (Slesarev et al., 1993), but the crystal structure of an N-terminal 61 kDa of topoisomerase V (Topo-61) (Taneja et al., 2006) revealed a completely new fold without similarity to other topoisomerases or any other known protein. Furthermore, the orientation of the putative active site residues is also different from other type I topoisomerases, suggesting a different mechanism of cleavage and religation of DNA. These observations, together with the lack of sequence similarity, indicated that topoisomerase V defines a new subtype of type I enzymes (Forterre, 2006). Topoisomerase V was identified in Methanopyrus kandleri, an extremophile isolated from a deep-water ‘black smoker’ chimney in the Gulf of California (Huber et al., 1989). The enzyme is active at very high temperatures (122°C) and high salt concentrations (0.65 M NaCl and 3.1 M potassium glutamate). The unusual characteristic of topoisomerase V is that it has both topoisomerase and DNA repair activities in the same polypeptide (Belova et al., 2001). Based on the sequence analysis of topoisomerase V, it has been predicted that the protein contains 24 helix-hairpin-helix (HhH) DNA binding motifs arranged as 12 (HhH)2 domains around the N-terminal topoisomerase domain (Belova et al., 2002) (Figure 1A). Some of these (HhH)2 domains are involved in the apurinic/apyrimidinic (AP) site- processing activity, but the exact location of the repair active site is not known yet. Topoisomerase V can relax both positively and negatively supercoiled DNA without the need for metal cations or high energy cofactors. Single molecule experiments have shown that topoisomerase V relaxes DNA by a constrained swiveling mechanism, relaxing around 12 turns of DNA per relaxation cycle (Taneja et al., 2007). Type IB enzymes, which also use a constrained swiveling mechanism for DNA relaxation, relax around 19 turns of DNA per relaxation cycle (Koster et al., 2005). The structure of Topo-61 showed that the topoisomerase domain is mainly alpha helical and that the first four (HhH)2 domains curl around the topoisomerase domain (Taneja et al., 2006) (Figure 1B). The topoisomerase and (HhH)2 domains are joined by a long bent helix, termed the “linker helix”. Three of the five putative active site residues are present in a helix-turn-helix (HTH) domain and the other two are present in an intervening loop and a helix. The active site residues are buried by the first (HhH)2 domain and it has been suggested that large conformational changes will be needed for the DNA to access the active site of topoisomerase V (Taneja et al., 2007). Here we present data that shows that the N- terminal 31 kDa fragment of topoisomerase V (Topo-31) has topoisomerase activity, consistent with previous predictions based on the structure. In addition, we show that the Topo-44 fragment (N-terminal 44 kDa fragment of topoisomerase V) can form a stable protein-DNA complex, emphasizing the need of the (HhH)2 domains for binding DNA. We determined a crystal structure of (Topo-31) fragment, which has only the topoisomerase domain, and three different crystal structures of the Topo-44 fragment, which includes the topoisomerase domain and three tandem HhH motifs. In all structures, the topoisomerase domain is very similar. In contrast, the structures of Topo-44 show conformational changes in the linker helix resulting in variable orientations of the (HhH)2 domains when compared Rajan et al. Page 2 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript to the Topo-61 structure. Phosphate ions are present in the vicinity of the topoisomerase active site in two of the Topo-44 structures. Some of the catalytic residues interact with the phosphate ions and may mimic contacts with DNA. These observations suggest that the movement of the (HhH)2 domains is mediated by the linker helix and helps expose the topoisomerase active site to facilitate DNA binding. In addition, the location of the phosphate ions suggests a possible path for the DNA and the way the active site residues interact with it. Results The topoisomerase domain can relax DNA DNA relaxation assays using different topoisomerase V fragments showed that the topoisomerase domain alone is capable of relaxing DNA. Topoisomerase V fragments with different numbers of (HhH)2 domains, Topo-31, Topo-44, and Topo-78, were studied using relaxation assays. Topo-31 has no (HhH)2 domains, Topo-44 has one full and one partial (HhH)2 domain, while Topo-78 has eight full (HhH)2 domains, including a putative DNA repair domain. In addition to standard conditions, the effect of different pH conditions and presence of magnesium ions were also tested. The experiments show that Topo-31 is capable of relaxing DNA, despite the absence of the (HhH)2 domains (Figure 2B). A pH profile analysis for the DNA relaxation assays showed that Topo-78 relaxes DNA over a wider pH range (pH 5 to 9), while Topo-31 and Topo-44 relax DNA optimally at pH 5 (Figure 2A, 2B, 2C). In addition, magnesium is not required for the reaction, but stimulates it at all pH values (Figure 2B, 2C). Topo-78 can relax DNA to the same extent with lower amounts of protein (0.1 μg/reaction) compared to Topo-44 (~1.5 μg/reaction) and Topo-31 (~9 μg/reaction). This could be due to the enhanced DNA binding facilitated by the (HhH)2 domains. Together, these results suggest that, even though the (HhH)2 domains are dispensable for topoisomerase activity, they enhance DNA relaxation activity. In addition, the pH dependence of the DNA relaxation activity indicates that the reaction is likely to involve side chains with ionizable groups in the low pH range, such as glutamates. Finally, the magnesium independence of the reactions confirms that even the smallest fragments do not require metals for activity, although magnesium has a stimulatory effect. This may be due to favorable interactions of the cations with DNA. The (HhH)2 domains enhance DNA binding affinity EMSA experiments with different fragments of topoisomerase V and DNA showed that (HhH)2 domains could help in the formation of a stable protein-DNA complex. Various topoisomerase V fragments (Topo-31, Topo-44, and Topo-78) and single and double stranded DNA were analyzed by EMSA experiments. Topo-44 and Topo-78 formed stable complexes with a 39mer double stranded DNA (Figure 2D), while no DNA binding was observed for the Topo-31 fragment (data not shown). These observations indicate that (HhH)2 domains are necessary for a stable protein-DNA complex and that as few as one and half (HhH)2 domains are enough for formation of a stable protein-DNA complex. EMSA with single stranded DNA showed that Topo-31 and Topo-44 cannot bind to single stranded DNA, while Topo-78 can bind to single stranded DNA (data not shown). Overall Structures The topoisomerase domain of topoisomerase V is a helical-rich compact domain that has no structural similarity to any other known protein. The only recognizable structural element is a HTH that contains some of the active site residues. Not surprisingly, the topoisomerase domain of the four structures (Topo-31, Topo-44 (Forms I, II, and III)) superimpose very well on each other and also to that from the Topo-61 structure. In the Topo-31 structure, two surface loops, residues 39-49 and 120-124, adopt a different conformation compared to the Rajan et al. Page 3 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Topo-61 and Topo-44 structures. These two loops are not always visible in the Topo-44 structures, suggesting that they are mobile regions. The r.m.s.d. for the superposition of the topoisomerase core domain of all the new structures on to the Topo-61 structure range from 0.2 Å to 0.7 Å if the two mobile surface loops are not included (Figure 3A). In general, the topoisomerase domain remains unchanged and is identical in all structures. The (HhH)2 domains also remain largely unchanged, with r.m.s.d. for the superposition of only the (HhH)2 domains from the three Topo-44 crystal forms and equivalent domains in the Topo-61 structure ranging from 0.31 Å to 0.56 Å. The five crystallographically independent structures of Topo-44 (Form I, Form II A and B monomers, and Form III A and B monomers) were compared with each other and to the two crystallographically independent Topo-61 monomers to understand the conformational changes in the protein. The r.m.s.d. for the superposition of all the Topo-44 structures (residues 3-375) on to the Topo-61 fragment or on each other vary between 0.9 Å and 2.7 Å, with the majority above 1.5 Å, showing that in general the structures have slightly different conformations. As mentioned above, the different domains behave as rigid or almost rigid subunits and the only change in the structure is the relative orientation between the topoisomerase and the (HhH)2 domains. The change in orientation of the domains starts at the linker helix (residues 269-295), which acts as a hinge region, and follows into the (HhH)2 domains. At the start of the linker helix, the structures superimpose very well for all five Topo-44 and two Topo-61 structures. In the middle of the linker helix there is a kink after which the linker helix from all the structures shows different orientations (Figure 3B). The flexibility of the linker helix is also evident by the fact that the linker helix in the B subunit of Form III crystals appears in two alternate conformations. The change in the relative orientation of the (HhH)2 and topoisomerase domains (Figure 3C and 3D), suggests that these domains can adopt different orientations and these movements might be necessary for the DNA to access the active site. The topoisomerase domain has a positively charged groove adjacent to the active site The structure of the Topo-31 as well as the structures of the Topo-44 fragment reveals the presence of a positively charged groove in the protein that encompasses the active site region (shown later in Figure 6C). This charged groove had been observed before in the structure of the Topo-61 fragment, although several (HhH)2 motifs partially obstruct it (Taneja et al., 2006). The structure of the Topo-31 confirms the presence of the groove even in the absence of the (HhH)2 motifs. The groove is long and can be deep in some areas. It includes regions of the HTH motifs and extends all the way to the linker helix. All the residues forming the active site pentad point towards the groove. The active site tyrosine, Tyr226, is found near one of the ends of the groove, a region where it widens. The positively charged character of the groove and its presence by the active site strongly suggest that it may be involved in DNA binding. Phosphate ions bind in the groove near the topoisomerase active site An interesting observation stemming from the Form II and Form III Topo-44 structures is the presence of phosphate ions near the positively charged DNA binding groove. All three Topo-44 crystal forms were crystallized in the presence of phosphate-citrate buffer, but only Form II and Form III structures showed phosphate ions bound to the protein, which were assigned based on electron density consistent with a tetrahedral phosphate ion (Figure 4A). Form II and Form III crystals include 1–1.2 M guanidium hydrochloride in the crystallization solution. The high resolution Form III structure shows clear density for three guanidium ions bound to the protein, two very well ordered and one with weak density. The presence of guanidium hydrochloride in the crystals appears to trigger a conformational change allowing the binding of phosphate ions to the protein. It is interesting to note that Rajan et al. Page 4 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Form I crystals did not show any bound phosphate albeit its presence in the crystallization condition. This could be due to the absence of guanidium hydrochloride to trigger the binding of phosphate ions as observed in Form II and Form III structures. There are three phosphate ions in the B subunit and none in the A subunit of the Form II Topo-44 structure. Two of the phosphates are in the topoisomerase active site and one of them forms close contacts with the putative active site residues in the topoisomerase domain (Figure 4B). Form III crystal has seven phosphate ions, three in each subunit and one between both the subunits. In the Form III structure, the phosphate ion near the active site Tyr226 is absent, but it shows several new locations for phosphate ions, especially in the positively charged groove containing the topoisomerase active site (Figure 5A). An overlay of the A and B subunits of the Topo-44 Form III structure with the B subunit of Topo-44 Form II structure shows eight unique phosphate ions (Figure 5A). It clearly shows that there are more phosphate ions bound in the positively charged groove compared to other regions of the protein. Taking into account all structures, there are five unique phosphate ion binding sites in the putative DNA binding groove and an additional one near its end and close to the start of the linker helix. Several pairs of phosphates in the groove are separated by a distance of around 7 Å (Figure 5B), which would be consistent with the phosphate-phosphate distance in adjacent nucleotides in a DNA double helix. One of the phosphates (P1) is found near the active site tyrosine and is coordinated by Tyr226 and Arg131, two residues that have been implicated in cleavage and religation of the DNA (Taneja et al., 2006), and by Glu215, whose charge may be important for interactions with DNA (R.R. and A.M., unpublished observations). The side chains of the tyrosine and the glutamate residues are in contact with Arg144 and His200, the other putative active site residues, and these interactions may help to orient them for the catalytic reaction. Adjacent to P1, there is a second phosphate (P2) at a distance of 7.5 Å which is trapped between the topoisomerase domain and an HhH motif. P2 is coordinated by Arg131, an active site residue, in addition to Arg108 from the topoisomerase domain and Arg293 and serines 322 and 324 from the second HhH motif (Figure 6C). Three more phosphates are found in the groove (P3, P4, and P5) coordinated mainly by positively charged residues, such as Arg37, Lys47, Arg108, Lys134, and Arg135 from the topoisomerase domain and also residues from the linker helix such as Tyr289 and Arg293. In general, some of the side chains can contact more than one phosphate. The distance between P3 and P4 and P4 and P5 is 6.8 Å and 6.5 Å respectively. A final phosphate (P6) is located at the start of the linker helix and on the edge of the groove (Figure 5A). Discussion Topoisomerase V is active at very high temperatures (122°C) and high salt concentrations. DNA relaxation assays with various topoisomerase V fragments (Topo-44 and Topo-61) show that a temperature above 60° C is required for optimal activity, although longer fragments of topoisomerase V can relax DNA at lower temperatures (Taneja et al., 2007). Topo-44 was first identified by limited proteolytic digestion of the full length topoisomerase V protein (Belova et al., 2002) at 80°C. In contrast, Topo-61 is the shortest fragment showing topoisomerase activity when the proteolytic reaction is performed at 37°C (Belova et al., 2002). The N-terminal Topo-31 fragment, which contains neither HhH motifs nor the linker helix, was identified as the smallest region spanning the topoisomerase domain from the crystal structure of Topo-61 fragment (Taneja et al., 2006) and it was suggested that it could represent the minimal domain capable of relaxing DNA. Relaxation experiments with this minimal domain show that this is indeed the case, although the activity is not as robust as with longer fragments. As expected, Topo-31 does not require magnesium for activity, but magnesium enhances it, as is the case for type IB topoisomerases, which also uses a Rajan et al. Page 5 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript swiveling mechanism for DNA relaxation (Stewart et al., 1996). The optimal pH for activity for the Topo-31 and the Topo-44 fragments is around 5. This pH dependence is not observed for the Topo-78 fragment. The DNA relaxation by shorter fragments of topoisomerase V at pH 5 could point to the involvement of some ionizable side chains in the relaxation activity. It could also be simply due to the effects of various side chains on DNA binding. Further experiments with different active site mutations in both longer and shorter fragments of topoisomerase V will be required to probe the pH dependence of the relaxation reaction by shorter topoisomerase V fragments. Gel shift experiments show that Topo-44 and also longer fragments (Topo-78) can bind double stranded DNA. Surprisingly, Topo-31 does not show DNA binding activity in these assays even though it is still capable of relaxing DNA. It appears that the presence of the (HhH)2 domains stabilizes the DNA/protein complex. One possibility is that the (HhH)2 domains could play a similar role to the cap domain present in type IB enzymes, which helps to encircle the DNA during the swiveling reaction (Redinbo et al., 1998). In addition, both short fragments of topoisomerase V do not bind single stranded DNA, whereas Topo-78 can form a stable complex with single stranded DNA (data not shown). (HhH)2 domains binding to single stranded DNA has been observed before. For instance, the N-terminal 8 kDa of mammalian polymerase β, which contains a single HhH motif, binds to single stranded DNA through both helices (Kumar et al., 1990; Liu et al., 1994). The exact mode of single stranded DNA binding by Topo-78 or the possible role in relaxation or repair activities is not yet clear. The structure of Topo-61 showed that the topoisomerase active site of topoisomerase V is buried by one of the (HhH)2 domains suggesting that conformational changes are essential for the protein to bind DNA. The present structures of Topo-44 reinforce this observation and show that the (HhH)2 domains can change their position relative to the topoisomerase domain and that this change is mediated by the movement of the linker helix. The (HhH)2 domains act as rigid individual units, as evidenced by the fact that in different structures they show the same structure and relative orientation of the two HhH motifs. The topoisomerase domain also appears to be rigid showing the same structure even in the total absence of the rest of the protein. The linker helix (residues 269-295), which is a long bent helix, serves as a hinge for the movement of the (HhH)2 domains away from the rigid topoisomerase domain, possibly by responding to interactions with double stranded DNA. This movement has to be quite large. The Topo-44 structures in the absence of DNA capture the regions that move, but do not show the full extent of the movement or indicate the way the HhH motifs interact with DNA. As mentioned before, topoisomerase V binds double stranded DNA and has a groove wide enough to accommodate double strand DNA (Figure 6C). The presence of an HTH domain normally associated with DNA binding, the positively charged nature of it, and several phosphates bound along it suggest that this groove could be involved in DNA binding. In addition, the active site is found in this groove and some residues form part of the HTH domain. Previously, DNA was modeled bound to the topoisomerase domain (Taneja et al., 2006) based on the structures of HTH domains in complex with DNA but there was no evidence to support it. Using the phosphates present in the groove in the current structures, it is possible to refine this model. A superposition of the B subunit of Form II and the A and B subunits of Form III Topo-44 structures shows five different phosphate ions in the positively charged groove which are separated by a distance of around 7 Å, consistent with the distance of consecutive phosphates in B DNA of ~6.4 Å. A sixth phosphate ion is found outside the groove near the linker helix. A double stranded DNA molecule was modeled into the groove based on the positions of the phosphate ions (Figure 6). Only five out of the six phosphates could be placed on the DNA molecule, as one of them was inconsistent with a Rajan et al. Page 6 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript double stranded molecule. Phosphate ions P3, P4, and P5 would correspond to three adjacent phosphates in one DNA strand, while P1, located near the active site, would belong to the opposite strand. A final phosphate (P6) is away from the groove and near the linker helix (Figure 6A, 6B). The fit to the four inner phosphates is excellent and the DNA can be accommodated in the groove of the Topo-31 structure without the need for any major rearrangements of the protein backbone. The fifth phosphate (P6) does not fit as well and a better fit would require movement of either the protein or the DNA, but the change would be relatively modest. Several side chains would need to move, but these changes would also be minor. The major change needed to accommodate the DNA in the structures with the (HhH)2 domains present is the movement of the (HhH)2 domains away from the topoisomerase domain (Figure 6B). The movement of (HhH)2 domains should be feasible as is evident from the Topo-44 structures showing different orientations of the (HhH)2 domains. The location of the (HhH)2 domains after DNA binding is not evident, but one possibility is that they would help enclose the DNA to form a clamp around it, similar to the arrangement in type IB enzymes. In the model of the topoisomerase domain in complex with DNA, the active site residues are in close contact with the backbone of DNA. The catalytic Tyr226 is pointing towards the phosphate of the DNA backbone; Arg131 and Arg144 are positioned to stabilize the protein- DNA covalent complex. Surprisingly Glu215 also appears to interact directly with the DNA phosphate backbone. The other active site residues like His200 and Lys 218 are also near the DNA. The active site is located near the end of the groove, where it widens. At this end, the DNA fits loosely in the groove, which is spacious to accommodate the movement of the strands. The ‘constrained swiveling’ mechanism employed by type IB and IC enzymes necessitates rotation of one strand about the other after forming the covalent protein-DNA intermediate. The position of the active site at the wider end of the putative DNA binding groove would facilitate the rotation of the DNA strand at this end, while holding the rest of the DNA in place through extensive interactions along the groove. Even though type IB and IC enzymes have a similar overall mechanism of action, the structures of fragments of topoisomerase V suggest many differences. Type IB enzymes have two domains which come together to form a C-shaped clamp around the DNA (Perry et al., 2006; Redinbo et al., 1998; Stewart et al., 1998) The protein has an open stage where these domains are separate and this helps in the entry and release of the DNA from the protein active site. A wide DNA binding cavity is not observed in the topoisomerase V structures. Instead, the structures show a positively charged groove which is always present in the protein and does not require domain rearrangements to form. DNA can access this groove after a conformational change involving the movement of the (HhH)2 domains exposing the active site. The (HhH)2 domains could help enclose DNA during the swiveling of the DNA, forming a similar enclosure to the one observed for type IB enzymes. It is not known whether all HhH motifs contact DNA simultaneously, but this appears unlikely without a major rearrangement of the motifs. It is likely that only some of the HhH motifs contact DNA at any given time or that some of the motifs do not have the capacity to bind DNA. Finally, similar to type IB enzymes (Cheng and Shuman, 1998), the putative domain enclosing the DNA is dispensable for activity, although it enhances the relaxation activity markedly. Thus, it is likely that type IB and IC enzymes have several overall similarities in the way that they interact with DNA, but the atomic details are markedly different. There are still many details of the atomic mechanism of type IC topoisomerases that need to be understood. The present functional and structural studies provide new information about topoisomerase V including the observations that the Topo-31 is the minimal fragment capable of DNA relaxation, the (HhH)2 domains enhance binding of the protein to DNA, the changes in relative orientation of the domains is mediated by the linker helix, and several Rajan et al. Page 7 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript phosphate ions bind in a positively charged groove. Furthermore, the position of the phosphate ions in the groove helped in the placement of DNA in the topoisomerase domain and this provides an initial model of how topoisomerase V interacts with DNA. Thus the present study helps to establish the role of different domains more clearly, to illustrate a mechanism to drive the conformational changes needed for activity, and to suggest a possible manner of binding DNA. Additional work on structures of protein/DNA complexes and intermediates in the swiveling reaction are needed to understand the way this new type of topoisomerases interacts with DNA to perform a complex reaction. Experimental Procedures Protein purification The N-terminal 31 kDa (Topo-31: residues 1-269), and 44 kDa (Topo-44: residues 1 to 380) fragments of topoisomerase V protein were cloned into the pET15b plasmid and transformed into Escherichia coli BL21 Rosetta (DE3) cells. The N-terminal 78 kDa (Topo-78: residues 1 to 685) fragment of topoisomerase V protein was cloned into the pET14b plasmid (Belova et al., 2002) and transformed into Escherichia coli BL21(DE3) cells. For protein production, cells were grown at 37° C in LB medium containing 100 μg/ml ampicillin and 100 μg/ml chloramphenicol for Rosetta cells and LB medium with 100 μg/ml ampicillin for BL21(DE3) cells to an optical density (OD600) of 0.6. The cells were then cooled down on ice, followed by the addition of isopropyl β-D-1-thiogalactopyranoside (IPTG) to a final 0.5 mM concentration, and grown overnight at 16° C. Cells were harvested and resuspended in 50 mM Tris pH 8, 500 mM NaCl, 0.5 mM EDTA, 1 mM DTT, flash frozen in liquid nitrogen and stored at −80° C. After thawing the pellet, pepstatin (1μg/ml), benzamidine (1mM), PMSF (1mM), and Brij 58 (0.1%) were added to the cells and the protein was purified as described earlier (Taneja et al., 2006) The protein was further purified by anion exchange and gel filtration chromatography. Pure protein was concentrated and stored in 50 mM Tris pH 8, 250 mM NaCl, and 1 mM DTT. The seleno- methionine substituted Topo-44 was prepared from cells grown in a minimal medium supplemented with nutrients and salts (Doublie, 1997); protein purification followed the same procedure as for the native protein except that 5mM DTT was used in all the purification steps and for storage. Relaxation assays Relaxation assays with the different topoisomerase V fragments were carried out at pH values ranging from 4 to 10. The pH of the buffers was adjusted at 65 °C to account for the change in pH at higher temperature. The different buffers used were: sodium acetate for pH 4 and 5, MES for pH 6, HEPES for pH 7, TRIS for pH 8, CHES for pH 9, and CAPS for pH 10. Topoisomerase activity assays were performed by incubating varying amounts of protein (Topo-31, Topo-44 or Topo-78) with 0.2 μg negatively supercoiled pUC19 DNA in 50 mM of the required buffer, 30 mM NaCl, 0.2 mM or 5 mM EDTA or 1 mM MgCl2. The reactions were carried out at 65 °C for 15 min and terminated by cooling and addition of SDS to a final 1% concentration. The products were resolved on a 1% agarose gel and visualized by ethidium bromide staining. Electrophoretic Mobility Shift Assay For Electrophoretic Mobility Shift Assay (EMSA), 4 μM of a 39mer double stranded DNA oligonucleotide (5′ GCGACGCGAGGCTGGATGGCCTTCCCCATTATGATTCTT3′) was incubated with different concentrations of topoisomerase V fragments in 50 mM sodium acetate pH 5, 30 mM NaCl, 1 mM MgCl2 at 65 °C for 30 minutes. Glycerol was added to the reaction mixture to a final concentration of 8% and the products were separated on a 4 % acrylamide native gel. The gel was stained with ethidium bromide to detect the DNA. When Rajan et al. Page 8 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript a stable protein-DNA complex was formed, there was an upward shift in the band indicating a higher molecular weight complex. Crystallization Topo-31 crystals were grown using the sitting drop vapor diffusion method equilibrated against, 23% PEG 6000, 0.1 M Na citrate pH 5.5, at 22°C. For data collection, the Topo-31 crystals were cryo-protected by adding glycerol to the mother liquor to a final 20% concentration. Topo-44 was crystallized by the hanging drop vapor diffusion method under three different crystallization conditions (Forms I, II, and III). Crystal Form I grew under 0.1 M phosphate citrate pH 5, 0.2 M NaCl, 15% PEG 3350 and 8% dioxane. The crystals were cryo-protected by increasing the PEG concentration to 30%. Form II crystals grew under 0.1 M phosphate citrate pH 5, 0.2 M NaCl, 16% PEG 8000 and 1M guanidium hydrochloride. For cryo-protection, they were transferred to a solution with 1.5X reservoir solution and 20% 2,3 butanediol or 20% DMSO for 10 seconds and immediately flash frozen under liquid nitrogen. Form III crystals grew under 0.1 M phosphate citrate pH 5.5, 0.15 M sodium sulfate, 0.01 M MgCl2, 1 M guanidium hydrochloride, and 28 % PEG 3350. The crystals were grown at 30°C and were cryo-protected by increasing the PEG concentration to 40%. Further details of crystallization are presented in the Supplementary Information. Data collection and structure determination Diffraction data were collected at the Dupont Northwestern Dow and Life Science Collaborative Access Team stations (DND and LS CAT) at the Advanced Photon Source in Argonne National Laboratory. Data collection and refinement statistics are shown in Table I. All data were processed and integrated using XDS (Kabsch, 1993) and scaled with SCALA (Collaborative-Computational-Project-4, 1994). Data on the Topo-31 crystals were collected to 2.4 Å resolution. The structure was solved by Molecular Replacement (McCoy et al., 2007) using the topoisomerase domain from the Topo-61 structure (residues 1-266) (Taneja et al., 2006) as the search model. It was refined with refmac5 (Murshudov et al., 1997) and Phenix (Afonine et al., 2005) to a final Rwork of 20.0 % and Rfree of 24.8 %. Topo-44 Form I crystals diffract to 1.8 Å. The structure of Form I crystals was solved by Molecular Replacement (McCoy et al., 2007) using the topoisomerase domain from the Topo-61 structure as the search model. Model rebuilding was performed using coot (Emsley and Cowtan, 2004), and refinement using refmac5 (Murshudov et al., 1997). The final Rwork and Rfree are 17.5 % and 22.0 % respectively. For Topo-44 Form II and Form III crystals, seleno-methionine derivatized crystals were used for single-wavelength anomalous dispersion (SAD) phasing. AutoSharp (Vonrhein et al., 2007) was used for locating the selenium atoms; model building was done using coot (Emsley and Cowtan, 2004), and refinement was carried out using refmac5 (Murshudov et al., 1997) Three phosphate ions were noticed in the Form II structure; two of which present in the topoisomerase active site and are separated by a distance of ~7.5 Å. The structure was refined to a final Rwork of 24.1 % and Rfree of 28.9 %. Topo-44 Form III crystals diffracted to 1.4 Å. The final Rwork and Rfree are 16.5 % and 18.4%, respectively. An interesting observation is the presence of both phosphate and guanidium ions in the Form III Topo-44 structure. The linker helix and part of the first HhH motif of the B monomer show alternate conformations and were built as two separate chains with occupancy of 0.5 each. Further details on data collection and structure determination are given in the Supplementary Information. All figures were made with Pymol (DeLano, 2002) and the electrostatic surfaces were calculated with APBS (Baker et al., 2001). Rajan et al. Page 9 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We acknowledge staff and instrumentation support from the Keck Biophysics Facility and the Center for Structural Biology at Northwestern University, and DND and LS-CAT at the Advanced Photon Source (APS) at Argonne National Laboratory. Support from the R.H. Lurie Comprehensive Cancer Center of Northwestern University to the Structural Biology Facility is also acknowledged. DND-CAT is supported by Dupont, DOW and the NSF. LS-CAT was supported by the Michigan Economic Development Corporation and the Michigan Technology Tri-Corridor. Use of the APS is supported by the Department of Energy (DOE). Research was supported by NIH grant GM51350 (to AM). References Afonine PV, Grosse-Kunstleve RW, Adams PD. A robust bulk-solvent correction and anisotropic scaling procedure. Acta Crystallogr D Biol Crystallogr. 2005; 61:850–855. [PubMed: 15983406] Baker NA, Sept D, Joseph S, Holst MJ, McCammon JA. Electrostatics of nanosystems: application to microtubules and the ribosome. Proc Natl Acad Sci U S A. 2001; 98:10037–10041. [PubMed: 11517324] Baker NM, Rajan R, Mondragon A. Structural studies of type I topoisomerases. Nucleic Acids Res. 2009; 37:693–701. [PubMed: 19106140] Belova GI, Prasad R, Kozyavkin SA, Lake JA, Wilson SH, Slesarev AI. A type IB topoisomerase with DNA repair activities. Proc Natl Acad Sci U S A. 2001; 98:6015–6020. [PubMed: 11353838] Belova GI, Prasad R, Nazimov IV, Wilson SH, Slesarev AI. The domain organization and properties of individual domains of DNA topoisomerase V, a type 1B topoisomerase with DNA repair activities. J Biol Chem. 2002; 277:4959–4965. [PubMed: 11733530] Champoux JJ. DNA Topoisomerases: Structure, Function, and Mechanism. Annu Rev Biochem. 2001; 70:369–413. [PubMed: 11395412] Cheng C, Shuman S. A catalytic domain of eukaryotic DNA topoisomerase I. J Biol Chem. 1998; 273:11589–11595. [PubMed: 9565576] Collaborative-Computational-Project-4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D. 1994; 50:760–763. [PubMed: 15299374] Davis IW, Murray LW, Richardson JS, Richardson DC. MOLPROBITY: structure validation and all- atom contact analysis for nucleic acids and their complexes. Nucleic Acids Res. 2004; 32:W615– 619. [PubMed: 15215462] DeLano, WL. The PyMol Molecular Graphics System. San Carlos, CA: DeLano Scientific; 2002. Diederichs K, Karplus PA. Improved R-factors for diffraction data analysis in macromolecular crystallography. Nat Struct Biol. 1997; 4:269–275. [PubMed: 9095194] Doublie S. Preparation of selenomethionyl proteins for phase determination. Methods Enzymol. 1997; 276:523–530. [PubMed: 9048379] Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] Forterre P. DNA topoisomerase V: a new fold of mysterious origin. Trends Biotechnol. 2006; 24:245– 247. [PubMed: 16650908] Forterre P, Gribaldo S, Gadelle D, Serre MC. Origin and evolution of DNA topoisomerases. Biochimie. 2007; 89:427–446. [PubMed: 17293019] Gellert M. DNA Topoisomerases. Annu Rev Biochem. 1981; 50:879–910. [PubMed: 6267993] Huber R, Kurr M, Jannasch HW, Stetter KO. A novel group of abyssal methanogenic archaebacteria (Methanopyrus) growing at 110 °C. Nature. 1989; 342:833–834. Kabsch W. Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Crystallogr. 1993; 26:795–800. Rajan et al. Page 10 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Koster DA, Croquette V, Dekker C, Shuman S, Dekker NH. Friction and torque govern the relaxation of DNA supercoils by eukaryotic topoisomerase IB. Nature. 2005; 434:671–674. [PubMed: 15800630] Kumar A, Widen SG, Williams KR, Kedar P, Karpel RL, Wilson SH. Studies of the domain structure of mammalian DNA polymerase beta. Identification of a discrete template binding domain. J Biol Chem. 1990; 265:2124–2131. [PubMed: 2404980] Liu D, DeRose EF, Prasad R, Wilson SH, Mullen GP. Assignments of 1H, 15N, and 13C resonances for the backbone and side chains of the N-terminal domain of DNA polymerase beta. Determination of the secondary structure and tertiary contacts. Biochemistry. 1994; 33:9537– 9545. [PubMed: 8068628] Maxwell A. DNA gyrase as a drug target. Biochem Soc Trans. 1999; 27:48–53. [PubMed: 10093705] McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software. J Appl Crystallogr. 2007; 40:658–674. [PubMed: 19461840] Murshudov GN, Vagin AA, Dodson EJ. Refinement of macromolecular structures by the maximum- likelihood method. Acta Crystallogr D. 1997; 53:240–255. [PubMed: 15299926] Perry K, Hwang Y, Bushman FD, Van Duyne GD. Structural basis for specificity in the poxvirus topoisomerase. Mol Cell. 2006; 23:343–354. [PubMed: 16885024] Pommier Y. Diversity of DNA topoisomerases I and inhibitors. Biochimie. 1998; 80:255–270. [PubMed: 9615865] Redinbo MR, Stewart L, Kuhn P, Champoux JJ, Hol WG. Crystal structures of human topoisomerase I in covalent and noncovalent complexes with DNA. Science. 1998; 279:1504–1513. [PubMed: 9488644] Rothenberg ML. Topoisomerase I inhibitors: review and update. Ann Oncol. 1997; 8:837–855. [PubMed: 9358934] Schoeffler AJ, Berger JM. DNA topoisomerases: harnessing and constraining energy to govern chromosome topology. Q Rev Biophys. 2008; 41:41–101. [PubMed: 18755053] Slesarev AI, Stetter KO, Lake JA, Gellert M, Krah R, Kozyavkin SA. DNA topoisomerase V is a relative of eukaryotic topoisomerase I from a hyperthermophilic prokaryote. Nature. 1993; 364:735–737. [PubMed: 8395022] Stewart L, Ireton GC, Parker LH, Madden KR, Champoux JJ. Biochemical and biophysical analyses of recombinant forms of human topoisomerase I. J Biol Chem. 1996; 271:7593–7601. [PubMed: 8631793] Stewart L, Redinbo MR, Qiu X, Hol WG, Champoux JJ. A model for the mechanism of human topoisomerase I. Science. 1998; 279:1534–1541. [PubMed: 9488652] Taneja B, Patel A, Slesarev A, Mondragon A. Structure of the N-terminal fragment of topoisomerase V reveals a new family of topoisomerases. EMBO J. 2006; 25:398–408. [PubMed: 16395333] Taneja B, Schnurr B, Slesarev A, Marko JF, Mondragon A. Topoisomerase V relaxes supercoiled DNA by a constrained swiveling mechanism. Proc Natl Acad Sci U S A. 2007; 104:14670–14675. [PubMed: 17804808] Vonrhein C, Blanc E, Roversi P, Bricogne G. Automated structure solution with autoSHARP. Methods Mol Biol. 2007; 364:215–230. [PubMed: 17172768] Wang HK, Morris-Natschke SL, Lee KH. Recent advances in the discovery and development of topoisomerase inhibitors as antitumor agents. Med Res Rev. 1997; 17:367–425. [PubMed: 9211397] Rajan et al. Page 11 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Organization of topoisomerase V Topoisomerase V is a multi-domain protein consisting of 24 helix-hairpin-helix (HhH) DNA binding motifs arranged as 12 (HhH)2 domains following the N-terminal topoisomerase domain. A) Schematic diagram of various topoisomerase V fragments. The topoisomerase domain is shown in red, the (HhH)2 domains are shown in alternating colors of cyan and yellow. The (HhH)2 domains with repair activity are shown in green. All fragments shown have topoisomerase activity, but only the full length protein and the Topo78 fragment have repair activity. B) Crystal structure of Topo-61 fragment (Taneja et al., 2006). The coloring scheme is the same as in Figure 1A, except that the linker helix is shown in grey. Rajan et al. Page 12 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. DNA relaxation activity and EMSA for Topo-31, Topo-44 and Topo-78 fragments of topoisomerase V A) pH profile of the DNA relaxation activity of Topo-78 and Topo-44 fragments. 0.2 μg of pUC19 DNA were incubated with 0.1 μg of Topo-78 or 1.5 μg of Topo-44 proteins at 65°C for 15 minutes for the relaxation reaction. The reaction buffer contained 50 mM of the appropriate buffer, 30 mM NaCl and 0.2 mM EDTA. Topo-78 relaxes DNA at a wider pH range (5 to 9) than Topo-44, which relaxes DNA efficiently only at pH 5. DNA relaxation activity of Topo-31 (B) and Topo-44 (C) fragments in the absence and presence of MgCl2. 0.2 μg of pUC19 DNA were incubated with 9 μg of Topo-31 or 1.5 μg of Topo-44 proteins at 65°C for 15 minutes for the relaxation reaction. The reaction buffer contained 50 mM of the appropriate buffer, 30 mM NaCl and 5 mM EDTA or 1 mM MgCl2. Both Topo-31 and Topo-44 fragments can relax DNA in the absence of MgCl2, but MgCl2 enhances the DNA relaxation activity of the topoisomerase V fragments. The black triangle in panels A, B and C represents increasing pH from 4 to 10 by one pH unit. D) EMSA of Topo-44 and Topo-78 fragments with a 39mer double stranded DNA. Both Topo-44 and Topo-78 form stable complexes with DNA, although Topo-78 seems to saturate DNA binding while Topo-44 does not. In addition, Topo-44 shows some cleavage of the DNA (bottom free DNA band), while the cleavage is not apparent in Topo-78. The numbers at the bottom represent the molar ratio of protein to DNA. Rajan et al. Page 13 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Structure of Topo-44 fragments A) Overlay of Form I (green), Form II (B subunit: blue), and Form III (B subunit: magenta) structures onto the Topo-61 structure (B subunit: orange). The topoisomerase domains superimpose very well for all the structures, while the linker helix and (HhH)2 domains show differences in orientation. B) Overlay of the linker helices of Form I, II, and III structures with that of Topo-61. The color scheme is same for all the figures unless mentioned otherwise. Note that the linker helices have the same orientation at the start and they change as they move further down the helix. C) Superposition of Form I, II, and III Topo-44 structures with that of Topo-61. Only the (HhH)2 domains are colored while the remaining parts are shown in gray for clarity. The active site residues are shown as orange sticks. Note that the (HhH)2 domains adopt different orientations in all the structures. D) Orientation of the (HhH)2 domains of Form I, II and Topo-61 structures. In Form I and II structures, the (HhH)2 domains are moved away from the topoisomerase domain. For clarity, the (HhH)2 domains of Form III are not shown. In panels C and D, the topoisomerase domains were superposed to emphasize the different orientation of the other domains. Rajan et al. Page 14 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. Phosphate ions present near the active site of the Topo-44 structure A) Stereo view of a Form III difference electron density map calculated with a model not including the phosphates. The electron density is contoured at 3.7σ and shows the tetrahedral shape of the phosphate ions. The active site residues are shown in stick. B) Stereo view of the interaction of the phosphate ions with the putative active site residues. The B subunit of Form II structure was superimposed onto the B subunit of Form III structure and the phosphates ions from both structures are shown together with the Form II B subunit protein backbone. The interactions made by the phosphate ion with the active site residues and the corresponding distances in Å are represented as black dotted lines. Rajan et al. Page 15 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Representation of the unique phosphate ions from Form II and Form III Topo-44 structures A) An overlay of the A (magenta) and B (brown) subunits of the Form III structure and B (blue) subunit of Form II Topo-44 structures. The positions of eight unique phosphate ions (orange spheres) are shown. Note that most phosphate ions are found along the DNA binding groove of the topoisomerase domain. B) The phosphate ions in the DNA binding groove are separated by distances of around 7 Å. The protein backbone is that of the B subunit of Form III structure. The active site residues are represented as sticks and distances in Å between adjacent phosphate ions are shown as black dotted lines. Rajan et al. Page 16 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. Model showing DNA bound to the topoisomerase domain A) Model of a 17-mer double stranded DNA bound to the Topo-31 structure (teal). The DNA is represented as green sticks, where as phosphate ions are represented as orange sticks. DNA binds along the DNA binding groove and five of the eight phosphate ions noted in the Topo-44 structures coincide with the DNA backbone. B) Model of Topo-44 (Form II, B subunit: blue) binding to 17-mer double stranded DNA. Note that the linker helix and the (HhH)2 domains interfere with DNA binding to the topoisomerase domain and are likely to move away to allow binding. C) Electrostatic surface representation of the Topo-31 structure. The positively charged DNA binding groove is clearly visible and the phosphate ions are bound in this groove. The orientation corresponds to a 90° rotation of the one shown in Figure 6A in the direction of the arrow. Note that the DNA binding groove goes from one end of the molecule to the other and it is narrower at one end (start of the linker helix) and wider at the other end. The putative active site residues (green sticks) are located at the wider end of the groove. Other residues lining the groove and interacting with the phosphate ions are shown as cyan sticks. D) Electrostatic surface representation of Topo-31 with phosphate ions (orange) and DNA (green). Three phosphate ions (P3, P4, and P5) coincide with the phosphates of one of the DNA strands, where as P1 coincides with a phosphate of the opposite DNA strand. The model shows that the DNA binding groove of topoisomerase V is wide enough to bind DNA and that the movement of linker helix and (HhH)2 domains Rajan et al. Page 17 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript are required to accommodate the DNA. The electrostatic potential was calculated with a dielectric constant of 80 for solvent and 2 for protein. The surface is colored with a blue to red gradient from +10 to −10 KbT/ec. Rajan et al. Page 18 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Rajan et al. Page 19 Table 1 Data collection and refinement statistics Topo-31 Topo-44 Form I Topo-44 Form II Topo-44 Form III Data Collection Space group C2221 C121 P41212 P212121 Cell dimensions a=106.7 Å, b=119.4 Å, c=63.7 Å a=104.2 Å, b=47.7 Å, c=81.2 Å (β=112.48) a=b=70.1 Å, c=349.6 Å a=63.6 Å, b=80.1 Å, c=137.2 Å Resolution (Å)a 79.56 – 2.4 (2.53 – 2.4) 75.05 – 1.82 (1.91 – 1.82) 29.5- 2.6 (2.72-2.6) 28.9-1.4 (1.46-1.4) Number of observed reflections 78,729 (11,538 134,411 (13,220) 227,408 (19,917) 1,157,917 (126,319) Number of unique reflections 16,259 (2,346) 32,998 (4,301) 28,151 (3,331) 136,662 (15,986) Completeness (%) 99.8 (99.8) 98.3 (88.6) 99.9 (100.0) 98.8 (95.5) Multiplicity 4.8 (4.9) 4.1 (3.1) 8.1 (6.0) 8.5 (7.9) Rmerge (%)b 4.7 (71.1) 4.0 (16.3) 7.4 (52.2) 4.5 (37.9) Rmeas (%)c 5.3 (79.6) 4.6 (19.4) 7.9 (57.2) 4.8 (40.5) ≪I>/σ(<I>)>d 20.5 (2.5) 23.0 (6.8) 19 (3.2) 27.5 (5.3) Refinement Resolution (Å) 79.56 - 2.4 (2.46 - 2.4) 28.06 -1.82 (1.87 – 1.82) 29.14 – 2.6 (2.67 – 2.6) 28.9 - 1.4 (1.44 - 1.4) Number of reflections working/test 15,419/821 31,317/1,673 26,710/1,438 129,802/6,859 Rwork (%)e 20.0(24.3) 17.5 (17.9) 24.1(36.6) 16.5 (19.3) Rfree(%)f 24.8 (31.1) 22.0 (24.8) 28.9 (45.1) 18.4 (22.1) Protein residues/atomsg 269/2,203 376/3212 727/5,970 738/7,511 Atoms in alternate conformations 0 258 (20 protein residues) 8 (1 protein residue) 2846 (157 protein residues) Water molecules 29 238 30 573 Other atoms - - 3 PO4 7 PO4, 3 Gmh, 3 Mg++, 2 Cl− B-factor (Å2) Protein atoms (chain) 68.4 22.8 A:53.8; B:58.2 A:13.4; B:14.9 Water molecules 59.1 29.3 40.0 23.7 r.m.s. deviations bond lengths (Å) 0.015 0.006 0.01 0.009 bond angles (°) 1.42 0.920 1.2 1.2 Ramachandran ploti Favored regions (%) 94.3 98.9 96.2 98.5 Outliers (%) 0.0 0.0 0.3 0 aNumbers in parenthesis correspond to highest resolution shell. bRmerge= Σ|I − <I>|/ΣI, where I is the observed intensity and <I> the average intensity obtained from multiple measurements. Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Rajan et al. Page 20 cRmeas as described in Diederichs and Karplus (Diederichs and Karplus, 1997). d≪I>/σ(<I>)> = Mean Ih over the standard deviation of the mean Ih averaged over all reflections in a resolution shell. eRwork= Σ ||Fo| − |Fc||/Σ|Fo|, where |Fo| is the observed structure factor amplitude and |Fc| the calculated structure factor amplitude. fRfree: Rfactor based on 5% of the data excluded from refinement. gTotal number of protein atoms, including those in alternate conformations. hGm: guanidinum ion. iAs reported by Molprobity (Davis et al., 2004). Structure. Author manuscript; available in PMC 2011 July 14.
3M6O
Crystal structure of Arabidopsis thaliana peptide deformylase 1B (AtPDF1B)
Trapping Conformational States Along Ligand-Binding Dynamics of Peptide Deformylase: The Impact of Induced Fit on Enzyme Catalysis Sonia Fieulaine1, Adrien Boularot1, Isabelle Artaud2,3, Michel Desmadril4,5, Fre´de´ric Dardel6,7, Thierry Meinnel1*, Carmela Giglione1* 1 CNRS, ISV, UPR2355, Gif-sur-Yvette, France, 2 Universite´ Paris Descartes, UMR8601, Paris, France, 3 CNRS, UMR8601, Paris, France, 4 Universite´ Paris-Sud, IBBMC, UMR8619, Orsay, France, 5 CNRS, IBBMC, UMR8619, Orsay, France, 6 Universite´ Paris Descartes, UMR8015, Paris, France, 7 CNRS, UMR8015, Paris, France Abstract For several decades, molecular recognition has been considered one of the most fundamental processes in biochemistry. For enzymes, substrate binding is often coupled to conformational changes that alter the local environment of the active site to align the reactive groups for efficient catalysis and to reach the transition state. Adaptive substrate recognition is a well-known concept; however, it has been poorly characterized at a structural level because of its dynamic nature. Here, we provide a detailed mechanism for an induced-fit process at atomic resolution. We take advantage of a slow, tight binding inhibitor-enzyme system, actinonin-peptide deformylase. Crystal structures of the initial open state and final closed state were solved, as well as those of several intermediate mimics captured during the process. Ligand-induced reshaping of a hydrophobic pocket drives closure of the active site, which is finally ‘‘zipped up’’ by additional binding interactions. Together with biochemical analyses, these data allow a coherent reconstruction of the sequence of events leading from the encounter complex to the key-lock binding state of the enzyme. A ‘‘movie’’ that reconstructs this entire process can be further extrapolated to catalysis. Citation: Fieulaine S, Boularot A, Artaud I, Desmadril M, Dardel F, et al. (2011) Trapping Conformational States Along Ligand-Binding Dynamics of Peptide Deformylase: The Impact of Induced Fit on Enzyme Catalysis. PLoS Biol 9(5): e1001066. doi:10.1371/journal.pbio.1001066 Academic Editor: Gregory A. Petsko, Brandeis University, United States of America Received January 7, 2011; Accepted April 14, 2011; Published May 24, 2011 Copyright:  2011 Fieulaine et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was supported by the Centre National de la Recherche Scientifique (CNRS, France), grant ANR-06-MIME-010-01 (Agence Nationale de la Recherche, France), and grant #4920 from the Association pour la Recherche sur le Cancer (Villejuif, France). SF was partly supported by a postdoctoral fellowship from the Fondation pour la Recherche Me´dicale (France). AB was supported by CNRS, France. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. Abbreviations: DSC, differential scanning calorimetry; Fo, formyl; PDF, peptide deformylase; r.m.s.d., root mean square deviation. * E-mail: Thierry.Meinnel@isv.cnrs-gif.fr (TM); Carmela.Giglione@isv.cnrs-gif.fr (CG) Introduction Flexibility of proteins around their active site is a central feature of molecular biochemistry [1–5]. Although this has been a central concept in biochemistry for half a century, the detailed mechanisms describing how the active enzyme conformation is achieved have remained largely elusive, as a consequence of their transient nature. Direct structural evidence and/or kinetic analyses have only recently emerged [6–10]. Three classic ‘‘textbook’’ models are used to describe the formation of the ligand-enzyme complex: (i) the Fischer’s ‘‘lock-and key’’ model, (ii) the Koshland’s induced-fit model, and (iii) the selected-shift model or conformational selection mechanism [6–8,11–13]. In the Fischer’s ‘‘lock-and key’’ model, the conformations of free and ligand-bound proteins are essentially the same. In the induced-fit model, ligand binding induces a conformational change in the protein, leading to the precise orientation of the catalytic groups and implying the existence of initial molecular matches that provide sufficient affinity prior to conformational adaptation [14]. In contrast, the selected-fit model assumes an equilibrium between multiple conformational states, in which the ligand is able to select and stabilize a complementary protein conformation. In this case, the conformational change precedes ligand binding, in contrast to the induced-fit model in which binding occurs first. The conformational selection and/or induced-fit processes have been shown to be involved in a number of enzymes [12,13,15,16]. For several of these studies, conformational selection is proposed because the experimental data support that, even in the absence of the ligand, the enzyme samples multiple conformational states, including the ligand-bound (active) state [6]. Although direct structural evidence and/or kinetic analyses have provided clues [6–8,12,13,16], how we can distinguish whether a protein binds its ligand in an induced- or selected-fit mechanism remains critical and often controversial. The enzyme-inhibitor interaction is a form of molecular recognition that is more amenable to investigation than the enzyme-substrate interaction as there is no chemical transforma- tion of the ligand during this process. In this context, slow, tight- binding inhibition is an interesting interaction process, as it closely mimics the substrate recognition process and has been shown to be commonly involved in adaptive conformational changes [12, 17,18]. In slow, tight-binding inhibition, the degree of inhibition at a fixed concentration of compound varies over time, leading to a curvature of the reaction progress curve over time during which PLoS Biology | www.plosbiology.org 1 May 2011 | Volume 9 | Issue 5 | e1001066 the uninhibited reaction progress curve is linear [19]. Indeed, the slow, tight-binding inhibition is a two-step mechanism that depends on the rate and strength of inhibitor interactions with the enzyme. Binding of the inhibitor (I) to the enzyme (E) leads to the rapid formation of a non-covalent enzyme-inhibitor complex (E:I) followed by monomolecular slower step (k5) in which the E:I is transformed into a more stable complex (E:I*) that relaxes and dissociates at a very slow rate, mainly inferred by the k6 value when k6,,k5,,k4, (Figure 1A; see also footnote f in Table 1). Although only a few studies have investigated the mechanisms of slow, tight-binding inhibitors, such molecules are favored for use as therapeutics, as they usually exhibit unique inhibitory properties, including selective potency and long-lasting effects [20–26]. Here, we explore the precise structural inhibitory mechanism of actinonin (Figure 1A; [27]), which is a slow, tight- binding inhibitor of peptide deformylase (PDF), a metal cation- dependent enzyme [28,29]. The function of the active-site metal is to activate the reactive water molecule involved in peptide hydrolysis [30]. PDF is the first enzyme in the N-terminal methionine excision pathway, an essential and ubiquitous process that contributes to the diversity of N-terminal amino acids [31,32]. Actinonin is a natural product with antibiotic activity that inhibits PDF by mimicking the structure of its natural substrates (nascent peptide chains starting with Fo-Met-Aaa, where Fo is a formyl group and Aaa is any amino acid) in their transition state (Figure 1B). The transition state inhibitor actinonin, as well as other structurally related inhibitors, has been shown to systemat- ically exhibit a ‘‘slow-binding’’ inhibition behavior (Figure 1A), regardless of the organism of origin of the PDF [29,33]. Using structural, biocomputing, and enzymatic analyses, we were able to (i) reveal that the free enzyme is in an open conformation and that actinonin induces transition of the enzyme into a closed conformation; (ii) show that there is no evidence for the occurrence of a closed conformation in the apostructure of the open enzyme, which, together with detailed kinetic analyses, makes the closed form fully compatible with an induced-fit model; and (iii) identify the sequence of molecular events leading to the final, bound, closed complex (E:I*). Moreover, using several rationally designed point mutants of the enzyme, ligand-induced intermediates, which mimic conformational states that normally would not be expected to accumulate with the wild-type (WT) enzyme, were trapped. These conformations recapitulate physical states that the WT enzyme must pass through during its overall transition from the apo-enzyme to the E:I* complex. ‘‘Freezing’’ of ligand-induced intermediate states provides direct evidence for an induced-fit mechanism and allows the reconstruction of a virtual ‘‘movie’’ that recapitulates this mechanism. Since PDF is one example of an enzyme remaining active in the crystalline state and because actinonin closely mimics the natural substrates bound to PDF in the transition state as shown previously with the Escherichia coli form (EcPDF; see Figure 1B) [34,35], we propose a model suggesting that induced fit also contributes to efficient catalysis. Results Slow, Tight Binding of the Transition-State Analog Actinonin to Peptide Deformylase In the present study, at the atomic level we explored the precise inhibitory mechanism of actinonin on Arabidopsis thaliana PDF1B (AtPDF), a close eukaryotic homologue of EcPDF (Figure S1) [36,37]. Measurements of the kinetic parameters of the second step of the binding mechanism (k5) revealed a timescale in the 10-s range (Table 1), which is consistent with the collective motion of a large domain [4,5]. This finding is supported by NMR studies [38,39], which showed that actinonin binding induces drastic changes in the heteronuclear single quantum coherence (HSQC) spectrum of EcPDF, since most resonances undergo significant shifts that affect a large part of the structure [40,41]. The existence of alternative conformational states of EcPDF is further supported by recent biophysical studies [42]. Previously reported snapshots of a series of different conformations of the enlarged and mobile loop—the so- called CD loop—of the dimeric PDF from Leptospira interrogans PDF (LiPDF) in the presence or absence of inhibitor led to the hypothesis of the existence of an equilibrium between a closed and open form of the CD-loop of PDF enzymes, suggesting a selected-shift model to the authors [43]. Taken together, these data suggest that the binding of actinonin to PDF is accompanied or preceded by conformational changes within the enzyme. Paradoxically, this proposal has not been currently supported by the available structural data. Indeed, free and complexed crystal structures have provided no evidence for any significant conformational change in PDF structure induced by the binding of ligand [35,43–47]. Tight inhibition in the closed state is associated with the KI* apparent equilibrium constant (Figure 1A). A KI* value (see Table 1 and Materials and Methods for the biochemical definition of KI*) of 0.9 nM for actinonin could be measured for AtPDF; that is, a value very similar to that obtained for bacterial PDFs, including EcPDF and Bacillus stearothermophilus PDF2 (BsPDF2, Table 1). Tightening of the initial encounter complex (E:I) resulted in a final complex (E:I*) in which the potency of actinonin (KI/KI*) was enhanced by more than two orders of magnitude and exhibited a very slow off-rate (k6, Table 1). The dissociation constant value of AtPDF for actinonin was also assessed using isothermal titration calorimetry (ITC) experiments (Table S1 and Figure S2A). The corresponding ITC titration curves (Figure S2A) are consistent with a very strong affinity of the ligand for the enzyme [48], enabling us to determine an accurate Kd. Moreover, these studies generated values similar to those measured by other means for AtPDF and EcPDF [42,49]. Author Summary The notion of induced fit when a protein binds its ligand— like a glove adapting to the shape of a hand—is a central concept of structural biochemistry introduced over 50 years ago. A detailed molecular demonstration of this phenomenon has eluded biochemists, however, largely due to the difficulty of capturing the steps of this very transient process: the ‘‘conformational change.’’ In this study, we were able to see this process by using X-ray diffraction to determine more than 10 distinct structures adopted by a single enzyme when it binds a ligand. To do this, we took advantage of the ‘‘slow, tight-binding’’ of a potent inhibitor to its specific target enzyme to trap intermediates in the binding process, which allowed us to monitor the action of an enzyme in real-time at atomic resolution. We showed the kinetics of the conformational change from an initial open state, including the encounter complex, to the final closed state of the enzyme. From these data and other biochemical and biophysical analyses, we make a coherent causal reconstruction of the sequence of events leading to inhibition of the enzyme’s activity. We also generated a movie that reconstructs the sequence of events during the encounter. Our data provide new insights into how enzymes achieve a catalytically competent conformation in which the reactive groups are brought into close proximity, resulting in catalysis. The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 2 May 2011 | Volume 9 | Issue 5 | e1001066 Ligand-Induced Conformational Closure of AtPDF in the Crystalline State Occurrence of a conformational change induced by drug binding was visualized via the resolution of several crystal structure forms of AtPDF, the free form and/or in a complex with actinonin (Table S2). The data reveal a structural switch between the two forms that can account for both the thermodynamic and kinetic data. The enzyme was observed in two states, a novel open apo-form and a closed, induced, actinonin-bound complex (Figure 1C). Binding of actinonin resulted in a tightening of the active site through the collective closure of the entire N-terminal portion of the protein (strands b1, b2, and b3; helix a1; and CD-loop, see Movies S1 and S2, Figure 1C, and Figure S1). The amplitude of the structural change was maximal for Pro60 (Figure S1), the Ca of which was shifted 4 A˚ upon actinonin binding. This collective movement involved the formation of a ‘‘super b-sheet’’ as the result of the large rearrangement of b-strands 4 and 5 relative to the rest of the structure in which actinonin forms an additional strand bridging the two b-sheets (b1 andb2) on either side of the active site (Figure 1D and Figure S1B). As actinonin is a peptide-like compound (see Introduction and Figure 1B), this behavior closely mimics what occurs in the natural protein substrates of PDF, which also form this strand-bridging interaction. This phenomenon also accounts for the strong stabilization of the protein by actinonin, which was also challenged by differential scanning calorimetry (DSC) experiments: the Tm of AtPDF increased from 61uC to 81uC upon binding of the inhibitor (Figure 1D, see also below). Thus far, this closure of the enzyme induced by actinonin is part of the rare structural evidence for the slow, tight-binding mechanism at an atomic scale. The open state, which has never been observed, was captured not only in the two molecules of the asymmetric subunit but also in different crystals and under two distinct crystallization conditions (Table S2 and Figure 2). All r.m.s.d. values were smaller than 0.25 A˚ . The closure is very unlikely to result from crystal packing constraints, as soaking the apo-AtPDF crystals in a solution containing actinonin induced the Figure 1. Slow, tight-binding inhibition of PDF by actinonin induces conformational change in the protein. (A) Inhibition by a two-step mechanism, involving a tightening of the initial enzyme-inhibitor complex (E?I) to form a more stable complex (E?I*), with the chemical structure of actinonin (I), the natural inhibitor of PDF enzymes (E). (B) Structures of EcPDF bound to actinonin (left) and to the transition state resulting from the cleavage of its substrate, Fo-Met-Ala-Ser (right) [34,35]. (C) Superimposition of free and actinonin-bound AtPDF indicated in green and purple, respectively. The three conserved motifs of the PDF enzymes family are indicated in orange and numbered I, II, and III. Molecules A of both models were superimposed, resulting in an r.m.s.d. of 0.9 A˚ for 100% of the Ca. Left inset, close-up comparison of the open and closed forms figured in the ribbon representation. (D) Baseline-corrected DSC thermograms of free and actinonin-bound WT AtPDF recorded under the same experimental conditions. doi:10.1371/journal.pbio.1001066.g001 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 3 May 2011 | Volume 9 | Issue 5 | e1001066 structural transition from the open to the closed state within the crystals without cracking them or altering their diffracting power. Thus, crystal packing is compatible with both states of the enzyme (Figure S3). Therefore, the open structure most likely corresponds to a stable state in solution. The closed final conformation was identical to that previously reported for PDF complexes obtained either with actinonin or with a product of the reaction [34,35,44,50], indicating that this structure is common for the ligands (compare Figures 1B and 2A, and Figure S4). Hydrogen bonding was also conserved, especially the bond between the backbone nitrogen of Ile42 (corresponding to Ile44 in EcPDF, see Figure 1B and Figure S5A) and the alkyl carbonyl chain of actinonin, which potently contributes to the formation of the super b-sheet (Movie S2 and Figure S1B, see also below). Between the open and closed states, the side chains of Ile42, Phe58, and Ile130 underwent significant structural changes (Figure 3A and D and Figure S6), corresponding to a hydophobic pocket rearrangement, with Ile42 being the most affected (Figure 3). Interestingly, Ile42 is the second residue of the conserved active-site motif G41IGLAAXG (motif 1) that was previously shown to be essential for activity [51]. To assess and visualize the differences between the two states, two independent structural parameters were measured: the r.m.s.d. value with respect to the open form and the aperture angle (dap), which measures the angle made between the N- and C-domains through three fixed-points, corresponding to the Ca of three conserved residues, each sitting in one of the three conserved motifs (Figure 2A). The bi-dimensional graph of these two parameters is a good representation of the closing motion snapshots (Figure 2B) shown in Movie S1. With this tool at this stage, two states could be defined: the closed (C) and open (O) states (Figure 2B). Evidence for a Pure Induced-Fit Mechanism in the Binding of Actinonin to AtPDF Recent quantitative analyses of both conformational selection and induced fit have led to an integrated continuum—a so-called ‘‘flux-description’’—of these two limiting mechanisms [16]. According to this model, conformation selection tends to be preferred at low ligand concentrations (mM range)—that is, using detailed kinetic studies—whereas induced fit dominates at high ligand and enzyme concentrations (mM range) obtained, for instance, in NMR or crystallographic approaches. Structural studies are most useful to reveal subpopulations of biological significance. We investigated the existence of lowly populated, alternative conformations of apoPDF. To probe the occurrence of alternate conformers in the crystalline state of PDF, the new Ringer program is the most suitable investigation tool [52,53]. Ringer searches for evidence of alternate rotamers by systematically sampling electron density maps—free of model bias—around the dihedral angles of protein side chains. Two independent WT open datasets of the apoenzyme, including a high-resolution set (1.3 A˚ ), were used in the analysis. Ringer analysis revealed the existence of only one rotamer of most side chains of either molecule in the asymmetric unit, including the three main residues primarily involved in conformation change—that is, Ile 42, Phe58, and Ile130 (Figure 4A). Ringer analysis showed evidence for unmodeled alternate conformers for very few residues, including Ile121 and Phe87, or Phe119 to a much lesser extent (Figure S7). There is therefore no evidence for the occurrence of a closed conformation in the apostructure of AtPDF, supporting the hypothesis that the conformational change was essentially induced by the binding of actinonin rather than from conformational selection among multiple states occurring in the crystalline state. To further investigate the mechanism involved, we followed a kinetic approach aimed at discriminating between induced fit and population shift at low ligand concentrations (sub-mM range) [12]. The experimentally observed pseudo-first-order rate constant for the approach to equilibrium between the free components and the binary AtPDF-actinonin complex (kobs) was measured and plotted as a function of actinonin concentration. This plot yielded a hyperbolic saturation curve with a positive slope, as fully expected for a pure induced-fit mechanism (Figure 4B and C). In contrast, if the enzyme sampled two or more conformational states, the curve would imply that the value of kobs decreases with increasing ligand concentration (see, for instance, curve C in Figure 1 in [12]). The same conclusion can be reached for EcPDF and BsPDF2 (Figure 4B and C) and was already reported by others for S. aureus PDF [29]. Together, these data indicate that a pure induced-fit mechanism triggered by the binding of actinonin appears to direct the conformational change both in solution and in the crystalline state. Single Variants at Gly41 Exhibit Strongly Reduced Actinonin-Binding Potency and Catalytic Efficiency When dealing with an induced-fit mechanism, knowledge of the initial O and final C state is crucial but does not provide direct information on the position of actinonin in the encounter complex or on the sequential mechanism of the transition process. We suspected that the conserved glycine-rich motif 1 (G41IGLAAXQ) could contribute to the flexibility required for the observed structural transition. Evidence for such flexibility comes from NMR analysis of EcPDF in which a few residues show exchange cross-peaks of an additional, alternative form [38]. The most strongly affected residues are Cys90, one of the metal ligands, its neighbor Leu91, and both of the alanines within the above conserved glycine-rich motif (Figure S1B), suggesting that EcPDF undergoes conformational dynamics in a similar region. To unravel the dynamics of the recognition process, we surmised that it should be possible to freeze the conformational Table 1. Comparison of the main kinetic and thermodynamic parameters describing the inhibition of PDF by actinonin. Parameter AtPDFa EcPDFa BsPDF2a,b KI (nM)d 140610 112610 185615 KI* (nM)c 0.960.5 1.360.2 2.960.8 KI/KI* 155615 86610 6467 k5 (s21) 6103d 6369 170620 7268 k6 (s21) 6104d 461 1962 1163 k4 (s21)e 140610 112610 185615 t1/2 (min)f 2965 661 1.160.2 aThe enzyme concentrations used in the assay were 100, 50, and 25 nM for AtPDF, EcPDF, and BsPDF2, respectively. bData from [49]. cPrior to kinetic analysis for determination of the KI* value, actinonin was incubated at the final concentration in the presence of the studied enzyme set for 10 min at 37uC. The kinetic assay was initiated by the addition of a small volume of the substrate. dFor determination of KI, k5, and k6 values, actinonin was not preincubated with the enzyme. The kinetic assay was initiated by the addition of the enzyme. ek4 corresponds to the kinetic constant of the dissociation of the primary enzyme-actinonin complex. It is assumed that the rate of complex association is diffusion-limited (see Table 7.3 in [19]), that is, k3—the kinetic constant of the association of the primary enzyme-actinonin complex—is 109 M21.s21. ft1/2 is 0.693(k4+k5+k6)/k4k6 (see case of induced fit and calculation in Table 1 of [12]). In this case, t1/2,0.693/k6 because k6,,k5,,k4. doi:10.1371/journal.pbio.1001066.t001 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 4 May 2011 | Volume 9 | Issue 5 | e1001066 Figure 2. Four distinct conformational states of PDF enzymes. (A) AtPDF and several other representative bacterial PDFs are superimposed. A zoom is displayed on the right of the panel. Superimpositions were realized using ‘‘module superpose’’ in the CCP4i package and the ‘‘secondary structure matching’’ tool. The extent of aperture/closure of PDF enzymes was assessed primarily by the measurement of the aperture angle (dap), the angle made between the Ca of three strictly conserved residues (C, H, and I) of all PDFs, each characterizing a secondary crucial structure module of the active site crevice, namely b4, a2, and b1 (see Figure S1C). Each single residue belongs to one of the three conserved motifs (motifs 2, 3, and 1, respectively) and corresponds respectively to Cys91, His137, and Ile42 in AtPDF. The dap was measured in each case (see B). (B) The dap values combined with those of the r.m.s.d. associated with the superimposition of the open structure of AtPDF allows the identification of four conformational states: open (O), intermediate (I), closed (C), and super-closed (S). We compared AtPDF1B (this work and PDB CODE 3CPM; brown, orange, and yellow in A and B; black in C), EcPDF (1BS7, free enzyme; 1BS6, with Met-Ala-Ser; 1G2A, with actinonin; magenta), BsPDF2 (1LQY, with actinonin; green), LiPDF (1SV2, free; 1SZZ, with actinonin; red), and PfPDF (1JYM, free; blue). doi:10.1371/journal.pbio.1001066.g002 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 5 May 2011 | Volume 9 | Issue 5 | e1001066 Figure 3. Effect of actinonin binding on the conformation of key residues in PDF. Conformation of key residues Ile42, Phe58, and Ile130 in the different complexes: (A) in unbound WT AtPDF, (B and C) in the structure of G41Q and G41M actinonin-bound variants, respectively, and (D) of actinonin-bound WT protein. In the final complex (D), a hydrogen bond is formed between actinonin and the peptidic bond, which links Gly41 and Ile42. During the deformylation reaction, which is catalyzed by the PDF enzyme, the N-terminal formyl-methionine fits into the S1’ pocket. The solvent-accessible surface of this pocket is represented here, and only the aliphatic chain of actinonin is shown, mimicking the N-terminal methionine. (E) Free WT enzyme with the S1’ pocket shown open in two orientations (top and bottom). (F and G) S1’ pocket in the G41Q and G41M variant structures, respectively, shown in two orientations (top and bottom). (H) After the complete conformational modifications of actinonin-bound WT protein induced by actinonin binding, the S1’ pocket is shown closed in two orientations (top and bottom). (I) The four models are superimposed; the ligand-binding site is magnified: unbound WT AtPDF; G41Q and G41M actinonin-bound enzyme; and WT actinonin-bound enzyme are indicated in brown, red, orange, and yellow, respectively. Actinonin is indicated by lines. (J) A detailed view of the AtPDF ligand-binding site for all the complexes, which are superimposed, as indicated in the same colors. Arrows indicate the direction of the closing movement within the enzyme, from the open, unbound state to the closed, bound state. doi:10.1371/journal.pbio.1001066.g003 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 6 May 2011 | Volume 9 | Issue 5 | e1001066 change along the pathway by introducing selected, minor variations within the above-mentioned crucial residues involved in the collective motion. In this respect, site-directed mutagenesis of AtPDF was performed on Gly41, Ile42, and Ile130. Single substitutions were made at Gly41 (G41A/Q/M), Ile42 (I42A/F/ N/W), and Ile130 (I130A/F), and the variants were purified and characterized. These mutant proteins showed no change in overall stability, as evidenced by DSC experiments (unpublished data). However, two variants of G41, G41Q and G41M, showed dramatic effects; the kcat/Km values were reduced by three orders of magnitude due to large decreases in the kcat values compared to the WT enzyme (Figure 5A and Table S1). The reduced kcat/Km values suggest an altered ability of these variants to attain the final enzyme-transition state complex and, as a result, to give rise to possible states different from the final E:I* complex. Substitutions at positions 42 and 130 only caused small reductions in the kcat values (Figure 5A, Figure S2C, and Table S1). The actinonin- binding potency of both G41 variants was also greatly reduced (Table S1 and Figure S2B). The time-dependent inhibition by actinonin of the most active variants was then studied (Table S3). Figure 4. Evidence for an induced fit in crystalline and solution states of AtPDF. (A) Absence of evidence for alternative conformers in the crystalline state of AtPDF. Ringer plots of electron density (r) versus x1 angle for representative residues of the 3D apo-structure of AtPDF. Data were obtained with the 3M6O, 3PN2, and 3PN3 datasets (2.0 and 1.3 A˚ resolution, respectively, see Table S1). The secondary peaks in the Ile residues are observed because Ile is a branched amino acid. To reveal an alternative conformation with Ile, three peaks should be observed. (B) kobs is a saturable function of actinonin with various PDFs, including AtPDF. Data obtained for kobs, the experimentally observed pseudo-first-order rate constant for the approach to equilibrium between the free components and the binary PDF-actinonin complex, were obtained at various concentrations of actinonin in the presence of EcPDF, AtPDF, and BsPDF2. A direct plot is shown. Inset, time-course measurement of deformylation as a function of varying actinonin concentrations. (C) Inverted plot of the data in panel B, which is expected to be a straight line if the kobs is ..k6 in the case of induced fit [19]. The correlation coefficient of each line is 1.00, 0.99, and 1.00 for AtPDF, BsPDF2, and EcPDF, respectively, indicative of the accuracy of the conclusion. doi:10.1371/journal.pbio.1001066.g004 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 7 May 2011 | Volume 9 | Issue 5 | e1001066 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 8 May 2011 | Volume 9 | Issue 5 | e1001066 The half-lives of the final complexes—as assessed by comparison of the 1/k6 values—were always significantly smaller (Table S3), suggesting that the conformational change induced by actinonin binding still occurred, but the C state is destabilized relative to the O state in the mutants compared to the WT. Accordingly, actinonin strongly stabilized almost all of the variants; Tm was increased by more than 20uC. This differs from the G41M and G41Q variants, which both showed increases in the Tm of only 12uC, consistent with reduced binding potency (Table S1). Conformational Changes of Gly41 Variants Are Affected On-Pathway The two most interesting variants, G41Q and G41M, could be crystallized under the same conditions as the WT protein. In the case of G41Q, the structure of the apo-protein did not show any modifications compared to the WT structure and remained in an O conformation (Figure 2B; ‘‘O’’ zone). In contrast, the 3D structure of the G41M variant showed that the asymmetric unit was composed of two molecules with distinct structures. One molecule (chain A) is in the O state and is similar to the structures of the WT and the G41Q variant (Figure 2B; zone ‘‘O’’). The second molecule (chain B) is in a C state, closer to that observed for the WT chain in the presence of actinonin (‘‘C’’), a so-called ‘‘superclosed’’ state (Figure 2B; zone ‘‘S’’), suggesting that the substitution modified the equilibrium between the two states in solution either (i) at the step of protein synthesis by providing two conformers, the inter-conversions of which are blocked due to steric hindrance brought by the new bulkier side-chain at position 41, or (ii) by dramatically unbalancing the free inter-conversion between the O and S conformers towards the S state. Ringer analysis indicates that in the free G41M variant, many residues show evidence for unmodeled alternate conformers—including positions 58, 42, and 130—in keeping with the second hypothesis. For all variants of position G41, addition of actinonin to the crystal (Figure 3 and Figure S6) induced a closure of the protein within the crystal. Nevertheless, as expected from in silico graphic modeling followed by energy minimization, the occurrence of a bulky side chain at position 41 prevented the completion of the closure in the presence of the ligand and, hence, the formation of the hydrogen bond between the backbone nitrogen of Ile42 and actinonin. This finding is consistent with the strongly reduced Tm of the complex of the variants with actinonin compared to WT as measured by DSC. Remarkably, both S and O forms of the G41M apo-structures in the asymmetric unit of the crystal yielded a unique intermediary structure (‘‘I’’ state) upon actinonin binding (r.m.s.d. between the molecules is ,0.25 A˚ ; see also Figure 2B, zone ‘‘I’’). In this case, it is likely that the induced-fit mechanism drives the equilibrium by capturing only the O population and closing it to an intermediary step, thus depleting the pool of O conformers that is shifted sequentially back from the remaining pool of S conformers and allows the complete binding of actinonin to the enzyme. In line with the rational design of the PDF mutants, the extent of the structural differences suggests that the underlying motions are dependent on the length of the side chain (Figure S8). Together, these data account for the reduced catalytic rate, as the hydrogen bond is strictly required for the substrate to be efficiently cleaved by PDFs (Figure S8A) [54]. Therefore, from both structural and kinetic analyses, each substitution most likely reproduces intermediates along the pathway that lead to the closure of PDF around its substrate (Figure S2B). Conformational Changes of Gly41 Variants Recapitulate Closing Intermediates Analysis of the structures allows us to propose the following sequence of atomic events (Figures 3 and 2B and Figure S6). To name the various sites of the ligand and subsites of PDF, we will use the usual nomenclature found in [55], which defines the various binding pockets of a protease, where P1’ is the first side chain at the C-terminal side of the cleavage site and its binding pocket is S1’, also referred to as the hydrophobic pocket in the case of PDF. First, actinonin aligns along the S1’ pocket to form the encounter complex, which shifts the Ile130 side chain to avoid steric hindrance in the S1’ pocket, promotes rotation of the Ile42 side chain, and finally rearranges the phenyl group of Phe58. These events achieve an optimal hydrophobic S1’ pocket conformation (Figure 3), and the concomitant closure leads to the formation of a hydrogen bond between the first carbonyl group of actinonin and the backbone nitrogen of Ile42. The initial N-O distance is reduced from 5 A˚ to 2.8 A˚ , which is an optimal value for hydrogen bonding (Movie S2 and Figure S8B). Thus, the primary driving force for the active site closure appears to be the P1’:S1’ hydrophobic interaction. The C state is ultimately locked by the super-b-sheet hydrogen bonds extending across the ligand, including those involving Ile42. The DDGbinding value (2.2– 2.4 kcal/mol, Figure S8B), as calculated from the Kd values for actinonin binding to wild-type (WT) and G41M and G41Q, is consistent with the loss of a hydrogen bond that also contributes to the conformational stability of the protein [56,57]. Thus, this bond contributes to the major binding free energy difference between the two complexes (3.1 kcal/mol; Figure S8B, Tables S1 and S3, and [29]). Interestingly, the above DDGbinding values also correlate with the DDGES values derived from the kcat/Km and kcat measurements [19]. This dataset strongly correlates with the Figure 5. Inhibition and enzymatic reactions progress through an induced fit pathway. (A) The catalytic parameters Km and kcat, for all AtPDF variants are provided as a percentage of the wild-type values (WT). Detailed values are presented in Table S2. (B) Schematic model for actinonin binding to AtPDF in favor of an induced-fit pathway. PDF might exist in at least two conformational states, open (O) or closed (C). The relative abundance of each conformation would vary, depending on the enzyme type. With AtPDF, it is likely that the most abundant form is the O one, which is the only form leading to a productive complex. The superclosed form (S) is likely to show reduced affinity for the ligand because of steric occlusion of the active site. At the initial stage, the inhibitor (shown in red) binds to AtPDF (indicated in brown) in the O conformation. To reach the final key-lock state (productive closed conformation, C), two major and extreme pathways can be used. According to the conformational selection pathway, the inhibitor selects the C conformation. This pathway, which is represented by the dashed arrow, does not occur within the crystal. In contrast, the G41Q and G41M mutants, by providing the structure of the enzyme in intermediate conformations (I), prove the existence of the so- called encounter complex and confirm that the inhibitor binds to the enzyme when it is in the O conformation. The ligand-binding site is then reorganized to yield the C enzyme conformation, that is, the key-lock state. Indeed, the inhibitor binds to the enzyme through the induced-fit pathway. Each timescale was calculated using the data available in the text and corresponds to t1/2 values deduced from the calculation of 0.693/ (kinetic constant of interest). The kcat value (k2) was used to assess the timescale of catalysis in panel C, whereas, in (B), k4 assesses the first step of inhibition, and k6 is used in the case of the slow step. For the SO conversion (left, B), the lifetime of the minor form of EcPDF was used to assess the order of magnitude (see text and [38]). (C) Schematic model for the deformylation reaction catalyzed by PDF. Since actinonin is a pseudo-peptidic inhibitor, it is likely that a peptidic substrate can bind to the PDF enzyme through an induced-fit pathway, as described in (B). The key-lock state represents a transition state in which the N-formylated substrate is deformylated to yield the final reaction product. doi:10.1371/journal.pbio.1001066.g005 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 9 May 2011 | Volume 9 | Issue 5 | e1001066 gyration and van der Waals radii of the side chain at position 41 as well as the N-O distance between the first carbonyl group of actinonin and the backbone nitrogen of Ile42 (Figure S8). These results suggest that the capacity of both G41M and G41Q variants to form the transition state is a consequence of their inability to reach the fully closed state. Thus, our study of the designed Gly41 mutant enzymes reveals that, in addition to the initial and final states observed for the WT enzyme, the conformations of the Gly41 variants correspond indeed to on-pathway intermediates, thus providing snapshots along the trajectory from the O to the C state of the enzyme (Figures 2B and 3). The 3D structure of the variants in the absence of ligand is similar to that of WT, and a strict correlation exists between the completeness of the conformational change and both binding potency and catalytic efficiency. This suggests that both events require complete protein closure to generate a productive complex. The strong stabilization of AtPDF by actinonin (Figure 1D) closely mimics what occurs with its natural substrates when it reaches the transition state [34,58]. Indeed, as expected, the enzyme facilitates the final C conformation by lowering its final energy [6]. Optimal arrangement of the S1’ pocket (Figure 3) proceeds along the reaction process towards the final C conformation, triggering the alignment of reactive groups in an optimal arrangement for ligand recognition. Upon binding, actinonin alters the thermodynamic landscape for the structural transition between the O and C states. This ligand is a potent inhibitor because it can trigger the above sequence of events similar to the substrate, but unlike the substrate, it is non- hydrolyzable. Thus, by mimicking the transition state and being non-hydrolyzable (Figure 1B), the final C complex is long lasting. Ligand-Induced Conformational Closure Is Initially Triggered by the Binding of the P1’ Group in the S1’ Pocket Given the similarity between actinonin and natural substrate binding, the very slow kinetics of inhibitor binding (10-s time-scale) remains puzzling compared to the 10 ms required for catalysis (deduced from the kcat). This finding could be explained as a conformational effect during the formation of the hydrogen bond, aligning the substrate as an additional beta-sheet and eventually stabilizing the entire enzyme-ligand complex. The significantly longer time needed to reach the most stable state compared to the substrate would most likely be due to the presence of the flexible and one carbon longer metal-binding group in actinonin (i.e., hydroxamate versus formyl, Figure 1B). This suggestion is in line with the overall data obtained when we investigated more deeply the role of the first carbonyl group of the ligand. This group is well known to exert a crucial effect in both productive and unproductive ligand binding (i.e., substrate and inhibitor) [54]. In this respect, we studied the binding of compound 6b (Figure S5B), a PDF ligand that does not exhibit a reactive group at this position [49]. We observed that this compound binds strongly to both EcPDF (KI* = 6366 nM) and AtPDF (KI* = 400635 nM) but, unlike actinonin, does not display slow, tight binding as KI* = KI. This impact on binding is consistent with the absence of the hydrogen bond involving the first carbonyl group of the ligand. The 3D structure of AtPDF was determined after soaking the compound in crystals of the free, open AtPDF form. Upon binding, 6b induced a complete conformational change, identical to that observed with actinonin (Figures 2B and 6A; ‘‘O’’ state). This result further suggests that the conformational change is not induced initially by the formation of this hydrogen bond and that the encounter complex is primarily driven by the fit within the S1’ pocket. This also reveals that the timescale of the large conformational change is several orders of magnitude faster than the kinetics of slow binding and fully compatible with both the first step of actinonin binding (k4 = 140 s21; see Table 1) and the catalytic rate of the substrate (kcat = 37 s21; see Table 1 and Table S3). The 3D structure also revealed that both the P1’ and the hydroxamate groups are bound similarly to the corresponding groups of actinonin (Figure 6B). As expected, no additional bonding occurs, especially around the backbone nitrogen of Ile42 (Figure 6C). Taken together, these data allow us to conclude that the conformational change observed upon ligand binding is triggered primarily by binding in the S1’ pocket. As revealed by the binding of 6b, the one carbon longer metal-binding group fits, immediately upon recognition of the P1’ group, in the S1’ pocket and forms a bidentate complex with the metal cation, mimicking the transition state as a result. Thus, the active site is very confined and rigid due to the presence and length of the hydroxamate group (compare right and left panels in Figure 1B). As a result, compared to the complex made with the substrate, it is likely that the formation of the hydrogen bond involving the carbonyl of actinonin and the backbone nitrogen of Ile42 becomes strongly rate-limiting (k5 = 0.044 s21; Table 1). Once this hydrogen link is locked, the uncleavable bond, mimicking the labile formyl group at the transition state, stabilizes the enzyme-inhibitor complex, making it long-lasting (k6 = 0.0006 s21; Table 1) and providing a mechanistic explanation for the slow-binding effect that involves both large and fine conformational changes. The large conformational change is similar to the one occurring with the substrate, whereas the second is more subtle and locks the hydrogen bond involving the backbone nitrogen of Ile42. The second step is rate-limiting with some transition state analogs such as actinonin (Figure 5B and C). Proper Positioning of the Carbonyl Group Is Required to Stabilize the Complex at S1’ Compound 21 corresponds to another interesting derivative designed to probe the impact of the peptide bond in PDF binding [49]. In addition to the hydroxamate group, this compound features both a hydrophobic benzyl group at P1’ and a reverse peptide bond. Compound 21 shows modest but significant inhibitory potency to AtPDF1B (KI* = 400637 nM), confirming the crucial role of the peptide bond in PDF binding. After soaking with crystals of apo-AtPDF, compound 21 could be detected in high-resolution electron density maps (Figure S9A). Unlike 6b, 21 did not bind the active site of the enzyme but an alternative pocket at the surface of the protein (Figure S9B). A docking study performed with EcPDF had previously revealed this alternative binding pocket (Figure S9C; [59]). The aforementioned data indicate that the occurrence of a S1’- binding group placed in the unfavorable context of a reverse peptide bond does not stably promote binding at the active site of AtPDF. Upon binding of 21, the 3D structure of both molecules of the asymmetric unit remain in an O conformation (r.m.s.d. ,0.2 A˚ with respect to the apo-structures in the ‘‘O’’ state). This finding suggests that only the binding of compounds entering the S1’ pocket, such as actinonin or 6b, induces conformational change, in keeping with the crucial role of the P1’ group if located in the frame of a classic peptide bond. Moreover, we noticed that the binding pocket of 21 was located on the rear side of the true S1’ pocket and induced a weak modification of the P1’ hosting platform (Figure S9D). Indeed, when crystals of the 21:AtPDF complex were soaked in actinonin, the final 3D structure no longer showed evidence of compound 21 occupancy greater than 5%. Instead, this structure revealed both actinonin and closing of the protein (Table S2). The r.m.s.d. between this structure and that The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 10 May 2011 | Volume 9 | Issue 5 | e1001066 obtained directly with actinonin was less than 0.2 A˚ ; the actinonin position was virtually identical, indicating that the protein had retained full capacity for binding actinonin and closing despite the presence of compound 21. We conclude that actinonin does compete with 21 because of the overlap at P1’ of AtPDF1B (Figure S9C). As the actinonin S1’ subsite strongly mimics that of a true substrate, this result also explains the inhibitory behavior of 21 towards AtPDF. Discussion Although PDF catalysis has been extensively studied and the mechanism has been elucidated [34], how the enzyme achieves the catalytically competent state remains unknown. Here, we provide insight on how the enzyme might reach a catalytically competent conformation, demonstrating that the reactive groups move into proximity to promote catalysis (Figures 2B and 5C). We suggest Figure 6. Effect of 6b binding on the conformation of key residues of PDF. Superimposition of free, 6b-, and actinonin-bound AtPDF indicated in brown, red, and yellow, respectively. (A) Molecule A in the three models was superimposed, resulting in an r.m.s.d. of 0.9 A˚ for 100% of the Ca. Actinonin is shown in yellow and 6b in red. (B) Conformation of key residues Ile42, Phe58, and Ile130 in the different complexes and in unbound WT AtPDF. Actinonin is shown in yellow and 6b in red. (C) A detailed view of the AtPDF ligand-binding site for both actinonin and 6b complexes, which are indicated by sticks and are superimposed. The two ligands are colored in pale and dark grey, respectively. The hydrogen bond made by actinonin only is shown. doi:10.1371/journal.pbio.1001066.g006 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 11 May 2011 | Volume 9 | Issue 5 | e1001066 that the motions of the catalytic centre starting with free ligand- PDF favor a final configuration that is optimal for binding and/or catalysis (Figures 1B, 2B, and 5B and C). In our model, we propose that free PDF might exist in at least two conformational states, that is, open (O) or super-closed (S). The relative abundance of each conformation varies by enzyme type and incubation conditions, explaining why both conformations have not been trapped thus far. In the case of AtPDF, it is likely that the most abundant form corresponds to an O state, which is the form that leads to a productive complex. Indeed, in the NMR spectra for EcPDF, a few residues show exchange cross-peaks from an additional, alternative form [38]. The most strongly affected residues are Cys90, one of the metal ligands, its neighbor Leu91, as well as Ala47 and Ala48 on the facing strand. This suggests that EcPDF exists in at least two conformations (‘‘S’’ and ‘‘C’’; see Figure 2B), which undergo slow interconversion on the NMR timescale. The 3D structure of the major conformation (75%, lifetime 300 ms) could be solved at high resolution, but the structure of the minor form (25%, lifetime 100 ms), which exhibits very weak signals, could not be solved [38]. This conformation appears to correspond to that of the complex obtained with the product of the reaction (Met-Ala-Ser). A very similar situation—although more balanced between the two states—appears to occur in the case of variant G41M, suggesting that a mechanism involving conformational selection followed by induced fit is a general model for PDF and that AtPDF is a specific case where population shift virtually does not occur as the free enzyme is completely in the O conformation. This is also in line with data obtained with L. interrogans PDF (LiPDF), which reveal conformers in both the S and C states (see Figure 2B) and suggest a population-shift mechanism [43]. It is interesting to note that LiPDF is a poorly active PDF [60]. According to the representation shown in Figure 2B, Plasmodium falciparum PDF (PfPDF), a poorly active PDF [61], was retrieved only in the S state. Finally, weak decompaction of the structure of Bacillus cereus and Staphylococcus aureus PDFs in the presence of actinonin have been described [45,46]. These examples suggest that the enzyme is trapped in the S conformer in the free state and converts to the C conformer when bound to actinonin, suggesting that the S conformer is overrepresented in solution compared to the O state, unlike AtPDF. This study of AtPDF—including 10 different crystal structures of apo- and complexed enzyme variants—reveals the 3D structure of a PDF in at least four distinct states. This includes the O form, the occurrence of which is crucial for catalysis, as it is the active form. Here, we propose that the transition from the O to the C state is directly induced by the ligand. Indeed, the O form, which is captured in the crystal, undergoes closure directly upon ligand binding in our soaking experiments. Progression to this closure involves intermediary states (‘‘I’’) similar to those observed with variants G41Q and G41M in the presence of actinonin (see Figure 2B). Extrapolating the situation to catalysis, which occurs in the crystalline states of PDF, it is likely that hydrolysis of the substrate frees the enzyme in its S state, which in turn needs to open to accommodate a new substrate (Figures 2B and 5C). This is well illustrated in the 3D structure of EcPDF complexed with a product of the reaction, obtained after co-crystallization of the enzyme with the substrate in a closed conformation [34]. The S free form is likely to exhibit a slower on-rate for the ligand (k3) compared to the O form because of steric occlusion of the active site (Figure S10). In support of this hypothesis, recent data show that the 3D structure of a C-terminally truncated, poorly active version of AtPDF is in the C conformation in the unbound state, although crystallized under conditions identical to ours [62,63]. This structure is similar to that of chain B, one of the two molecules of the asymmetric subunit of variant G41M (Figure 2B). This suggests that alterations remote from the active site significantly unbalance the equilibrium between the two conform- ers, thus altering the efficiency of the reaction (Figure 5C). As the S version corresponds to a significantly less active version of AtPDF compared to that reported in our present work, this further confirms that, compared to the O state, the S state has a significantly weaker propensity to bind substrate or a close mimic ligand, such as actinonin. Comparison of the 3D structures of the free-closed and the ligand-bound-closed forms reveals some differences responsible for the slight steric reduction of the active site of free-closed AtPDF1B with respect to that of the actinonin- AtPDF1B complex (Figure S10A), including the side chain of Ile42 burying the S1’ binding pocket (Figure S10B). Overall, these data suggest that an S form might exist under the free state but that it would feature a k3 value with respect to the ligand that is significantly weaker than that of the O form, which would strongly slow down the reaction or the binding as a result. With the interaction scheme proposed in our model (Figure 5B and C), the ligand/substrate binds more easily to the O form and induces the optimal conformation of the enzyme to reach the transition state, thus allowing the reaction to be efficiently catalyzed. In the final model (Figure 5C), there is both conformational selection and induced fit subsequently involved in line with the recently proposed existence of such mixed mechanisms for other enzymes [15,16]. Nevertheless, in our model (Figure 5C), we suggest that induced fit is the primary mechanism, as it provides energy input from the ligand, which eventually drives the enzyme towards the productive key-lock complex. Unambig- uous distinction between the relative contributions of the two mechanisms is deduced from the observation that kobs is a saturable function of actinonin with various PDF, including EcPDF, BsPDF, AtPDF (Figure 4B and C), and S. aureus PDF [12,16,29,49]. Using crystallographic reconstruction analysis involving enzyme variants, motions of small mobile loops and movie reconstructions of snapshots of catalytic events have been previously documented [1–3,64–66], often by visualizing the binding of unnatural inhibitors and not necessarily mimicking closely the substrate and transition state as actinonin does [67,68]. However, only a few examples make use of soaking conditions of a crystal to promote the motion and show the importance of induced fit [1,69]. None of these data show a motion of the amplitude revealed here with PDF and a large stabilization of the complex involving the formation of the four-stranded b-sheet superstructure and the entire N-domain of the enzyme. Compared to previous crystallographic analyses, our work integrates biophysical, computational, and kinetic analyses to reconstruct the whole picture, allowing a better understanding of the slow-binding mechanism. While our work primarily focused on an induced-fit mechanism of enzyme inhibition and catalysis, it should be emphasized that this phenomenon is also applicable to the broader area of receptor-ligand interactions. For example, in all cases where conformational change mechanisms have been proposed for kinase inhibitors without supporting experimental data [12,26], further experimental work must be provided to clarify the precise mechanism. We expect this will have important implications on how one conducts future drug-discovery efforts against such enzymes [70]. Materials and Methods Protein Expression and Purification Expression and purification of mature Arabidopsis thaliana PDF1B and all variants (i.e., AtPDF) were derived from the previously The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 12 May 2011 | Volume 9 | Issue 5 | e1001066 described protocol [37]: the lysis supernatant after sonication was applied on a Q-Sepharose column (GE Healthcare; buffers A and B as described containing 5 mM NiCl2) followed by Superdex-75 chromatography (GE Healthcare) using buffer C consisting of buffer A supplemented with 0.1 M NaCl. For crystallization experiments, the protein was purified further. The sample was concentrated on an Amicon Ultra-15 centrifugal filter unit (Millipore Corp.) with a 5-kDa cut-off and applied to a MonoQ HR5/5 column (GE Healthcare) previously equilibrated in buffer A (50 mM Hepes, pH 7.5, and 5 mM NiCl2). Elution was performed with a 50-mL gradient from 0% to 100% buffer B. The buffer of the pooled purified AtPDF1B was exchanged using a PD-10 desalting column (GE Healthcare) to yield a protein solution in 50 mM Hepes, pH 7.5, 0.1 M NaCl, and 5 mM NiCl2 (buffer C). The protein was concentrated on an Amicon Ultra-15 centrifugal filter unit. The resulting AtPDF1B preparation was frozen in aliquots and stored at 280uC (for crystallization purposes) or diluted 2-fold in 100% glycerol and stored at 220uC (for enzymatic purposes). The typical yield was 5–10 mg AtPDF per liter of culture. All purification procedures were performed at 4uC. Samples of the collected fractions were analyzed by SDS-PAGE on 12% acrylamide gels, and protein concentrations were estimated from the calculated extinction coefficients for each variant. Site-directed mutagenesis of AtPDF sequence in plasmid pQdef1bDN [36] was carried out using the QuickChange Site- Directed Mutagenesis Kit (Stratagene). Enzymology Assay of PDF activity was coupled to formate dehydrogenase, where the absorbance of NADH at 340 nm was measured at 37uC as previously described [71]. For measurements of classical kinetic parameters (i.e., Km and kcat), the reaction was initiated by addition of the substrate Fo-Met-Ala-Ser to the mixture containing purified enzyme in the presence of 1 mM NiCl2. The kinetics parameters were derived from iterative non-linear least square calculations using the Michaelis-Menten equation based on the experimental data (Sigma-Plot; Kinetics module). For determination of kinetic parameters related to actinonin, the reaction mixture contained 750 mM NiCl2. In some cases, the mixture containing PDF and actinonin was incubated for 15 min at 37uC before kinetic analysis, which was initiated by the addition of substrate. The same protocol was used to determine the dissociation constant of actinonin [KI* = k4/(k3+k3k5/k6)], but the initial reaction velocities were measured with varying concentrations of Fo-Met-Ala-Ser and actinonin. The data were then calculated according to the method of Henderson, which can be used to determine the dissociation constant of the tight-binding competitive enzyme inhibitor [28,49,72] by varying both the inhibitor and substrate concentra- tions. To determine KI, k5, and k6, the reaction was initiated by the addition of enzyme as previously described [29,49]. KI*app measurements were used for comparative studies of AtPDF variants (Table S3) at a concentration of 2 mM substrate by varying the concentration of actinonin. KI*app is the slope of the v[Actinonin]/v0 line curve. kobs was fitted from the kinetic data without preincubation with vI = vs + (v0 2 vs)e2kobst where vI is the observed velocity at a given concentration of inhibitor I, v0 is the velocity, and vs is the steady-state velocity [18]. From the set of values obtained at various concentrations of I, k5 and k6 could be derived using kobs = k6 + k5[I]/(KI + [I]). By choosing a set of values with kobs..k6, 1/kobs = 1/k5(KI/[I] +1) and 1/kobs = f(1/[I]) is expected to be a straight line in case of induced fit whose positive slope corresponds to 1/k5. k6 was derived from equation k6 = k5/ (KI/KI*21) [18,19]. Microcalorimetry ITC experiments were performed using a VP-ITC isothermal titration calorimeter (Microcal Corp.). Experiments were per- formed at 37uC. For each experiment, injections of 10 mL actinonin (180 mM) were added using a computer-controlled 300 mL microsyringe at intervals of 240 s into the Ni-AtPDF variant solution (5 to 10 mM, cell volume = 2.1 mL) dissolved in buffer C with stirring at 310 rpm. A theoretical titration curve was fitted to the experimental data using the ORIGIN software (Microcal). This software uses the relationship between the heat generated after each injection and DHu (enthalpy change in kcal/ mol), KA (the association binding constant in M21), n (number of binding sites per monomer), total protein concentration, and free and total ligand concentrations. The thermal stability of the WT and variants of Ni-AtPDF1B was studied by DSC using VP-DSC calorimetry (Microcal Corp.). DSC measurements were made with 10 mM protein solutions in buffer C. The actinonin concentration was 20 mM. The same buffer was used as a reference. All solutions were degassed just before loading into the calorimeter. Scanning was performed at 1uC/min. The temperature dependence of the partial molar capacity (Cp) was expressed in kcal/K after subtracting the buffer signal using Origin(R) software. Crystallization and Soaking Experiments Crystallization conditions were screened by a robot using the sitting drop vapor diffusion method. Crystals were obtained and optimized at 20uC with 15%–20% PEG-3350 and either 0.1 or 0.2 M zinc acetate. The drops were formed by mixing 2 mL of a solution containing 2 to 4 mg/mL protein and 2 mL of the crystallization solution. Crystals were soaked for 24 h by adding actinonin to the crystallization drops at a final concentration of 5 mM. Cryoprotection was achieved by placing crystals for 30 s in a solution that was composed of 20% PEG-3350 and 0.2 M zinc acetate, supplemented with 5%, 10%, and 15% glycerol. Crystals were then directly flash frozen in liquid nitrogen using cryoloops (Hampton Research). Crystals were also grown under conditions described for the C-terminally deleted, weakly active version of AtPDF [63]. X-Ray Diffraction Data Collection Data collections were performed at 100 K at the European Synchrotron Radiation Facility (Grenoble, France) on station ID29, FIP-BM30A, ID14-1, and ID23-2, and at SOLEIL (Gif-sur- Yvette, France) on station PROXIMA1. In each case, a single crystal was used to collect a complete dataset. Data were processed and scaled using XDS software [73]. Two crystal forms were encountered with different cell parameters. In each case, b parameter was nearly equal to a, and data could be indexed into two space groups, P212121 or P43212. The data are shown in Table S2. Structure Determination and Refinement The structure of free AtPDF was solved by molecular replacement with Phaser [74] followed by a rigid-body refinement by CNS [75] using coordinates from the Plasmodium falciparum PDF (PDB code 1RL4) [76] as a search model. The structures of actinonin-bound proteins—that is, WT and mutants—were solved using rigid-body refinement by CNS of the free AtPDF structure. The ten final models were obtained by manual rebuilding using TURBO-FRODO [77] and combined with refinement of only calculated phases using CNS and Refmac [78] software. No non- crystallographic symmetries were used. Quality control of the three models was performed using the PROCHECK program The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 13 May 2011 | Volume 9 | Issue 5 | e1001066 [79]. To probe for alternative conformers, Ringer was used [53]. Ringer is a program to detect molecular motions by automatic X- ray electron density sampling, and can be accessed at http:// ucxray.berkeley.edu/ringer.htm. Accession Numbers PDB codes for the PDF structures presented within this manuscript are as follows: 3M6O, 3PN2, 3M6P, 3O3J, 3PN3, 3PN4, 3PN5, 3M6Q, 3PN6, and 3M6R. UniProtKB accession numbers for other PDF studied are P0A6K3 (EcPDF) and O31410 (BsPDF). Supporting Information Figure S1 Alignment of PDF sequences and secondary structures. (A) PDF1B from Arabidopsis thaliana (AtPDF1B) is compared with bacterial type 1B (EcPDF and LiPDF), pathogenic protozoa (PfPDF1B), eukaryotic mitochondrial PDF1A from A. thaliana (AtPDF1A), and bacterial type 2 (BsPDF2). This figure was created with ENDscript [80]. The sequence alignment was realized with the algorithm muscle included in ENDscript, and modified according to the superimposition of structures. The blue frames indicate conserved residues, white characters in red boxes indicate strict identity, and red characters in yellow boxes indicate homology. The secondary structures at the top (a-helices, 310 helices, b-strands, and b-turns are shown by medium squiggles, small squiggles, arrows, and TT letters, respectively) were predicted by DSSP [81]. Relative accessibility (acc) of subunit A is shown by a blue-colored bar below sequence. White is buried, cyan is intermediate, and blue with red borders is highly exposed. A red box means that relative accessibility is not calculated for the residue, because it is truncated. Hydropathy (hyd) is calculated from the sequence according to [82]. It is shown by a second bar below accessibility: pink is hydrophobic, grey is intermediate, and cyan is hydrophilic. Motifs 1 (41GwGwAAXQ48), 2 (89EGCLS93), and 3 (133HEwDH137), where w is a hydrophobic amino acid, are labeled by red stars below the sequence alignment. To simplify the nomenclature, AtPDF1B is referred to as AtPDF throughout the text. (B) Topology cartoon of AtPDF, free (left) or actinonin bound (right), in the same color code as (A). Actinonin (represented by the yellow arrow) binding to the ligand binding site allows the linkage of the two distinct b-sheets into one single b-sheet, by mimicking an additional b-strand. PDB sum (http://www.ebi.ac. uk/thornton-srv/databases/pdbsum/) was used. (C) 3-D structure of AtPDF is represented showing the position of the residues discussed in the text, indicated in red. (EPS) Figure S2 Microcalorimetric titration of AtPDF with actinonin. Data were obtained at 37uC by an automated sequence of 28 injections of 180 mM actinonin from a 300 ml syringe into the reaction cell, which contain 9.85 mM AtPDF. The volume of each reaction was 10 ml, and injections were made at 240 s intervals. Top, raw data from the titration. Each peak corresponds to the injection. Bottom, the peaks in the upper panel were integrated with ORIGIN software and the values were plotted versus injection number. Each point corresponds to the heat in mcal generated by the reaction upon each injection. The solid line is the curve fit to the data by the Origin program. This fit yields values for Kd. Experiments were done with wild type protein and others variants, and gave similar raw data and curve fit. (A) WT; (B) variant G41M; (C) variant I42W. (EPS) Figure S3 Binding of actinonin to AtPDF does barely modify the crystal packing. (A) Crystal pack of the two complexes: open, free complex (left) and bound to actinonin (right) (B). Non-crystallo- graphic contacts into asymmetric unit are not modified by closing movement of the protein due to actinonin binding, except for zinc atom number 6. This metal ion is coordinated by side chains of Asp40 and Glu63, and water molecules, Asp40 and Glu63 being hydrogen bonded by side chain of Lys38 of the other subunit of the asymmetric unit. With the closing movement of the protein into the crystal, Ca of Asp40 shifted by 3.1 A˚ and its side chain flipped by 90u. Therefore, it does no longer participate to the coordination shell of this Zn2+ ion. However, it is still hydrogen bonded by Lys38 from chain B. (EPS) Figure S4 Binding of actinonin to AtPDF closely mimics both actinonin and product binding to EcPDF. Superimposition of EcPDF and AtPDF bound to either actinonin (1LRU PDB code, panel A) or Met-Ala-Ser (1BS6 PDB code, panel B), the product of the reaction. The r.m.s.d. value is 1.11 A˚ for 151 Ca superimposed. (EPS) Figure S5 The ligand binding site of AtPDF. This picture shows the residues of AtPDF that are in contact with actinonin (left) and 6b (right) according to the 3-D structure; this should be compared to the similar scheme shown in Figure 1B for EcPDF. (EPS) Figure S6 Electronic densities of the moving side-chains and of actinonin at the binding site in some variants of AtPDF. Actinonin and selected residues (G/Q/M41, I42, F58, and I130) are drawn in stick and are shown in their FO–FC electron density omit maps contoured at 2s, in free wild-type AtPDF (two crystallization conditions, WT1 and WT2), and ligand-bound WT (actinonin, 6b and 21), G41Q, and G41M variants. (EPS) Figure S7 Only few residues show alternative conformation in AtPDF. Alternative conformers in the crystalline state of AtPDF. Ringer plots of electron density (r) versus x1 angle for representative residues of the 3-D apostructure of AtPDF. Data were obtained with the 3M6O dataset (see Table S1). The secondary peaks in the Ile residues are observed because Ile is a branched amino acid. To evidence an alternative conformation with Ile, three peaks should be observed. (EPS) Figure S8 Impact of induced fit on the binding free energy of actinonin depends on the capacity to stabilize a hydrogen bond with PDF. (A) The gyration radii [83] of the side chain occurring at position 41 is displayed with black squares and compared to the kcat/Km values (grey bars). (B) The distance between the NH of I42 and the CO of actinonin was measured in each case. The percentage of the distance required to make a hydrogen bond (2.8 A˚ ) is reported (dark squares). The difference of binding free energy (DDGbinding) between the open, free state and the variants closed complexes of the G41 variants are displayed as grey bars. The values were calculated as follows. For the WT, it corresponds to the RT ln(KI*/KI) value [29], where R is the ideal gas constant and T is the temperature in Kelvin. RT is 0.616 kcal.mol21 at 37uC. For the G41M and G41Q variants, the DDGbinding corresponds to RT ln(KI-G41variant/KD-WT). The obtained values are similar to that obtained if the kcat/Km substitutes the KD value in the calculation (DDGbinding = RT ln(kcat/Km –G41variant/kcat/Km –WT). (EPS) Figure S9 Compound 21 does not bind AtPDF1B at S1’. (A) 21 is shown in ball-and-stick format in its FO–FC electron density omit The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 14 May 2011 | Volume 9 | Issue 5 | e1001066 map contoured at 2s. (B) Binding site of 21 into AtPDF1B is detailed. Red and blue residues indicate residues that accommo- date the ‘‘phenylalanine’’ and ‘‘trimethyl’’ groups of 21, respectively. (C) Overall view of 21 binding site (left). Molecular surface of AtPDF is represented, as well as 21 in ball-and-stick format. Residues belonging to the 21 binding pocket are colored in orange. For comparison, molecular surface of EcPDF (PDB code 1G2A) in the same orientation is also represented, with residues forming the new ligand binding pocket colored in orange. Actinonin is represented in ball-and-stick format and is seen through the molecular surface of each PDF. (D) Ball-and-stick representation of the interaction network around compound 21. The metal cation is shown as a grey sphere. (EPS) Figure S10 Poorly active versions of AtPDF are in a closed conformation incompatible with actinonin binding. (A) Free and close AtPDF were superimposed as in Figure 1C and are figured in brown and yellow, respectively. Both the G41M (chain B, shown in orange) and the free C-deleted weakly active AtPDF versions ([63], colored in purple, PDB entry code 3CPM) were superim- posed, to the two structures, showing that they both fit better to the ligand-bound full-length close form than to the free open form, but that the closure is further pronounced, burying the entrance to a ligand. (B) Close-up showing that the shape of the S1’ pocket of the poorly active closed versions make it poorly available to P1’ recognition (see circled Ile142 and Ile130 side chains). (EPS) Table S1 Catalytic properties of AtPDF. Nm, not measurable; ND, not determined; WT, is wild-type. aKinetic constants were determined using the coupled assay as indicated in Materials and Methods with substrate Fo-Met-Ala-Ser, in the presence of 100 nM enzyme variant and 750 mM NiCl2, at 37uC. The relative value of kcat/Km for wild-type AtPDF was set at 100%. bData correspond to the binding constant of actinonin as obtained either from ITC or from enzymatic analysis when indicated with an asterisk. cData from Table S3. dGyration radii are from [83]. (DOC) Table S2 Crystallographic data and refinement statistics. Values in parentheses are for the outer resolution shell. aRsym (I) = ShklSi|Ihkl,i 2 ,Ihkl.|/ShklSi|Ihkl,i|, where ,Ihkl. is the mean intensity of the multiple Ihkl,i observations for symmetry-related reflections. bRwork = 1006(Shkl|Fobs 2 Fcalc|/Shkl|Fobs|). Rfree is a test set including ,5% of the data. cPercentage of residues in most-favored/additionally allowed/generously allowed/disal- lowed regions of the Ramachandran plot. dCompound 21 was added first, and actinonin afterwards. (DOC) Table S3 Kinetic parameters for inhibition of some AtPDF variants by actinonin. The enzyme concentration used in the assay was 100 nM. Prior to kinetic analysis for determination of KI*app values, actinonin was incubated in the presence of each variant set at the final concentration for 10 min at 37uC; kinetic assay was started by adding a small volume of the substrate. For determination of KI, k5, and k6 values, actinonin was not pre- incubated with enzyme and kinetic assay was started by adding the enzyme. (DOCX) Movie S1 Dynamics of actinonin binding to peptide deformylase and closure of the active site. (WMV) Movie S2 Progressive motions of the main side chains at the active site and final locking of the hydrogen bond. (WMV) Acknowledgments We are strongly indebted to James Fraser and Tom Alber (University of California, Berkeley, USA) for introducing us to Ringer before the release of the freely available downloadable version. We thank Benoıˆt Gigant, Virginie Gueguen-Chaignon, Solange Morera, and Philippe Peynot (CNRS, Gif-sur-Yvette, France) for help with data processing and access to the crystallization facilities. We also thank Magali Nicaise-Aumont (IBBMC, Orsay, France), who performed the microcalorimetry experi- ments. We are grateful to the staff of the European Synchrotron Radiation Facility (ESRF) and SOLEIL beamlines for their help during data collection. Author Contributions The author(s) have made the following declarations about their contributions: Conceived and designed the experiments: SF CG TM. Performed the experiments: AB SF. Analyzed the data: FD MD SF CG TM. Contributed reagents/materials/analysis tools: IA MD CG TM. Wrote the paper: CG TM. References 1. Knowles JR (1991) Enzyme catalysis: not different, just better. Nature 350: 121–124. 2. Hammes GG (2002) Multiple conformational changes in enzyme catalysis. Biochemistry 41: 8221–8228. 3. Benkovic SJ, Hammes-Schiffer S (2003) A perspective on enzyme catalysis. Science 301: 1196–1202. 4. Henzler-Wildman K, Kern D (2007) Dynamic personalities of proteins. Nature 450: 964–972. 5. Teilum K, Olsen JG, Kragelund BB (2009) Functional aspects of protein flexibility. Cell Mol Life Sci. 6. Sullivan SM, Holyoak T (2008) Enzymes with lid-gated active sites must operate by an induced fit mechanism instead of conformational selection. Proc Natl Acad Sci U S A 105: 13829–13834. 7. Weikl TR, von Deuster C (2009) Selected-fit versus induced-fit protein binding: kinetic differences and mutational analysis. Proteins 75: 104–110. 8. Johnson KA (2008) Role of induced fit in enzyme specificity: a molecular forward/reverse switch. J Biol Chem 283: 26297–26301. 9. Bourgeois D, Royant A (2005) Advances in kinetic protein crystallography. Curr Opin Struct Biol 15: 538–547. 10. Katona G, Carpentier P, Niviere V, Amara P, Adam V, et al. (2007) Raman- assisted crystallography reveals end-on peroxide intermediates in a nonheme iron enzyme. Science 316: 449–453. 11. Koshland DE (1958) Application of a theory of enzyme specificity to protein synthesis. Proc Natl Acad Sci U S A 44: 98–104. 12. Tummino PJ, Copeland RA (2008) Residence time of receptor-ligand complexes and its effect on biological function. Biochemistry 47: 5481–5492. 13. Boehr DD, Nussinov R, Wright PE (2009) The role of dynamic conformational ensembles in biomolecular recognition. Nat Chem Biol 5: 789–796. 14. Bosshard HR (2001) Molecular recognition by induced fit: how fit is the concept? News Physiol Sci 16: 171–173. 15. Benkovic SJ, Hammes GG, Hammes-Schiffer S (2008) Free-energy landscape of enzyme catalysis. Biochemistry 47: 3317–3321. 16. Hammes GG, Chang YC, Oas TG (2009) Conformational selection or induced fit: a flux description of reaction mechanism. Proc Natl Acad Sci U S A 106: 13737–13741. 17. Fersht AR (1998) Structure and mechanism in protein science. New York: W.H. Feeman & Co. 18. Morrison JF, Walsh CT (1988) The behavior and significance of slow-binding enzyme inhibitors. Adv Enzymol Relat Areas Mol Biol 61: 201–301. 19. Copeland RA (2005) Evaluation of enzyme inhibitors in drug discovery: a guide for medicinal chemists and pharmacologists. New Jersey: John Wiley & Sons. 296 p. 20. Dash C, Vathipadiekal V, George SP, Rao M (2002) Slow-tight binding inhibition of xylanase by an aspartic protease inhibitor: kinetic parameters and conformational changes that determine the affinity and selectivity of the bifunctional nature of the inhibitor. J Biol Chem 277: 17978–17986. 21. Mac Sweeney A, Lange R, Fernandes RP, Schulz H, Dale GE, et al. (2005) The crystal structure of E.coli 1-deoxy-D-xylulose-5-phosphate reductoisomerase in a The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 15 May 2011 | Volume 9 | Issue 5 | e1001066 ternary complex with the antimalarial compound fosmidomycin and NADPH reveals a tight-binding closed enzyme conformation. J Mol Biol 345: 115–127. 22. Chou CJ, Herman D, Gottesfeld JM (2008) Pimelic diphenylamide 106 is a slow, tight-binding inhibitor of class I histone deacetylases. J Biol Chem 283: 35402–35409. 23. Barb AW, Jiang L, Raetz CR, Zhou P (2007) Structure of the deacetylase LpxC bound to the antibiotic CHIR-090: Time-dependent inhibition and specificity in ligand binding. Proc Natl Acad Sci U S A 104: 18433–18438. 24. Dunford JE, Kwaasi AA, Rogers MJ, Barnett BL, Ebetino FH, et al. (2008) Structure-activity relationships among the nitrogen containing bisphosphonates in clinical use and other analogues: time-dependent inhibition of human farnesyl pyrophosphate synthase. J Med Chem 51: 2187–2195. 25. Bateman RL, Ashworth J, Witte JF, Baker LJ, Bhanumoorthy P, et al. (2007) Slow-onset inhibition of fumarylacetoacetate hydrolase by phosphinate mimics of the tetrahedral intermediate: kinetics, crystal structure and pharmacokinetics. Biochem J 402: 251–260. 26. Copeland RA, Pompliano DL, Meek TD (2006) Drug-target residence time and its implications for lead optimization. Nat Rev Drug Discov 5: 730–739. 27. Gordon JJ, Kelly BK, Miller GA (1962) Actinonin: an antibiotic substance produced by an actinomycete. Nature 195: 701–702. 28. Chen DZ, Patel DV, Hackbarth CJ, Wang W, Dreyer G, et al. (2000) Actinonin, a naturally occurring antibacterial agent, is a potent deformylase inhibitor. Biochemistry 39: 1256–1262. 29. Van Aller GS, Nandigama R, Petit CM, Dewolf WE, Jr., Quinn CJ, et al. (2005) Mechanism of time-dependent inhibition of polypeptide deformylase by actinonin. Biochemistry 44: 253–260. 30. Rajagopalan PT, Grimme S, Pei D (2000) Characterization of cobalt(II)- substituted peptide deformylase: function of the metal ion and the catalytic residue Glu-133. Biochemistry 39: 791–799. 31. Schmitt E, Guillon JM, Meinnel T, Mechulam Y, Dardel F, et al. (1996) Molecular recognition governing the initiation of translation in Escherichia coli. A review. Biochimie 78: 543–554. 32. Giglione C, Fieulaine S, Meinnel T (2009) Cotranslational processing mechanisms: towards a dynamic 3D model. Trends Biochem Sci 34: 417–426. 33. Nguyen KT, Hu X, Pei D (2004) Slow-binding inhibition of peptide deformylase by cyclic peptidomimetics as revealed by a new spectrophotometric assay. Bioorg Chem 32: 178–191. 34. Becker A, Schlichting I, Kabsch W, Groche D, Schultz S, et al. (1998) Iron center, substrate recognition and mechanism of peptide deformylase. Nat Struct Biol 5: 1053–1058. 35. Guilloteau JP, Mathieu M, Giglione C, Blanc V, Dupuy A, et al. (2002) The crystal structures of four peptide deformylases bound to the antibiotic actinonin reveal two distinct types: a platform for the structure-based design of antibacterial agents. J Mol Biol 320: 951–962. 36. Giglione C, Serero A, Pierre M, Boisson B, Meinnel T (2000) Identification of eukaryotic peptide deformylases reveals universality of N-terminal protein processing mechanisms. EMBO J 19: 5916–5929. 37. Serero A, Giglione C, Meinnel T (2001) Distinctive features of the two classes of eukaryotic peptide deformylases. J Mol Biol 314: 695–708. 38. Dardel F, Ragusa S, Lazennec C, Blanquet S, Meinnel T (1998) Solution structure of nickel-peptide deformylase. J Mol Biol 280: 501–513. 39. Meinnel T, Blanquet S, Dardel F (1996) A new subclass of the zinc metalloproteases superfamily revealed by the solution structure of peptide deformylase. J Mol Biol 262: 375–386. 40. Larue V, Seijo B, Tisne C, Dardel F (2009) 1H, 13C and 15N NMR assignments of the E. coli peptide deformylase in complex with a natural inhibitor called actinonin. Biomolecular NMR Assignments 3: 153–155. 41. Amero CD, Byerly DW, McElroy CA, Simmons A, Foster MP (2009) Ligand- induced changes in the structure and dynamics of Escherichia coli peptide deformylase. Biochemistry 48: 7595–7607. 42. Berg AK, Srivastava DK (2009) Delineation of alternative conformational states in Escherichia coli peptide deformylase via thermodynamic studies for the binding of actinonin. Biochemistry 48: 1584–1594. 43. Zhou Z, Song X, Gong W (2005) Novel conformational states of peptide deformylase from pathogenic bacterium Leptospira interrogans: implications for population shift. J Biol Chem 280: 42391–42396. 44. Clements JM, Beckett RP, Brown A, Catlin G, Lobell M, et al. (2001) Antibiotic activity and characterization of BB-3497, a novel peptide deformylase inhibitor. Antimicrob Agents Chemother 45: 563–570. 45. Yoon HJ, Kim HL, Lee SK, Kim HW, Lee JY, et al. (2004) Crystal structure of peptide deformylase from Staphylococcus aureus in complex with actinonin, a naturally occurring antibacterial agent. Proteins 57: 639–642. 46. Moon JH, Park JK, Kim EE (2005) Structure analysis of peptide deformylase from Bacillus cereus. Proteins 61: 217–220. 47. Park J, Fu H, Pei D (2004) Peptidyl aldehydes as slow-binding inhibitors of dual- specificity phosphatases. Bioorg Med Chem Lett 14: 685–687. 48. Velazquez-Campoy A, Ohtaka H, Nezami A, Muzammil S, Freire E (2004) Isothermal titration calorimetry. Curr Protoc Cell Biol Chapter 17: Unit 17 18. 49. Boularot A, Giglione C, Petit S, Duroc Y, Sousa RA, et al. (2007) Discovery and refinement of a new structural class of potent peptide deformylase inhibitors. J Med Chem 50: 10–20. 50. Hackbarth CJ, Chen DZ, Lewis JG, Clark K, Mangold JB, et al. (2002) N-alkyl urea hydroxamic acids as a new class of peptide deformylase inhibitors with antibacterial activity. Antimicrob Agents Chemother 46: 2752–2764. 51. Ragusa S, Mouchet P, Lazennec C, Dive V, Meinnel T (1999) Substrate recognition and selectivity of peptide deformylase. Similarities and differences with metzincins and thermolysin. J Mol Biol 289: 1445–1457. 52. Fraser JS, Clarkson MW, Degnan SC, Erion R, Kern D, et al. (2009) Hidden alternative structures of proline isomerase essential for catalysis. Nature 462: 669–673. 53. Lang PT, Ng HL, Fraser JS, Corn JE, Echols N, et al. (2010) Automated electron-density sampling reveals widespread conformational polymorphism in proteins. Protein Sci 19: 1420–1431. 54. Meinnel T, Patiny L, Ragusa S, Blanquet S (1999) Design and synthesis of substrate analogue inhibitors of peptide deformylase. Biochemistry 38: 4287–4295. 55. Schechter I, Berger A (1967) On the size of the active site in proteases. I. Papain. Bochem Biophys Res Commun 27: 157–162. 56. Fersht AR, Shi JP, Knill-Jones J, Lowe DM, Wilkinson AJ, et al. (1985) Hydrogen bonding and biological specificity analysed by protein engineering. Nature 314: 235–238. 57. Takano K, Yamagata Y, Kubota M, Funahashi J, Fujii S, et al. (1999) Contribution of hydrogen bonds to the conformational stability of human lysozyme: calorimetry and X-ray analysis of six Ser —. Ala mutants. Biochemistry 38: 6623–6629. 58. Yuan Z, Trias J, White RJ (2001) Deformylase as a novel antibacterial target. Drug Discov Today 6: 954–961. 59. Wang Q, Zhang D, Wang J, Cai Z, Xu W (2006) Docking studies of Nickel- Peptide deformylase (PDF) inhibitors: exploring the new binding pockets. Biophys Chem 122: 43–49. 60. Li Y, Chen Z, Gong W (2002) Enzymatic properties of a new peptide deformylase from pathogenic bacterium Leptospira interrogans. Biochem Biophys Res Commun 295: 884–889. 61. Bracchi-Ricard V, Nguyen KT, Zhou Y, Rajagopalan PT, Chakrabarti D, et al. (2001) Characterization of an eukaryotic peptide deformylase from Plasmodium falciparum. Arch Biochem Biophys 396: 162–170. 62. Dirk LM, Williams MA, Houtz RL (2001) Eukaryotic peptide deformylases. Nuclear-encoded and chloroplast-targeted enzymes in Arabidopsis. Plant Physiol 127: 97–107. 63. Dirk LM, Schmidt JJ, Cai Y, Barnes JC, Hanger KM, et al. (2008) Insights into the substrate specificity of plant peptide deformylase, an essential enzyme with potential for the development of novel biotechnology applications in agriculture. Biochem J 413: 417–427. 64. Lee JE, Smith GD, Horvatin C, Huang DJ, Cornell KA, et al. (2005) Structural snapshots of MTA/AdoHcy nucleosidase along the reaction coordinate provide insights into enzyme and nucleoside flexibility during catalysis. J Mol Biol 352: 559–574. 65. Wang Y, Liu L, Wei Z, Cheng Z, Lin Y, et al. (2006) Seeing the process of histidine phosphorylation in human bisphosphoglycerate mutase. J Biol Chem 281: 39642–39648. 66. Parker JB, Bianchet MA, Krosky DJ, Friedman JI, Amzel LM, et al. (2007) Enzymatic capture of an extrahelical thymine in the search for uracil in DNA. Nature 449: 433–437. 67. Towler P, Staker B, Prasad SG, Menon S, Tang J, et al. (2004) ACE2 X-ray structures reveal a large hinge-bending motion important for inhibitor binding and catalysis. J Biol Chem 279: 17996–18007. 68. Teague SJ (2003) Implications of protein flexibility for drug discovery. Nat Rev Drug Discov 2: 527–541. 69. Geremia S, Campagnolo M, Schinzel R, Johnson LN (2002) Enzymatic catalysis in crystals of Escherichia coli maltodextrin phosphorylase. J Mol Biol 322: 413–423. 70. Pargellis C, Tong L, Churchill L, Cirillo PF, Gilmore T, et al. (2002) Inhibition of p38 MAP kinase by utilizing a novel allosteric binding site. Nat Struct Biol 9: 268–272. 71. Lazennec C, Meinnel T (1997) Formate dehydrogenase-coupled spectrophoto- metric assay of peptide deformylase. Anal Biochem 244: 180–182. 72. Henderson PJ (1972) A linear equation that describes the steady-state kinetics of enzymes and subcellular particles interacting with tightly bound inhibitors. Biochem J 127: 321–333. 73. Kabsch W (1993) Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Cryst 26: 795–800. 74. McCoy AJ, Grosse-Kunstleve RW, Storoni LC, Read RJ (2005) Likelihood- enhanced fast translation functions. Acta Crystallogr D Biol Crystallogr 61: 458–464. 75. Brunger AT, Adams PD, Clore GM, DeLano WL, Gros P, et al. (1998) Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr 54(Pt 5): 905–921. 76. Robien MA, Nguyen KT, Kumar A, Hirsh I, Turley S, et al. (2004) An improved crystal form of Plasmodium falciparum peptide deformylase. Protein Sci 13: 1155–1163. 77. Roussel A, Cambillau C (1989) TURBO-FRODO. In: Graphics S, ed. Silicon Graphics geometry partners directory Mountain View, CA. pp 77–78. 78. Murshudov GN, Vagin AA, Dodson EJ (1997) Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystal- logr 53: 240–255. 79. Laskowski RA, Moss DS, Thornton JM (1993) Main-chain bond lengths and bond angles in protein structures. J Mol Biol 231: 1049–1067. The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 16 May 2011 | Volume 9 | Issue 5 | e1001066 80. Gouet P, Courcelle E, Stuart DI, Metoz F (1999) ESPript: analysis of multiple sequence alignments in PostScript. Bioinformatics 15: 305–308. 81. Kabsch W, Sander C (1983) Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features. Biopolymers 22: 2577–2637. 82. Kyte J, Doolittle RF (1982) A simple method for displaying the hydropathic character of a protein. J Mol Biol 157: 105–132. 83. Levitt M (1976) A simplified representation of protein conformations for rapid simulation of protein folding. J Mol Biol 104: 59–107. The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 17 May 2011 | Volume 9 | Issue 5 | e1001066
3M6P
Crystal structure of Arabidopsis thaliana peptide deformylase 1B (AtPDF1B) in complex with actinonin
Trapping Conformational States Along Ligand-Binding Dynamics of Peptide Deformylase: The Impact of Induced Fit on Enzyme Catalysis Sonia Fieulaine1, Adrien Boularot1, Isabelle Artaud2,3, Michel Desmadril4,5, Fre´de´ric Dardel6,7, Thierry Meinnel1*, Carmela Giglione1* 1 CNRS, ISV, UPR2355, Gif-sur-Yvette, France, 2 Universite´ Paris Descartes, UMR8601, Paris, France, 3 CNRS, UMR8601, Paris, France, 4 Universite´ Paris-Sud, IBBMC, UMR8619, Orsay, France, 5 CNRS, IBBMC, UMR8619, Orsay, France, 6 Universite´ Paris Descartes, UMR8015, Paris, France, 7 CNRS, UMR8015, Paris, France Abstract For several decades, molecular recognition has been considered one of the most fundamental processes in biochemistry. For enzymes, substrate binding is often coupled to conformational changes that alter the local environment of the active site to align the reactive groups for efficient catalysis and to reach the transition state. Adaptive substrate recognition is a well-known concept; however, it has been poorly characterized at a structural level because of its dynamic nature. Here, we provide a detailed mechanism for an induced-fit process at atomic resolution. We take advantage of a slow, tight binding inhibitor-enzyme system, actinonin-peptide deformylase. Crystal structures of the initial open state and final closed state were solved, as well as those of several intermediate mimics captured during the process. Ligand-induced reshaping of a hydrophobic pocket drives closure of the active site, which is finally ‘‘zipped up’’ by additional binding interactions. Together with biochemical analyses, these data allow a coherent reconstruction of the sequence of events leading from the encounter complex to the key-lock binding state of the enzyme. A ‘‘movie’’ that reconstructs this entire process can be further extrapolated to catalysis. Citation: Fieulaine S, Boularot A, Artaud I, Desmadril M, Dardel F, et al. (2011) Trapping Conformational States Along Ligand-Binding Dynamics of Peptide Deformylase: The Impact of Induced Fit on Enzyme Catalysis. PLoS Biol 9(5): e1001066. doi:10.1371/journal.pbio.1001066 Academic Editor: Gregory A. Petsko, Brandeis University, United States of America Received January 7, 2011; Accepted April 14, 2011; Published May 24, 2011 Copyright:  2011 Fieulaine et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was supported by the Centre National de la Recherche Scientifique (CNRS, France), grant ANR-06-MIME-010-01 (Agence Nationale de la Recherche, France), and grant #4920 from the Association pour la Recherche sur le Cancer (Villejuif, France). SF was partly supported by a postdoctoral fellowship from the Fondation pour la Recherche Me´dicale (France). AB was supported by CNRS, France. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. Abbreviations: DSC, differential scanning calorimetry; Fo, formyl; PDF, peptide deformylase; r.m.s.d., root mean square deviation. * E-mail: Thierry.Meinnel@isv.cnrs-gif.fr (TM); Carmela.Giglione@isv.cnrs-gif.fr (CG) Introduction Flexibility of proteins around their active site is a central feature of molecular biochemistry [1–5]. Although this has been a central concept in biochemistry for half a century, the detailed mechanisms describing how the active enzyme conformation is achieved have remained largely elusive, as a consequence of their transient nature. Direct structural evidence and/or kinetic analyses have only recently emerged [6–10]. Three classic ‘‘textbook’’ models are used to describe the formation of the ligand-enzyme complex: (i) the Fischer’s ‘‘lock-and key’’ model, (ii) the Koshland’s induced-fit model, and (iii) the selected-shift model or conformational selection mechanism [6–8,11–13]. In the Fischer’s ‘‘lock-and key’’ model, the conformations of free and ligand-bound proteins are essentially the same. In the induced-fit model, ligand binding induces a conformational change in the protein, leading to the precise orientation of the catalytic groups and implying the existence of initial molecular matches that provide sufficient affinity prior to conformational adaptation [14]. In contrast, the selected-fit model assumes an equilibrium between multiple conformational states, in which the ligand is able to select and stabilize a complementary protein conformation. In this case, the conformational change precedes ligand binding, in contrast to the induced-fit model in which binding occurs first. The conformational selection and/or induced-fit processes have been shown to be involved in a number of enzymes [12,13,15,16]. For several of these studies, conformational selection is proposed because the experimental data support that, even in the absence of the ligand, the enzyme samples multiple conformational states, including the ligand-bound (active) state [6]. Although direct structural evidence and/or kinetic analyses have provided clues [6–8,12,13,16], how we can distinguish whether a protein binds its ligand in an induced- or selected-fit mechanism remains critical and often controversial. The enzyme-inhibitor interaction is a form of molecular recognition that is more amenable to investigation than the enzyme-substrate interaction as there is no chemical transforma- tion of the ligand during this process. In this context, slow, tight- binding inhibition is an interesting interaction process, as it closely mimics the substrate recognition process and has been shown to be commonly involved in adaptive conformational changes [12, 17,18]. In slow, tight-binding inhibition, the degree of inhibition at a fixed concentration of compound varies over time, leading to a curvature of the reaction progress curve over time during which PLoS Biology | www.plosbiology.org 1 May 2011 | Volume 9 | Issue 5 | e1001066 the uninhibited reaction progress curve is linear [19]. Indeed, the slow, tight-binding inhibition is a two-step mechanism that depends on the rate and strength of inhibitor interactions with the enzyme. Binding of the inhibitor (I) to the enzyme (E) leads to the rapid formation of a non-covalent enzyme-inhibitor complex (E:I) followed by monomolecular slower step (k5) in which the E:I is transformed into a more stable complex (E:I*) that relaxes and dissociates at a very slow rate, mainly inferred by the k6 value when k6,,k5,,k4, (Figure 1A; see also footnote f in Table 1). Although only a few studies have investigated the mechanisms of slow, tight-binding inhibitors, such molecules are favored for use as therapeutics, as they usually exhibit unique inhibitory properties, including selective potency and long-lasting effects [20–26]. Here, we explore the precise structural inhibitory mechanism of actinonin (Figure 1A; [27]), which is a slow, tight- binding inhibitor of peptide deformylase (PDF), a metal cation- dependent enzyme [28,29]. The function of the active-site metal is to activate the reactive water molecule involved in peptide hydrolysis [30]. PDF is the first enzyme in the N-terminal methionine excision pathway, an essential and ubiquitous process that contributes to the diversity of N-terminal amino acids [31,32]. Actinonin is a natural product with antibiotic activity that inhibits PDF by mimicking the structure of its natural substrates (nascent peptide chains starting with Fo-Met-Aaa, where Fo is a formyl group and Aaa is any amino acid) in their transition state (Figure 1B). The transition state inhibitor actinonin, as well as other structurally related inhibitors, has been shown to systemat- ically exhibit a ‘‘slow-binding’’ inhibition behavior (Figure 1A), regardless of the organism of origin of the PDF [29,33]. Using structural, biocomputing, and enzymatic analyses, we were able to (i) reveal that the free enzyme is in an open conformation and that actinonin induces transition of the enzyme into a closed conformation; (ii) show that there is no evidence for the occurrence of a closed conformation in the apostructure of the open enzyme, which, together with detailed kinetic analyses, makes the closed form fully compatible with an induced-fit model; and (iii) identify the sequence of molecular events leading to the final, bound, closed complex (E:I*). Moreover, using several rationally designed point mutants of the enzyme, ligand-induced intermediates, which mimic conformational states that normally would not be expected to accumulate with the wild-type (WT) enzyme, were trapped. These conformations recapitulate physical states that the WT enzyme must pass through during its overall transition from the apo-enzyme to the E:I* complex. ‘‘Freezing’’ of ligand-induced intermediate states provides direct evidence for an induced-fit mechanism and allows the reconstruction of a virtual ‘‘movie’’ that recapitulates this mechanism. Since PDF is one example of an enzyme remaining active in the crystalline state and because actinonin closely mimics the natural substrates bound to PDF in the transition state as shown previously with the Escherichia coli form (EcPDF; see Figure 1B) [34,35], we propose a model suggesting that induced fit also contributes to efficient catalysis. Results Slow, Tight Binding of the Transition-State Analog Actinonin to Peptide Deformylase In the present study, at the atomic level we explored the precise inhibitory mechanism of actinonin on Arabidopsis thaliana PDF1B (AtPDF), a close eukaryotic homologue of EcPDF (Figure S1) [36,37]. Measurements of the kinetic parameters of the second step of the binding mechanism (k5) revealed a timescale in the 10-s range (Table 1), which is consistent with the collective motion of a large domain [4,5]. This finding is supported by NMR studies [38,39], which showed that actinonin binding induces drastic changes in the heteronuclear single quantum coherence (HSQC) spectrum of EcPDF, since most resonances undergo significant shifts that affect a large part of the structure [40,41]. The existence of alternative conformational states of EcPDF is further supported by recent biophysical studies [42]. Previously reported snapshots of a series of different conformations of the enlarged and mobile loop—the so- called CD loop—of the dimeric PDF from Leptospira interrogans PDF (LiPDF) in the presence or absence of inhibitor led to the hypothesis of the existence of an equilibrium between a closed and open form of the CD-loop of PDF enzymes, suggesting a selected-shift model to the authors [43]. Taken together, these data suggest that the binding of actinonin to PDF is accompanied or preceded by conformational changes within the enzyme. Paradoxically, this proposal has not been currently supported by the available structural data. Indeed, free and complexed crystal structures have provided no evidence for any significant conformational change in PDF structure induced by the binding of ligand [35,43–47]. Tight inhibition in the closed state is associated with the KI* apparent equilibrium constant (Figure 1A). A KI* value (see Table 1 and Materials and Methods for the biochemical definition of KI*) of 0.9 nM for actinonin could be measured for AtPDF; that is, a value very similar to that obtained for bacterial PDFs, including EcPDF and Bacillus stearothermophilus PDF2 (BsPDF2, Table 1). Tightening of the initial encounter complex (E:I) resulted in a final complex (E:I*) in which the potency of actinonin (KI/KI*) was enhanced by more than two orders of magnitude and exhibited a very slow off-rate (k6, Table 1). The dissociation constant value of AtPDF for actinonin was also assessed using isothermal titration calorimetry (ITC) experiments (Table S1 and Figure S2A). The corresponding ITC titration curves (Figure S2A) are consistent with a very strong affinity of the ligand for the enzyme [48], enabling us to determine an accurate Kd. Moreover, these studies generated values similar to those measured by other means for AtPDF and EcPDF [42,49]. Author Summary The notion of induced fit when a protein binds its ligand— like a glove adapting to the shape of a hand—is a central concept of structural biochemistry introduced over 50 years ago. A detailed molecular demonstration of this phenomenon has eluded biochemists, however, largely due to the difficulty of capturing the steps of this very transient process: the ‘‘conformational change.’’ In this study, we were able to see this process by using X-ray diffraction to determine more than 10 distinct structures adopted by a single enzyme when it binds a ligand. To do this, we took advantage of the ‘‘slow, tight-binding’’ of a potent inhibitor to its specific target enzyme to trap intermediates in the binding process, which allowed us to monitor the action of an enzyme in real-time at atomic resolution. We showed the kinetics of the conformational change from an initial open state, including the encounter complex, to the final closed state of the enzyme. From these data and other biochemical and biophysical analyses, we make a coherent causal reconstruction of the sequence of events leading to inhibition of the enzyme’s activity. We also generated a movie that reconstructs the sequence of events during the encounter. Our data provide new insights into how enzymes achieve a catalytically competent conformation in which the reactive groups are brought into close proximity, resulting in catalysis. The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 2 May 2011 | Volume 9 | Issue 5 | e1001066 Ligand-Induced Conformational Closure of AtPDF in the Crystalline State Occurrence of a conformational change induced by drug binding was visualized via the resolution of several crystal structure forms of AtPDF, the free form and/or in a complex with actinonin (Table S2). The data reveal a structural switch between the two forms that can account for both the thermodynamic and kinetic data. The enzyme was observed in two states, a novel open apo-form and a closed, induced, actinonin-bound complex (Figure 1C). Binding of actinonin resulted in a tightening of the active site through the collective closure of the entire N-terminal portion of the protein (strands b1, b2, and b3; helix a1; and CD-loop, see Movies S1 and S2, Figure 1C, and Figure S1). The amplitude of the structural change was maximal for Pro60 (Figure S1), the Ca of which was shifted 4 A˚ upon actinonin binding. This collective movement involved the formation of a ‘‘super b-sheet’’ as the result of the large rearrangement of b-strands 4 and 5 relative to the rest of the structure in which actinonin forms an additional strand bridging the two b-sheets (b1 andb2) on either side of the active site (Figure 1D and Figure S1B). As actinonin is a peptide-like compound (see Introduction and Figure 1B), this behavior closely mimics what occurs in the natural protein substrates of PDF, which also form this strand-bridging interaction. This phenomenon also accounts for the strong stabilization of the protein by actinonin, which was also challenged by differential scanning calorimetry (DSC) experiments: the Tm of AtPDF increased from 61uC to 81uC upon binding of the inhibitor (Figure 1D, see also below). Thus far, this closure of the enzyme induced by actinonin is part of the rare structural evidence for the slow, tight-binding mechanism at an atomic scale. The open state, which has never been observed, was captured not only in the two molecules of the asymmetric subunit but also in different crystals and under two distinct crystallization conditions (Table S2 and Figure 2). All r.m.s.d. values were smaller than 0.25 A˚ . The closure is very unlikely to result from crystal packing constraints, as soaking the apo-AtPDF crystals in a solution containing actinonin induced the Figure 1. Slow, tight-binding inhibition of PDF by actinonin induces conformational change in the protein. (A) Inhibition by a two-step mechanism, involving a tightening of the initial enzyme-inhibitor complex (E?I) to form a more stable complex (E?I*), with the chemical structure of actinonin (I), the natural inhibitor of PDF enzymes (E). (B) Structures of EcPDF bound to actinonin (left) and to the transition state resulting from the cleavage of its substrate, Fo-Met-Ala-Ser (right) [34,35]. (C) Superimposition of free and actinonin-bound AtPDF indicated in green and purple, respectively. The three conserved motifs of the PDF enzymes family are indicated in orange and numbered I, II, and III. Molecules A of both models were superimposed, resulting in an r.m.s.d. of 0.9 A˚ for 100% of the Ca. Left inset, close-up comparison of the open and closed forms figured in the ribbon representation. (D) Baseline-corrected DSC thermograms of free and actinonin-bound WT AtPDF recorded under the same experimental conditions. doi:10.1371/journal.pbio.1001066.g001 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 3 May 2011 | Volume 9 | Issue 5 | e1001066 structural transition from the open to the closed state within the crystals without cracking them or altering their diffracting power. Thus, crystal packing is compatible with both states of the enzyme (Figure S3). Therefore, the open structure most likely corresponds to a stable state in solution. The closed final conformation was identical to that previously reported for PDF complexes obtained either with actinonin or with a product of the reaction [34,35,44,50], indicating that this structure is common for the ligands (compare Figures 1B and 2A, and Figure S4). Hydrogen bonding was also conserved, especially the bond between the backbone nitrogen of Ile42 (corresponding to Ile44 in EcPDF, see Figure 1B and Figure S5A) and the alkyl carbonyl chain of actinonin, which potently contributes to the formation of the super b-sheet (Movie S2 and Figure S1B, see also below). Between the open and closed states, the side chains of Ile42, Phe58, and Ile130 underwent significant structural changes (Figure 3A and D and Figure S6), corresponding to a hydophobic pocket rearrangement, with Ile42 being the most affected (Figure 3). Interestingly, Ile42 is the second residue of the conserved active-site motif G41IGLAAXG (motif 1) that was previously shown to be essential for activity [51]. To assess and visualize the differences between the two states, two independent structural parameters were measured: the r.m.s.d. value with respect to the open form and the aperture angle (dap), which measures the angle made between the N- and C-domains through three fixed-points, corresponding to the Ca of three conserved residues, each sitting in one of the three conserved motifs (Figure 2A). The bi-dimensional graph of these two parameters is a good representation of the closing motion snapshots (Figure 2B) shown in Movie S1. With this tool at this stage, two states could be defined: the closed (C) and open (O) states (Figure 2B). Evidence for a Pure Induced-Fit Mechanism in the Binding of Actinonin to AtPDF Recent quantitative analyses of both conformational selection and induced fit have led to an integrated continuum—a so-called ‘‘flux-description’’—of these two limiting mechanisms [16]. According to this model, conformation selection tends to be preferred at low ligand concentrations (mM range)—that is, using detailed kinetic studies—whereas induced fit dominates at high ligand and enzyme concentrations (mM range) obtained, for instance, in NMR or crystallographic approaches. Structural studies are most useful to reveal subpopulations of biological significance. We investigated the existence of lowly populated, alternative conformations of apoPDF. To probe the occurrence of alternate conformers in the crystalline state of PDF, the new Ringer program is the most suitable investigation tool [52,53]. Ringer searches for evidence of alternate rotamers by systematically sampling electron density maps—free of model bias—around the dihedral angles of protein side chains. Two independent WT open datasets of the apoenzyme, including a high-resolution set (1.3 A˚ ), were used in the analysis. Ringer analysis revealed the existence of only one rotamer of most side chains of either molecule in the asymmetric unit, including the three main residues primarily involved in conformation change—that is, Ile 42, Phe58, and Ile130 (Figure 4A). Ringer analysis showed evidence for unmodeled alternate conformers for very few residues, including Ile121 and Phe87, or Phe119 to a much lesser extent (Figure S7). There is therefore no evidence for the occurrence of a closed conformation in the apostructure of AtPDF, supporting the hypothesis that the conformational change was essentially induced by the binding of actinonin rather than from conformational selection among multiple states occurring in the crystalline state. To further investigate the mechanism involved, we followed a kinetic approach aimed at discriminating between induced fit and population shift at low ligand concentrations (sub-mM range) [12]. The experimentally observed pseudo-first-order rate constant for the approach to equilibrium between the free components and the binary AtPDF-actinonin complex (kobs) was measured and plotted as a function of actinonin concentration. This plot yielded a hyperbolic saturation curve with a positive slope, as fully expected for a pure induced-fit mechanism (Figure 4B and C). In contrast, if the enzyme sampled two or more conformational states, the curve would imply that the value of kobs decreases with increasing ligand concentration (see, for instance, curve C in Figure 1 in [12]). The same conclusion can be reached for EcPDF and BsPDF2 (Figure 4B and C) and was already reported by others for S. aureus PDF [29]. Together, these data indicate that a pure induced-fit mechanism triggered by the binding of actinonin appears to direct the conformational change both in solution and in the crystalline state. Single Variants at Gly41 Exhibit Strongly Reduced Actinonin-Binding Potency and Catalytic Efficiency When dealing with an induced-fit mechanism, knowledge of the initial O and final C state is crucial but does not provide direct information on the position of actinonin in the encounter complex or on the sequential mechanism of the transition process. We suspected that the conserved glycine-rich motif 1 (G41IGLAAXQ) could contribute to the flexibility required for the observed structural transition. Evidence for such flexibility comes from NMR analysis of EcPDF in which a few residues show exchange cross-peaks of an additional, alternative form [38]. The most strongly affected residues are Cys90, one of the metal ligands, its neighbor Leu91, and both of the alanines within the above conserved glycine-rich motif (Figure S1B), suggesting that EcPDF undergoes conformational dynamics in a similar region. To unravel the dynamics of the recognition process, we surmised that it should be possible to freeze the conformational Table 1. Comparison of the main kinetic and thermodynamic parameters describing the inhibition of PDF by actinonin. Parameter AtPDFa EcPDFa BsPDF2a,b KI (nM)d 140610 112610 185615 KI* (nM)c 0.960.5 1.360.2 2.960.8 KI/KI* 155615 86610 6467 k5 (s21) 6103d 6369 170620 7268 k6 (s21) 6104d 461 1962 1163 k4 (s21)e 140610 112610 185615 t1/2 (min)f 2965 661 1.160.2 aThe enzyme concentrations used in the assay were 100, 50, and 25 nM for AtPDF, EcPDF, and BsPDF2, respectively. bData from [49]. cPrior to kinetic analysis for determination of the KI* value, actinonin was incubated at the final concentration in the presence of the studied enzyme set for 10 min at 37uC. The kinetic assay was initiated by the addition of a small volume of the substrate. dFor determination of KI, k5, and k6 values, actinonin was not preincubated with the enzyme. The kinetic assay was initiated by the addition of the enzyme. ek4 corresponds to the kinetic constant of the dissociation of the primary enzyme-actinonin complex. It is assumed that the rate of complex association is diffusion-limited (see Table 7.3 in [19]), that is, k3—the kinetic constant of the association of the primary enzyme-actinonin complex—is 109 M21.s21. ft1/2 is 0.693(k4+k5+k6)/k4k6 (see case of induced fit and calculation in Table 1 of [12]). In this case, t1/2,0.693/k6 because k6,,k5,,k4. doi:10.1371/journal.pbio.1001066.t001 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 4 May 2011 | Volume 9 | Issue 5 | e1001066 Figure 2. Four distinct conformational states of PDF enzymes. (A) AtPDF and several other representative bacterial PDFs are superimposed. A zoom is displayed on the right of the panel. Superimpositions were realized using ‘‘module superpose’’ in the CCP4i package and the ‘‘secondary structure matching’’ tool. The extent of aperture/closure of PDF enzymes was assessed primarily by the measurement of the aperture angle (dap), the angle made between the Ca of three strictly conserved residues (C, H, and I) of all PDFs, each characterizing a secondary crucial structure module of the active site crevice, namely b4, a2, and b1 (see Figure S1C). Each single residue belongs to one of the three conserved motifs (motifs 2, 3, and 1, respectively) and corresponds respectively to Cys91, His137, and Ile42 in AtPDF. The dap was measured in each case (see B). (B) The dap values combined with those of the r.m.s.d. associated with the superimposition of the open structure of AtPDF allows the identification of four conformational states: open (O), intermediate (I), closed (C), and super-closed (S). We compared AtPDF1B (this work and PDB CODE 3CPM; brown, orange, and yellow in A and B; black in C), EcPDF (1BS7, free enzyme; 1BS6, with Met-Ala-Ser; 1G2A, with actinonin; magenta), BsPDF2 (1LQY, with actinonin; green), LiPDF (1SV2, free; 1SZZ, with actinonin; red), and PfPDF (1JYM, free; blue). doi:10.1371/journal.pbio.1001066.g002 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 5 May 2011 | Volume 9 | Issue 5 | e1001066 Figure 3. Effect of actinonin binding on the conformation of key residues in PDF. Conformation of key residues Ile42, Phe58, and Ile130 in the different complexes: (A) in unbound WT AtPDF, (B and C) in the structure of G41Q and G41M actinonin-bound variants, respectively, and (D) of actinonin-bound WT protein. In the final complex (D), a hydrogen bond is formed between actinonin and the peptidic bond, which links Gly41 and Ile42. During the deformylation reaction, which is catalyzed by the PDF enzyme, the N-terminal formyl-methionine fits into the S1’ pocket. The solvent-accessible surface of this pocket is represented here, and only the aliphatic chain of actinonin is shown, mimicking the N-terminal methionine. (E) Free WT enzyme with the S1’ pocket shown open in two orientations (top and bottom). (F and G) S1’ pocket in the G41Q and G41M variant structures, respectively, shown in two orientations (top and bottom). (H) After the complete conformational modifications of actinonin-bound WT protein induced by actinonin binding, the S1’ pocket is shown closed in two orientations (top and bottom). (I) The four models are superimposed; the ligand-binding site is magnified: unbound WT AtPDF; G41Q and G41M actinonin-bound enzyme; and WT actinonin-bound enzyme are indicated in brown, red, orange, and yellow, respectively. Actinonin is indicated by lines. (J) A detailed view of the AtPDF ligand-binding site for all the complexes, which are superimposed, as indicated in the same colors. Arrows indicate the direction of the closing movement within the enzyme, from the open, unbound state to the closed, bound state. doi:10.1371/journal.pbio.1001066.g003 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 6 May 2011 | Volume 9 | Issue 5 | e1001066 change along the pathway by introducing selected, minor variations within the above-mentioned crucial residues involved in the collective motion. In this respect, site-directed mutagenesis of AtPDF was performed on Gly41, Ile42, and Ile130. Single substitutions were made at Gly41 (G41A/Q/M), Ile42 (I42A/F/ N/W), and Ile130 (I130A/F), and the variants were purified and characterized. These mutant proteins showed no change in overall stability, as evidenced by DSC experiments (unpublished data). However, two variants of G41, G41Q and G41M, showed dramatic effects; the kcat/Km values were reduced by three orders of magnitude due to large decreases in the kcat values compared to the WT enzyme (Figure 5A and Table S1). The reduced kcat/Km values suggest an altered ability of these variants to attain the final enzyme-transition state complex and, as a result, to give rise to possible states different from the final E:I* complex. Substitutions at positions 42 and 130 only caused small reductions in the kcat values (Figure 5A, Figure S2C, and Table S1). The actinonin- binding potency of both G41 variants was also greatly reduced (Table S1 and Figure S2B). The time-dependent inhibition by actinonin of the most active variants was then studied (Table S3). Figure 4. Evidence for an induced fit in crystalline and solution states of AtPDF. (A) Absence of evidence for alternative conformers in the crystalline state of AtPDF. Ringer plots of electron density (r) versus x1 angle for representative residues of the 3D apo-structure of AtPDF. Data were obtained with the 3M6O, 3PN2, and 3PN3 datasets (2.0 and 1.3 A˚ resolution, respectively, see Table S1). The secondary peaks in the Ile residues are observed because Ile is a branched amino acid. To reveal an alternative conformation with Ile, three peaks should be observed. (B) kobs is a saturable function of actinonin with various PDFs, including AtPDF. Data obtained for kobs, the experimentally observed pseudo-first-order rate constant for the approach to equilibrium between the free components and the binary PDF-actinonin complex, were obtained at various concentrations of actinonin in the presence of EcPDF, AtPDF, and BsPDF2. A direct plot is shown. Inset, time-course measurement of deformylation as a function of varying actinonin concentrations. (C) Inverted plot of the data in panel B, which is expected to be a straight line if the kobs is ..k6 in the case of induced fit [19]. The correlation coefficient of each line is 1.00, 0.99, and 1.00 for AtPDF, BsPDF2, and EcPDF, respectively, indicative of the accuracy of the conclusion. doi:10.1371/journal.pbio.1001066.g004 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 7 May 2011 | Volume 9 | Issue 5 | e1001066 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 8 May 2011 | Volume 9 | Issue 5 | e1001066 The half-lives of the final complexes—as assessed by comparison of the 1/k6 values—were always significantly smaller (Table S3), suggesting that the conformational change induced by actinonin binding still occurred, but the C state is destabilized relative to the O state in the mutants compared to the WT. Accordingly, actinonin strongly stabilized almost all of the variants; Tm was increased by more than 20uC. This differs from the G41M and G41Q variants, which both showed increases in the Tm of only 12uC, consistent with reduced binding potency (Table S1). Conformational Changes of Gly41 Variants Are Affected On-Pathway The two most interesting variants, G41Q and G41M, could be crystallized under the same conditions as the WT protein. In the case of G41Q, the structure of the apo-protein did not show any modifications compared to the WT structure and remained in an O conformation (Figure 2B; ‘‘O’’ zone). In contrast, the 3D structure of the G41M variant showed that the asymmetric unit was composed of two molecules with distinct structures. One molecule (chain A) is in the O state and is similar to the structures of the WT and the G41Q variant (Figure 2B; zone ‘‘O’’). The second molecule (chain B) is in a C state, closer to that observed for the WT chain in the presence of actinonin (‘‘C’’), a so-called ‘‘superclosed’’ state (Figure 2B; zone ‘‘S’’), suggesting that the substitution modified the equilibrium between the two states in solution either (i) at the step of protein synthesis by providing two conformers, the inter-conversions of which are blocked due to steric hindrance brought by the new bulkier side-chain at position 41, or (ii) by dramatically unbalancing the free inter-conversion between the O and S conformers towards the S state. Ringer analysis indicates that in the free G41M variant, many residues show evidence for unmodeled alternate conformers—including positions 58, 42, and 130—in keeping with the second hypothesis. For all variants of position G41, addition of actinonin to the crystal (Figure 3 and Figure S6) induced a closure of the protein within the crystal. Nevertheless, as expected from in silico graphic modeling followed by energy minimization, the occurrence of a bulky side chain at position 41 prevented the completion of the closure in the presence of the ligand and, hence, the formation of the hydrogen bond between the backbone nitrogen of Ile42 and actinonin. This finding is consistent with the strongly reduced Tm of the complex of the variants with actinonin compared to WT as measured by DSC. Remarkably, both S and O forms of the G41M apo-structures in the asymmetric unit of the crystal yielded a unique intermediary structure (‘‘I’’ state) upon actinonin binding (r.m.s.d. between the molecules is ,0.25 A˚ ; see also Figure 2B, zone ‘‘I’’). In this case, it is likely that the induced-fit mechanism drives the equilibrium by capturing only the O population and closing it to an intermediary step, thus depleting the pool of O conformers that is shifted sequentially back from the remaining pool of S conformers and allows the complete binding of actinonin to the enzyme. In line with the rational design of the PDF mutants, the extent of the structural differences suggests that the underlying motions are dependent on the length of the side chain (Figure S8). Together, these data account for the reduced catalytic rate, as the hydrogen bond is strictly required for the substrate to be efficiently cleaved by PDFs (Figure S8A) [54]. Therefore, from both structural and kinetic analyses, each substitution most likely reproduces intermediates along the pathway that lead to the closure of PDF around its substrate (Figure S2B). Conformational Changes of Gly41 Variants Recapitulate Closing Intermediates Analysis of the structures allows us to propose the following sequence of atomic events (Figures 3 and 2B and Figure S6). To name the various sites of the ligand and subsites of PDF, we will use the usual nomenclature found in [55], which defines the various binding pockets of a protease, where P1’ is the first side chain at the C-terminal side of the cleavage site and its binding pocket is S1’, also referred to as the hydrophobic pocket in the case of PDF. First, actinonin aligns along the S1’ pocket to form the encounter complex, which shifts the Ile130 side chain to avoid steric hindrance in the S1’ pocket, promotes rotation of the Ile42 side chain, and finally rearranges the phenyl group of Phe58. These events achieve an optimal hydrophobic S1’ pocket conformation (Figure 3), and the concomitant closure leads to the formation of a hydrogen bond between the first carbonyl group of actinonin and the backbone nitrogen of Ile42. The initial N-O distance is reduced from 5 A˚ to 2.8 A˚ , which is an optimal value for hydrogen bonding (Movie S2 and Figure S8B). Thus, the primary driving force for the active site closure appears to be the P1’:S1’ hydrophobic interaction. The C state is ultimately locked by the super-b-sheet hydrogen bonds extending across the ligand, including those involving Ile42. The DDGbinding value (2.2– 2.4 kcal/mol, Figure S8B), as calculated from the Kd values for actinonin binding to wild-type (WT) and G41M and G41Q, is consistent with the loss of a hydrogen bond that also contributes to the conformational stability of the protein [56,57]. Thus, this bond contributes to the major binding free energy difference between the two complexes (3.1 kcal/mol; Figure S8B, Tables S1 and S3, and [29]). Interestingly, the above DDGbinding values also correlate with the DDGES values derived from the kcat/Km and kcat measurements [19]. This dataset strongly correlates with the Figure 5. Inhibition and enzymatic reactions progress through an induced fit pathway. (A) The catalytic parameters Km and kcat, for all AtPDF variants are provided as a percentage of the wild-type values (WT). Detailed values are presented in Table S2. (B) Schematic model for actinonin binding to AtPDF in favor of an induced-fit pathway. PDF might exist in at least two conformational states, open (O) or closed (C). The relative abundance of each conformation would vary, depending on the enzyme type. With AtPDF, it is likely that the most abundant form is the O one, which is the only form leading to a productive complex. The superclosed form (S) is likely to show reduced affinity for the ligand because of steric occlusion of the active site. At the initial stage, the inhibitor (shown in red) binds to AtPDF (indicated in brown) in the O conformation. To reach the final key-lock state (productive closed conformation, C), two major and extreme pathways can be used. According to the conformational selection pathway, the inhibitor selects the C conformation. This pathway, which is represented by the dashed arrow, does not occur within the crystal. In contrast, the G41Q and G41M mutants, by providing the structure of the enzyme in intermediate conformations (I), prove the existence of the so- called encounter complex and confirm that the inhibitor binds to the enzyme when it is in the O conformation. The ligand-binding site is then reorganized to yield the C enzyme conformation, that is, the key-lock state. Indeed, the inhibitor binds to the enzyme through the induced-fit pathway. Each timescale was calculated using the data available in the text and corresponds to t1/2 values deduced from the calculation of 0.693/ (kinetic constant of interest). The kcat value (k2) was used to assess the timescale of catalysis in panel C, whereas, in (B), k4 assesses the first step of inhibition, and k6 is used in the case of the slow step. For the SO conversion (left, B), the lifetime of the minor form of EcPDF was used to assess the order of magnitude (see text and [38]). (C) Schematic model for the deformylation reaction catalyzed by PDF. Since actinonin is a pseudo-peptidic inhibitor, it is likely that a peptidic substrate can bind to the PDF enzyme through an induced-fit pathway, as described in (B). The key-lock state represents a transition state in which the N-formylated substrate is deformylated to yield the final reaction product. doi:10.1371/journal.pbio.1001066.g005 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 9 May 2011 | Volume 9 | Issue 5 | e1001066 gyration and van der Waals radii of the side chain at position 41 as well as the N-O distance between the first carbonyl group of actinonin and the backbone nitrogen of Ile42 (Figure S8). These results suggest that the capacity of both G41M and G41Q variants to form the transition state is a consequence of their inability to reach the fully closed state. Thus, our study of the designed Gly41 mutant enzymes reveals that, in addition to the initial and final states observed for the WT enzyme, the conformations of the Gly41 variants correspond indeed to on-pathway intermediates, thus providing snapshots along the trajectory from the O to the C state of the enzyme (Figures 2B and 3). The 3D structure of the variants in the absence of ligand is similar to that of WT, and a strict correlation exists between the completeness of the conformational change and both binding potency and catalytic efficiency. This suggests that both events require complete protein closure to generate a productive complex. The strong stabilization of AtPDF by actinonin (Figure 1D) closely mimics what occurs with its natural substrates when it reaches the transition state [34,58]. Indeed, as expected, the enzyme facilitates the final C conformation by lowering its final energy [6]. Optimal arrangement of the S1’ pocket (Figure 3) proceeds along the reaction process towards the final C conformation, triggering the alignment of reactive groups in an optimal arrangement for ligand recognition. Upon binding, actinonin alters the thermodynamic landscape for the structural transition between the O and C states. This ligand is a potent inhibitor because it can trigger the above sequence of events similar to the substrate, but unlike the substrate, it is non- hydrolyzable. Thus, by mimicking the transition state and being non-hydrolyzable (Figure 1B), the final C complex is long lasting. Ligand-Induced Conformational Closure Is Initially Triggered by the Binding of the P1’ Group in the S1’ Pocket Given the similarity between actinonin and natural substrate binding, the very slow kinetics of inhibitor binding (10-s time-scale) remains puzzling compared to the 10 ms required for catalysis (deduced from the kcat). This finding could be explained as a conformational effect during the formation of the hydrogen bond, aligning the substrate as an additional beta-sheet and eventually stabilizing the entire enzyme-ligand complex. The significantly longer time needed to reach the most stable state compared to the substrate would most likely be due to the presence of the flexible and one carbon longer metal-binding group in actinonin (i.e., hydroxamate versus formyl, Figure 1B). This suggestion is in line with the overall data obtained when we investigated more deeply the role of the first carbonyl group of the ligand. This group is well known to exert a crucial effect in both productive and unproductive ligand binding (i.e., substrate and inhibitor) [54]. In this respect, we studied the binding of compound 6b (Figure S5B), a PDF ligand that does not exhibit a reactive group at this position [49]. We observed that this compound binds strongly to both EcPDF (KI* = 6366 nM) and AtPDF (KI* = 400635 nM) but, unlike actinonin, does not display slow, tight binding as KI* = KI. This impact on binding is consistent with the absence of the hydrogen bond involving the first carbonyl group of the ligand. The 3D structure of AtPDF was determined after soaking the compound in crystals of the free, open AtPDF form. Upon binding, 6b induced a complete conformational change, identical to that observed with actinonin (Figures 2B and 6A; ‘‘O’’ state). This result further suggests that the conformational change is not induced initially by the formation of this hydrogen bond and that the encounter complex is primarily driven by the fit within the S1’ pocket. This also reveals that the timescale of the large conformational change is several orders of magnitude faster than the kinetics of slow binding and fully compatible with both the first step of actinonin binding (k4 = 140 s21; see Table 1) and the catalytic rate of the substrate (kcat = 37 s21; see Table 1 and Table S3). The 3D structure also revealed that both the P1’ and the hydroxamate groups are bound similarly to the corresponding groups of actinonin (Figure 6B). As expected, no additional bonding occurs, especially around the backbone nitrogen of Ile42 (Figure 6C). Taken together, these data allow us to conclude that the conformational change observed upon ligand binding is triggered primarily by binding in the S1’ pocket. As revealed by the binding of 6b, the one carbon longer metal-binding group fits, immediately upon recognition of the P1’ group, in the S1’ pocket and forms a bidentate complex with the metal cation, mimicking the transition state as a result. Thus, the active site is very confined and rigid due to the presence and length of the hydroxamate group (compare right and left panels in Figure 1B). As a result, compared to the complex made with the substrate, it is likely that the formation of the hydrogen bond involving the carbonyl of actinonin and the backbone nitrogen of Ile42 becomes strongly rate-limiting (k5 = 0.044 s21; Table 1). Once this hydrogen link is locked, the uncleavable bond, mimicking the labile formyl group at the transition state, stabilizes the enzyme-inhibitor complex, making it long-lasting (k6 = 0.0006 s21; Table 1) and providing a mechanistic explanation for the slow-binding effect that involves both large and fine conformational changes. The large conformational change is similar to the one occurring with the substrate, whereas the second is more subtle and locks the hydrogen bond involving the backbone nitrogen of Ile42. The second step is rate-limiting with some transition state analogs such as actinonin (Figure 5B and C). Proper Positioning of the Carbonyl Group Is Required to Stabilize the Complex at S1’ Compound 21 corresponds to another interesting derivative designed to probe the impact of the peptide bond in PDF binding [49]. In addition to the hydroxamate group, this compound features both a hydrophobic benzyl group at P1’ and a reverse peptide bond. Compound 21 shows modest but significant inhibitory potency to AtPDF1B (KI* = 400637 nM), confirming the crucial role of the peptide bond in PDF binding. After soaking with crystals of apo-AtPDF, compound 21 could be detected in high-resolution electron density maps (Figure S9A). Unlike 6b, 21 did not bind the active site of the enzyme but an alternative pocket at the surface of the protein (Figure S9B). A docking study performed with EcPDF had previously revealed this alternative binding pocket (Figure S9C; [59]). The aforementioned data indicate that the occurrence of a S1’- binding group placed in the unfavorable context of a reverse peptide bond does not stably promote binding at the active site of AtPDF. Upon binding of 21, the 3D structure of both molecules of the asymmetric unit remain in an O conformation (r.m.s.d. ,0.2 A˚ with respect to the apo-structures in the ‘‘O’’ state). This finding suggests that only the binding of compounds entering the S1’ pocket, such as actinonin or 6b, induces conformational change, in keeping with the crucial role of the P1’ group if located in the frame of a classic peptide bond. Moreover, we noticed that the binding pocket of 21 was located on the rear side of the true S1’ pocket and induced a weak modification of the P1’ hosting platform (Figure S9D). Indeed, when crystals of the 21:AtPDF complex were soaked in actinonin, the final 3D structure no longer showed evidence of compound 21 occupancy greater than 5%. Instead, this structure revealed both actinonin and closing of the protein (Table S2). The r.m.s.d. between this structure and that The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 10 May 2011 | Volume 9 | Issue 5 | e1001066 obtained directly with actinonin was less than 0.2 A˚ ; the actinonin position was virtually identical, indicating that the protein had retained full capacity for binding actinonin and closing despite the presence of compound 21. We conclude that actinonin does compete with 21 because of the overlap at P1’ of AtPDF1B (Figure S9C). As the actinonin S1’ subsite strongly mimics that of a true substrate, this result also explains the inhibitory behavior of 21 towards AtPDF. Discussion Although PDF catalysis has been extensively studied and the mechanism has been elucidated [34], how the enzyme achieves the catalytically competent state remains unknown. Here, we provide insight on how the enzyme might reach a catalytically competent conformation, demonstrating that the reactive groups move into proximity to promote catalysis (Figures 2B and 5C). We suggest Figure 6. Effect of 6b binding on the conformation of key residues of PDF. Superimposition of free, 6b-, and actinonin-bound AtPDF indicated in brown, red, and yellow, respectively. (A) Molecule A in the three models was superimposed, resulting in an r.m.s.d. of 0.9 A˚ for 100% of the Ca. Actinonin is shown in yellow and 6b in red. (B) Conformation of key residues Ile42, Phe58, and Ile130 in the different complexes and in unbound WT AtPDF. Actinonin is shown in yellow and 6b in red. (C) A detailed view of the AtPDF ligand-binding site for both actinonin and 6b complexes, which are indicated by sticks and are superimposed. The two ligands are colored in pale and dark grey, respectively. The hydrogen bond made by actinonin only is shown. doi:10.1371/journal.pbio.1001066.g006 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 11 May 2011 | Volume 9 | Issue 5 | e1001066 that the motions of the catalytic centre starting with free ligand- PDF favor a final configuration that is optimal for binding and/or catalysis (Figures 1B, 2B, and 5B and C). In our model, we propose that free PDF might exist in at least two conformational states, that is, open (O) or super-closed (S). The relative abundance of each conformation varies by enzyme type and incubation conditions, explaining why both conformations have not been trapped thus far. In the case of AtPDF, it is likely that the most abundant form corresponds to an O state, which is the form that leads to a productive complex. Indeed, in the NMR spectra for EcPDF, a few residues show exchange cross-peaks from an additional, alternative form [38]. The most strongly affected residues are Cys90, one of the metal ligands, its neighbor Leu91, as well as Ala47 and Ala48 on the facing strand. This suggests that EcPDF exists in at least two conformations (‘‘S’’ and ‘‘C’’; see Figure 2B), which undergo slow interconversion on the NMR timescale. The 3D structure of the major conformation (75%, lifetime 300 ms) could be solved at high resolution, but the structure of the minor form (25%, lifetime 100 ms), which exhibits very weak signals, could not be solved [38]. This conformation appears to correspond to that of the complex obtained with the product of the reaction (Met-Ala-Ser). A very similar situation—although more balanced between the two states—appears to occur in the case of variant G41M, suggesting that a mechanism involving conformational selection followed by induced fit is a general model for PDF and that AtPDF is a specific case where population shift virtually does not occur as the free enzyme is completely in the O conformation. This is also in line with data obtained with L. interrogans PDF (LiPDF), which reveal conformers in both the S and C states (see Figure 2B) and suggest a population-shift mechanism [43]. It is interesting to note that LiPDF is a poorly active PDF [60]. According to the representation shown in Figure 2B, Plasmodium falciparum PDF (PfPDF), a poorly active PDF [61], was retrieved only in the S state. Finally, weak decompaction of the structure of Bacillus cereus and Staphylococcus aureus PDFs in the presence of actinonin have been described [45,46]. These examples suggest that the enzyme is trapped in the S conformer in the free state and converts to the C conformer when bound to actinonin, suggesting that the S conformer is overrepresented in solution compared to the O state, unlike AtPDF. This study of AtPDF—including 10 different crystal structures of apo- and complexed enzyme variants—reveals the 3D structure of a PDF in at least four distinct states. This includes the O form, the occurrence of which is crucial for catalysis, as it is the active form. Here, we propose that the transition from the O to the C state is directly induced by the ligand. Indeed, the O form, which is captured in the crystal, undergoes closure directly upon ligand binding in our soaking experiments. Progression to this closure involves intermediary states (‘‘I’’) similar to those observed with variants G41Q and G41M in the presence of actinonin (see Figure 2B). Extrapolating the situation to catalysis, which occurs in the crystalline states of PDF, it is likely that hydrolysis of the substrate frees the enzyme in its S state, which in turn needs to open to accommodate a new substrate (Figures 2B and 5C). This is well illustrated in the 3D structure of EcPDF complexed with a product of the reaction, obtained after co-crystallization of the enzyme with the substrate in a closed conformation [34]. The S free form is likely to exhibit a slower on-rate for the ligand (k3) compared to the O form because of steric occlusion of the active site (Figure S10). In support of this hypothesis, recent data show that the 3D structure of a C-terminally truncated, poorly active version of AtPDF is in the C conformation in the unbound state, although crystallized under conditions identical to ours [62,63]. This structure is similar to that of chain B, one of the two molecules of the asymmetric subunit of variant G41M (Figure 2B). This suggests that alterations remote from the active site significantly unbalance the equilibrium between the two conform- ers, thus altering the efficiency of the reaction (Figure 5C). As the S version corresponds to a significantly less active version of AtPDF compared to that reported in our present work, this further confirms that, compared to the O state, the S state has a significantly weaker propensity to bind substrate or a close mimic ligand, such as actinonin. Comparison of the 3D structures of the free-closed and the ligand-bound-closed forms reveals some differences responsible for the slight steric reduction of the active site of free-closed AtPDF1B with respect to that of the actinonin- AtPDF1B complex (Figure S10A), including the side chain of Ile42 burying the S1’ binding pocket (Figure S10B). Overall, these data suggest that an S form might exist under the free state but that it would feature a k3 value with respect to the ligand that is significantly weaker than that of the O form, which would strongly slow down the reaction or the binding as a result. With the interaction scheme proposed in our model (Figure 5B and C), the ligand/substrate binds more easily to the O form and induces the optimal conformation of the enzyme to reach the transition state, thus allowing the reaction to be efficiently catalyzed. In the final model (Figure 5C), there is both conformational selection and induced fit subsequently involved in line with the recently proposed existence of such mixed mechanisms for other enzymes [15,16]. Nevertheless, in our model (Figure 5C), we suggest that induced fit is the primary mechanism, as it provides energy input from the ligand, which eventually drives the enzyme towards the productive key-lock complex. Unambig- uous distinction between the relative contributions of the two mechanisms is deduced from the observation that kobs is a saturable function of actinonin with various PDF, including EcPDF, BsPDF, AtPDF (Figure 4B and C), and S. aureus PDF [12,16,29,49]. Using crystallographic reconstruction analysis involving enzyme variants, motions of small mobile loops and movie reconstructions of snapshots of catalytic events have been previously documented [1–3,64–66], often by visualizing the binding of unnatural inhibitors and not necessarily mimicking closely the substrate and transition state as actinonin does [67,68]. However, only a few examples make use of soaking conditions of a crystal to promote the motion and show the importance of induced fit [1,69]. None of these data show a motion of the amplitude revealed here with PDF and a large stabilization of the complex involving the formation of the four-stranded b-sheet superstructure and the entire N-domain of the enzyme. Compared to previous crystallographic analyses, our work integrates biophysical, computational, and kinetic analyses to reconstruct the whole picture, allowing a better understanding of the slow-binding mechanism. While our work primarily focused on an induced-fit mechanism of enzyme inhibition and catalysis, it should be emphasized that this phenomenon is also applicable to the broader area of receptor-ligand interactions. For example, in all cases where conformational change mechanisms have been proposed for kinase inhibitors without supporting experimental data [12,26], further experimental work must be provided to clarify the precise mechanism. We expect this will have important implications on how one conducts future drug-discovery efforts against such enzymes [70]. Materials and Methods Protein Expression and Purification Expression and purification of mature Arabidopsis thaliana PDF1B and all variants (i.e., AtPDF) were derived from the previously The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 12 May 2011 | Volume 9 | Issue 5 | e1001066 described protocol [37]: the lysis supernatant after sonication was applied on a Q-Sepharose column (GE Healthcare; buffers A and B as described containing 5 mM NiCl2) followed by Superdex-75 chromatography (GE Healthcare) using buffer C consisting of buffer A supplemented with 0.1 M NaCl. For crystallization experiments, the protein was purified further. The sample was concentrated on an Amicon Ultra-15 centrifugal filter unit (Millipore Corp.) with a 5-kDa cut-off and applied to a MonoQ HR5/5 column (GE Healthcare) previously equilibrated in buffer A (50 mM Hepes, pH 7.5, and 5 mM NiCl2). Elution was performed with a 50-mL gradient from 0% to 100% buffer B. The buffer of the pooled purified AtPDF1B was exchanged using a PD-10 desalting column (GE Healthcare) to yield a protein solution in 50 mM Hepes, pH 7.5, 0.1 M NaCl, and 5 mM NiCl2 (buffer C). The protein was concentrated on an Amicon Ultra-15 centrifugal filter unit. The resulting AtPDF1B preparation was frozen in aliquots and stored at 280uC (for crystallization purposes) or diluted 2-fold in 100% glycerol and stored at 220uC (for enzymatic purposes). The typical yield was 5–10 mg AtPDF per liter of culture. All purification procedures were performed at 4uC. Samples of the collected fractions were analyzed by SDS-PAGE on 12% acrylamide gels, and protein concentrations were estimated from the calculated extinction coefficients for each variant. Site-directed mutagenesis of AtPDF sequence in plasmid pQdef1bDN [36] was carried out using the QuickChange Site- Directed Mutagenesis Kit (Stratagene). Enzymology Assay of PDF activity was coupled to formate dehydrogenase, where the absorbance of NADH at 340 nm was measured at 37uC as previously described [71]. For measurements of classical kinetic parameters (i.e., Km and kcat), the reaction was initiated by addition of the substrate Fo-Met-Ala-Ser to the mixture containing purified enzyme in the presence of 1 mM NiCl2. The kinetics parameters were derived from iterative non-linear least square calculations using the Michaelis-Menten equation based on the experimental data (Sigma-Plot; Kinetics module). For determination of kinetic parameters related to actinonin, the reaction mixture contained 750 mM NiCl2. In some cases, the mixture containing PDF and actinonin was incubated for 15 min at 37uC before kinetic analysis, which was initiated by the addition of substrate. The same protocol was used to determine the dissociation constant of actinonin [KI* = k4/(k3+k3k5/k6)], but the initial reaction velocities were measured with varying concentrations of Fo-Met-Ala-Ser and actinonin. The data were then calculated according to the method of Henderson, which can be used to determine the dissociation constant of the tight-binding competitive enzyme inhibitor [28,49,72] by varying both the inhibitor and substrate concentra- tions. To determine KI, k5, and k6, the reaction was initiated by the addition of enzyme as previously described [29,49]. KI*app measurements were used for comparative studies of AtPDF variants (Table S3) at a concentration of 2 mM substrate by varying the concentration of actinonin. KI*app is the slope of the v[Actinonin]/v0 line curve. kobs was fitted from the kinetic data without preincubation with vI = vs + (v0 2 vs)e2kobst where vI is the observed velocity at a given concentration of inhibitor I, v0 is the velocity, and vs is the steady-state velocity [18]. From the set of values obtained at various concentrations of I, k5 and k6 could be derived using kobs = k6 + k5[I]/(KI + [I]). By choosing a set of values with kobs..k6, 1/kobs = 1/k5(KI/[I] +1) and 1/kobs = f(1/[I]) is expected to be a straight line in case of induced fit whose positive slope corresponds to 1/k5. k6 was derived from equation k6 = k5/ (KI/KI*21) [18,19]. Microcalorimetry ITC experiments were performed using a VP-ITC isothermal titration calorimeter (Microcal Corp.). Experiments were per- formed at 37uC. For each experiment, injections of 10 mL actinonin (180 mM) were added using a computer-controlled 300 mL microsyringe at intervals of 240 s into the Ni-AtPDF variant solution (5 to 10 mM, cell volume = 2.1 mL) dissolved in buffer C with stirring at 310 rpm. A theoretical titration curve was fitted to the experimental data using the ORIGIN software (Microcal). This software uses the relationship between the heat generated after each injection and DHu (enthalpy change in kcal/ mol), KA (the association binding constant in M21), n (number of binding sites per monomer), total protein concentration, and free and total ligand concentrations. The thermal stability of the WT and variants of Ni-AtPDF1B was studied by DSC using VP-DSC calorimetry (Microcal Corp.). DSC measurements were made with 10 mM protein solutions in buffer C. The actinonin concentration was 20 mM. The same buffer was used as a reference. All solutions were degassed just before loading into the calorimeter. Scanning was performed at 1uC/min. The temperature dependence of the partial molar capacity (Cp) was expressed in kcal/K after subtracting the buffer signal using Origin(R) software. Crystallization and Soaking Experiments Crystallization conditions were screened by a robot using the sitting drop vapor diffusion method. Crystals were obtained and optimized at 20uC with 15%–20% PEG-3350 and either 0.1 or 0.2 M zinc acetate. The drops were formed by mixing 2 mL of a solution containing 2 to 4 mg/mL protein and 2 mL of the crystallization solution. Crystals were soaked for 24 h by adding actinonin to the crystallization drops at a final concentration of 5 mM. Cryoprotection was achieved by placing crystals for 30 s in a solution that was composed of 20% PEG-3350 and 0.2 M zinc acetate, supplemented with 5%, 10%, and 15% glycerol. Crystals were then directly flash frozen in liquid nitrogen using cryoloops (Hampton Research). Crystals were also grown under conditions described for the C-terminally deleted, weakly active version of AtPDF [63]. X-Ray Diffraction Data Collection Data collections were performed at 100 K at the European Synchrotron Radiation Facility (Grenoble, France) on station ID29, FIP-BM30A, ID14-1, and ID23-2, and at SOLEIL (Gif-sur- Yvette, France) on station PROXIMA1. In each case, a single crystal was used to collect a complete dataset. Data were processed and scaled using XDS software [73]. Two crystal forms were encountered with different cell parameters. In each case, b parameter was nearly equal to a, and data could be indexed into two space groups, P212121 or P43212. The data are shown in Table S2. Structure Determination and Refinement The structure of free AtPDF was solved by molecular replacement with Phaser [74] followed by a rigid-body refinement by CNS [75] using coordinates from the Plasmodium falciparum PDF (PDB code 1RL4) [76] as a search model. The structures of actinonin-bound proteins—that is, WT and mutants—were solved using rigid-body refinement by CNS of the free AtPDF structure. The ten final models were obtained by manual rebuilding using TURBO-FRODO [77] and combined with refinement of only calculated phases using CNS and Refmac [78] software. No non- crystallographic symmetries were used. Quality control of the three models was performed using the PROCHECK program The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 13 May 2011 | Volume 9 | Issue 5 | e1001066 [79]. To probe for alternative conformers, Ringer was used [53]. Ringer is a program to detect molecular motions by automatic X- ray electron density sampling, and can be accessed at http:// ucxray.berkeley.edu/ringer.htm. Accession Numbers PDB codes for the PDF structures presented within this manuscript are as follows: 3M6O, 3PN2, 3M6P, 3O3J, 3PN3, 3PN4, 3PN5, 3M6Q, 3PN6, and 3M6R. UniProtKB accession numbers for other PDF studied are P0A6K3 (EcPDF) and O31410 (BsPDF). Supporting Information Figure S1 Alignment of PDF sequences and secondary structures. (A) PDF1B from Arabidopsis thaliana (AtPDF1B) is compared with bacterial type 1B (EcPDF and LiPDF), pathogenic protozoa (PfPDF1B), eukaryotic mitochondrial PDF1A from A. thaliana (AtPDF1A), and bacterial type 2 (BsPDF2). This figure was created with ENDscript [80]. The sequence alignment was realized with the algorithm muscle included in ENDscript, and modified according to the superimposition of structures. The blue frames indicate conserved residues, white characters in red boxes indicate strict identity, and red characters in yellow boxes indicate homology. The secondary structures at the top (a-helices, 310 helices, b-strands, and b-turns are shown by medium squiggles, small squiggles, arrows, and TT letters, respectively) were predicted by DSSP [81]. Relative accessibility (acc) of subunit A is shown by a blue-colored bar below sequence. White is buried, cyan is intermediate, and blue with red borders is highly exposed. A red box means that relative accessibility is not calculated for the residue, because it is truncated. Hydropathy (hyd) is calculated from the sequence according to [82]. It is shown by a second bar below accessibility: pink is hydrophobic, grey is intermediate, and cyan is hydrophilic. Motifs 1 (41GwGwAAXQ48), 2 (89EGCLS93), and 3 (133HEwDH137), where w is a hydrophobic amino acid, are labeled by red stars below the sequence alignment. To simplify the nomenclature, AtPDF1B is referred to as AtPDF throughout the text. (B) Topology cartoon of AtPDF, free (left) or actinonin bound (right), in the same color code as (A). Actinonin (represented by the yellow arrow) binding to the ligand binding site allows the linkage of the two distinct b-sheets into one single b-sheet, by mimicking an additional b-strand. PDB sum (http://www.ebi.ac. uk/thornton-srv/databases/pdbsum/) was used. (C) 3-D structure of AtPDF is represented showing the position of the residues discussed in the text, indicated in red. (EPS) Figure S2 Microcalorimetric titration of AtPDF with actinonin. Data were obtained at 37uC by an automated sequence of 28 injections of 180 mM actinonin from a 300 ml syringe into the reaction cell, which contain 9.85 mM AtPDF. The volume of each reaction was 10 ml, and injections were made at 240 s intervals. Top, raw data from the titration. Each peak corresponds to the injection. Bottom, the peaks in the upper panel were integrated with ORIGIN software and the values were plotted versus injection number. Each point corresponds to the heat in mcal generated by the reaction upon each injection. The solid line is the curve fit to the data by the Origin program. This fit yields values for Kd. Experiments were done with wild type protein and others variants, and gave similar raw data and curve fit. (A) WT; (B) variant G41M; (C) variant I42W. (EPS) Figure S3 Binding of actinonin to AtPDF does barely modify the crystal packing. (A) Crystal pack of the two complexes: open, free complex (left) and bound to actinonin (right) (B). Non-crystallo- graphic contacts into asymmetric unit are not modified by closing movement of the protein due to actinonin binding, except for zinc atom number 6. This metal ion is coordinated by side chains of Asp40 and Glu63, and water molecules, Asp40 and Glu63 being hydrogen bonded by side chain of Lys38 of the other subunit of the asymmetric unit. With the closing movement of the protein into the crystal, Ca of Asp40 shifted by 3.1 A˚ and its side chain flipped by 90u. Therefore, it does no longer participate to the coordination shell of this Zn2+ ion. However, it is still hydrogen bonded by Lys38 from chain B. (EPS) Figure S4 Binding of actinonin to AtPDF closely mimics both actinonin and product binding to EcPDF. Superimposition of EcPDF and AtPDF bound to either actinonin (1LRU PDB code, panel A) or Met-Ala-Ser (1BS6 PDB code, panel B), the product of the reaction. The r.m.s.d. value is 1.11 A˚ for 151 Ca superimposed. (EPS) Figure S5 The ligand binding site of AtPDF. This picture shows the residues of AtPDF that are in contact with actinonin (left) and 6b (right) according to the 3-D structure; this should be compared to the similar scheme shown in Figure 1B for EcPDF. (EPS) Figure S6 Electronic densities of the moving side-chains and of actinonin at the binding site in some variants of AtPDF. Actinonin and selected residues (G/Q/M41, I42, F58, and I130) are drawn in stick and are shown in their FO–FC electron density omit maps contoured at 2s, in free wild-type AtPDF (two crystallization conditions, WT1 and WT2), and ligand-bound WT (actinonin, 6b and 21), G41Q, and G41M variants. (EPS) Figure S7 Only few residues show alternative conformation in AtPDF. Alternative conformers in the crystalline state of AtPDF. Ringer plots of electron density (r) versus x1 angle for representative residues of the 3-D apostructure of AtPDF. Data were obtained with the 3M6O dataset (see Table S1). The secondary peaks in the Ile residues are observed because Ile is a branched amino acid. To evidence an alternative conformation with Ile, three peaks should be observed. (EPS) Figure S8 Impact of induced fit on the binding free energy of actinonin depends on the capacity to stabilize a hydrogen bond with PDF. (A) The gyration radii [83] of the side chain occurring at position 41 is displayed with black squares and compared to the kcat/Km values (grey bars). (B) The distance between the NH of I42 and the CO of actinonin was measured in each case. The percentage of the distance required to make a hydrogen bond (2.8 A˚ ) is reported (dark squares). The difference of binding free energy (DDGbinding) between the open, free state and the variants closed complexes of the G41 variants are displayed as grey bars. The values were calculated as follows. For the WT, it corresponds to the RT ln(KI*/KI) value [29], where R is the ideal gas constant and T is the temperature in Kelvin. RT is 0.616 kcal.mol21 at 37uC. For the G41M and G41Q variants, the DDGbinding corresponds to RT ln(KI-G41variant/KD-WT). The obtained values are similar to that obtained if the kcat/Km substitutes the KD value in the calculation (DDGbinding = RT ln(kcat/Km –G41variant/kcat/Km –WT). (EPS) Figure S9 Compound 21 does not bind AtPDF1B at S1’. (A) 21 is shown in ball-and-stick format in its FO–FC electron density omit The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 14 May 2011 | Volume 9 | Issue 5 | e1001066 map contoured at 2s. (B) Binding site of 21 into AtPDF1B is detailed. Red and blue residues indicate residues that accommo- date the ‘‘phenylalanine’’ and ‘‘trimethyl’’ groups of 21, respectively. (C) Overall view of 21 binding site (left). Molecular surface of AtPDF is represented, as well as 21 in ball-and-stick format. Residues belonging to the 21 binding pocket are colored in orange. For comparison, molecular surface of EcPDF (PDB code 1G2A) in the same orientation is also represented, with residues forming the new ligand binding pocket colored in orange. Actinonin is represented in ball-and-stick format and is seen through the molecular surface of each PDF. (D) Ball-and-stick representation of the interaction network around compound 21. The metal cation is shown as a grey sphere. (EPS) Figure S10 Poorly active versions of AtPDF are in a closed conformation incompatible with actinonin binding. (A) Free and close AtPDF were superimposed as in Figure 1C and are figured in brown and yellow, respectively. Both the G41M (chain B, shown in orange) and the free C-deleted weakly active AtPDF versions ([63], colored in purple, PDB entry code 3CPM) were superim- posed, to the two structures, showing that they both fit better to the ligand-bound full-length close form than to the free open form, but that the closure is further pronounced, burying the entrance to a ligand. (B) Close-up showing that the shape of the S1’ pocket of the poorly active closed versions make it poorly available to P1’ recognition (see circled Ile142 and Ile130 side chains). (EPS) Table S1 Catalytic properties of AtPDF. Nm, not measurable; ND, not determined; WT, is wild-type. aKinetic constants were determined using the coupled assay as indicated in Materials and Methods with substrate Fo-Met-Ala-Ser, in the presence of 100 nM enzyme variant and 750 mM NiCl2, at 37uC. The relative value of kcat/Km for wild-type AtPDF was set at 100%. bData correspond to the binding constant of actinonin as obtained either from ITC or from enzymatic analysis when indicated with an asterisk. cData from Table S3. dGyration radii are from [83]. (DOC) Table S2 Crystallographic data and refinement statistics. Values in parentheses are for the outer resolution shell. aRsym (I) = ShklSi|Ihkl,i 2 ,Ihkl.|/ShklSi|Ihkl,i|, where ,Ihkl. is the mean intensity of the multiple Ihkl,i observations for symmetry-related reflections. bRwork = 1006(Shkl|Fobs 2 Fcalc|/Shkl|Fobs|). Rfree is a test set including ,5% of the data. cPercentage of residues in most-favored/additionally allowed/generously allowed/disal- lowed regions of the Ramachandran plot. dCompound 21 was added first, and actinonin afterwards. (DOC) Table S3 Kinetic parameters for inhibition of some AtPDF variants by actinonin. The enzyme concentration used in the assay was 100 nM. Prior to kinetic analysis for determination of KI*app values, actinonin was incubated in the presence of each variant set at the final concentration for 10 min at 37uC; kinetic assay was started by adding a small volume of the substrate. For determination of KI, k5, and k6 values, actinonin was not pre- incubated with enzyme and kinetic assay was started by adding the enzyme. (DOCX) Movie S1 Dynamics of actinonin binding to peptide deformylase and closure of the active site. (WMV) Movie S2 Progressive motions of the main side chains at the active site and final locking of the hydrogen bond. (WMV) Acknowledgments We are strongly indebted to James Fraser and Tom Alber (University of California, Berkeley, USA) for introducing us to Ringer before the release of the freely available downloadable version. We thank Benoıˆt Gigant, Virginie Gueguen-Chaignon, Solange Morera, and Philippe Peynot (CNRS, Gif-sur-Yvette, France) for help with data processing and access to the crystallization facilities. We also thank Magali Nicaise-Aumont (IBBMC, Orsay, France), who performed the microcalorimetry experi- ments. We are grateful to the staff of the European Synchrotron Radiation Facility (ESRF) and SOLEIL beamlines for their help during data collection. Author Contributions The author(s) have made the following declarations about their contributions: Conceived and designed the experiments: SF CG TM. Performed the experiments: AB SF. Analyzed the data: FD MD SF CG TM. Contributed reagents/materials/analysis tools: IA MD CG TM. Wrote the paper: CG TM. References 1. Knowles JR (1991) Enzyme catalysis: not different, just better. Nature 350: 121–124. 2. Hammes GG (2002) Multiple conformational changes in enzyme catalysis. Biochemistry 41: 8221–8228. 3. Benkovic SJ, Hammes-Schiffer S (2003) A perspective on enzyme catalysis. Science 301: 1196–1202. 4. Henzler-Wildman K, Kern D (2007) Dynamic personalities of proteins. Nature 450: 964–972. 5. Teilum K, Olsen JG, Kragelund BB (2009) Functional aspects of protein flexibility. Cell Mol Life Sci. 6. Sullivan SM, Holyoak T (2008) Enzymes with lid-gated active sites must operate by an induced fit mechanism instead of conformational selection. Proc Natl Acad Sci U S A 105: 13829–13834. 7. Weikl TR, von Deuster C (2009) Selected-fit versus induced-fit protein binding: kinetic differences and mutational analysis. Proteins 75: 104–110. 8. Johnson KA (2008) Role of induced fit in enzyme specificity: a molecular forward/reverse switch. J Biol Chem 283: 26297–26301. 9. Bourgeois D, Royant A (2005) Advances in kinetic protein crystallography. Curr Opin Struct Biol 15: 538–547. 10. Katona G, Carpentier P, Niviere V, Amara P, Adam V, et al. (2007) Raman- assisted crystallography reveals end-on peroxide intermediates in a nonheme iron enzyme. Science 316: 449–453. 11. Koshland DE (1958) Application of a theory of enzyme specificity to protein synthesis. Proc Natl Acad Sci U S A 44: 98–104. 12. Tummino PJ, Copeland RA (2008) Residence time of receptor-ligand complexes and its effect on biological function. Biochemistry 47: 5481–5492. 13. Boehr DD, Nussinov R, Wright PE (2009) The role of dynamic conformational ensembles in biomolecular recognition. Nat Chem Biol 5: 789–796. 14. Bosshard HR (2001) Molecular recognition by induced fit: how fit is the concept? News Physiol Sci 16: 171–173. 15. Benkovic SJ, Hammes GG, Hammes-Schiffer S (2008) Free-energy landscape of enzyme catalysis. Biochemistry 47: 3317–3321. 16. Hammes GG, Chang YC, Oas TG (2009) Conformational selection or induced fit: a flux description of reaction mechanism. Proc Natl Acad Sci U S A 106: 13737–13741. 17. Fersht AR (1998) Structure and mechanism in protein science. New York: W.H. Feeman & Co. 18. Morrison JF, Walsh CT (1988) The behavior and significance of slow-binding enzyme inhibitors. Adv Enzymol Relat Areas Mol Biol 61: 201–301. 19. Copeland RA (2005) Evaluation of enzyme inhibitors in drug discovery: a guide for medicinal chemists and pharmacologists. New Jersey: John Wiley & Sons. 296 p. 20. Dash C, Vathipadiekal V, George SP, Rao M (2002) Slow-tight binding inhibition of xylanase by an aspartic protease inhibitor: kinetic parameters and conformational changes that determine the affinity and selectivity of the bifunctional nature of the inhibitor. J Biol Chem 277: 17978–17986. 21. Mac Sweeney A, Lange R, Fernandes RP, Schulz H, Dale GE, et al. (2005) The crystal structure of E.coli 1-deoxy-D-xylulose-5-phosphate reductoisomerase in a The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 15 May 2011 | Volume 9 | Issue 5 | e1001066 ternary complex with the antimalarial compound fosmidomycin and NADPH reveals a tight-binding closed enzyme conformation. J Mol Biol 345: 115–127. 22. Chou CJ, Herman D, Gottesfeld JM (2008) Pimelic diphenylamide 106 is a slow, tight-binding inhibitor of class I histone deacetylases. J Biol Chem 283: 35402–35409. 23. Barb AW, Jiang L, Raetz CR, Zhou P (2007) Structure of the deacetylase LpxC bound to the antibiotic CHIR-090: Time-dependent inhibition and specificity in ligand binding. Proc Natl Acad Sci U S A 104: 18433–18438. 24. Dunford JE, Kwaasi AA, Rogers MJ, Barnett BL, Ebetino FH, et al. (2008) Structure-activity relationships among the nitrogen containing bisphosphonates in clinical use and other analogues: time-dependent inhibition of human farnesyl pyrophosphate synthase. J Med Chem 51: 2187–2195. 25. Bateman RL, Ashworth J, Witte JF, Baker LJ, Bhanumoorthy P, et al. (2007) Slow-onset inhibition of fumarylacetoacetate hydrolase by phosphinate mimics of the tetrahedral intermediate: kinetics, crystal structure and pharmacokinetics. Biochem J 402: 251–260. 26. Copeland RA, Pompliano DL, Meek TD (2006) Drug-target residence time and its implications for lead optimization. Nat Rev Drug Discov 5: 730–739. 27. Gordon JJ, Kelly BK, Miller GA (1962) Actinonin: an antibiotic substance produced by an actinomycete. Nature 195: 701–702. 28. Chen DZ, Patel DV, Hackbarth CJ, Wang W, Dreyer G, et al. (2000) Actinonin, a naturally occurring antibacterial agent, is a potent deformylase inhibitor. Biochemistry 39: 1256–1262. 29. Van Aller GS, Nandigama R, Petit CM, Dewolf WE, Jr., Quinn CJ, et al. (2005) Mechanism of time-dependent inhibition of polypeptide deformylase by actinonin. Biochemistry 44: 253–260. 30. Rajagopalan PT, Grimme S, Pei D (2000) Characterization of cobalt(II)- substituted peptide deformylase: function of the metal ion and the catalytic residue Glu-133. Biochemistry 39: 791–799. 31. Schmitt E, Guillon JM, Meinnel T, Mechulam Y, Dardel F, et al. (1996) Molecular recognition governing the initiation of translation in Escherichia coli. A review. Biochimie 78: 543–554. 32. Giglione C, Fieulaine S, Meinnel T (2009) Cotranslational processing mechanisms: towards a dynamic 3D model. Trends Biochem Sci 34: 417–426. 33. Nguyen KT, Hu X, Pei D (2004) Slow-binding inhibition of peptide deformylase by cyclic peptidomimetics as revealed by a new spectrophotometric assay. Bioorg Chem 32: 178–191. 34. Becker A, Schlichting I, Kabsch W, Groche D, Schultz S, et al. (1998) Iron center, substrate recognition and mechanism of peptide deformylase. Nat Struct Biol 5: 1053–1058. 35. Guilloteau JP, Mathieu M, Giglione C, Blanc V, Dupuy A, et al. (2002) The crystal structures of four peptide deformylases bound to the antibiotic actinonin reveal two distinct types: a platform for the structure-based design of antibacterial agents. J Mol Biol 320: 951–962. 36. Giglione C, Serero A, Pierre M, Boisson B, Meinnel T (2000) Identification of eukaryotic peptide deformylases reveals universality of N-terminal protein processing mechanisms. EMBO J 19: 5916–5929. 37. Serero A, Giglione C, Meinnel T (2001) Distinctive features of the two classes of eukaryotic peptide deformylases. J Mol Biol 314: 695–708. 38. Dardel F, Ragusa S, Lazennec C, Blanquet S, Meinnel T (1998) Solution structure of nickel-peptide deformylase. J Mol Biol 280: 501–513. 39. Meinnel T, Blanquet S, Dardel F (1996) A new subclass of the zinc metalloproteases superfamily revealed by the solution structure of peptide deformylase. J Mol Biol 262: 375–386. 40. Larue V, Seijo B, Tisne C, Dardel F (2009) 1H, 13C and 15N NMR assignments of the E. coli peptide deformylase in complex with a natural inhibitor called actinonin. Biomolecular NMR Assignments 3: 153–155. 41. Amero CD, Byerly DW, McElroy CA, Simmons A, Foster MP (2009) Ligand- induced changes in the structure and dynamics of Escherichia coli peptide deformylase. Biochemistry 48: 7595–7607. 42. Berg AK, Srivastava DK (2009) Delineation of alternative conformational states in Escherichia coli peptide deformylase via thermodynamic studies for the binding of actinonin. Biochemistry 48: 1584–1594. 43. Zhou Z, Song X, Gong W (2005) Novel conformational states of peptide deformylase from pathogenic bacterium Leptospira interrogans: implications for population shift. J Biol Chem 280: 42391–42396. 44. Clements JM, Beckett RP, Brown A, Catlin G, Lobell M, et al. (2001) Antibiotic activity and characterization of BB-3497, a novel peptide deformylase inhibitor. Antimicrob Agents Chemother 45: 563–570. 45. Yoon HJ, Kim HL, Lee SK, Kim HW, Lee JY, et al. (2004) Crystal structure of peptide deformylase from Staphylococcus aureus in complex with actinonin, a naturally occurring antibacterial agent. Proteins 57: 639–642. 46. Moon JH, Park JK, Kim EE (2005) Structure analysis of peptide deformylase from Bacillus cereus. Proteins 61: 217–220. 47. Park J, Fu H, Pei D (2004) Peptidyl aldehydes as slow-binding inhibitors of dual- specificity phosphatases. Bioorg Med Chem Lett 14: 685–687. 48. Velazquez-Campoy A, Ohtaka H, Nezami A, Muzammil S, Freire E (2004) Isothermal titration calorimetry. Curr Protoc Cell Biol Chapter 17: Unit 17 18. 49. Boularot A, Giglione C, Petit S, Duroc Y, Sousa RA, et al. (2007) Discovery and refinement of a new structural class of potent peptide deformylase inhibitors. J Med Chem 50: 10–20. 50. Hackbarth CJ, Chen DZ, Lewis JG, Clark K, Mangold JB, et al. (2002) N-alkyl urea hydroxamic acids as a new class of peptide deformylase inhibitors with antibacterial activity. Antimicrob Agents Chemother 46: 2752–2764. 51. Ragusa S, Mouchet P, Lazennec C, Dive V, Meinnel T (1999) Substrate recognition and selectivity of peptide deformylase. Similarities and differences with metzincins and thermolysin. J Mol Biol 289: 1445–1457. 52. Fraser JS, Clarkson MW, Degnan SC, Erion R, Kern D, et al. (2009) Hidden alternative structures of proline isomerase essential for catalysis. Nature 462: 669–673. 53. Lang PT, Ng HL, Fraser JS, Corn JE, Echols N, et al. (2010) Automated electron-density sampling reveals widespread conformational polymorphism in proteins. Protein Sci 19: 1420–1431. 54. Meinnel T, Patiny L, Ragusa S, Blanquet S (1999) Design and synthesis of substrate analogue inhibitors of peptide deformylase. Biochemistry 38: 4287–4295. 55. Schechter I, Berger A (1967) On the size of the active site in proteases. I. Papain. Bochem Biophys Res Commun 27: 157–162. 56. Fersht AR, Shi JP, Knill-Jones J, Lowe DM, Wilkinson AJ, et al. (1985) Hydrogen bonding and biological specificity analysed by protein engineering. Nature 314: 235–238. 57. Takano K, Yamagata Y, Kubota M, Funahashi J, Fujii S, et al. (1999) Contribution of hydrogen bonds to the conformational stability of human lysozyme: calorimetry and X-ray analysis of six Ser —. Ala mutants. Biochemistry 38: 6623–6629. 58. Yuan Z, Trias J, White RJ (2001) Deformylase as a novel antibacterial target. Drug Discov Today 6: 954–961. 59. Wang Q, Zhang D, Wang J, Cai Z, Xu W (2006) Docking studies of Nickel- Peptide deformylase (PDF) inhibitors: exploring the new binding pockets. Biophys Chem 122: 43–49. 60. Li Y, Chen Z, Gong W (2002) Enzymatic properties of a new peptide deformylase from pathogenic bacterium Leptospira interrogans. Biochem Biophys Res Commun 295: 884–889. 61. Bracchi-Ricard V, Nguyen KT, Zhou Y, Rajagopalan PT, Chakrabarti D, et al. (2001) Characterization of an eukaryotic peptide deformylase from Plasmodium falciparum. Arch Biochem Biophys 396: 162–170. 62. Dirk LM, Williams MA, Houtz RL (2001) Eukaryotic peptide deformylases. Nuclear-encoded and chloroplast-targeted enzymes in Arabidopsis. Plant Physiol 127: 97–107. 63. Dirk LM, Schmidt JJ, Cai Y, Barnes JC, Hanger KM, et al. (2008) Insights into the substrate specificity of plant peptide deformylase, an essential enzyme with potential for the development of novel biotechnology applications in agriculture. Biochem J 413: 417–427. 64. Lee JE, Smith GD, Horvatin C, Huang DJ, Cornell KA, et al. (2005) Structural snapshots of MTA/AdoHcy nucleosidase along the reaction coordinate provide insights into enzyme and nucleoside flexibility during catalysis. J Mol Biol 352: 559–574. 65. Wang Y, Liu L, Wei Z, Cheng Z, Lin Y, et al. (2006) Seeing the process of histidine phosphorylation in human bisphosphoglycerate mutase. J Biol Chem 281: 39642–39648. 66. Parker JB, Bianchet MA, Krosky DJ, Friedman JI, Amzel LM, et al. (2007) Enzymatic capture of an extrahelical thymine in the search for uracil in DNA. Nature 449: 433–437. 67. Towler P, Staker B, Prasad SG, Menon S, Tang J, et al. (2004) ACE2 X-ray structures reveal a large hinge-bending motion important for inhibitor binding and catalysis. J Biol Chem 279: 17996–18007. 68. Teague SJ (2003) Implications of protein flexibility for drug discovery. Nat Rev Drug Discov 2: 527–541. 69. Geremia S, Campagnolo M, Schinzel R, Johnson LN (2002) Enzymatic catalysis in crystals of Escherichia coli maltodextrin phosphorylase. J Mol Biol 322: 413–423. 70. Pargellis C, Tong L, Churchill L, Cirillo PF, Gilmore T, et al. (2002) Inhibition of p38 MAP kinase by utilizing a novel allosteric binding site. Nat Struct Biol 9: 268–272. 71. Lazennec C, Meinnel T (1997) Formate dehydrogenase-coupled spectrophoto- metric assay of peptide deformylase. Anal Biochem 244: 180–182. 72. Henderson PJ (1972) A linear equation that describes the steady-state kinetics of enzymes and subcellular particles interacting with tightly bound inhibitors. Biochem J 127: 321–333. 73. Kabsch W (1993) Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Cryst 26: 795–800. 74. McCoy AJ, Grosse-Kunstleve RW, Storoni LC, Read RJ (2005) Likelihood- enhanced fast translation functions. Acta Crystallogr D Biol Crystallogr 61: 458–464. 75. Brunger AT, Adams PD, Clore GM, DeLano WL, Gros P, et al. (1998) Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr 54(Pt 5): 905–921. 76. Robien MA, Nguyen KT, Kumar A, Hirsh I, Turley S, et al. (2004) An improved crystal form of Plasmodium falciparum peptide deformylase. Protein Sci 13: 1155–1163. 77. Roussel A, Cambillau C (1989) TURBO-FRODO. In: Graphics S, ed. Silicon Graphics geometry partners directory Mountain View, CA. pp 77–78. 78. Murshudov GN, Vagin AA, Dodson EJ (1997) Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystal- logr 53: 240–255. 79. Laskowski RA, Moss DS, Thornton JM (1993) Main-chain bond lengths and bond angles in protein structures. J Mol Biol 231: 1049–1067. The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 16 May 2011 | Volume 9 | Issue 5 | e1001066 80. Gouet P, Courcelle E, Stuart DI, Metoz F (1999) ESPript: analysis of multiple sequence alignments in PostScript. Bioinformatics 15: 305–308. 81. Kabsch W, Sander C (1983) Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features. Biopolymers 22: 2577–2637. 82. Kyte J, Doolittle RF (1982) A simple method for displaying the hydropathic character of a protein. J Mol Biol 157: 105–132. 83. Levitt M (1976) A simplified representation of protein conformations for rapid simulation of protein folding. J Mol Biol 104: 59–107. The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 17 May 2011 | Volume 9 | Issue 5 | e1001066
3M6Q
Crystal structure of Arabidopsis thaliana peptide deformylase 1B (AtPDF1B) G41Q mutant in complex with actinonin
Trapping Conformational States Along Ligand-Binding Dynamics of Peptide Deformylase: The Impact of Induced Fit on Enzyme Catalysis Sonia Fieulaine1, Adrien Boularot1, Isabelle Artaud2,3, Michel Desmadril4,5, Fre´de´ric Dardel6,7, Thierry Meinnel1*, Carmela Giglione1* 1 CNRS, ISV, UPR2355, Gif-sur-Yvette, France, 2 Universite´ Paris Descartes, UMR8601, Paris, France, 3 CNRS, UMR8601, Paris, France, 4 Universite´ Paris-Sud, IBBMC, UMR8619, Orsay, France, 5 CNRS, IBBMC, UMR8619, Orsay, France, 6 Universite´ Paris Descartes, UMR8015, Paris, France, 7 CNRS, UMR8015, Paris, France Abstract For several decades, molecular recognition has been considered one of the most fundamental processes in biochemistry. For enzymes, substrate binding is often coupled to conformational changes that alter the local environment of the active site to align the reactive groups for efficient catalysis and to reach the transition state. Adaptive substrate recognition is a well-known concept; however, it has been poorly characterized at a structural level because of its dynamic nature. Here, we provide a detailed mechanism for an induced-fit process at atomic resolution. We take advantage of a slow, tight binding inhibitor-enzyme system, actinonin-peptide deformylase. Crystal structures of the initial open state and final closed state were solved, as well as those of several intermediate mimics captured during the process. Ligand-induced reshaping of a hydrophobic pocket drives closure of the active site, which is finally ‘‘zipped up’’ by additional binding interactions. Together with biochemical analyses, these data allow a coherent reconstruction of the sequence of events leading from the encounter complex to the key-lock binding state of the enzyme. A ‘‘movie’’ that reconstructs this entire process can be further extrapolated to catalysis. Citation: Fieulaine S, Boularot A, Artaud I, Desmadril M, Dardel F, et al. (2011) Trapping Conformational States Along Ligand-Binding Dynamics of Peptide Deformylase: The Impact of Induced Fit on Enzyme Catalysis. PLoS Biol 9(5): e1001066. doi:10.1371/journal.pbio.1001066 Academic Editor: Gregory A. Petsko, Brandeis University, United States of America Received January 7, 2011; Accepted April 14, 2011; Published May 24, 2011 Copyright:  2011 Fieulaine et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was supported by the Centre National de la Recherche Scientifique (CNRS, France), grant ANR-06-MIME-010-01 (Agence Nationale de la Recherche, France), and grant #4920 from the Association pour la Recherche sur le Cancer (Villejuif, France). SF was partly supported by a postdoctoral fellowship from the Fondation pour la Recherche Me´dicale (France). AB was supported by CNRS, France. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. Abbreviations: DSC, differential scanning calorimetry; Fo, formyl; PDF, peptide deformylase; r.m.s.d., root mean square deviation. * E-mail: Thierry.Meinnel@isv.cnrs-gif.fr (TM); Carmela.Giglione@isv.cnrs-gif.fr (CG) Introduction Flexibility of proteins around their active site is a central feature of molecular biochemistry [1–5]. Although this has been a central concept in biochemistry for half a century, the detailed mechanisms describing how the active enzyme conformation is achieved have remained largely elusive, as a consequence of their transient nature. Direct structural evidence and/or kinetic analyses have only recently emerged [6–10]. Three classic ‘‘textbook’’ models are used to describe the formation of the ligand-enzyme complex: (i) the Fischer’s ‘‘lock-and key’’ model, (ii) the Koshland’s induced-fit model, and (iii) the selected-shift model or conformational selection mechanism [6–8,11–13]. In the Fischer’s ‘‘lock-and key’’ model, the conformations of free and ligand-bound proteins are essentially the same. In the induced-fit model, ligand binding induces a conformational change in the protein, leading to the precise orientation of the catalytic groups and implying the existence of initial molecular matches that provide sufficient affinity prior to conformational adaptation [14]. In contrast, the selected-fit model assumes an equilibrium between multiple conformational states, in which the ligand is able to select and stabilize a complementary protein conformation. In this case, the conformational change precedes ligand binding, in contrast to the induced-fit model in which binding occurs first. The conformational selection and/or induced-fit processes have been shown to be involved in a number of enzymes [12,13,15,16]. For several of these studies, conformational selection is proposed because the experimental data support that, even in the absence of the ligand, the enzyme samples multiple conformational states, including the ligand-bound (active) state [6]. Although direct structural evidence and/or kinetic analyses have provided clues [6–8,12,13,16], how we can distinguish whether a protein binds its ligand in an induced- or selected-fit mechanism remains critical and often controversial. The enzyme-inhibitor interaction is a form of molecular recognition that is more amenable to investigation than the enzyme-substrate interaction as there is no chemical transforma- tion of the ligand during this process. In this context, slow, tight- binding inhibition is an interesting interaction process, as it closely mimics the substrate recognition process and has been shown to be commonly involved in adaptive conformational changes [12, 17,18]. In slow, tight-binding inhibition, the degree of inhibition at a fixed concentration of compound varies over time, leading to a curvature of the reaction progress curve over time during which PLoS Biology | www.plosbiology.org 1 May 2011 | Volume 9 | Issue 5 | e1001066 the uninhibited reaction progress curve is linear [19]. Indeed, the slow, tight-binding inhibition is a two-step mechanism that depends on the rate and strength of inhibitor interactions with the enzyme. Binding of the inhibitor (I) to the enzyme (E) leads to the rapid formation of a non-covalent enzyme-inhibitor complex (E:I) followed by monomolecular slower step (k5) in which the E:I is transformed into a more stable complex (E:I*) that relaxes and dissociates at a very slow rate, mainly inferred by the k6 value when k6,,k5,,k4, (Figure 1A; see also footnote f in Table 1). Although only a few studies have investigated the mechanisms of slow, tight-binding inhibitors, such molecules are favored for use as therapeutics, as they usually exhibit unique inhibitory properties, including selective potency and long-lasting effects [20–26]. Here, we explore the precise structural inhibitory mechanism of actinonin (Figure 1A; [27]), which is a slow, tight- binding inhibitor of peptide deformylase (PDF), a metal cation- dependent enzyme [28,29]. The function of the active-site metal is to activate the reactive water molecule involved in peptide hydrolysis [30]. PDF is the first enzyme in the N-terminal methionine excision pathway, an essential and ubiquitous process that contributes to the diversity of N-terminal amino acids [31,32]. Actinonin is a natural product with antibiotic activity that inhibits PDF by mimicking the structure of its natural substrates (nascent peptide chains starting with Fo-Met-Aaa, where Fo is a formyl group and Aaa is any amino acid) in their transition state (Figure 1B). The transition state inhibitor actinonin, as well as other structurally related inhibitors, has been shown to systemat- ically exhibit a ‘‘slow-binding’’ inhibition behavior (Figure 1A), regardless of the organism of origin of the PDF [29,33]. Using structural, biocomputing, and enzymatic analyses, we were able to (i) reveal that the free enzyme is in an open conformation and that actinonin induces transition of the enzyme into a closed conformation; (ii) show that there is no evidence for the occurrence of a closed conformation in the apostructure of the open enzyme, which, together with detailed kinetic analyses, makes the closed form fully compatible with an induced-fit model; and (iii) identify the sequence of molecular events leading to the final, bound, closed complex (E:I*). Moreover, using several rationally designed point mutants of the enzyme, ligand-induced intermediates, which mimic conformational states that normally would not be expected to accumulate with the wild-type (WT) enzyme, were trapped. These conformations recapitulate physical states that the WT enzyme must pass through during its overall transition from the apo-enzyme to the E:I* complex. ‘‘Freezing’’ of ligand-induced intermediate states provides direct evidence for an induced-fit mechanism and allows the reconstruction of a virtual ‘‘movie’’ that recapitulates this mechanism. Since PDF is one example of an enzyme remaining active in the crystalline state and because actinonin closely mimics the natural substrates bound to PDF in the transition state as shown previously with the Escherichia coli form (EcPDF; see Figure 1B) [34,35], we propose a model suggesting that induced fit also contributes to efficient catalysis. Results Slow, Tight Binding of the Transition-State Analog Actinonin to Peptide Deformylase In the present study, at the atomic level we explored the precise inhibitory mechanism of actinonin on Arabidopsis thaliana PDF1B (AtPDF), a close eukaryotic homologue of EcPDF (Figure S1) [36,37]. Measurements of the kinetic parameters of the second step of the binding mechanism (k5) revealed a timescale in the 10-s range (Table 1), which is consistent with the collective motion of a large domain [4,5]. This finding is supported by NMR studies [38,39], which showed that actinonin binding induces drastic changes in the heteronuclear single quantum coherence (HSQC) spectrum of EcPDF, since most resonances undergo significant shifts that affect a large part of the structure [40,41]. The existence of alternative conformational states of EcPDF is further supported by recent biophysical studies [42]. Previously reported snapshots of a series of different conformations of the enlarged and mobile loop—the so- called CD loop—of the dimeric PDF from Leptospira interrogans PDF (LiPDF) in the presence or absence of inhibitor led to the hypothesis of the existence of an equilibrium between a closed and open form of the CD-loop of PDF enzymes, suggesting a selected-shift model to the authors [43]. Taken together, these data suggest that the binding of actinonin to PDF is accompanied or preceded by conformational changes within the enzyme. Paradoxically, this proposal has not been currently supported by the available structural data. Indeed, free and complexed crystal structures have provided no evidence for any significant conformational change in PDF structure induced by the binding of ligand [35,43–47]. Tight inhibition in the closed state is associated with the KI* apparent equilibrium constant (Figure 1A). A KI* value (see Table 1 and Materials and Methods for the biochemical definition of KI*) of 0.9 nM for actinonin could be measured for AtPDF; that is, a value very similar to that obtained for bacterial PDFs, including EcPDF and Bacillus stearothermophilus PDF2 (BsPDF2, Table 1). Tightening of the initial encounter complex (E:I) resulted in a final complex (E:I*) in which the potency of actinonin (KI/KI*) was enhanced by more than two orders of magnitude and exhibited a very slow off-rate (k6, Table 1). The dissociation constant value of AtPDF for actinonin was also assessed using isothermal titration calorimetry (ITC) experiments (Table S1 and Figure S2A). The corresponding ITC titration curves (Figure S2A) are consistent with a very strong affinity of the ligand for the enzyme [48], enabling us to determine an accurate Kd. Moreover, these studies generated values similar to those measured by other means for AtPDF and EcPDF [42,49]. Author Summary The notion of induced fit when a protein binds its ligand— like a glove adapting to the shape of a hand—is a central concept of structural biochemistry introduced over 50 years ago. A detailed molecular demonstration of this phenomenon has eluded biochemists, however, largely due to the difficulty of capturing the steps of this very transient process: the ‘‘conformational change.’’ In this study, we were able to see this process by using X-ray diffraction to determine more than 10 distinct structures adopted by a single enzyme when it binds a ligand. To do this, we took advantage of the ‘‘slow, tight-binding’’ of a potent inhibitor to its specific target enzyme to trap intermediates in the binding process, which allowed us to monitor the action of an enzyme in real-time at atomic resolution. We showed the kinetics of the conformational change from an initial open state, including the encounter complex, to the final closed state of the enzyme. From these data and other biochemical and biophysical analyses, we make a coherent causal reconstruction of the sequence of events leading to inhibition of the enzyme’s activity. We also generated a movie that reconstructs the sequence of events during the encounter. Our data provide new insights into how enzymes achieve a catalytically competent conformation in which the reactive groups are brought into close proximity, resulting in catalysis. The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 2 May 2011 | Volume 9 | Issue 5 | e1001066 Ligand-Induced Conformational Closure of AtPDF in the Crystalline State Occurrence of a conformational change induced by drug binding was visualized via the resolution of several crystal structure forms of AtPDF, the free form and/or in a complex with actinonin (Table S2). The data reveal a structural switch between the two forms that can account for both the thermodynamic and kinetic data. The enzyme was observed in two states, a novel open apo-form and a closed, induced, actinonin-bound complex (Figure 1C). Binding of actinonin resulted in a tightening of the active site through the collective closure of the entire N-terminal portion of the protein (strands b1, b2, and b3; helix a1; and CD-loop, see Movies S1 and S2, Figure 1C, and Figure S1). The amplitude of the structural change was maximal for Pro60 (Figure S1), the Ca of which was shifted 4 A˚ upon actinonin binding. This collective movement involved the formation of a ‘‘super b-sheet’’ as the result of the large rearrangement of b-strands 4 and 5 relative to the rest of the structure in which actinonin forms an additional strand bridging the two b-sheets (b1 andb2) on either side of the active site (Figure 1D and Figure S1B). As actinonin is a peptide-like compound (see Introduction and Figure 1B), this behavior closely mimics what occurs in the natural protein substrates of PDF, which also form this strand-bridging interaction. This phenomenon also accounts for the strong stabilization of the protein by actinonin, which was also challenged by differential scanning calorimetry (DSC) experiments: the Tm of AtPDF increased from 61uC to 81uC upon binding of the inhibitor (Figure 1D, see also below). Thus far, this closure of the enzyme induced by actinonin is part of the rare structural evidence for the slow, tight-binding mechanism at an atomic scale. The open state, which has never been observed, was captured not only in the two molecules of the asymmetric subunit but also in different crystals and under two distinct crystallization conditions (Table S2 and Figure 2). All r.m.s.d. values were smaller than 0.25 A˚ . The closure is very unlikely to result from crystal packing constraints, as soaking the apo-AtPDF crystals in a solution containing actinonin induced the Figure 1. Slow, tight-binding inhibition of PDF by actinonin induces conformational change in the protein. (A) Inhibition by a two-step mechanism, involving a tightening of the initial enzyme-inhibitor complex (E?I) to form a more stable complex (E?I*), with the chemical structure of actinonin (I), the natural inhibitor of PDF enzymes (E). (B) Structures of EcPDF bound to actinonin (left) and to the transition state resulting from the cleavage of its substrate, Fo-Met-Ala-Ser (right) [34,35]. (C) Superimposition of free and actinonin-bound AtPDF indicated in green and purple, respectively. The three conserved motifs of the PDF enzymes family are indicated in orange and numbered I, II, and III. Molecules A of both models were superimposed, resulting in an r.m.s.d. of 0.9 A˚ for 100% of the Ca. Left inset, close-up comparison of the open and closed forms figured in the ribbon representation. (D) Baseline-corrected DSC thermograms of free and actinonin-bound WT AtPDF recorded under the same experimental conditions. doi:10.1371/journal.pbio.1001066.g001 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 3 May 2011 | Volume 9 | Issue 5 | e1001066 structural transition from the open to the closed state within the crystals without cracking them or altering their diffracting power. Thus, crystal packing is compatible with both states of the enzyme (Figure S3). Therefore, the open structure most likely corresponds to a stable state in solution. The closed final conformation was identical to that previously reported for PDF complexes obtained either with actinonin or with a product of the reaction [34,35,44,50], indicating that this structure is common for the ligands (compare Figures 1B and 2A, and Figure S4). Hydrogen bonding was also conserved, especially the bond between the backbone nitrogen of Ile42 (corresponding to Ile44 in EcPDF, see Figure 1B and Figure S5A) and the alkyl carbonyl chain of actinonin, which potently contributes to the formation of the super b-sheet (Movie S2 and Figure S1B, see also below). Between the open and closed states, the side chains of Ile42, Phe58, and Ile130 underwent significant structural changes (Figure 3A and D and Figure S6), corresponding to a hydophobic pocket rearrangement, with Ile42 being the most affected (Figure 3). Interestingly, Ile42 is the second residue of the conserved active-site motif G41IGLAAXG (motif 1) that was previously shown to be essential for activity [51]. To assess and visualize the differences between the two states, two independent structural parameters were measured: the r.m.s.d. value with respect to the open form and the aperture angle (dap), which measures the angle made between the N- and C-domains through three fixed-points, corresponding to the Ca of three conserved residues, each sitting in one of the three conserved motifs (Figure 2A). The bi-dimensional graph of these two parameters is a good representation of the closing motion snapshots (Figure 2B) shown in Movie S1. With this tool at this stage, two states could be defined: the closed (C) and open (O) states (Figure 2B). Evidence for a Pure Induced-Fit Mechanism in the Binding of Actinonin to AtPDF Recent quantitative analyses of both conformational selection and induced fit have led to an integrated continuum—a so-called ‘‘flux-description’’—of these two limiting mechanisms [16]. According to this model, conformation selection tends to be preferred at low ligand concentrations (mM range)—that is, using detailed kinetic studies—whereas induced fit dominates at high ligand and enzyme concentrations (mM range) obtained, for instance, in NMR or crystallographic approaches. Structural studies are most useful to reveal subpopulations of biological significance. We investigated the existence of lowly populated, alternative conformations of apoPDF. To probe the occurrence of alternate conformers in the crystalline state of PDF, the new Ringer program is the most suitable investigation tool [52,53]. Ringer searches for evidence of alternate rotamers by systematically sampling electron density maps—free of model bias—around the dihedral angles of protein side chains. Two independent WT open datasets of the apoenzyme, including a high-resolution set (1.3 A˚ ), were used in the analysis. Ringer analysis revealed the existence of only one rotamer of most side chains of either molecule in the asymmetric unit, including the three main residues primarily involved in conformation change—that is, Ile 42, Phe58, and Ile130 (Figure 4A). Ringer analysis showed evidence for unmodeled alternate conformers for very few residues, including Ile121 and Phe87, or Phe119 to a much lesser extent (Figure S7). There is therefore no evidence for the occurrence of a closed conformation in the apostructure of AtPDF, supporting the hypothesis that the conformational change was essentially induced by the binding of actinonin rather than from conformational selection among multiple states occurring in the crystalline state. To further investigate the mechanism involved, we followed a kinetic approach aimed at discriminating between induced fit and population shift at low ligand concentrations (sub-mM range) [12]. The experimentally observed pseudo-first-order rate constant for the approach to equilibrium between the free components and the binary AtPDF-actinonin complex (kobs) was measured and plotted as a function of actinonin concentration. This plot yielded a hyperbolic saturation curve with a positive slope, as fully expected for a pure induced-fit mechanism (Figure 4B and C). In contrast, if the enzyme sampled two or more conformational states, the curve would imply that the value of kobs decreases with increasing ligand concentration (see, for instance, curve C in Figure 1 in [12]). The same conclusion can be reached for EcPDF and BsPDF2 (Figure 4B and C) and was already reported by others for S. aureus PDF [29]. Together, these data indicate that a pure induced-fit mechanism triggered by the binding of actinonin appears to direct the conformational change both in solution and in the crystalline state. Single Variants at Gly41 Exhibit Strongly Reduced Actinonin-Binding Potency and Catalytic Efficiency When dealing with an induced-fit mechanism, knowledge of the initial O and final C state is crucial but does not provide direct information on the position of actinonin in the encounter complex or on the sequential mechanism of the transition process. We suspected that the conserved glycine-rich motif 1 (G41IGLAAXQ) could contribute to the flexibility required for the observed structural transition. Evidence for such flexibility comes from NMR analysis of EcPDF in which a few residues show exchange cross-peaks of an additional, alternative form [38]. The most strongly affected residues are Cys90, one of the metal ligands, its neighbor Leu91, and both of the alanines within the above conserved glycine-rich motif (Figure S1B), suggesting that EcPDF undergoes conformational dynamics in a similar region. To unravel the dynamics of the recognition process, we surmised that it should be possible to freeze the conformational Table 1. Comparison of the main kinetic and thermodynamic parameters describing the inhibition of PDF by actinonin. Parameter AtPDFa EcPDFa BsPDF2a,b KI (nM)d 140610 112610 185615 KI* (nM)c 0.960.5 1.360.2 2.960.8 KI/KI* 155615 86610 6467 k5 (s21) 6103d 6369 170620 7268 k6 (s21) 6104d 461 1962 1163 k4 (s21)e 140610 112610 185615 t1/2 (min)f 2965 661 1.160.2 aThe enzyme concentrations used in the assay were 100, 50, and 25 nM for AtPDF, EcPDF, and BsPDF2, respectively. bData from [49]. cPrior to kinetic analysis for determination of the KI* value, actinonin was incubated at the final concentration in the presence of the studied enzyme set for 10 min at 37uC. The kinetic assay was initiated by the addition of a small volume of the substrate. dFor determination of KI, k5, and k6 values, actinonin was not preincubated with the enzyme. The kinetic assay was initiated by the addition of the enzyme. ek4 corresponds to the kinetic constant of the dissociation of the primary enzyme-actinonin complex. It is assumed that the rate of complex association is diffusion-limited (see Table 7.3 in [19]), that is, k3—the kinetic constant of the association of the primary enzyme-actinonin complex—is 109 M21.s21. ft1/2 is 0.693(k4+k5+k6)/k4k6 (see case of induced fit and calculation in Table 1 of [12]). In this case, t1/2,0.693/k6 because k6,,k5,,k4. doi:10.1371/journal.pbio.1001066.t001 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 4 May 2011 | Volume 9 | Issue 5 | e1001066 Figure 2. Four distinct conformational states of PDF enzymes. (A) AtPDF and several other representative bacterial PDFs are superimposed. A zoom is displayed on the right of the panel. Superimpositions were realized using ‘‘module superpose’’ in the CCP4i package and the ‘‘secondary structure matching’’ tool. The extent of aperture/closure of PDF enzymes was assessed primarily by the measurement of the aperture angle (dap), the angle made between the Ca of three strictly conserved residues (C, H, and I) of all PDFs, each characterizing a secondary crucial structure module of the active site crevice, namely b4, a2, and b1 (see Figure S1C). Each single residue belongs to one of the three conserved motifs (motifs 2, 3, and 1, respectively) and corresponds respectively to Cys91, His137, and Ile42 in AtPDF. The dap was measured in each case (see B). (B) The dap values combined with those of the r.m.s.d. associated with the superimposition of the open structure of AtPDF allows the identification of four conformational states: open (O), intermediate (I), closed (C), and super-closed (S). We compared AtPDF1B (this work and PDB CODE 3CPM; brown, orange, and yellow in A and B; black in C), EcPDF (1BS7, free enzyme; 1BS6, with Met-Ala-Ser; 1G2A, with actinonin; magenta), BsPDF2 (1LQY, with actinonin; green), LiPDF (1SV2, free; 1SZZ, with actinonin; red), and PfPDF (1JYM, free; blue). doi:10.1371/journal.pbio.1001066.g002 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 5 May 2011 | Volume 9 | Issue 5 | e1001066 Figure 3. Effect of actinonin binding on the conformation of key residues in PDF. Conformation of key residues Ile42, Phe58, and Ile130 in the different complexes: (A) in unbound WT AtPDF, (B and C) in the structure of G41Q and G41M actinonin-bound variants, respectively, and (D) of actinonin-bound WT protein. In the final complex (D), a hydrogen bond is formed between actinonin and the peptidic bond, which links Gly41 and Ile42. During the deformylation reaction, which is catalyzed by the PDF enzyme, the N-terminal formyl-methionine fits into the S1’ pocket. The solvent-accessible surface of this pocket is represented here, and only the aliphatic chain of actinonin is shown, mimicking the N-terminal methionine. (E) Free WT enzyme with the S1’ pocket shown open in two orientations (top and bottom). (F and G) S1’ pocket in the G41Q and G41M variant structures, respectively, shown in two orientations (top and bottom). (H) After the complete conformational modifications of actinonin-bound WT protein induced by actinonin binding, the S1’ pocket is shown closed in two orientations (top and bottom). (I) The four models are superimposed; the ligand-binding site is magnified: unbound WT AtPDF; G41Q and G41M actinonin-bound enzyme; and WT actinonin-bound enzyme are indicated in brown, red, orange, and yellow, respectively. Actinonin is indicated by lines. (J) A detailed view of the AtPDF ligand-binding site for all the complexes, which are superimposed, as indicated in the same colors. Arrows indicate the direction of the closing movement within the enzyme, from the open, unbound state to the closed, bound state. doi:10.1371/journal.pbio.1001066.g003 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 6 May 2011 | Volume 9 | Issue 5 | e1001066 change along the pathway by introducing selected, minor variations within the above-mentioned crucial residues involved in the collective motion. In this respect, site-directed mutagenesis of AtPDF was performed on Gly41, Ile42, and Ile130. Single substitutions were made at Gly41 (G41A/Q/M), Ile42 (I42A/F/ N/W), and Ile130 (I130A/F), and the variants were purified and characterized. These mutant proteins showed no change in overall stability, as evidenced by DSC experiments (unpublished data). However, two variants of G41, G41Q and G41M, showed dramatic effects; the kcat/Km values were reduced by three orders of magnitude due to large decreases in the kcat values compared to the WT enzyme (Figure 5A and Table S1). The reduced kcat/Km values suggest an altered ability of these variants to attain the final enzyme-transition state complex and, as a result, to give rise to possible states different from the final E:I* complex. Substitutions at positions 42 and 130 only caused small reductions in the kcat values (Figure 5A, Figure S2C, and Table S1). The actinonin- binding potency of both G41 variants was also greatly reduced (Table S1 and Figure S2B). The time-dependent inhibition by actinonin of the most active variants was then studied (Table S3). Figure 4. Evidence for an induced fit in crystalline and solution states of AtPDF. (A) Absence of evidence for alternative conformers in the crystalline state of AtPDF. Ringer plots of electron density (r) versus x1 angle for representative residues of the 3D apo-structure of AtPDF. Data were obtained with the 3M6O, 3PN2, and 3PN3 datasets (2.0 and 1.3 A˚ resolution, respectively, see Table S1). The secondary peaks in the Ile residues are observed because Ile is a branched amino acid. To reveal an alternative conformation with Ile, three peaks should be observed. (B) kobs is a saturable function of actinonin with various PDFs, including AtPDF. Data obtained for kobs, the experimentally observed pseudo-first-order rate constant for the approach to equilibrium between the free components and the binary PDF-actinonin complex, were obtained at various concentrations of actinonin in the presence of EcPDF, AtPDF, and BsPDF2. A direct plot is shown. Inset, time-course measurement of deformylation as a function of varying actinonin concentrations. (C) Inverted plot of the data in panel B, which is expected to be a straight line if the kobs is ..k6 in the case of induced fit [19]. The correlation coefficient of each line is 1.00, 0.99, and 1.00 for AtPDF, BsPDF2, and EcPDF, respectively, indicative of the accuracy of the conclusion. doi:10.1371/journal.pbio.1001066.g004 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 7 May 2011 | Volume 9 | Issue 5 | e1001066 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 8 May 2011 | Volume 9 | Issue 5 | e1001066 The half-lives of the final complexes—as assessed by comparison of the 1/k6 values—were always significantly smaller (Table S3), suggesting that the conformational change induced by actinonin binding still occurred, but the C state is destabilized relative to the O state in the mutants compared to the WT. Accordingly, actinonin strongly stabilized almost all of the variants; Tm was increased by more than 20uC. This differs from the G41M and G41Q variants, which both showed increases in the Tm of only 12uC, consistent with reduced binding potency (Table S1). Conformational Changes of Gly41 Variants Are Affected On-Pathway The two most interesting variants, G41Q and G41M, could be crystallized under the same conditions as the WT protein. In the case of G41Q, the structure of the apo-protein did not show any modifications compared to the WT structure and remained in an O conformation (Figure 2B; ‘‘O’’ zone). In contrast, the 3D structure of the G41M variant showed that the asymmetric unit was composed of two molecules with distinct structures. One molecule (chain A) is in the O state and is similar to the structures of the WT and the G41Q variant (Figure 2B; zone ‘‘O’’). The second molecule (chain B) is in a C state, closer to that observed for the WT chain in the presence of actinonin (‘‘C’’), a so-called ‘‘superclosed’’ state (Figure 2B; zone ‘‘S’’), suggesting that the substitution modified the equilibrium between the two states in solution either (i) at the step of protein synthesis by providing two conformers, the inter-conversions of which are blocked due to steric hindrance brought by the new bulkier side-chain at position 41, or (ii) by dramatically unbalancing the free inter-conversion between the O and S conformers towards the S state. Ringer analysis indicates that in the free G41M variant, many residues show evidence for unmodeled alternate conformers—including positions 58, 42, and 130—in keeping with the second hypothesis. For all variants of position G41, addition of actinonin to the crystal (Figure 3 and Figure S6) induced a closure of the protein within the crystal. Nevertheless, as expected from in silico graphic modeling followed by energy minimization, the occurrence of a bulky side chain at position 41 prevented the completion of the closure in the presence of the ligand and, hence, the formation of the hydrogen bond between the backbone nitrogen of Ile42 and actinonin. This finding is consistent with the strongly reduced Tm of the complex of the variants with actinonin compared to WT as measured by DSC. Remarkably, both S and O forms of the G41M apo-structures in the asymmetric unit of the crystal yielded a unique intermediary structure (‘‘I’’ state) upon actinonin binding (r.m.s.d. between the molecules is ,0.25 A˚ ; see also Figure 2B, zone ‘‘I’’). In this case, it is likely that the induced-fit mechanism drives the equilibrium by capturing only the O population and closing it to an intermediary step, thus depleting the pool of O conformers that is shifted sequentially back from the remaining pool of S conformers and allows the complete binding of actinonin to the enzyme. In line with the rational design of the PDF mutants, the extent of the structural differences suggests that the underlying motions are dependent on the length of the side chain (Figure S8). Together, these data account for the reduced catalytic rate, as the hydrogen bond is strictly required for the substrate to be efficiently cleaved by PDFs (Figure S8A) [54]. Therefore, from both structural and kinetic analyses, each substitution most likely reproduces intermediates along the pathway that lead to the closure of PDF around its substrate (Figure S2B). Conformational Changes of Gly41 Variants Recapitulate Closing Intermediates Analysis of the structures allows us to propose the following sequence of atomic events (Figures 3 and 2B and Figure S6). To name the various sites of the ligand and subsites of PDF, we will use the usual nomenclature found in [55], which defines the various binding pockets of a protease, where P1’ is the first side chain at the C-terminal side of the cleavage site and its binding pocket is S1’, also referred to as the hydrophobic pocket in the case of PDF. First, actinonin aligns along the S1’ pocket to form the encounter complex, which shifts the Ile130 side chain to avoid steric hindrance in the S1’ pocket, promotes rotation of the Ile42 side chain, and finally rearranges the phenyl group of Phe58. These events achieve an optimal hydrophobic S1’ pocket conformation (Figure 3), and the concomitant closure leads to the formation of a hydrogen bond between the first carbonyl group of actinonin and the backbone nitrogen of Ile42. The initial N-O distance is reduced from 5 A˚ to 2.8 A˚ , which is an optimal value for hydrogen bonding (Movie S2 and Figure S8B). Thus, the primary driving force for the active site closure appears to be the P1’:S1’ hydrophobic interaction. The C state is ultimately locked by the super-b-sheet hydrogen bonds extending across the ligand, including those involving Ile42. The DDGbinding value (2.2– 2.4 kcal/mol, Figure S8B), as calculated from the Kd values for actinonin binding to wild-type (WT) and G41M and G41Q, is consistent with the loss of a hydrogen bond that also contributes to the conformational stability of the protein [56,57]. Thus, this bond contributes to the major binding free energy difference between the two complexes (3.1 kcal/mol; Figure S8B, Tables S1 and S3, and [29]). Interestingly, the above DDGbinding values also correlate with the DDGES values derived from the kcat/Km and kcat measurements [19]. This dataset strongly correlates with the Figure 5. Inhibition and enzymatic reactions progress through an induced fit pathway. (A) The catalytic parameters Km and kcat, for all AtPDF variants are provided as a percentage of the wild-type values (WT). Detailed values are presented in Table S2. (B) Schematic model for actinonin binding to AtPDF in favor of an induced-fit pathway. PDF might exist in at least two conformational states, open (O) or closed (C). The relative abundance of each conformation would vary, depending on the enzyme type. With AtPDF, it is likely that the most abundant form is the O one, which is the only form leading to a productive complex. The superclosed form (S) is likely to show reduced affinity for the ligand because of steric occlusion of the active site. At the initial stage, the inhibitor (shown in red) binds to AtPDF (indicated in brown) in the O conformation. To reach the final key-lock state (productive closed conformation, C), two major and extreme pathways can be used. According to the conformational selection pathway, the inhibitor selects the C conformation. This pathway, which is represented by the dashed arrow, does not occur within the crystal. In contrast, the G41Q and G41M mutants, by providing the structure of the enzyme in intermediate conformations (I), prove the existence of the so- called encounter complex and confirm that the inhibitor binds to the enzyme when it is in the O conformation. The ligand-binding site is then reorganized to yield the C enzyme conformation, that is, the key-lock state. Indeed, the inhibitor binds to the enzyme through the induced-fit pathway. Each timescale was calculated using the data available in the text and corresponds to t1/2 values deduced from the calculation of 0.693/ (kinetic constant of interest). The kcat value (k2) was used to assess the timescale of catalysis in panel C, whereas, in (B), k4 assesses the first step of inhibition, and k6 is used in the case of the slow step. For the SO conversion (left, B), the lifetime of the minor form of EcPDF was used to assess the order of magnitude (see text and [38]). (C) Schematic model for the deformylation reaction catalyzed by PDF. Since actinonin is a pseudo-peptidic inhibitor, it is likely that a peptidic substrate can bind to the PDF enzyme through an induced-fit pathway, as described in (B). The key-lock state represents a transition state in which the N-formylated substrate is deformylated to yield the final reaction product. doi:10.1371/journal.pbio.1001066.g005 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 9 May 2011 | Volume 9 | Issue 5 | e1001066 gyration and van der Waals radii of the side chain at position 41 as well as the N-O distance between the first carbonyl group of actinonin and the backbone nitrogen of Ile42 (Figure S8). These results suggest that the capacity of both G41M and G41Q variants to form the transition state is a consequence of their inability to reach the fully closed state. Thus, our study of the designed Gly41 mutant enzymes reveals that, in addition to the initial and final states observed for the WT enzyme, the conformations of the Gly41 variants correspond indeed to on-pathway intermediates, thus providing snapshots along the trajectory from the O to the C state of the enzyme (Figures 2B and 3). The 3D structure of the variants in the absence of ligand is similar to that of WT, and a strict correlation exists between the completeness of the conformational change and both binding potency and catalytic efficiency. This suggests that both events require complete protein closure to generate a productive complex. The strong stabilization of AtPDF by actinonin (Figure 1D) closely mimics what occurs with its natural substrates when it reaches the transition state [34,58]. Indeed, as expected, the enzyme facilitates the final C conformation by lowering its final energy [6]. Optimal arrangement of the S1’ pocket (Figure 3) proceeds along the reaction process towards the final C conformation, triggering the alignment of reactive groups in an optimal arrangement for ligand recognition. Upon binding, actinonin alters the thermodynamic landscape for the structural transition between the O and C states. This ligand is a potent inhibitor because it can trigger the above sequence of events similar to the substrate, but unlike the substrate, it is non- hydrolyzable. Thus, by mimicking the transition state and being non-hydrolyzable (Figure 1B), the final C complex is long lasting. Ligand-Induced Conformational Closure Is Initially Triggered by the Binding of the P1’ Group in the S1’ Pocket Given the similarity between actinonin and natural substrate binding, the very slow kinetics of inhibitor binding (10-s time-scale) remains puzzling compared to the 10 ms required for catalysis (deduced from the kcat). This finding could be explained as a conformational effect during the formation of the hydrogen bond, aligning the substrate as an additional beta-sheet and eventually stabilizing the entire enzyme-ligand complex. The significantly longer time needed to reach the most stable state compared to the substrate would most likely be due to the presence of the flexible and one carbon longer metal-binding group in actinonin (i.e., hydroxamate versus formyl, Figure 1B). This suggestion is in line with the overall data obtained when we investigated more deeply the role of the first carbonyl group of the ligand. This group is well known to exert a crucial effect in both productive and unproductive ligand binding (i.e., substrate and inhibitor) [54]. In this respect, we studied the binding of compound 6b (Figure S5B), a PDF ligand that does not exhibit a reactive group at this position [49]. We observed that this compound binds strongly to both EcPDF (KI* = 6366 nM) and AtPDF (KI* = 400635 nM) but, unlike actinonin, does not display slow, tight binding as KI* = KI. This impact on binding is consistent with the absence of the hydrogen bond involving the first carbonyl group of the ligand. The 3D structure of AtPDF was determined after soaking the compound in crystals of the free, open AtPDF form. Upon binding, 6b induced a complete conformational change, identical to that observed with actinonin (Figures 2B and 6A; ‘‘O’’ state). This result further suggests that the conformational change is not induced initially by the formation of this hydrogen bond and that the encounter complex is primarily driven by the fit within the S1’ pocket. This also reveals that the timescale of the large conformational change is several orders of magnitude faster than the kinetics of slow binding and fully compatible with both the first step of actinonin binding (k4 = 140 s21; see Table 1) and the catalytic rate of the substrate (kcat = 37 s21; see Table 1 and Table S3). The 3D structure also revealed that both the P1’ and the hydroxamate groups are bound similarly to the corresponding groups of actinonin (Figure 6B). As expected, no additional bonding occurs, especially around the backbone nitrogen of Ile42 (Figure 6C). Taken together, these data allow us to conclude that the conformational change observed upon ligand binding is triggered primarily by binding in the S1’ pocket. As revealed by the binding of 6b, the one carbon longer metal-binding group fits, immediately upon recognition of the P1’ group, in the S1’ pocket and forms a bidentate complex with the metal cation, mimicking the transition state as a result. Thus, the active site is very confined and rigid due to the presence and length of the hydroxamate group (compare right and left panels in Figure 1B). As a result, compared to the complex made with the substrate, it is likely that the formation of the hydrogen bond involving the carbonyl of actinonin and the backbone nitrogen of Ile42 becomes strongly rate-limiting (k5 = 0.044 s21; Table 1). Once this hydrogen link is locked, the uncleavable bond, mimicking the labile formyl group at the transition state, stabilizes the enzyme-inhibitor complex, making it long-lasting (k6 = 0.0006 s21; Table 1) and providing a mechanistic explanation for the slow-binding effect that involves both large and fine conformational changes. The large conformational change is similar to the one occurring with the substrate, whereas the second is more subtle and locks the hydrogen bond involving the backbone nitrogen of Ile42. The second step is rate-limiting with some transition state analogs such as actinonin (Figure 5B and C). Proper Positioning of the Carbonyl Group Is Required to Stabilize the Complex at S1’ Compound 21 corresponds to another interesting derivative designed to probe the impact of the peptide bond in PDF binding [49]. In addition to the hydroxamate group, this compound features both a hydrophobic benzyl group at P1’ and a reverse peptide bond. Compound 21 shows modest but significant inhibitory potency to AtPDF1B (KI* = 400637 nM), confirming the crucial role of the peptide bond in PDF binding. After soaking with crystals of apo-AtPDF, compound 21 could be detected in high-resolution electron density maps (Figure S9A). Unlike 6b, 21 did not bind the active site of the enzyme but an alternative pocket at the surface of the protein (Figure S9B). A docking study performed with EcPDF had previously revealed this alternative binding pocket (Figure S9C; [59]). The aforementioned data indicate that the occurrence of a S1’- binding group placed in the unfavorable context of a reverse peptide bond does not stably promote binding at the active site of AtPDF. Upon binding of 21, the 3D structure of both molecules of the asymmetric unit remain in an O conformation (r.m.s.d. ,0.2 A˚ with respect to the apo-structures in the ‘‘O’’ state). This finding suggests that only the binding of compounds entering the S1’ pocket, such as actinonin or 6b, induces conformational change, in keeping with the crucial role of the P1’ group if located in the frame of a classic peptide bond. Moreover, we noticed that the binding pocket of 21 was located on the rear side of the true S1’ pocket and induced a weak modification of the P1’ hosting platform (Figure S9D). Indeed, when crystals of the 21:AtPDF complex were soaked in actinonin, the final 3D structure no longer showed evidence of compound 21 occupancy greater than 5%. Instead, this structure revealed both actinonin and closing of the protein (Table S2). The r.m.s.d. between this structure and that The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 10 May 2011 | Volume 9 | Issue 5 | e1001066 obtained directly with actinonin was less than 0.2 A˚ ; the actinonin position was virtually identical, indicating that the protein had retained full capacity for binding actinonin and closing despite the presence of compound 21. We conclude that actinonin does compete with 21 because of the overlap at P1’ of AtPDF1B (Figure S9C). As the actinonin S1’ subsite strongly mimics that of a true substrate, this result also explains the inhibitory behavior of 21 towards AtPDF. Discussion Although PDF catalysis has been extensively studied and the mechanism has been elucidated [34], how the enzyme achieves the catalytically competent state remains unknown. Here, we provide insight on how the enzyme might reach a catalytically competent conformation, demonstrating that the reactive groups move into proximity to promote catalysis (Figures 2B and 5C). We suggest Figure 6. Effect of 6b binding on the conformation of key residues of PDF. Superimposition of free, 6b-, and actinonin-bound AtPDF indicated in brown, red, and yellow, respectively. (A) Molecule A in the three models was superimposed, resulting in an r.m.s.d. of 0.9 A˚ for 100% of the Ca. Actinonin is shown in yellow and 6b in red. (B) Conformation of key residues Ile42, Phe58, and Ile130 in the different complexes and in unbound WT AtPDF. Actinonin is shown in yellow and 6b in red. (C) A detailed view of the AtPDF ligand-binding site for both actinonin and 6b complexes, which are indicated by sticks and are superimposed. The two ligands are colored in pale and dark grey, respectively. The hydrogen bond made by actinonin only is shown. doi:10.1371/journal.pbio.1001066.g006 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 11 May 2011 | Volume 9 | Issue 5 | e1001066 that the motions of the catalytic centre starting with free ligand- PDF favor a final configuration that is optimal for binding and/or catalysis (Figures 1B, 2B, and 5B and C). In our model, we propose that free PDF might exist in at least two conformational states, that is, open (O) or super-closed (S). The relative abundance of each conformation varies by enzyme type and incubation conditions, explaining why both conformations have not been trapped thus far. In the case of AtPDF, it is likely that the most abundant form corresponds to an O state, which is the form that leads to a productive complex. Indeed, in the NMR spectra for EcPDF, a few residues show exchange cross-peaks from an additional, alternative form [38]. The most strongly affected residues are Cys90, one of the metal ligands, its neighbor Leu91, as well as Ala47 and Ala48 on the facing strand. This suggests that EcPDF exists in at least two conformations (‘‘S’’ and ‘‘C’’; see Figure 2B), which undergo slow interconversion on the NMR timescale. The 3D structure of the major conformation (75%, lifetime 300 ms) could be solved at high resolution, but the structure of the minor form (25%, lifetime 100 ms), which exhibits very weak signals, could not be solved [38]. This conformation appears to correspond to that of the complex obtained with the product of the reaction (Met-Ala-Ser). A very similar situation—although more balanced between the two states—appears to occur in the case of variant G41M, suggesting that a mechanism involving conformational selection followed by induced fit is a general model for PDF and that AtPDF is a specific case where population shift virtually does not occur as the free enzyme is completely in the O conformation. This is also in line with data obtained with L. interrogans PDF (LiPDF), which reveal conformers in both the S and C states (see Figure 2B) and suggest a population-shift mechanism [43]. It is interesting to note that LiPDF is a poorly active PDF [60]. According to the representation shown in Figure 2B, Plasmodium falciparum PDF (PfPDF), a poorly active PDF [61], was retrieved only in the S state. Finally, weak decompaction of the structure of Bacillus cereus and Staphylococcus aureus PDFs in the presence of actinonin have been described [45,46]. These examples suggest that the enzyme is trapped in the S conformer in the free state and converts to the C conformer when bound to actinonin, suggesting that the S conformer is overrepresented in solution compared to the O state, unlike AtPDF. This study of AtPDF—including 10 different crystal structures of apo- and complexed enzyme variants—reveals the 3D structure of a PDF in at least four distinct states. This includes the O form, the occurrence of which is crucial for catalysis, as it is the active form. Here, we propose that the transition from the O to the C state is directly induced by the ligand. Indeed, the O form, which is captured in the crystal, undergoes closure directly upon ligand binding in our soaking experiments. Progression to this closure involves intermediary states (‘‘I’’) similar to those observed with variants G41Q and G41M in the presence of actinonin (see Figure 2B). Extrapolating the situation to catalysis, which occurs in the crystalline states of PDF, it is likely that hydrolysis of the substrate frees the enzyme in its S state, which in turn needs to open to accommodate a new substrate (Figures 2B and 5C). This is well illustrated in the 3D structure of EcPDF complexed with a product of the reaction, obtained after co-crystallization of the enzyme with the substrate in a closed conformation [34]. The S free form is likely to exhibit a slower on-rate for the ligand (k3) compared to the O form because of steric occlusion of the active site (Figure S10). In support of this hypothesis, recent data show that the 3D structure of a C-terminally truncated, poorly active version of AtPDF is in the C conformation in the unbound state, although crystallized under conditions identical to ours [62,63]. This structure is similar to that of chain B, one of the two molecules of the asymmetric subunit of variant G41M (Figure 2B). This suggests that alterations remote from the active site significantly unbalance the equilibrium between the two conform- ers, thus altering the efficiency of the reaction (Figure 5C). As the S version corresponds to a significantly less active version of AtPDF compared to that reported in our present work, this further confirms that, compared to the O state, the S state has a significantly weaker propensity to bind substrate or a close mimic ligand, such as actinonin. Comparison of the 3D structures of the free-closed and the ligand-bound-closed forms reveals some differences responsible for the slight steric reduction of the active site of free-closed AtPDF1B with respect to that of the actinonin- AtPDF1B complex (Figure S10A), including the side chain of Ile42 burying the S1’ binding pocket (Figure S10B). Overall, these data suggest that an S form might exist under the free state but that it would feature a k3 value with respect to the ligand that is significantly weaker than that of the O form, which would strongly slow down the reaction or the binding as a result. With the interaction scheme proposed in our model (Figure 5B and C), the ligand/substrate binds more easily to the O form and induces the optimal conformation of the enzyme to reach the transition state, thus allowing the reaction to be efficiently catalyzed. In the final model (Figure 5C), there is both conformational selection and induced fit subsequently involved in line with the recently proposed existence of such mixed mechanisms for other enzymes [15,16]. Nevertheless, in our model (Figure 5C), we suggest that induced fit is the primary mechanism, as it provides energy input from the ligand, which eventually drives the enzyme towards the productive key-lock complex. Unambig- uous distinction between the relative contributions of the two mechanisms is deduced from the observation that kobs is a saturable function of actinonin with various PDF, including EcPDF, BsPDF, AtPDF (Figure 4B and C), and S. aureus PDF [12,16,29,49]. Using crystallographic reconstruction analysis involving enzyme variants, motions of small mobile loops and movie reconstructions of snapshots of catalytic events have been previously documented [1–3,64–66], often by visualizing the binding of unnatural inhibitors and not necessarily mimicking closely the substrate and transition state as actinonin does [67,68]. However, only a few examples make use of soaking conditions of a crystal to promote the motion and show the importance of induced fit [1,69]. None of these data show a motion of the amplitude revealed here with PDF and a large stabilization of the complex involving the formation of the four-stranded b-sheet superstructure and the entire N-domain of the enzyme. Compared to previous crystallographic analyses, our work integrates biophysical, computational, and kinetic analyses to reconstruct the whole picture, allowing a better understanding of the slow-binding mechanism. While our work primarily focused on an induced-fit mechanism of enzyme inhibition and catalysis, it should be emphasized that this phenomenon is also applicable to the broader area of receptor-ligand interactions. For example, in all cases where conformational change mechanisms have been proposed for kinase inhibitors without supporting experimental data [12,26], further experimental work must be provided to clarify the precise mechanism. We expect this will have important implications on how one conducts future drug-discovery efforts against such enzymes [70]. Materials and Methods Protein Expression and Purification Expression and purification of mature Arabidopsis thaliana PDF1B and all variants (i.e., AtPDF) were derived from the previously The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 12 May 2011 | Volume 9 | Issue 5 | e1001066 described protocol [37]: the lysis supernatant after sonication was applied on a Q-Sepharose column (GE Healthcare; buffers A and B as described containing 5 mM NiCl2) followed by Superdex-75 chromatography (GE Healthcare) using buffer C consisting of buffer A supplemented with 0.1 M NaCl. For crystallization experiments, the protein was purified further. The sample was concentrated on an Amicon Ultra-15 centrifugal filter unit (Millipore Corp.) with a 5-kDa cut-off and applied to a MonoQ HR5/5 column (GE Healthcare) previously equilibrated in buffer A (50 mM Hepes, pH 7.5, and 5 mM NiCl2). Elution was performed with a 50-mL gradient from 0% to 100% buffer B. The buffer of the pooled purified AtPDF1B was exchanged using a PD-10 desalting column (GE Healthcare) to yield a protein solution in 50 mM Hepes, pH 7.5, 0.1 M NaCl, and 5 mM NiCl2 (buffer C). The protein was concentrated on an Amicon Ultra-15 centrifugal filter unit. The resulting AtPDF1B preparation was frozen in aliquots and stored at 280uC (for crystallization purposes) or diluted 2-fold in 100% glycerol and stored at 220uC (for enzymatic purposes). The typical yield was 5–10 mg AtPDF per liter of culture. All purification procedures were performed at 4uC. Samples of the collected fractions were analyzed by SDS-PAGE on 12% acrylamide gels, and protein concentrations were estimated from the calculated extinction coefficients for each variant. Site-directed mutagenesis of AtPDF sequence in plasmid pQdef1bDN [36] was carried out using the QuickChange Site- Directed Mutagenesis Kit (Stratagene). Enzymology Assay of PDF activity was coupled to formate dehydrogenase, where the absorbance of NADH at 340 nm was measured at 37uC as previously described [71]. For measurements of classical kinetic parameters (i.e., Km and kcat), the reaction was initiated by addition of the substrate Fo-Met-Ala-Ser to the mixture containing purified enzyme in the presence of 1 mM NiCl2. The kinetics parameters were derived from iterative non-linear least square calculations using the Michaelis-Menten equation based on the experimental data (Sigma-Plot; Kinetics module). For determination of kinetic parameters related to actinonin, the reaction mixture contained 750 mM NiCl2. In some cases, the mixture containing PDF and actinonin was incubated for 15 min at 37uC before kinetic analysis, which was initiated by the addition of substrate. The same protocol was used to determine the dissociation constant of actinonin [KI* = k4/(k3+k3k5/k6)], but the initial reaction velocities were measured with varying concentrations of Fo-Met-Ala-Ser and actinonin. The data were then calculated according to the method of Henderson, which can be used to determine the dissociation constant of the tight-binding competitive enzyme inhibitor [28,49,72] by varying both the inhibitor and substrate concentra- tions. To determine KI, k5, and k6, the reaction was initiated by the addition of enzyme as previously described [29,49]. KI*app measurements were used for comparative studies of AtPDF variants (Table S3) at a concentration of 2 mM substrate by varying the concentration of actinonin. KI*app is the slope of the v[Actinonin]/v0 line curve. kobs was fitted from the kinetic data without preincubation with vI = vs + (v0 2 vs)e2kobst where vI is the observed velocity at a given concentration of inhibitor I, v0 is the velocity, and vs is the steady-state velocity [18]. From the set of values obtained at various concentrations of I, k5 and k6 could be derived using kobs = k6 + k5[I]/(KI + [I]). By choosing a set of values with kobs..k6, 1/kobs = 1/k5(KI/[I] +1) and 1/kobs = f(1/[I]) is expected to be a straight line in case of induced fit whose positive slope corresponds to 1/k5. k6 was derived from equation k6 = k5/ (KI/KI*21) [18,19]. Microcalorimetry ITC experiments were performed using a VP-ITC isothermal titration calorimeter (Microcal Corp.). Experiments were per- formed at 37uC. For each experiment, injections of 10 mL actinonin (180 mM) were added using a computer-controlled 300 mL microsyringe at intervals of 240 s into the Ni-AtPDF variant solution (5 to 10 mM, cell volume = 2.1 mL) dissolved in buffer C with stirring at 310 rpm. A theoretical titration curve was fitted to the experimental data using the ORIGIN software (Microcal). This software uses the relationship between the heat generated after each injection and DHu (enthalpy change in kcal/ mol), KA (the association binding constant in M21), n (number of binding sites per monomer), total protein concentration, and free and total ligand concentrations. The thermal stability of the WT and variants of Ni-AtPDF1B was studied by DSC using VP-DSC calorimetry (Microcal Corp.). DSC measurements were made with 10 mM protein solutions in buffer C. The actinonin concentration was 20 mM. The same buffer was used as a reference. All solutions were degassed just before loading into the calorimeter. Scanning was performed at 1uC/min. The temperature dependence of the partial molar capacity (Cp) was expressed in kcal/K after subtracting the buffer signal using Origin(R) software. Crystallization and Soaking Experiments Crystallization conditions were screened by a robot using the sitting drop vapor diffusion method. Crystals were obtained and optimized at 20uC with 15%–20% PEG-3350 and either 0.1 or 0.2 M zinc acetate. The drops were formed by mixing 2 mL of a solution containing 2 to 4 mg/mL protein and 2 mL of the crystallization solution. Crystals were soaked for 24 h by adding actinonin to the crystallization drops at a final concentration of 5 mM. Cryoprotection was achieved by placing crystals for 30 s in a solution that was composed of 20% PEG-3350 and 0.2 M zinc acetate, supplemented with 5%, 10%, and 15% glycerol. Crystals were then directly flash frozen in liquid nitrogen using cryoloops (Hampton Research). Crystals were also grown under conditions described for the C-terminally deleted, weakly active version of AtPDF [63]. X-Ray Diffraction Data Collection Data collections were performed at 100 K at the European Synchrotron Radiation Facility (Grenoble, France) on station ID29, FIP-BM30A, ID14-1, and ID23-2, and at SOLEIL (Gif-sur- Yvette, France) on station PROXIMA1. In each case, a single crystal was used to collect a complete dataset. Data were processed and scaled using XDS software [73]. Two crystal forms were encountered with different cell parameters. In each case, b parameter was nearly equal to a, and data could be indexed into two space groups, P212121 or P43212. The data are shown in Table S2. Structure Determination and Refinement The structure of free AtPDF was solved by molecular replacement with Phaser [74] followed by a rigid-body refinement by CNS [75] using coordinates from the Plasmodium falciparum PDF (PDB code 1RL4) [76] as a search model. The structures of actinonin-bound proteins—that is, WT and mutants—were solved using rigid-body refinement by CNS of the free AtPDF structure. The ten final models were obtained by manual rebuilding using TURBO-FRODO [77] and combined with refinement of only calculated phases using CNS and Refmac [78] software. No non- crystallographic symmetries were used. Quality control of the three models was performed using the PROCHECK program The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 13 May 2011 | Volume 9 | Issue 5 | e1001066 [79]. To probe for alternative conformers, Ringer was used [53]. Ringer is a program to detect molecular motions by automatic X- ray electron density sampling, and can be accessed at http:// ucxray.berkeley.edu/ringer.htm. Accession Numbers PDB codes for the PDF structures presented within this manuscript are as follows: 3M6O, 3PN2, 3M6P, 3O3J, 3PN3, 3PN4, 3PN5, 3M6Q, 3PN6, and 3M6R. UniProtKB accession numbers for other PDF studied are P0A6K3 (EcPDF) and O31410 (BsPDF). Supporting Information Figure S1 Alignment of PDF sequences and secondary structures. (A) PDF1B from Arabidopsis thaliana (AtPDF1B) is compared with bacterial type 1B (EcPDF and LiPDF), pathogenic protozoa (PfPDF1B), eukaryotic mitochondrial PDF1A from A. thaliana (AtPDF1A), and bacterial type 2 (BsPDF2). This figure was created with ENDscript [80]. The sequence alignment was realized with the algorithm muscle included in ENDscript, and modified according to the superimposition of structures. The blue frames indicate conserved residues, white characters in red boxes indicate strict identity, and red characters in yellow boxes indicate homology. The secondary structures at the top (a-helices, 310 helices, b-strands, and b-turns are shown by medium squiggles, small squiggles, arrows, and TT letters, respectively) were predicted by DSSP [81]. Relative accessibility (acc) of subunit A is shown by a blue-colored bar below sequence. White is buried, cyan is intermediate, and blue with red borders is highly exposed. A red box means that relative accessibility is not calculated for the residue, because it is truncated. Hydropathy (hyd) is calculated from the sequence according to [82]. It is shown by a second bar below accessibility: pink is hydrophobic, grey is intermediate, and cyan is hydrophilic. Motifs 1 (41GwGwAAXQ48), 2 (89EGCLS93), and 3 (133HEwDH137), where w is a hydrophobic amino acid, are labeled by red stars below the sequence alignment. To simplify the nomenclature, AtPDF1B is referred to as AtPDF throughout the text. (B) Topology cartoon of AtPDF, free (left) or actinonin bound (right), in the same color code as (A). Actinonin (represented by the yellow arrow) binding to the ligand binding site allows the linkage of the two distinct b-sheets into one single b-sheet, by mimicking an additional b-strand. PDB sum (http://www.ebi.ac. uk/thornton-srv/databases/pdbsum/) was used. (C) 3-D structure of AtPDF is represented showing the position of the residues discussed in the text, indicated in red. (EPS) Figure S2 Microcalorimetric titration of AtPDF with actinonin. Data were obtained at 37uC by an automated sequence of 28 injections of 180 mM actinonin from a 300 ml syringe into the reaction cell, which contain 9.85 mM AtPDF. The volume of each reaction was 10 ml, and injections were made at 240 s intervals. Top, raw data from the titration. Each peak corresponds to the injection. Bottom, the peaks in the upper panel were integrated with ORIGIN software and the values were plotted versus injection number. Each point corresponds to the heat in mcal generated by the reaction upon each injection. The solid line is the curve fit to the data by the Origin program. This fit yields values for Kd. Experiments were done with wild type protein and others variants, and gave similar raw data and curve fit. (A) WT; (B) variant G41M; (C) variant I42W. (EPS) Figure S3 Binding of actinonin to AtPDF does barely modify the crystal packing. (A) Crystal pack of the two complexes: open, free complex (left) and bound to actinonin (right) (B). Non-crystallo- graphic contacts into asymmetric unit are not modified by closing movement of the protein due to actinonin binding, except for zinc atom number 6. This metal ion is coordinated by side chains of Asp40 and Glu63, and water molecules, Asp40 and Glu63 being hydrogen bonded by side chain of Lys38 of the other subunit of the asymmetric unit. With the closing movement of the protein into the crystal, Ca of Asp40 shifted by 3.1 A˚ and its side chain flipped by 90u. Therefore, it does no longer participate to the coordination shell of this Zn2+ ion. However, it is still hydrogen bonded by Lys38 from chain B. (EPS) Figure S4 Binding of actinonin to AtPDF closely mimics both actinonin and product binding to EcPDF. Superimposition of EcPDF and AtPDF bound to either actinonin (1LRU PDB code, panel A) or Met-Ala-Ser (1BS6 PDB code, panel B), the product of the reaction. The r.m.s.d. value is 1.11 A˚ for 151 Ca superimposed. (EPS) Figure S5 The ligand binding site of AtPDF. This picture shows the residues of AtPDF that are in contact with actinonin (left) and 6b (right) according to the 3-D structure; this should be compared to the similar scheme shown in Figure 1B for EcPDF. (EPS) Figure S6 Electronic densities of the moving side-chains and of actinonin at the binding site in some variants of AtPDF. Actinonin and selected residues (G/Q/M41, I42, F58, and I130) are drawn in stick and are shown in their FO–FC electron density omit maps contoured at 2s, in free wild-type AtPDF (two crystallization conditions, WT1 and WT2), and ligand-bound WT (actinonin, 6b and 21), G41Q, and G41M variants. (EPS) Figure S7 Only few residues show alternative conformation in AtPDF. Alternative conformers in the crystalline state of AtPDF. Ringer plots of electron density (r) versus x1 angle for representative residues of the 3-D apostructure of AtPDF. Data were obtained with the 3M6O dataset (see Table S1). The secondary peaks in the Ile residues are observed because Ile is a branched amino acid. To evidence an alternative conformation with Ile, three peaks should be observed. (EPS) Figure S8 Impact of induced fit on the binding free energy of actinonin depends on the capacity to stabilize a hydrogen bond with PDF. (A) The gyration radii [83] of the side chain occurring at position 41 is displayed with black squares and compared to the kcat/Km values (grey bars). (B) The distance between the NH of I42 and the CO of actinonin was measured in each case. The percentage of the distance required to make a hydrogen bond (2.8 A˚ ) is reported (dark squares). The difference of binding free energy (DDGbinding) between the open, free state and the variants closed complexes of the G41 variants are displayed as grey bars. The values were calculated as follows. For the WT, it corresponds to the RT ln(KI*/KI) value [29], where R is the ideal gas constant and T is the temperature in Kelvin. RT is 0.616 kcal.mol21 at 37uC. For the G41M and G41Q variants, the DDGbinding corresponds to RT ln(KI-G41variant/KD-WT). The obtained values are similar to that obtained if the kcat/Km substitutes the KD value in the calculation (DDGbinding = RT ln(kcat/Km –G41variant/kcat/Km –WT). (EPS) Figure S9 Compound 21 does not bind AtPDF1B at S1’. (A) 21 is shown in ball-and-stick format in its FO–FC electron density omit The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 14 May 2011 | Volume 9 | Issue 5 | e1001066 map contoured at 2s. (B) Binding site of 21 into AtPDF1B is detailed. Red and blue residues indicate residues that accommo- date the ‘‘phenylalanine’’ and ‘‘trimethyl’’ groups of 21, respectively. (C) Overall view of 21 binding site (left). Molecular surface of AtPDF is represented, as well as 21 in ball-and-stick format. Residues belonging to the 21 binding pocket are colored in orange. For comparison, molecular surface of EcPDF (PDB code 1G2A) in the same orientation is also represented, with residues forming the new ligand binding pocket colored in orange. Actinonin is represented in ball-and-stick format and is seen through the molecular surface of each PDF. (D) Ball-and-stick representation of the interaction network around compound 21. The metal cation is shown as a grey sphere. (EPS) Figure S10 Poorly active versions of AtPDF are in a closed conformation incompatible with actinonin binding. (A) Free and close AtPDF were superimposed as in Figure 1C and are figured in brown and yellow, respectively. Both the G41M (chain B, shown in orange) and the free C-deleted weakly active AtPDF versions ([63], colored in purple, PDB entry code 3CPM) were superim- posed, to the two structures, showing that they both fit better to the ligand-bound full-length close form than to the free open form, but that the closure is further pronounced, burying the entrance to a ligand. (B) Close-up showing that the shape of the S1’ pocket of the poorly active closed versions make it poorly available to P1’ recognition (see circled Ile142 and Ile130 side chains). (EPS) Table S1 Catalytic properties of AtPDF. Nm, not measurable; ND, not determined; WT, is wild-type. aKinetic constants were determined using the coupled assay as indicated in Materials and Methods with substrate Fo-Met-Ala-Ser, in the presence of 100 nM enzyme variant and 750 mM NiCl2, at 37uC. The relative value of kcat/Km for wild-type AtPDF was set at 100%. bData correspond to the binding constant of actinonin as obtained either from ITC or from enzymatic analysis when indicated with an asterisk. cData from Table S3. dGyration radii are from [83]. (DOC) Table S2 Crystallographic data and refinement statistics. Values in parentheses are for the outer resolution shell. aRsym (I) = ShklSi|Ihkl,i 2 ,Ihkl.|/ShklSi|Ihkl,i|, where ,Ihkl. is the mean intensity of the multiple Ihkl,i observations for symmetry-related reflections. bRwork = 1006(Shkl|Fobs 2 Fcalc|/Shkl|Fobs|). Rfree is a test set including ,5% of the data. cPercentage of residues in most-favored/additionally allowed/generously allowed/disal- lowed regions of the Ramachandran plot. dCompound 21 was added first, and actinonin afterwards. (DOC) Table S3 Kinetic parameters for inhibition of some AtPDF variants by actinonin. The enzyme concentration used in the assay was 100 nM. Prior to kinetic analysis for determination of KI*app values, actinonin was incubated in the presence of each variant set at the final concentration for 10 min at 37uC; kinetic assay was started by adding a small volume of the substrate. For determination of KI, k5, and k6 values, actinonin was not pre- incubated with enzyme and kinetic assay was started by adding the enzyme. (DOCX) Movie S1 Dynamics of actinonin binding to peptide deformylase and closure of the active site. (WMV) Movie S2 Progressive motions of the main side chains at the active site and final locking of the hydrogen bond. (WMV) Acknowledgments We are strongly indebted to James Fraser and Tom Alber (University of California, Berkeley, USA) for introducing us to Ringer before the release of the freely available downloadable version. We thank Benoıˆt Gigant, Virginie Gueguen-Chaignon, Solange Morera, and Philippe Peynot (CNRS, Gif-sur-Yvette, France) for help with data processing and access to the crystallization facilities. We also thank Magali Nicaise-Aumont (IBBMC, Orsay, France), who performed the microcalorimetry experi- ments. We are grateful to the staff of the European Synchrotron Radiation Facility (ESRF) and SOLEIL beamlines for their help during data collection. Author Contributions The author(s) have made the following declarations about their contributions: Conceived and designed the experiments: SF CG TM. Performed the experiments: AB SF. Analyzed the data: FD MD SF CG TM. Contributed reagents/materials/analysis tools: IA MD CG TM. Wrote the paper: CG TM. References 1. Knowles JR (1991) Enzyme catalysis: not different, just better. Nature 350: 121–124. 2. Hammes GG (2002) Multiple conformational changes in enzyme catalysis. Biochemistry 41: 8221–8228. 3. Benkovic SJ, Hammes-Schiffer S (2003) A perspective on enzyme catalysis. Science 301: 1196–1202. 4. Henzler-Wildman K, Kern D (2007) Dynamic personalities of proteins. Nature 450: 964–972. 5. Teilum K, Olsen JG, Kragelund BB (2009) Functional aspects of protein flexibility. Cell Mol Life Sci. 6. Sullivan SM, Holyoak T (2008) Enzymes with lid-gated active sites must operate by an induced fit mechanism instead of conformational selection. Proc Natl Acad Sci U S A 105: 13829–13834. 7. Weikl TR, von Deuster C (2009) Selected-fit versus induced-fit protein binding: kinetic differences and mutational analysis. Proteins 75: 104–110. 8. Johnson KA (2008) Role of induced fit in enzyme specificity: a molecular forward/reverse switch. J Biol Chem 283: 26297–26301. 9. Bourgeois D, Royant A (2005) Advances in kinetic protein crystallography. Curr Opin Struct Biol 15: 538–547. 10. Katona G, Carpentier P, Niviere V, Amara P, Adam V, et al. (2007) Raman- assisted crystallography reveals end-on peroxide intermediates in a nonheme iron enzyme. Science 316: 449–453. 11. Koshland DE (1958) Application of a theory of enzyme specificity to protein synthesis. Proc Natl Acad Sci U S A 44: 98–104. 12. Tummino PJ, Copeland RA (2008) Residence time of receptor-ligand complexes and its effect on biological function. Biochemistry 47: 5481–5492. 13. Boehr DD, Nussinov R, Wright PE (2009) The role of dynamic conformational ensembles in biomolecular recognition. Nat Chem Biol 5: 789–796. 14. Bosshard HR (2001) Molecular recognition by induced fit: how fit is the concept? News Physiol Sci 16: 171–173. 15. Benkovic SJ, Hammes GG, Hammes-Schiffer S (2008) Free-energy landscape of enzyme catalysis. Biochemistry 47: 3317–3321. 16. Hammes GG, Chang YC, Oas TG (2009) Conformational selection or induced fit: a flux description of reaction mechanism. Proc Natl Acad Sci U S A 106: 13737–13741. 17. Fersht AR (1998) Structure and mechanism in protein science. New York: W.H. Feeman & Co. 18. Morrison JF, Walsh CT (1988) The behavior and significance of slow-binding enzyme inhibitors. Adv Enzymol Relat Areas Mol Biol 61: 201–301. 19. Copeland RA (2005) Evaluation of enzyme inhibitors in drug discovery: a guide for medicinal chemists and pharmacologists. New Jersey: John Wiley & Sons. 296 p. 20. Dash C, Vathipadiekal V, George SP, Rao M (2002) Slow-tight binding inhibition of xylanase by an aspartic protease inhibitor: kinetic parameters and conformational changes that determine the affinity and selectivity of the bifunctional nature of the inhibitor. J Biol Chem 277: 17978–17986. 21. Mac Sweeney A, Lange R, Fernandes RP, Schulz H, Dale GE, et al. (2005) The crystal structure of E.coli 1-deoxy-D-xylulose-5-phosphate reductoisomerase in a The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 15 May 2011 | Volume 9 | Issue 5 | e1001066 ternary complex with the antimalarial compound fosmidomycin and NADPH reveals a tight-binding closed enzyme conformation. J Mol Biol 345: 115–127. 22. Chou CJ, Herman D, Gottesfeld JM (2008) Pimelic diphenylamide 106 is a slow, tight-binding inhibitor of class I histone deacetylases. J Biol Chem 283: 35402–35409. 23. Barb AW, Jiang L, Raetz CR, Zhou P (2007) Structure of the deacetylase LpxC bound to the antibiotic CHIR-090: Time-dependent inhibition and specificity in ligand binding. Proc Natl Acad Sci U S A 104: 18433–18438. 24. Dunford JE, Kwaasi AA, Rogers MJ, Barnett BL, Ebetino FH, et al. (2008) Structure-activity relationships among the nitrogen containing bisphosphonates in clinical use and other analogues: time-dependent inhibition of human farnesyl pyrophosphate synthase. J Med Chem 51: 2187–2195. 25. Bateman RL, Ashworth J, Witte JF, Baker LJ, Bhanumoorthy P, et al. (2007) Slow-onset inhibition of fumarylacetoacetate hydrolase by phosphinate mimics of the tetrahedral intermediate: kinetics, crystal structure and pharmacokinetics. Biochem J 402: 251–260. 26. Copeland RA, Pompliano DL, Meek TD (2006) Drug-target residence time and its implications for lead optimization. Nat Rev Drug Discov 5: 730–739. 27. Gordon JJ, Kelly BK, Miller GA (1962) Actinonin: an antibiotic substance produced by an actinomycete. Nature 195: 701–702. 28. Chen DZ, Patel DV, Hackbarth CJ, Wang W, Dreyer G, et al. (2000) Actinonin, a naturally occurring antibacterial agent, is a potent deformylase inhibitor. Biochemistry 39: 1256–1262. 29. Van Aller GS, Nandigama R, Petit CM, Dewolf WE, Jr., Quinn CJ, et al. (2005) Mechanism of time-dependent inhibition of polypeptide deformylase by actinonin. Biochemistry 44: 253–260. 30. Rajagopalan PT, Grimme S, Pei D (2000) Characterization of cobalt(II)- substituted peptide deformylase: function of the metal ion and the catalytic residue Glu-133. Biochemistry 39: 791–799. 31. Schmitt E, Guillon JM, Meinnel T, Mechulam Y, Dardel F, et al. (1996) Molecular recognition governing the initiation of translation in Escherichia coli. A review. Biochimie 78: 543–554. 32. Giglione C, Fieulaine S, Meinnel T (2009) Cotranslational processing mechanisms: towards a dynamic 3D model. Trends Biochem Sci 34: 417–426. 33. Nguyen KT, Hu X, Pei D (2004) Slow-binding inhibition of peptide deformylase by cyclic peptidomimetics as revealed by a new spectrophotometric assay. Bioorg Chem 32: 178–191. 34. Becker A, Schlichting I, Kabsch W, Groche D, Schultz S, et al. (1998) Iron center, substrate recognition and mechanism of peptide deformylase. Nat Struct Biol 5: 1053–1058. 35. Guilloteau JP, Mathieu M, Giglione C, Blanc V, Dupuy A, et al. (2002) The crystal structures of four peptide deformylases bound to the antibiotic actinonin reveal two distinct types: a platform for the structure-based design of antibacterial agents. J Mol Biol 320: 951–962. 36. Giglione C, Serero A, Pierre M, Boisson B, Meinnel T (2000) Identification of eukaryotic peptide deformylases reveals universality of N-terminal protein processing mechanisms. EMBO J 19: 5916–5929. 37. Serero A, Giglione C, Meinnel T (2001) Distinctive features of the two classes of eukaryotic peptide deformylases. J Mol Biol 314: 695–708. 38. Dardel F, Ragusa S, Lazennec C, Blanquet S, Meinnel T (1998) Solution structure of nickel-peptide deformylase. J Mol Biol 280: 501–513. 39. Meinnel T, Blanquet S, Dardel F (1996) A new subclass of the zinc metalloproteases superfamily revealed by the solution structure of peptide deformylase. J Mol Biol 262: 375–386. 40. Larue V, Seijo B, Tisne C, Dardel F (2009) 1H, 13C and 15N NMR assignments of the E. coli peptide deformylase in complex with a natural inhibitor called actinonin. Biomolecular NMR Assignments 3: 153–155. 41. Amero CD, Byerly DW, McElroy CA, Simmons A, Foster MP (2009) Ligand- induced changes in the structure and dynamics of Escherichia coli peptide deformylase. Biochemistry 48: 7595–7607. 42. Berg AK, Srivastava DK (2009) Delineation of alternative conformational states in Escherichia coli peptide deformylase via thermodynamic studies for the binding of actinonin. Biochemistry 48: 1584–1594. 43. Zhou Z, Song X, Gong W (2005) Novel conformational states of peptide deformylase from pathogenic bacterium Leptospira interrogans: implications for population shift. J Biol Chem 280: 42391–42396. 44. Clements JM, Beckett RP, Brown A, Catlin G, Lobell M, et al. (2001) Antibiotic activity and characterization of BB-3497, a novel peptide deformylase inhibitor. Antimicrob Agents Chemother 45: 563–570. 45. Yoon HJ, Kim HL, Lee SK, Kim HW, Lee JY, et al. (2004) Crystal structure of peptide deformylase from Staphylococcus aureus in complex with actinonin, a naturally occurring antibacterial agent. Proteins 57: 639–642. 46. Moon JH, Park JK, Kim EE (2005) Structure analysis of peptide deformylase from Bacillus cereus. Proteins 61: 217–220. 47. Park J, Fu H, Pei D (2004) Peptidyl aldehydes as slow-binding inhibitors of dual- specificity phosphatases. Bioorg Med Chem Lett 14: 685–687. 48. Velazquez-Campoy A, Ohtaka H, Nezami A, Muzammil S, Freire E (2004) Isothermal titration calorimetry. Curr Protoc Cell Biol Chapter 17: Unit 17 18. 49. Boularot A, Giglione C, Petit S, Duroc Y, Sousa RA, et al. (2007) Discovery and refinement of a new structural class of potent peptide deformylase inhibitors. J Med Chem 50: 10–20. 50. Hackbarth CJ, Chen DZ, Lewis JG, Clark K, Mangold JB, et al. (2002) N-alkyl urea hydroxamic acids as a new class of peptide deformylase inhibitors with antibacterial activity. Antimicrob Agents Chemother 46: 2752–2764. 51. Ragusa S, Mouchet P, Lazennec C, Dive V, Meinnel T (1999) Substrate recognition and selectivity of peptide deformylase. Similarities and differences with metzincins and thermolysin. J Mol Biol 289: 1445–1457. 52. Fraser JS, Clarkson MW, Degnan SC, Erion R, Kern D, et al. (2009) Hidden alternative structures of proline isomerase essential for catalysis. Nature 462: 669–673. 53. Lang PT, Ng HL, Fraser JS, Corn JE, Echols N, et al. (2010) Automated electron-density sampling reveals widespread conformational polymorphism in proteins. Protein Sci 19: 1420–1431. 54. Meinnel T, Patiny L, Ragusa S, Blanquet S (1999) Design and synthesis of substrate analogue inhibitors of peptide deformylase. Biochemistry 38: 4287–4295. 55. Schechter I, Berger A (1967) On the size of the active site in proteases. I. Papain. Bochem Biophys Res Commun 27: 157–162. 56. Fersht AR, Shi JP, Knill-Jones J, Lowe DM, Wilkinson AJ, et al. (1985) Hydrogen bonding and biological specificity analysed by protein engineering. Nature 314: 235–238. 57. Takano K, Yamagata Y, Kubota M, Funahashi J, Fujii S, et al. (1999) Contribution of hydrogen bonds to the conformational stability of human lysozyme: calorimetry and X-ray analysis of six Ser —. Ala mutants. Biochemistry 38: 6623–6629. 58. Yuan Z, Trias J, White RJ (2001) Deformylase as a novel antibacterial target. Drug Discov Today 6: 954–961. 59. Wang Q, Zhang D, Wang J, Cai Z, Xu W (2006) Docking studies of Nickel- Peptide deformylase (PDF) inhibitors: exploring the new binding pockets. Biophys Chem 122: 43–49. 60. Li Y, Chen Z, Gong W (2002) Enzymatic properties of a new peptide deformylase from pathogenic bacterium Leptospira interrogans. Biochem Biophys Res Commun 295: 884–889. 61. Bracchi-Ricard V, Nguyen KT, Zhou Y, Rajagopalan PT, Chakrabarti D, et al. (2001) Characterization of an eukaryotic peptide deformylase from Plasmodium falciparum. Arch Biochem Biophys 396: 162–170. 62. Dirk LM, Williams MA, Houtz RL (2001) Eukaryotic peptide deformylases. Nuclear-encoded and chloroplast-targeted enzymes in Arabidopsis. Plant Physiol 127: 97–107. 63. Dirk LM, Schmidt JJ, Cai Y, Barnes JC, Hanger KM, et al. (2008) Insights into the substrate specificity of plant peptide deformylase, an essential enzyme with potential for the development of novel biotechnology applications in agriculture. Biochem J 413: 417–427. 64. Lee JE, Smith GD, Horvatin C, Huang DJ, Cornell KA, et al. (2005) Structural snapshots of MTA/AdoHcy nucleosidase along the reaction coordinate provide insights into enzyme and nucleoside flexibility during catalysis. J Mol Biol 352: 559–574. 65. Wang Y, Liu L, Wei Z, Cheng Z, Lin Y, et al. (2006) Seeing the process of histidine phosphorylation in human bisphosphoglycerate mutase. J Biol Chem 281: 39642–39648. 66. Parker JB, Bianchet MA, Krosky DJ, Friedman JI, Amzel LM, et al. (2007) Enzymatic capture of an extrahelical thymine in the search for uracil in DNA. Nature 449: 433–437. 67. Towler P, Staker B, Prasad SG, Menon S, Tang J, et al. (2004) ACE2 X-ray structures reveal a large hinge-bending motion important for inhibitor binding and catalysis. J Biol Chem 279: 17996–18007. 68. Teague SJ (2003) Implications of protein flexibility for drug discovery. Nat Rev Drug Discov 2: 527–541. 69. Geremia S, Campagnolo M, Schinzel R, Johnson LN (2002) Enzymatic catalysis in crystals of Escherichia coli maltodextrin phosphorylase. J Mol Biol 322: 413–423. 70. Pargellis C, Tong L, Churchill L, Cirillo PF, Gilmore T, et al. (2002) Inhibition of p38 MAP kinase by utilizing a novel allosteric binding site. Nat Struct Biol 9: 268–272. 71. Lazennec C, Meinnel T (1997) Formate dehydrogenase-coupled spectrophoto- metric assay of peptide deformylase. Anal Biochem 244: 180–182. 72. Henderson PJ (1972) A linear equation that describes the steady-state kinetics of enzymes and subcellular particles interacting with tightly bound inhibitors. Biochem J 127: 321–333. 73. Kabsch W (1993) Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Cryst 26: 795–800. 74. McCoy AJ, Grosse-Kunstleve RW, Storoni LC, Read RJ (2005) Likelihood- enhanced fast translation functions. Acta Crystallogr D Biol Crystallogr 61: 458–464. 75. Brunger AT, Adams PD, Clore GM, DeLano WL, Gros P, et al. (1998) Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr 54(Pt 5): 905–921. 76. Robien MA, Nguyen KT, Kumar A, Hirsh I, Turley S, et al. (2004) An improved crystal form of Plasmodium falciparum peptide deformylase. Protein Sci 13: 1155–1163. 77. Roussel A, Cambillau C (1989) TURBO-FRODO. In: Graphics S, ed. Silicon Graphics geometry partners directory Mountain View, CA. pp 77–78. 78. Murshudov GN, Vagin AA, Dodson EJ (1997) Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystal- logr 53: 240–255. 79. Laskowski RA, Moss DS, Thornton JM (1993) Main-chain bond lengths and bond angles in protein structures. J Mol Biol 231: 1049–1067. The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 16 May 2011 | Volume 9 | Issue 5 | e1001066 80. Gouet P, Courcelle E, Stuart DI, Metoz F (1999) ESPript: analysis of multiple sequence alignments in PostScript. Bioinformatics 15: 305–308. 81. Kabsch W, Sander C (1983) Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features. Biopolymers 22: 2577–2637. 82. Kyte J, Doolittle RF (1982) A simple method for displaying the hydropathic character of a protein. J Mol Biol 157: 105–132. 83. Levitt M (1976) A simplified representation of protein conformations for rapid simulation of protein folding. J Mol Biol 104: 59–107. The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 17 May 2011 | Volume 9 | Issue 5 | e1001066
3M6R
Crystal structure of Arabidopsis thaliana peptide deformylase 1B (AtPDF1B) G41M mutant in complex with actinonin
Trapping Conformational States Along Ligand-Binding Dynamics of Peptide Deformylase: The Impact of Induced Fit on Enzyme Catalysis Sonia Fieulaine1, Adrien Boularot1, Isabelle Artaud2,3, Michel Desmadril4,5, Fre´de´ric Dardel6,7, Thierry Meinnel1*, Carmela Giglione1* 1 CNRS, ISV, UPR2355, Gif-sur-Yvette, France, 2 Universite´ Paris Descartes, UMR8601, Paris, France, 3 CNRS, UMR8601, Paris, France, 4 Universite´ Paris-Sud, IBBMC, UMR8619, Orsay, France, 5 CNRS, IBBMC, UMR8619, Orsay, France, 6 Universite´ Paris Descartes, UMR8015, Paris, France, 7 CNRS, UMR8015, Paris, France Abstract For several decades, molecular recognition has been considered one of the most fundamental processes in biochemistry. For enzymes, substrate binding is often coupled to conformational changes that alter the local environment of the active site to align the reactive groups for efficient catalysis and to reach the transition state. Adaptive substrate recognition is a well-known concept; however, it has been poorly characterized at a structural level because of its dynamic nature. Here, we provide a detailed mechanism for an induced-fit process at atomic resolution. We take advantage of a slow, tight binding inhibitor-enzyme system, actinonin-peptide deformylase. Crystal structures of the initial open state and final closed state were solved, as well as those of several intermediate mimics captured during the process. Ligand-induced reshaping of a hydrophobic pocket drives closure of the active site, which is finally ‘‘zipped up’’ by additional binding interactions. Together with biochemical analyses, these data allow a coherent reconstruction of the sequence of events leading from the encounter complex to the key-lock binding state of the enzyme. A ‘‘movie’’ that reconstructs this entire process can be further extrapolated to catalysis. Citation: Fieulaine S, Boularot A, Artaud I, Desmadril M, Dardel F, et al. (2011) Trapping Conformational States Along Ligand-Binding Dynamics of Peptide Deformylase: The Impact of Induced Fit on Enzyme Catalysis. PLoS Biol 9(5): e1001066. doi:10.1371/journal.pbio.1001066 Academic Editor: Gregory A. Petsko, Brandeis University, United States of America Received January 7, 2011; Accepted April 14, 2011; Published May 24, 2011 Copyright:  2011 Fieulaine et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was supported by the Centre National de la Recherche Scientifique (CNRS, France), grant ANR-06-MIME-010-01 (Agence Nationale de la Recherche, France), and grant #4920 from the Association pour la Recherche sur le Cancer (Villejuif, France). SF was partly supported by a postdoctoral fellowship from the Fondation pour la Recherche Me´dicale (France). AB was supported by CNRS, France. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. Abbreviations: DSC, differential scanning calorimetry; Fo, formyl; PDF, peptide deformylase; r.m.s.d., root mean square deviation. * E-mail: Thierry.Meinnel@isv.cnrs-gif.fr (TM); Carmela.Giglione@isv.cnrs-gif.fr (CG) Introduction Flexibility of proteins around their active site is a central feature of molecular biochemistry [1–5]. Although this has been a central concept in biochemistry for half a century, the detailed mechanisms describing how the active enzyme conformation is achieved have remained largely elusive, as a consequence of their transient nature. Direct structural evidence and/or kinetic analyses have only recently emerged [6–10]. Three classic ‘‘textbook’’ models are used to describe the formation of the ligand-enzyme complex: (i) the Fischer’s ‘‘lock-and key’’ model, (ii) the Koshland’s induced-fit model, and (iii) the selected-shift model or conformational selection mechanism [6–8,11–13]. In the Fischer’s ‘‘lock-and key’’ model, the conformations of free and ligand-bound proteins are essentially the same. In the induced-fit model, ligand binding induces a conformational change in the protein, leading to the precise orientation of the catalytic groups and implying the existence of initial molecular matches that provide sufficient affinity prior to conformational adaptation [14]. In contrast, the selected-fit model assumes an equilibrium between multiple conformational states, in which the ligand is able to select and stabilize a complementary protein conformation. In this case, the conformational change precedes ligand binding, in contrast to the induced-fit model in which binding occurs first. The conformational selection and/or induced-fit processes have been shown to be involved in a number of enzymes [12,13,15,16]. For several of these studies, conformational selection is proposed because the experimental data support that, even in the absence of the ligand, the enzyme samples multiple conformational states, including the ligand-bound (active) state [6]. Although direct structural evidence and/or kinetic analyses have provided clues [6–8,12,13,16], how we can distinguish whether a protein binds its ligand in an induced- or selected-fit mechanism remains critical and often controversial. The enzyme-inhibitor interaction is a form of molecular recognition that is more amenable to investigation than the enzyme-substrate interaction as there is no chemical transforma- tion of the ligand during this process. In this context, slow, tight- binding inhibition is an interesting interaction process, as it closely mimics the substrate recognition process and has been shown to be commonly involved in adaptive conformational changes [12, 17,18]. In slow, tight-binding inhibition, the degree of inhibition at a fixed concentration of compound varies over time, leading to a curvature of the reaction progress curve over time during which PLoS Biology | www.plosbiology.org 1 May 2011 | Volume 9 | Issue 5 | e1001066 the uninhibited reaction progress curve is linear [19]. Indeed, the slow, tight-binding inhibition is a two-step mechanism that depends on the rate and strength of inhibitor interactions with the enzyme. Binding of the inhibitor (I) to the enzyme (E) leads to the rapid formation of a non-covalent enzyme-inhibitor complex (E:I) followed by monomolecular slower step (k5) in which the E:I is transformed into a more stable complex (E:I*) that relaxes and dissociates at a very slow rate, mainly inferred by the k6 value when k6,,k5,,k4, (Figure 1A; see also footnote f in Table 1). Although only a few studies have investigated the mechanisms of slow, tight-binding inhibitors, such molecules are favored for use as therapeutics, as they usually exhibit unique inhibitory properties, including selective potency and long-lasting effects [20–26]. Here, we explore the precise structural inhibitory mechanism of actinonin (Figure 1A; [27]), which is a slow, tight- binding inhibitor of peptide deformylase (PDF), a metal cation- dependent enzyme [28,29]. The function of the active-site metal is to activate the reactive water molecule involved in peptide hydrolysis [30]. PDF is the first enzyme in the N-terminal methionine excision pathway, an essential and ubiquitous process that contributes to the diversity of N-terminal amino acids [31,32]. Actinonin is a natural product with antibiotic activity that inhibits PDF by mimicking the structure of its natural substrates (nascent peptide chains starting with Fo-Met-Aaa, where Fo is a formyl group and Aaa is any amino acid) in their transition state (Figure 1B). The transition state inhibitor actinonin, as well as other structurally related inhibitors, has been shown to systemat- ically exhibit a ‘‘slow-binding’’ inhibition behavior (Figure 1A), regardless of the organism of origin of the PDF [29,33]. Using structural, biocomputing, and enzymatic analyses, we were able to (i) reveal that the free enzyme is in an open conformation and that actinonin induces transition of the enzyme into a closed conformation; (ii) show that there is no evidence for the occurrence of a closed conformation in the apostructure of the open enzyme, which, together with detailed kinetic analyses, makes the closed form fully compatible with an induced-fit model; and (iii) identify the sequence of molecular events leading to the final, bound, closed complex (E:I*). Moreover, using several rationally designed point mutants of the enzyme, ligand-induced intermediates, which mimic conformational states that normally would not be expected to accumulate with the wild-type (WT) enzyme, were trapped. These conformations recapitulate physical states that the WT enzyme must pass through during its overall transition from the apo-enzyme to the E:I* complex. ‘‘Freezing’’ of ligand-induced intermediate states provides direct evidence for an induced-fit mechanism and allows the reconstruction of a virtual ‘‘movie’’ that recapitulates this mechanism. Since PDF is one example of an enzyme remaining active in the crystalline state and because actinonin closely mimics the natural substrates bound to PDF in the transition state as shown previously with the Escherichia coli form (EcPDF; see Figure 1B) [34,35], we propose a model suggesting that induced fit also contributes to efficient catalysis. Results Slow, Tight Binding of the Transition-State Analog Actinonin to Peptide Deformylase In the present study, at the atomic level we explored the precise inhibitory mechanism of actinonin on Arabidopsis thaliana PDF1B (AtPDF), a close eukaryotic homologue of EcPDF (Figure S1) [36,37]. Measurements of the kinetic parameters of the second step of the binding mechanism (k5) revealed a timescale in the 10-s range (Table 1), which is consistent with the collective motion of a large domain [4,5]. This finding is supported by NMR studies [38,39], which showed that actinonin binding induces drastic changes in the heteronuclear single quantum coherence (HSQC) spectrum of EcPDF, since most resonances undergo significant shifts that affect a large part of the structure [40,41]. The existence of alternative conformational states of EcPDF is further supported by recent biophysical studies [42]. Previously reported snapshots of a series of different conformations of the enlarged and mobile loop—the so- called CD loop—of the dimeric PDF from Leptospira interrogans PDF (LiPDF) in the presence or absence of inhibitor led to the hypothesis of the existence of an equilibrium between a closed and open form of the CD-loop of PDF enzymes, suggesting a selected-shift model to the authors [43]. Taken together, these data suggest that the binding of actinonin to PDF is accompanied or preceded by conformational changes within the enzyme. Paradoxically, this proposal has not been currently supported by the available structural data. Indeed, free and complexed crystal structures have provided no evidence for any significant conformational change in PDF structure induced by the binding of ligand [35,43–47]. Tight inhibition in the closed state is associated with the KI* apparent equilibrium constant (Figure 1A). A KI* value (see Table 1 and Materials and Methods for the biochemical definition of KI*) of 0.9 nM for actinonin could be measured for AtPDF; that is, a value very similar to that obtained for bacterial PDFs, including EcPDF and Bacillus stearothermophilus PDF2 (BsPDF2, Table 1). Tightening of the initial encounter complex (E:I) resulted in a final complex (E:I*) in which the potency of actinonin (KI/KI*) was enhanced by more than two orders of magnitude and exhibited a very slow off-rate (k6, Table 1). The dissociation constant value of AtPDF for actinonin was also assessed using isothermal titration calorimetry (ITC) experiments (Table S1 and Figure S2A). The corresponding ITC titration curves (Figure S2A) are consistent with a very strong affinity of the ligand for the enzyme [48], enabling us to determine an accurate Kd. Moreover, these studies generated values similar to those measured by other means for AtPDF and EcPDF [42,49]. Author Summary The notion of induced fit when a protein binds its ligand— like a glove adapting to the shape of a hand—is a central concept of structural biochemistry introduced over 50 years ago. A detailed molecular demonstration of this phenomenon has eluded biochemists, however, largely due to the difficulty of capturing the steps of this very transient process: the ‘‘conformational change.’’ In this study, we were able to see this process by using X-ray diffraction to determine more than 10 distinct structures adopted by a single enzyme when it binds a ligand. To do this, we took advantage of the ‘‘slow, tight-binding’’ of a potent inhibitor to its specific target enzyme to trap intermediates in the binding process, which allowed us to monitor the action of an enzyme in real-time at atomic resolution. We showed the kinetics of the conformational change from an initial open state, including the encounter complex, to the final closed state of the enzyme. From these data and other biochemical and biophysical analyses, we make a coherent causal reconstruction of the sequence of events leading to inhibition of the enzyme’s activity. We also generated a movie that reconstructs the sequence of events during the encounter. Our data provide new insights into how enzymes achieve a catalytically competent conformation in which the reactive groups are brought into close proximity, resulting in catalysis. The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 2 May 2011 | Volume 9 | Issue 5 | e1001066 Ligand-Induced Conformational Closure of AtPDF in the Crystalline State Occurrence of a conformational change induced by drug binding was visualized via the resolution of several crystal structure forms of AtPDF, the free form and/or in a complex with actinonin (Table S2). The data reveal a structural switch between the two forms that can account for both the thermodynamic and kinetic data. The enzyme was observed in two states, a novel open apo-form and a closed, induced, actinonin-bound complex (Figure 1C). Binding of actinonin resulted in a tightening of the active site through the collective closure of the entire N-terminal portion of the protein (strands b1, b2, and b3; helix a1; and CD-loop, see Movies S1 and S2, Figure 1C, and Figure S1). The amplitude of the structural change was maximal for Pro60 (Figure S1), the Ca of which was shifted 4 A˚ upon actinonin binding. This collective movement involved the formation of a ‘‘super b-sheet’’ as the result of the large rearrangement of b-strands 4 and 5 relative to the rest of the structure in which actinonin forms an additional strand bridging the two b-sheets (b1 andb2) on either side of the active site (Figure 1D and Figure S1B). As actinonin is a peptide-like compound (see Introduction and Figure 1B), this behavior closely mimics what occurs in the natural protein substrates of PDF, which also form this strand-bridging interaction. This phenomenon also accounts for the strong stabilization of the protein by actinonin, which was also challenged by differential scanning calorimetry (DSC) experiments: the Tm of AtPDF increased from 61uC to 81uC upon binding of the inhibitor (Figure 1D, see also below). Thus far, this closure of the enzyme induced by actinonin is part of the rare structural evidence for the slow, tight-binding mechanism at an atomic scale. The open state, which has never been observed, was captured not only in the two molecules of the asymmetric subunit but also in different crystals and under two distinct crystallization conditions (Table S2 and Figure 2). All r.m.s.d. values were smaller than 0.25 A˚ . The closure is very unlikely to result from crystal packing constraints, as soaking the apo-AtPDF crystals in a solution containing actinonin induced the Figure 1. Slow, tight-binding inhibition of PDF by actinonin induces conformational change in the protein. (A) Inhibition by a two-step mechanism, involving a tightening of the initial enzyme-inhibitor complex (E?I) to form a more stable complex (E?I*), with the chemical structure of actinonin (I), the natural inhibitor of PDF enzymes (E). (B) Structures of EcPDF bound to actinonin (left) and to the transition state resulting from the cleavage of its substrate, Fo-Met-Ala-Ser (right) [34,35]. (C) Superimposition of free and actinonin-bound AtPDF indicated in green and purple, respectively. The three conserved motifs of the PDF enzymes family are indicated in orange and numbered I, II, and III. Molecules A of both models were superimposed, resulting in an r.m.s.d. of 0.9 A˚ for 100% of the Ca. Left inset, close-up comparison of the open and closed forms figured in the ribbon representation. (D) Baseline-corrected DSC thermograms of free and actinonin-bound WT AtPDF recorded under the same experimental conditions. doi:10.1371/journal.pbio.1001066.g001 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 3 May 2011 | Volume 9 | Issue 5 | e1001066 structural transition from the open to the closed state within the crystals without cracking them or altering their diffracting power. Thus, crystal packing is compatible with both states of the enzyme (Figure S3). Therefore, the open structure most likely corresponds to a stable state in solution. The closed final conformation was identical to that previously reported for PDF complexes obtained either with actinonin or with a product of the reaction [34,35,44,50], indicating that this structure is common for the ligands (compare Figures 1B and 2A, and Figure S4). Hydrogen bonding was also conserved, especially the bond between the backbone nitrogen of Ile42 (corresponding to Ile44 in EcPDF, see Figure 1B and Figure S5A) and the alkyl carbonyl chain of actinonin, which potently contributes to the formation of the super b-sheet (Movie S2 and Figure S1B, see also below). Between the open and closed states, the side chains of Ile42, Phe58, and Ile130 underwent significant structural changes (Figure 3A and D and Figure S6), corresponding to a hydophobic pocket rearrangement, with Ile42 being the most affected (Figure 3). Interestingly, Ile42 is the second residue of the conserved active-site motif G41IGLAAXG (motif 1) that was previously shown to be essential for activity [51]. To assess and visualize the differences between the two states, two independent structural parameters were measured: the r.m.s.d. value with respect to the open form and the aperture angle (dap), which measures the angle made between the N- and C-domains through three fixed-points, corresponding to the Ca of three conserved residues, each sitting in one of the three conserved motifs (Figure 2A). The bi-dimensional graph of these two parameters is a good representation of the closing motion snapshots (Figure 2B) shown in Movie S1. With this tool at this stage, two states could be defined: the closed (C) and open (O) states (Figure 2B). Evidence for a Pure Induced-Fit Mechanism in the Binding of Actinonin to AtPDF Recent quantitative analyses of both conformational selection and induced fit have led to an integrated continuum—a so-called ‘‘flux-description’’—of these two limiting mechanisms [16]. According to this model, conformation selection tends to be preferred at low ligand concentrations (mM range)—that is, using detailed kinetic studies—whereas induced fit dominates at high ligand and enzyme concentrations (mM range) obtained, for instance, in NMR or crystallographic approaches. Structural studies are most useful to reveal subpopulations of biological significance. We investigated the existence of lowly populated, alternative conformations of apoPDF. To probe the occurrence of alternate conformers in the crystalline state of PDF, the new Ringer program is the most suitable investigation tool [52,53]. Ringer searches for evidence of alternate rotamers by systematically sampling electron density maps—free of model bias—around the dihedral angles of protein side chains. Two independent WT open datasets of the apoenzyme, including a high-resolution set (1.3 A˚ ), were used in the analysis. Ringer analysis revealed the existence of only one rotamer of most side chains of either molecule in the asymmetric unit, including the three main residues primarily involved in conformation change—that is, Ile 42, Phe58, and Ile130 (Figure 4A). Ringer analysis showed evidence for unmodeled alternate conformers for very few residues, including Ile121 and Phe87, or Phe119 to a much lesser extent (Figure S7). There is therefore no evidence for the occurrence of a closed conformation in the apostructure of AtPDF, supporting the hypothesis that the conformational change was essentially induced by the binding of actinonin rather than from conformational selection among multiple states occurring in the crystalline state. To further investigate the mechanism involved, we followed a kinetic approach aimed at discriminating between induced fit and population shift at low ligand concentrations (sub-mM range) [12]. The experimentally observed pseudo-first-order rate constant for the approach to equilibrium between the free components and the binary AtPDF-actinonin complex (kobs) was measured and plotted as a function of actinonin concentration. This plot yielded a hyperbolic saturation curve with a positive slope, as fully expected for a pure induced-fit mechanism (Figure 4B and C). In contrast, if the enzyme sampled two or more conformational states, the curve would imply that the value of kobs decreases with increasing ligand concentration (see, for instance, curve C in Figure 1 in [12]). The same conclusion can be reached for EcPDF and BsPDF2 (Figure 4B and C) and was already reported by others for S. aureus PDF [29]. Together, these data indicate that a pure induced-fit mechanism triggered by the binding of actinonin appears to direct the conformational change both in solution and in the crystalline state. Single Variants at Gly41 Exhibit Strongly Reduced Actinonin-Binding Potency and Catalytic Efficiency When dealing with an induced-fit mechanism, knowledge of the initial O and final C state is crucial but does not provide direct information on the position of actinonin in the encounter complex or on the sequential mechanism of the transition process. We suspected that the conserved glycine-rich motif 1 (G41IGLAAXQ) could contribute to the flexibility required for the observed structural transition. Evidence for such flexibility comes from NMR analysis of EcPDF in which a few residues show exchange cross-peaks of an additional, alternative form [38]. The most strongly affected residues are Cys90, one of the metal ligands, its neighbor Leu91, and both of the alanines within the above conserved glycine-rich motif (Figure S1B), suggesting that EcPDF undergoes conformational dynamics in a similar region. To unravel the dynamics of the recognition process, we surmised that it should be possible to freeze the conformational Table 1. Comparison of the main kinetic and thermodynamic parameters describing the inhibition of PDF by actinonin. Parameter AtPDFa EcPDFa BsPDF2a,b KI (nM)d 140610 112610 185615 KI* (nM)c 0.960.5 1.360.2 2.960.8 KI/KI* 155615 86610 6467 k5 (s21) 6103d 6369 170620 7268 k6 (s21) 6104d 461 1962 1163 k4 (s21)e 140610 112610 185615 t1/2 (min)f 2965 661 1.160.2 aThe enzyme concentrations used in the assay were 100, 50, and 25 nM for AtPDF, EcPDF, and BsPDF2, respectively. bData from [49]. cPrior to kinetic analysis for determination of the KI* value, actinonin was incubated at the final concentration in the presence of the studied enzyme set for 10 min at 37uC. The kinetic assay was initiated by the addition of a small volume of the substrate. dFor determination of KI, k5, and k6 values, actinonin was not preincubated with the enzyme. The kinetic assay was initiated by the addition of the enzyme. ek4 corresponds to the kinetic constant of the dissociation of the primary enzyme-actinonin complex. It is assumed that the rate of complex association is diffusion-limited (see Table 7.3 in [19]), that is, k3—the kinetic constant of the association of the primary enzyme-actinonin complex—is 109 M21.s21. ft1/2 is 0.693(k4+k5+k6)/k4k6 (see case of induced fit and calculation in Table 1 of [12]). In this case, t1/2,0.693/k6 because k6,,k5,,k4. doi:10.1371/journal.pbio.1001066.t001 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 4 May 2011 | Volume 9 | Issue 5 | e1001066 Figure 2. Four distinct conformational states of PDF enzymes. (A) AtPDF and several other representative bacterial PDFs are superimposed. A zoom is displayed on the right of the panel. Superimpositions were realized using ‘‘module superpose’’ in the CCP4i package and the ‘‘secondary structure matching’’ tool. The extent of aperture/closure of PDF enzymes was assessed primarily by the measurement of the aperture angle (dap), the angle made between the Ca of three strictly conserved residues (C, H, and I) of all PDFs, each characterizing a secondary crucial structure module of the active site crevice, namely b4, a2, and b1 (see Figure S1C). Each single residue belongs to one of the three conserved motifs (motifs 2, 3, and 1, respectively) and corresponds respectively to Cys91, His137, and Ile42 in AtPDF. The dap was measured in each case (see B). (B) The dap values combined with those of the r.m.s.d. associated with the superimposition of the open structure of AtPDF allows the identification of four conformational states: open (O), intermediate (I), closed (C), and super-closed (S). We compared AtPDF1B (this work and PDB CODE 3CPM; brown, orange, and yellow in A and B; black in C), EcPDF (1BS7, free enzyme; 1BS6, with Met-Ala-Ser; 1G2A, with actinonin; magenta), BsPDF2 (1LQY, with actinonin; green), LiPDF (1SV2, free; 1SZZ, with actinonin; red), and PfPDF (1JYM, free; blue). doi:10.1371/journal.pbio.1001066.g002 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 5 May 2011 | Volume 9 | Issue 5 | e1001066 Figure 3. Effect of actinonin binding on the conformation of key residues in PDF. Conformation of key residues Ile42, Phe58, and Ile130 in the different complexes: (A) in unbound WT AtPDF, (B and C) in the structure of G41Q and G41M actinonin-bound variants, respectively, and (D) of actinonin-bound WT protein. In the final complex (D), a hydrogen bond is formed between actinonin and the peptidic bond, which links Gly41 and Ile42. During the deformylation reaction, which is catalyzed by the PDF enzyme, the N-terminal formyl-methionine fits into the S1’ pocket. The solvent-accessible surface of this pocket is represented here, and only the aliphatic chain of actinonin is shown, mimicking the N-terminal methionine. (E) Free WT enzyme with the S1’ pocket shown open in two orientations (top and bottom). (F and G) S1’ pocket in the G41Q and G41M variant structures, respectively, shown in two orientations (top and bottom). (H) After the complete conformational modifications of actinonin-bound WT protein induced by actinonin binding, the S1’ pocket is shown closed in two orientations (top and bottom). (I) The four models are superimposed; the ligand-binding site is magnified: unbound WT AtPDF; G41Q and G41M actinonin-bound enzyme; and WT actinonin-bound enzyme are indicated in brown, red, orange, and yellow, respectively. Actinonin is indicated by lines. (J) A detailed view of the AtPDF ligand-binding site for all the complexes, which are superimposed, as indicated in the same colors. Arrows indicate the direction of the closing movement within the enzyme, from the open, unbound state to the closed, bound state. doi:10.1371/journal.pbio.1001066.g003 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 6 May 2011 | Volume 9 | Issue 5 | e1001066 change along the pathway by introducing selected, minor variations within the above-mentioned crucial residues involved in the collective motion. In this respect, site-directed mutagenesis of AtPDF was performed on Gly41, Ile42, and Ile130. Single substitutions were made at Gly41 (G41A/Q/M), Ile42 (I42A/F/ N/W), and Ile130 (I130A/F), and the variants were purified and characterized. These mutant proteins showed no change in overall stability, as evidenced by DSC experiments (unpublished data). However, two variants of G41, G41Q and G41M, showed dramatic effects; the kcat/Km values were reduced by three orders of magnitude due to large decreases in the kcat values compared to the WT enzyme (Figure 5A and Table S1). The reduced kcat/Km values suggest an altered ability of these variants to attain the final enzyme-transition state complex and, as a result, to give rise to possible states different from the final E:I* complex. Substitutions at positions 42 and 130 only caused small reductions in the kcat values (Figure 5A, Figure S2C, and Table S1). The actinonin- binding potency of both G41 variants was also greatly reduced (Table S1 and Figure S2B). The time-dependent inhibition by actinonin of the most active variants was then studied (Table S3). Figure 4. Evidence for an induced fit in crystalline and solution states of AtPDF. (A) Absence of evidence for alternative conformers in the crystalline state of AtPDF. Ringer plots of electron density (r) versus x1 angle for representative residues of the 3D apo-structure of AtPDF. Data were obtained with the 3M6O, 3PN2, and 3PN3 datasets (2.0 and 1.3 A˚ resolution, respectively, see Table S1). The secondary peaks in the Ile residues are observed because Ile is a branched amino acid. To reveal an alternative conformation with Ile, three peaks should be observed. (B) kobs is a saturable function of actinonin with various PDFs, including AtPDF. Data obtained for kobs, the experimentally observed pseudo-first-order rate constant for the approach to equilibrium between the free components and the binary PDF-actinonin complex, were obtained at various concentrations of actinonin in the presence of EcPDF, AtPDF, and BsPDF2. A direct plot is shown. Inset, time-course measurement of deformylation as a function of varying actinonin concentrations. (C) Inverted plot of the data in panel B, which is expected to be a straight line if the kobs is ..k6 in the case of induced fit [19]. The correlation coefficient of each line is 1.00, 0.99, and 1.00 for AtPDF, BsPDF2, and EcPDF, respectively, indicative of the accuracy of the conclusion. doi:10.1371/journal.pbio.1001066.g004 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 7 May 2011 | Volume 9 | Issue 5 | e1001066 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 8 May 2011 | Volume 9 | Issue 5 | e1001066 The half-lives of the final complexes—as assessed by comparison of the 1/k6 values—were always significantly smaller (Table S3), suggesting that the conformational change induced by actinonin binding still occurred, but the C state is destabilized relative to the O state in the mutants compared to the WT. Accordingly, actinonin strongly stabilized almost all of the variants; Tm was increased by more than 20uC. This differs from the G41M and G41Q variants, which both showed increases in the Tm of only 12uC, consistent with reduced binding potency (Table S1). Conformational Changes of Gly41 Variants Are Affected On-Pathway The two most interesting variants, G41Q and G41M, could be crystallized under the same conditions as the WT protein. In the case of G41Q, the structure of the apo-protein did not show any modifications compared to the WT structure and remained in an O conformation (Figure 2B; ‘‘O’’ zone). In contrast, the 3D structure of the G41M variant showed that the asymmetric unit was composed of two molecules with distinct structures. One molecule (chain A) is in the O state and is similar to the structures of the WT and the G41Q variant (Figure 2B; zone ‘‘O’’). The second molecule (chain B) is in a C state, closer to that observed for the WT chain in the presence of actinonin (‘‘C’’), a so-called ‘‘superclosed’’ state (Figure 2B; zone ‘‘S’’), suggesting that the substitution modified the equilibrium between the two states in solution either (i) at the step of protein synthesis by providing two conformers, the inter-conversions of which are blocked due to steric hindrance brought by the new bulkier side-chain at position 41, or (ii) by dramatically unbalancing the free inter-conversion between the O and S conformers towards the S state. Ringer analysis indicates that in the free G41M variant, many residues show evidence for unmodeled alternate conformers—including positions 58, 42, and 130—in keeping with the second hypothesis. For all variants of position G41, addition of actinonin to the crystal (Figure 3 and Figure S6) induced a closure of the protein within the crystal. Nevertheless, as expected from in silico graphic modeling followed by energy minimization, the occurrence of a bulky side chain at position 41 prevented the completion of the closure in the presence of the ligand and, hence, the formation of the hydrogen bond between the backbone nitrogen of Ile42 and actinonin. This finding is consistent with the strongly reduced Tm of the complex of the variants with actinonin compared to WT as measured by DSC. Remarkably, both S and O forms of the G41M apo-structures in the asymmetric unit of the crystal yielded a unique intermediary structure (‘‘I’’ state) upon actinonin binding (r.m.s.d. between the molecules is ,0.25 A˚ ; see also Figure 2B, zone ‘‘I’’). In this case, it is likely that the induced-fit mechanism drives the equilibrium by capturing only the O population and closing it to an intermediary step, thus depleting the pool of O conformers that is shifted sequentially back from the remaining pool of S conformers and allows the complete binding of actinonin to the enzyme. In line with the rational design of the PDF mutants, the extent of the structural differences suggests that the underlying motions are dependent on the length of the side chain (Figure S8). Together, these data account for the reduced catalytic rate, as the hydrogen bond is strictly required for the substrate to be efficiently cleaved by PDFs (Figure S8A) [54]. Therefore, from both structural and kinetic analyses, each substitution most likely reproduces intermediates along the pathway that lead to the closure of PDF around its substrate (Figure S2B). Conformational Changes of Gly41 Variants Recapitulate Closing Intermediates Analysis of the structures allows us to propose the following sequence of atomic events (Figures 3 and 2B and Figure S6). To name the various sites of the ligand and subsites of PDF, we will use the usual nomenclature found in [55], which defines the various binding pockets of a protease, where P1’ is the first side chain at the C-terminal side of the cleavage site and its binding pocket is S1’, also referred to as the hydrophobic pocket in the case of PDF. First, actinonin aligns along the S1’ pocket to form the encounter complex, which shifts the Ile130 side chain to avoid steric hindrance in the S1’ pocket, promotes rotation of the Ile42 side chain, and finally rearranges the phenyl group of Phe58. These events achieve an optimal hydrophobic S1’ pocket conformation (Figure 3), and the concomitant closure leads to the formation of a hydrogen bond between the first carbonyl group of actinonin and the backbone nitrogen of Ile42. The initial N-O distance is reduced from 5 A˚ to 2.8 A˚ , which is an optimal value for hydrogen bonding (Movie S2 and Figure S8B). Thus, the primary driving force for the active site closure appears to be the P1’:S1’ hydrophobic interaction. The C state is ultimately locked by the super-b-sheet hydrogen bonds extending across the ligand, including those involving Ile42. The DDGbinding value (2.2– 2.4 kcal/mol, Figure S8B), as calculated from the Kd values for actinonin binding to wild-type (WT) and G41M and G41Q, is consistent with the loss of a hydrogen bond that also contributes to the conformational stability of the protein [56,57]. Thus, this bond contributes to the major binding free energy difference between the two complexes (3.1 kcal/mol; Figure S8B, Tables S1 and S3, and [29]). Interestingly, the above DDGbinding values also correlate with the DDGES values derived from the kcat/Km and kcat measurements [19]. This dataset strongly correlates with the Figure 5. Inhibition and enzymatic reactions progress through an induced fit pathway. (A) The catalytic parameters Km and kcat, for all AtPDF variants are provided as a percentage of the wild-type values (WT). Detailed values are presented in Table S2. (B) Schematic model for actinonin binding to AtPDF in favor of an induced-fit pathway. PDF might exist in at least two conformational states, open (O) or closed (C). The relative abundance of each conformation would vary, depending on the enzyme type. With AtPDF, it is likely that the most abundant form is the O one, which is the only form leading to a productive complex. The superclosed form (S) is likely to show reduced affinity for the ligand because of steric occlusion of the active site. At the initial stage, the inhibitor (shown in red) binds to AtPDF (indicated in brown) in the O conformation. To reach the final key-lock state (productive closed conformation, C), two major and extreme pathways can be used. According to the conformational selection pathway, the inhibitor selects the C conformation. This pathway, which is represented by the dashed arrow, does not occur within the crystal. In contrast, the G41Q and G41M mutants, by providing the structure of the enzyme in intermediate conformations (I), prove the existence of the so- called encounter complex and confirm that the inhibitor binds to the enzyme when it is in the O conformation. The ligand-binding site is then reorganized to yield the C enzyme conformation, that is, the key-lock state. Indeed, the inhibitor binds to the enzyme through the induced-fit pathway. Each timescale was calculated using the data available in the text and corresponds to t1/2 values deduced from the calculation of 0.693/ (kinetic constant of interest). The kcat value (k2) was used to assess the timescale of catalysis in panel C, whereas, in (B), k4 assesses the first step of inhibition, and k6 is used in the case of the slow step. For the SO conversion (left, B), the lifetime of the minor form of EcPDF was used to assess the order of magnitude (see text and [38]). (C) Schematic model for the deformylation reaction catalyzed by PDF. Since actinonin is a pseudo-peptidic inhibitor, it is likely that a peptidic substrate can bind to the PDF enzyme through an induced-fit pathway, as described in (B). The key-lock state represents a transition state in which the N-formylated substrate is deformylated to yield the final reaction product. doi:10.1371/journal.pbio.1001066.g005 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 9 May 2011 | Volume 9 | Issue 5 | e1001066 gyration and van der Waals radii of the side chain at position 41 as well as the N-O distance between the first carbonyl group of actinonin and the backbone nitrogen of Ile42 (Figure S8). These results suggest that the capacity of both G41M and G41Q variants to form the transition state is a consequence of their inability to reach the fully closed state. Thus, our study of the designed Gly41 mutant enzymes reveals that, in addition to the initial and final states observed for the WT enzyme, the conformations of the Gly41 variants correspond indeed to on-pathway intermediates, thus providing snapshots along the trajectory from the O to the C state of the enzyme (Figures 2B and 3). The 3D structure of the variants in the absence of ligand is similar to that of WT, and a strict correlation exists between the completeness of the conformational change and both binding potency and catalytic efficiency. This suggests that both events require complete protein closure to generate a productive complex. The strong stabilization of AtPDF by actinonin (Figure 1D) closely mimics what occurs with its natural substrates when it reaches the transition state [34,58]. Indeed, as expected, the enzyme facilitates the final C conformation by lowering its final energy [6]. Optimal arrangement of the S1’ pocket (Figure 3) proceeds along the reaction process towards the final C conformation, triggering the alignment of reactive groups in an optimal arrangement for ligand recognition. Upon binding, actinonin alters the thermodynamic landscape for the structural transition between the O and C states. This ligand is a potent inhibitor because it can trigger the above sequence of events similar to the substrate, but unlike the substrate, it is non- hydrolyzable. Thus, by mimicking the transition state and being non-hydrolyzable (Figure 1B), the final C complex is long lasting. Ligand-Induced Conformational Closure Is Initially Triggered by the Binding of the P1’ Group in the S1’ Pocket Given the similarity between actinonin and natural substrate binding, the very slow kinetics of inhibitor binding (10-s time-scale) remains puzzling compared to the 10 ms required for catalysis (deduced from the kcat). This finding could be explained as a conformational effect during the formation of the hydrogen bond, aligning the substrate as an additional beta-sheet and eventually stabilizing the entire enzyme-ligand complex. The significantly longer time needed to reach the most stable state compared to the substrate would most likely be due to the presence of the flexible and one carbon longer metal-binding group in actinonin (i.e., hydroxamate versus formyl, Figure 1B). This suggestion is in line with the overall data obtained when we investigated more deeply the role of the first carbonyl group of the ligand. This group is well known to exert a crucial effect in both productive and unproductive ligand binding (i.e., substrate and inhibitor) [54]. In this respect, we studied the binding of compound 6b (Figure S5B), a PDF ligand that does not exhibit a reactive group at this position [49]. We observed that this compound binds strongly to both EcPDF (KI* = 6366 nM) and AtPDF (KI* = 400635 nM) but, unlike actinonin, does not display slow, tight binding as KI* = KI. This impact on binding is consistent with the absence of the hydrogen bond involving the first carbonyl group of the ligand. The 3D structure of AtPDF was determined after soaking the compound in crystals of the free, open AtPDF form. Upon binding, 6b induced a complete conformational change, identical to that observed with actinonin (Figures 2B and 6A; ‘‘O’’ state). This result further suggests that the conformational change is not induced initially by the formation of this hydrogen bond and that the encounter complex is primarily driven by the fit within the S1’ pocket. This also reveals that the timescale of the large conformational change is several orders of magnitude faster than the kinetics of slow binding and fully compatible with both the first step of actinonin binding (k4 = 140 s21; see Table 1) and the catalytic rate of the substrate (kcat = 37 s21; see Table 1 and Table S3). The 3D structure also revealed that both the P1’ and the hydroxamate groups are bound similarly to the corresponding groups of actinonin (Figure 6B). As expected, no additional bonding occurs, especially around the backbone nitrogen of Ile42 (Figure 6C). Taken together, these data allow us to conclude that the conformational change observed upon ligand binding is triggered primarily by binding in the S1’ pocket. As revealed by the binding of 6b, the one carbon longer metal-binding group fits, immediately upon recognition of the P1’ group, in the S1’ pocket and forms a bidentate complex with the metal cation, mimicking the transition state as a result. Thus, the active site is very confined and rigid due to the presence and length of the hydroxamate group (compare right and left panels in Figure 1B). As a result, compared to the complex made with the substrate, it is likely that the formation of the hydrogen bond involving the carbonyl of actinonin and the backbone nitrogen of Ile42 becomes strongly rate-limiting (k5 = 0.044 s21; Table 1). Once this hydrogen link is locked, the uncleavable bond, mimicking the labile formyl group at the transition state, stabilizes the enzyme-inhibitor complex, making it long-lasting (k6 = 0.0006 s21; Table 1) and providing a mechanistic explanation for the slow-binding effect that involves both large and fine conformational changes. The large conformational change is similar to the one occurring with the substrate, whereas the second is more subtle and locks the hydrogen bond involving the backbone nitrogen of Ile42. The second step is rate-limiting with some transition state analogs such as actinonin (Figure 5B and C). Proper Positioning of the Carbonyl Group Is Required to Stabilize the Complex at S1’ Compound 21 corresponds to another interesting derivative designed to probe the impact of the peptide bond in PDF binding [49]. In addition to the hydroxamate group, this compound features both a hydrophobic benzyl group at P1’ and a reverse peptide bond. Compound 21 shows modest but significant inhibitory potency to AtPDF1B (KI* = 400637 nM), confirming the crucial role of the peptide bond in PDF binding. After soaking with crystals of apo-AtPDF, compound 21 could be detected in high-resolution electron density maps (Figure S9A). Unlike 6b, 21 did not bind the active site of the enzyme but an alternative pocket at the surface of the protein (Figure S9B). A docking study performed with EcPDF had previously revealed this alternative binding pocket (Figure S9C; [59]). The aforementioned data indicate that the occurrence of a S1’- binding group placed in the unfavorable context of a reverse peptide bond does not stably promote binding at the active site of AtPDF. Upon binding of 21, the 3D structure of both molecules of the asymmetric unit remain in an O conformation (r.m.s.d. ,0.2 A˚ with respect to the apo-structures in the ‘‘O’’ state). This finding suggests that only the binding of compounds entering the S1’ pocket, such as actinonin or 6b, induces conformational change, in keeping with the crucial role of the P1’ group if located in the frame of a classic peptide bond. Moreover, we noticed that the binding pocket of 21 was located on the rear side of the true S1’ pocket and induced a weak modification of the P1’ hosting platform (Figure S9D). Indeed, when crystals of the 21:AtPDF complex were soaked in actinonin, the final 3D structure no longer showed evidence of compound 21 occupancy greater than 5%. Instead, this structure revealed both actinonin and closing of the protein (Table S2). The r.m.s.d. between this structure and that The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 10 May 2011 | Volume 9 | Issue 5 | e1001066 obtained directly with actinonin was less than 0.2 A˚ ; the actinonin position was virtually identical, indicating that the protein had retained full capacity for binding actinonin and closing despite the presence of compound 21. We conclude that actinonin does compete with 21 because of the overlap at P1’ of AtPDF1B (Figure S9C). As the actinonin S1’ subsite strongly mimics that of a true substrate, this result also explains the inhibitory behavior of 21 towards AtPDF. Discussion Although PDF catalysis has been extensively studied and the mechanism has been elucidated [34], how the enzyme achieves the catalytically competent state remains unknown. Here, we provide insight on how the enzyme might reach a catalytically competent conformation, demonstrating that the reactive groups move into proximity to promote catalysis (Figures 2B and 5C). We suggest Figure 6. Effect of 6b binding on the conformation of key residues of PDF. Superimposition of free, 6b-, and actinonin-bound AtPDF indicated in brown, red, and yellow, respectively. (A) Molecule A in the three models was superimposed, resulting in an r.m.s.d. of 0.9 A˚ for 100% of the Ca. Actinonin is shown in yellow and 6b in red. (B) Conformation of key residues Ile42, Phe58, and Ile130 in the different complexes and in unbound WT AtPDF. Actinonin is shown in yellow and 6b in red. (C) A detailed view of the AtPDF ligand-binding site for both actinonin and 6b complexes, which are indicated by sticks and are superimposed. The two ligands are colored in pale and dark grey, respectively. The hydrogen bond made by actinonin only is shown. doi:10.1371/journal.pbio.1001066.g006 The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 11 May 2011 | Volume 9 | Issue 5 | e1001066 that the motions of the catalytic centre starting with free ligand- PDF favor a final configuration that is optimal for binding and/or catalysis (Figures 1B, 2B, and 5B and C). In our model, we propose that free PDF might exist in at least two conformational states, that is, open (O) or super-closed (S). The relative abundance of each conformation varies by enzyme type and incubation conditions, explaining why both conformations have not been trapped thus far. In the case of AtPDF, it is likely that the most abundant form corresponds to an O state, which is the form that leads to a productive complex. Indeed, in the NMR spectra for EcPDF, a few residues show exchange cross-peaks from an additional, alternative form [38]. The most strongly affected residues are Cys90, one of the metal ligands, its neighbor Leu91, as well as Ala47 and Ala48 on the facing strand. This suggests that EcPDF exists in at least two conformations (‘‘S’’ and ‘‘C’’; see Figure 2B), which undergo slow interconversion on the NMR timescale. The 3D structure of the major conformation (75%, lifetime 300 ms) could be solved at high resolution, but the structure of the minor form (25%, lifetime 100 ms), which exhibits very weak signals, could not be solved [38]. This conformation appears to correspond to that of the complex obtained with the product of the reaction (Met-Ala-Ser). A very similar situation—although more balanced between the two states—appears to occur in the case of variant G41M, suggesting that a mechanism involving conformational selection followed by induced fit is a general model for PDF and that AtPDF is a specific case where population shift virtually does not occur as the free enzyme is completely in the O conformation. This is also in line with data obtained with L. interrogans PDF (LiPDF), which reveal conformers in both the S and C states (see Figure 2B) and suggest a population-shift mechanism [43]. It is interesting to note that LiPDF is a poorly active PDF [60]. According to the representation shown in Figure 2B, Plasmodium falciparum PDF (PfPDF), a poorly active PDF [61], was retrieved only in the S state. Finally, weak decompaction of the structure of Bacillus cereus and Staphylococcus aureus PDFs in the presence of actinonin have been described [45,46]. These examples suggest that the enzyme is trapped in the S conformer in the free state and converts to the C conformer when bound to actinonin, suggesting that the S conformer is overrepresented in solution compared to the O state, unlike AtPDF. This study of AtPDF—including 10 different crystal structures of apo- and complexed enzyme variants—reveals the 3D structure of a PDF in at least four distinct states. This includes the O form, the occurrence of which is crucial for catalysis, as it is the active form. Here, we propose that the transition from the O to the C state is directly induced by the ligand. Indeed, the O form, which is captured in the crystal, undergoes closure directly upon ligand binding in our soaking experiments. Progression to this closure involves intermediary states (‘‘I’’) similar to those observed with variants G41Q and G41M in the presence of actinonin (see Figure 2B). Extrapolating the situation to catalysis, which occurs in the crystalline states of PDF, it is likely that hydrolysis of the substrate frees the enzyme in its S state, which in turn needs to open to accommodate a new substrate (Figures 2B and 5C). This is well illustrated in the 3D structure of EcPDF complexed with a product of the reaction, obtained after co-crystallization of the enzyme with the substrate in a closed conformation [34]. The S free form is likely to exhibit a slower on-rate for the ligand (k3) compared to the O form because of steric occlusion of the active site (Figure S10). In support of this hypothesis, recent data show that the 3D structure of a C-terminally truncated, poorly active version of AtPDF is in the C conformation in the unbound state, although crystallized under conditions identical to ours [62,63]. This structure is similar to that of chain B, one of the two molecules of the asymmetric subunit of variant G41M (Figure 2B). This suggests that alterations remote from the active site significantly unbalance the equilibrium between the two conform- ers, thus altering the efficiency of the reaction (Figure 5C). As the S version corresponds to a significantly less active version of AtPDF compared to that reported in our present work, this further confirms that, compared to the O state, the S state has a significantly weaker propensity to bind substrate or a close mimic ligand, such as actinonin. Comparison of the 3D structures of the free-closed and the ligand-bound-closed forms reveals some differences responsible for the slight steric reduction of the active site of free-closed AtPDF1B with respect to that of the actinonin- AtPDF1B complex (Figure S10A), including the side chain of Ile42 burying the S1’ binding pocket (Figure S10B). Overall, these data suggest that an S form might exist under the free state but that it would feature a k3 value with respect to the ligand that is significantly weaker than that of the O form, which would strongly slow down the reaction or the binding as a result. With the interaction scheme proposed in our model (Figure 5B and C), the ligand/substrate binds more easily to the O form and induces the optimal conformation of the enzyme to reach the transition state, thus allowing the reaction to be efficiently catalyzed. In the final model (Figure 5C), there is both conformational selection and induced fit subsequently involved in line with the recently proposed existence of such mixed mechanisms for other enzymes [15,16]. Nevertheless, in our model (Figure 5C), we suggest that induced fit is the primary mechanism, as it provides energy input from the ligand, which eventually drives the enzyme towards the productive key-lock complex. Unambig- uous distinction between the relative contributions of the two mechanisms is deduced from the observation that kobs is a saturable function of actinonin with various PDF, including EcPDF, BsPDF, AtPDF (Figure 4B and C), and S. aureus PDF [12,16,29,49]. Using crystallographic reconstruction analysis involving enzyme variants, motions of small mobile loops and movie reconstructions of snapshots of catalytic events have been previously documented [1–3,64–66], often by visualizing the binding of unnatural inhibitors and not necessarily mimicking closely the substrate and transition state as actinonin does [67,68]. However, only a few examples make use of soaking conditions of a crystal to promote the motion and show the importance of induced fit [1,69]. None of these data show a motion of the amplitude revealed here with PDF and a large stabilization of the complex involving the formation of the four-stranded b-sheet superstructure and the entire N-domain of the enzyme. Compared to previous crystallographic analyses, our work integrates biophysical, computational, and kinetic analyses to reconstruct the whole picture, allowing a better understanding of the slow-binding mechanism. While our work primarily focused on an induced-fit mechanism of enzyme inhibition and catalysis, it should be emphasized that this phenomenon is also applicable to the broader area of receptor-ligand interactions. For example, in all cases where conformational change mechanisms have been proposed for kinase inhibitors without supporting experimental data [12,26], further experimental work must be provided to clarify the precise mechanism. We expect this will have important implications on how one conducts future drug-discovery efforts against such enzymes [70]. Materials and Methods Protein Expression and Purification Expression and purification of mature Arabidopsis thaliana PDF1B and all variants (i.e., AtPDF) were derived from the previously The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 12 May 2011 | Volume 9 | Issue 5 | e1001066 described protocol [37]: the lysis supernatant after sonication was applied on a Q-Sepharose column (GE Healthcare; buffers A and B as described containing 5 mM NiCl2) followed by Superdex-75 chromatography (GE Healthcare) using buffer C consisting of buffer A supplemented with 0.1 M NaCl. For crystallization experiments, the protein was purified further. The sample was concentrated on an Amicon Ultra-15 centrifugal filter unit (Millipore Corp.) with a 5-kDa cut-off and applied to a MonoQ HR5/5 column (GE Healthcare) previously equilibrated in buffer A (50 mM Hepes, pH 7.5, and 5 mM NiCl2). Elution was performed with a 50-mL gradient from 0% to 100% buffer B. The buffer of the pooled purified AtPDF1B was exchanged using a PD-10 desalting column (GE Healthcare) to yield a protein solution in 50 mM Hepes, pH 7.5, 0.1 M NaCl, and 5 mM NiCl2 (buffer C). The protein was concentrated on an Amicon Ultra-15 centrifugal filter unit. The resulting AtPDF1B preparation was frozen in aliquots and stored at 280uC (for crystallization purposes) or diluted 2-fold in 100% glycerol and stored at 220uC (for enzymatic purposes). The typical yield was 5–10 mg AtPDF per liter of culture. All purification procedures were performed at 4uC. Samples of the collected fractions were analyzed by SDS-PAGE on 12% acrylamide gels, and protein concentrations were estimated from the calculated extinction coefficients for each variant. Site-directed mutagenesis of AtPDF sequence in plasmid pQdef1bDN [36] was carried out using the QuickChange Site- Directed Mutagenesis Kit (Stratagene). Enzymology Assay of PDF activity was coupled to formate dehydrogenase, where the absorbance of NADH at 340 nm was measured at 37uC as previously described [71]. For measurements of classical kinetic parameters (i.e., Km and kcat), the reaction was initiated by addition of the substrate Fo-Met-Ala-Ser to the mixture containing purified enzyme in the presence of 1 mM NiCl2. The kinetics parameters were derived from iterative non-linear least square calculations using the Michaelis-Menten equation based on the experimental data (Sigma-Plot; Kinetics module). For determination of kinetic parameters related to actinonin, the reaction mixture contained 750 mM NiCl2. In some cases, the mixture containing PDF and actinonin was incubated for 15 min at 37uC before kinetic analysis, which was initiated by the addition of substrate. The same protocol was used to determine the dissociation constant of actinonin [KI* = k4/(k3+k3k5/k6)], but the initial reaction velocities were measured with varying concentrations of Fo-Met-Ala-Ser and actinonin. The data were then calculated according to the method of Henderson, which can be used to determine the dissociation constant of the tight-binding competitive enzyme inhibitor [28,49,72] by varying both the inhibitor and substrate concentra- tions. To determine KI, k5, and k6, the reaction was initiated by the addition of enzyme as previously described [29,49]. KI*app measurements were used for comparative studies of AtPDF variants (Table S3) at a concentration of 2 mM substrate by varying the concentration of actinonin. KI*app is the slope of the v[Actinonin]/v0 line curve. kobs was fitted from the kinetic data without preincubation with vI = vs + (v0 2 vs)e2kobst where vI is the observed velocity at a given concentration of inhibitor I, v0 is the velocity, and vs is the steady-state velocity [18]. From the set of values obtained at various concentrations of I, k5 and k6 could be derived using kobs = k6 + k5[I]/(KI + [I]). By choosing a set of values with kobs..k6, 1/kobs = 1/k5(KI/[I] +1) and 1/kobs = f(1/[I]) is expected to be a straight line in case of induced fit whose positive slope corresponds to 1/k5. k6 was derived from equation k6 = k5/ (KI/KI*21) [18,19]. Microcalorimetry ITC experiments were performed using a VP-ITC isothermal titration calorimeter (Microcal Corp.). Experiments were per- formed at 37uC. For each experiment, injections of 10 mL actinonin (180 mM) were added using a computer-controlled 300 mL microsyringe at intervals of 240 s into the Ni-AtPDF variant solution (5 to 10 mM, cell volume = 2.1 mL) dissolved in buffer C with stirring at 310 rpm. A theoretical titration curve was fitted to the experimental data using the ORIGIN software (Microcal). This software uses the relationship between the heat generated after each injection and DHu (enthalpy change in kcal/ mol), KA (the association binding constant in M21), n (number of binding sites per monomer), total protein concentration, and free and total ligand concentrations. The thermal stability of the WT and variants of Ni-AtPDF1B was studied by DSC using VP-DSC calorimetry (Microcal Corp.). DSC measurements were made with 10 mM protein solutions in buffer C. The actinonin concentration was 20 mM. The same buffer was used as a reference. All solutions were degassed just before loading into the calorimeter. Scanning was performed at 1uC/min. The temperature dependence of the partial molar capacity (Cp) was expressed in kcal/K after subtracting the buffer signal using Origin(R) software. Crystallization and Soaking Experiments Crystallization conditions were screened by a robot using the sitting drop vapor diffusion method. Crystals were obtained and optimized at 20uC with 15%–20% PEG-3350 and either 0.1 or 0.2 M zinc acetate. The drops were formed by mixing 2 mL of a solution containing 2 to 4 mg/mL protein and 2 mL of the crystallization solution. Crystals were soaked for 24 h by adding actinonin to the crystallization drops at a final concentration of 5 mM. Cryoprotection was achieved by placing crystals for 30 s in a solution that was composed of 20% PEG-3350 and 0.2 M zinc acetate, supplemented with 5%, 10%, and 15% glycerol. Crystals were then directly flash frozen in liquid nitrogen using cryoloops (Hampton Research). Crystals were also grown under conditions described for the C-terminally deleted, weakly active version of AtPDF [63]. X-Ray Diffraction Data Collection Data collections were performed at 100 K at the European Synchrotron Radiation Facility (Grenoble, France) on station ID29, FIP-BM30A, ID14-1, and ID23-2, and at SOLEIL (Gif-sur- Yvette, France) on station PROXIMA1. In each case, a single crystal was used to collect a complete dataset. Data were processed and scaled using XDS software [73]. Two crystal forms were encountered with different cell parameters. In each case, b parameter was nearly equal to a, and data could be indexed into two space groups, P212121 or P43212. The data are shown in Table S2. Structure Determination and Refinement The structure of free AtPDF was solved by molecular replacement with Phaser [74] followed by a rigid-body refinement by CNS [75] using coordinates from the Plasmodium falciparum PDF (PDB code 1RL4) [76] as a search model. The structures of actinonin-bound proteins—that is, WT and mutants—were solved using rigid-body refinement by CNS of the free AtPDF structure. The ten final models were obtained by manual rebuilding using TURBO-FRODO [77] and combined with refinement of only calculated phases using CNS and Refmac [78] software. No non- crystallographic symmetries were used. Quality control of the three models was performed using the PROCHECK program The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 13 May 2011 | Volume 9 | Issue 5 | e1001066 [79]. To probe for alternative conformers, Ringer was used [53]. Ringer is a program to detect molecular motions by automatic X- ray electron density sampling, and can be accessed at http:// ucxray.berkeley.edu/ringer.htm. Accession Numbers PDB codes for the PDF structures presented within this manuscript are as follows: 3M6O, 3PN2, 3M6P, 3O3J, 3PN3, 3PN4, 3PN5, 3M6Q, 3PN6, and 3M6R. UniProtKB accession numbers for other PDF studied are P0A6K3 (EcPDF) and O31410 (BsPDF). Supporting Information Figure S1 Alignment of PDF sequences and secondary structures. (A) PDF1B from Arabidopsis thaliana (AtPDF1B) is compared with bacterial type 1B (EcPDF and LiPDF), pathogenic protozoa (PfPDF1B), eukaryotic mitochondrial PDF1A from A. thaliana (AtPDF1A), and bacterial type 2 (BsPDF2). This figure was created with ENDscript [80]. The sequence alignment was realized with the algorithm muscle included in ENDscript, and modified according to the superimposition of structures. The blue frames indicate conserved residues, white characters in red boxes indicate strict identity, and red characters in yellow boxes indicate homology. The secondary structures at the top (a-helices, 310 helices, b-strands, and b-turns are shown by medium squiggles, small squiggles, arrows, and TT letters, respectively) were predicted by DSSP [81]. Relative accessibility (acc) of subunit A is shown by a blue-colored bar below sequence. White is buried, cyan is intermediate, and blue with red borders is highly exposed. A red box means that relative accessibility is not calculated for the residue, because it is truncated. Hydropathy (hyd) is calculated from the sequence according to [82]. It is shown by a second bar below accessibility: pink is hydrophobic, grey is intermediate, and cyan is hydrophilic. Motifs 1 (41GwGwAAXQ48), 2 (89EGCLS93), and 3 (133HEwDH137), where w is a hydrophobic amino acid, are labeled by red stars below the sequence alignment. To simplify the nomenclature, AtPDF1B is referred to as AtPDF throughout the text. (B) Topology cartoon of AtPDF, free (left) or actinonin bound (right), in the same color code as (A). Actinonin (represented by the yellow arrow) binding to the ligand binding site allows the linkage of the two distinct b-sheets into one single b-sheet, by mimicking an additional b-strand. PDB sum (http://www.ebi.ac. uk/thornton-srv/databases/pdbsum/) was used. (C) 3-D structure of AtPDF is represented showing the position of the residues discussed in the text, indicated in red. (EPS) Figure S2 Microcalorimetric titration of AtPDF with actinonin. Data were obtained at 37uC by an automated sequence of 28 injections of 180 mM actinonin from a 300 ml syringe into the reaction cell, which contain 9.85 mM AtPDF. The volume of each reaction was 10 ml, and injections were made at 240 s intervals. Top, raw data from the titration. Each peak corresponds to the injection. Bottom, the peaks in the upper panel were integrated with ORIGIN software and the values were plotted versus injection number. Each point corresponds to the heat in mcal generated by the reaction upon each injection. The solid line is the curve fit to the data by the Origin program. This fit yields values for Kd. Experiments were done with wild type protein and others variants, and gave similar raw data and curve fit. (A) WT; (B) variant G41M; (C) variant I42W. (EPS) Figure S3 Binding of actinonin to AtPDF does barely modify the crystal packing. (A) Crystal pack of the two complexes: open, free complex (left) and bound to actinonin (right) (B). Non-crystallo- graphic contacts into asymmetric unit are not modified by closing movement of the protein due to actinonin binding, except for zinc atom number 6. This metal ion is coordinated by side chains of Asp40 and Glu63, and water molecules, Asp40 and Glu63 being hydrogen bonded by side chain of Lys38 of the other subunit of the asymmetric unit. With the closing movement of the protein into the crystal, Ca of Asp40 shifted by 3.1 A˚ and its side chain flipped by 90u. Therefore, it does no longer participate to the coordination shell of this Zn2+ ion. However, it is still hydrogen bonded by Lys38 from chain B. (EPS) Figure S4 Binding of actinonin to AtPDF closely mimics both actinonin and product binding to EcPDF. Superimposition of EcPDF and AtPDF bound to either actinonin (1LRU PDB code, panel A) or Met-Ala-Ser (1BS6 PDB code, panel B), the product of the reaction. The r.m.s.d. value is 1.11 A˚ for 151 Ca superimposed. (EPS) Figure S5 The ligand binding site of AtPDF. This picture shows the residues of AtPDF that are in contact with actinonin (left) and 6b (right) according to the 3-D structure; this should be compared to the similar scheme shown in Figure 1B for EcPDF. (EPS) Figure S6 Electronic densities of the moving side-chains and of actinonin at the binding site in some variants of AtPDF. Actinonin and selected residues (G/Q/M41, I42, F58, and I130) are drawn in stick and are shown in their FO–FC electron density omit maps contoured at 2s, in free wild-type AtPDF (two crystallization conditions, WT1 and WT2), and ligand-bound WT (actinonin, 6b and 21), G41Q, and G41M variants. (EPS) Figure S7 Only few residues show alternative conformation in AtPDF. Alternative conformers in the crystalline state of AtPDF. Ringer plots of electron density (r) versus x1 angle for representative residues of the 3-D apostructure of AtPDF. Data were obtained with the 3M6O dataset (see Table S1). The secondary peaks in the Ile residues are observed because Ile is a branched amino acid. To evidence an alternative conformation with Ile, three peaks should be observed. (EPS) Figure S8 Impact of induced fit on the binding free energy of actinonin depends on the capacity to stabilize a hydrogen bond with PDF. (A) The gyration radii [83] of the side chain occurring at position 41 is displayed with black squares and compared to the kcat/Km values (grey bars). (B) The distance between the NH of I42 and the CO of actinonin was measured in each case. The percentage of the distance required to make a hydrogen bond (2.8 A˚ ) is reported (dark squares). The difference of binding free energy (DDGbinding) between the open, free state and the variants closed complexes of the G41 variants are displayed as grey bars. The values were calculated as follows. For the WT, it corresponds to the RT ln(KI*/KI) value [29], where R is the ideal gas constant and T is the temperature in Kelvin. RT is 0.616 kcal.mol21 at 37uC. For the G41M and G41Q variants, the DDGbinding corresponds to RT ln(KI-G41variant/KD-WT). The obtained values are similar to that obtained if the kcat/Km substitutes the KD value in the calculation (DDGbinding = RT ln(kcat/Km –G41variant/kcat/Km –WT). (EPS) Figure S9 Compound 21 does not bind AtPDF1B at S1’. (A) 21 is shown in ball-and-stick format in its FO–FC electron density omit The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 14 May 2011 | Volume 9 | Issue 5 | e1001066 map contoured at 2s. (B) Binding site of 21 into AtPDF1B is detailed. Red and blue residues indicate residues that accommo- date the ‘‘phenylalanine’’ and ‘‘trimethyl’’ groups of 21, respectively. (C) Overall view of 21 binding site (left). Molecular surface of AtPDF is represented, as well as 21 in ball-and-stick format. Residues belonging to the 21 binding pocket are colored in orange. For comparison, molecular surface of EcPDF (PDB code 1G2A) in the same orientation is also represented, with residues forming the new ligand binding pocket colored in orange. Actinonin is represented in ball-and-stick format and is seen through the molecular surface of each PDF. (D) Ball-and-stick representation of the interaction network around compound 21. The metal cation is shown as a grey sphere. (EPS) Figure S10 Poorly active versions of AtPDF are in a closed conformation incompatible with actinonin binding. (A) Free and close AtPDF were superimposed as in Figure 1C and are figured in brown and yellow, respectively. Both the G41M (chain B, shown in orange) and the free C-deleted weakly active AtPDF versions ([63], colored in purple, PDB entry code 3CPM) were superim- posed, to the two structures, showing that they both fit better to the ligand-bound full-length close form than to the free open form, but that the closure is further pronounced, burying the entrance to a ligand. (B) Close-up showing that the shape of the S1’ pocket of the poorly active closed versions make it poorly available to P1’ recognition (see circled Ile142 and Ile130 side chains). (EPS) Table S1 Catalytic properties of AtPDF. Nm, not measurable; ND, not determined; WT, is wild-type. aKinetic constants were determined using the coupled assay as indicated in Materials and Methods with substrate Fo-Met-Ala-Ser, in the presence of 100 nM enzyme variant and 750 mM NiCl2, at 37uC. The relative value of kcat/Km for wild-type AtPDF was set at 100%. bData correspond to the binding constant of actinonin as obtained either from ITC or from enzymatic analysis when indicated with an asterisk. cData from Table S3. dGyration radii are from [83]. (DOC) Table S2 Crystallographic data and refinement statistics. Values in parentheses are for the outer resolution shell. aRsym (I) = ShklSi|Ihkl,i 2 ,Ihkl.|/ShklSi|Ihkl,i|, where ,Ihkl. is the mean intensity of the multiple Ihkl,i observations for symmetry-related reflections. bRwork = 1006(Shkl|Fobs 2 Fcalc|/Shkl|Fobs|). Rfree is a test set including ,5% of the data. cPercentage of residues in most-favored/additionally allowed/generously allowed/disal- lowed regions of the Ramachandran plot. dCompound 21 was added first, and actinonin afterwards. (DOC) Table S3 Kinetic parameters for inhibition of some AtPDF variants by actinonin. The enzyme concentration used in the assay was 100 nM. Prior to kinetic analysis for determination of KI*app values, actinonin was incubated in the presence of each variant set at the final concentration for 10 min at 37uC; kinetic assay was started by adding a small volume of the substrate. For determination of KI, k5, and k6 values, actinonin was not pre- incubated with enzyme and kinetic assay was started by adding the enzyme. (DOCX) Movie S1 Dynamics of actinonin binding to peptide deformylase and closure of the active site. (WMV) Movie S2 Progressive motions of the main side chains at the active site and final locking of the hydrogen bond. (WMV) Acknowledgments We are strongly indebted to James Fraser and Tom Alber (University of California, Berkeley, USA) for introducing us to Ringer before the release of the freely available downloadable version. We thank Benoıˆt Gigant, Virginie Gueguen-Chaignon, Solange Morera, and Philippe Peynot (CNRS, Gif-sur-Yvette, France) for help with data processing and access to the crystallization facilities. We also thank Magali Nicaise-Aumont (IBBMC, Orsay, France), who performed the microcalorimetry experi- ments. We are grateful to the staff of the European Synchrotron Radiation Facility (ESRF) and SOLEIL beamlines for their help during data collection. Author Contributions The author(s) have made the following declarations about their contributions: Conceived and designed the experiments: SF CG TM. Performed the experiments: AB SF. Analyzed the data: FD MD SF CG TM. Contributed reagents/materials/analysis tools: IA MD CG TM. Wrote the paper: CG TM. References 1. Knowles JR (1991) Enzyme catalysis: not different, just better. Nature 350: 121–124. 2. Hammes GG (2002) Multiple conformational changes in enzyme catalysis. Biochemistry 41: 8221–8228. 3. Benkovic SJ, Hammes-Schiffer S (2003) A perspective on enzyme catalysis. Science 301: 1196–1202. 4. Henzler-Wildman K, Kern D (2007) Dynamic personalities of proteins. Nature 450: 964–972. 5. Teilum K, Olsen JG, Kragelund BB (2009) Functional aspects of protein flexibility. Cell Mol Life Sci. 6. Sullivan SM, Holyoak T (2008) Enzymes with lid-gated active sites must operate by an induced fit mechanism instead of conformational selection. Proc Natl Acad Sci U S A 105: 13829–13834. 7. Weikl TR, von Deuster C (2009) Selected-fit versus induced-fit protein binding: kinetic differences and mutational analysis. Proteins 75: 104–110. 8. Johnson KA (2008) Role of induced fit in enzyme specificity: a molecular forward/reverse switch. J Biol Chem 283: 26297–26301. 9. Bourgeois D, Royant A (2005) Advances in kinetic protein crystallography. Curr Opin Struct Biol 15: 538–547. 10. Katona G, Carpentier P, Niviere V, Amara P, Adam V, et al. (2007) Raman- assisted crystallography reveals end-on peroxide intermediates in a nonheme iron enzyme. Science 316: 449–453. 11. Koshland DE (1958) Application of a theory of enzyme specificity to protein synthesis. Proc Natl Acad Sci U S A 44: 98–104. 12. Tummino PJ, Copeland RA (2008) Residence time of receptor-ligand complexes and its effect on biological function. Biochemistry 47: 5481–5492. 13. Boehr DD, Nussinov R, Wright PE (2009) The role of dynamic conformational ensembles in biomolecular recognition. Nat Chem Biol 5: 789–796. 14. Bosshard HR (2001) Molecular recognition by induced fit: how fit is the concept? News Physiol Sci 16: 171–173. 15. Benkovic SJ, Hammes GG, Hammes-Schiffer S (2008) Free-energy landscape of enzyme catalysis. Biochemistry 47: 3317–3321. 16. Hammes GG, Chang YC, Oas TG (2009) Conformational selection or induced fit: a flux description of reaction mechanism. Proc Natl Acad Sci U S A 106: 13737–13741. 17. Fersht AR (1998) Structure and mechanism in protein science. New York: W.H. Feeman & Co. 18. Morrison JF, Walsh CT (1988) The behavior and significance of slow-binding enzyme inhibitors. Adv Enzymol Relat Areas Mol Biol 61: 201–301. 19. Copeland RA (2005) Evaluation of enzyme inhibitors in drug discovery: a guide for medicinal chemists and pharmacologists. New Jersey: John Wiley & Sons. 296 p. 20. Dash C, Vathipadiekal V, George SP, Rao M (2002) Slow-tight binding inhibition of xylanase by an aspartic protease inhibitor: kinetic parameters and conformational changes that determine the affinity and selectivity of the bifunctional nature of the inhibitor. J Biol Chem 277: 17978–17986. 21. Mac Sweeney A, Lange R, Fernandes RP, Schulz H, Dale GE, et al. (2005) The crystal structure of E.coli 1-deoxy-D-xylulose-5-phosphate reductoisomerase in a The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 15 May 2011 | Volume 9 | Issue 5 | e1001066 ternary complex with the antimalarial compound fosmidomycin and NADPH reveals a tight-binding closed enzyme conformation. J Mol Biol 345: 115–127. 22. Chou CJ, Herman D, Gottesfeld JM (2008) Pimelic diphenylamide 106 is a slow, tight-binding inhibitor of class I histone deacetylases. J Biol Chem 283: 35402–35409. 23. Barb AW, Jiang L, Raetz CR, Zhou P (2007) Structure of the deacetylase LpxC bound to the antibiotic CHIR-090: Time-dependent inhibition and specificity in ligand binding. Proc Natl Acad Sci U S A 104: 18433–18438. 24. Dunford JE, Kwaasi AA, Rogers MJ, Barnett BL, Ebetino FH, et al. (2008) Structure-activity relationships among the nitrogen containing bisphosphonates in clinical use and other analogues: time-dependent inhibition of human farnesyl pyrophosphate synthase. J Med Chem 51: 2187–2195. 25. Bateman RL, Ashworth J, Witte JF, Baker LJ, Bhanumoorthy P, et al. (2007) Slow-onset inhibition of fumarylacetoacetate hydrolase by phosphinate mimics of the tetrahedral intermediate: kinetics, crystal structure and pharmacokinetics. Biochem J 402: 251–260. 26. Copeland RA, Pompliano DL, Meek TD (2006) Drug-target residence time and its implications for lead optimization. Nat Rev Drug Discov 5: 730–739. 27. Gordon JJ, Kelly BK, Miller GA (1962) Actinonin: an antibiotic substance produced by an actinomycete. Nature 195: 701–702. 28. Chen DZ, Patel DV, Hackbarth CJ, Wang W, Dreyer G, et al. (2000) Actinonin, a naturally occurring antibacterial agent, is a potent deformylase inhibitor. Biochemistry 39: 1256–1262. 29. Van Aller GS, Nandigama R, Petit CM, Dewolf WE, Jr., Quinn CJ, et al. (2005) Mechanism of time-dependent inhibition of polypeptide deformylase by actinonin. Biochemistry 44: 253–260. 30. Rajagopalan PT, Grimme S, Pei D (2000) Characterization of cobalt(II)- substituted peptide deformylase: function of the metal ion and the catalytic residue Glu-133. Biochemistry 39: 791–799. 31. Schmitt E, Guillon JM, Meinnel T, Mechulam Y, Dardel F, et al. (1996) Molecular recognition governing the initiation of translation in Escherichia coli. A review. Biochimie 78: 543–554. 32. Giglione C, Fieulaine S, Meinnel T (2009) Cotranslational processing mechanisms: towards a dynamic 3D model. Trends Biochem Sci 34: 417–426. 33. Nguyen KT, Hu X, Pei D (2004) Slow-binding inhibition of peptide deformylase by cyclic peptidomimetics as revealed by a new spectrophotometric assay. Bioorg Chem 32: 178–191. 34. Becker A, Schlichting I, Kabsch W, Groche D, Schultz S, et al. (1998) Iron center, substrate recognition and mechanism of peptide deformylase. Nat Struct Biol 5: 1053–1058. 35. Guilloteau JP, Mathieu M, Giglione C, Blanc V, Dupuy A, et al. (2002) The crystal structures of four peptide deformylases bound to the antibiotic actinonin reveal two distinct types: a platform for the structure-based design of antibacterial agents. J Mol Biol 320: 951–962. 36. Giglione C, Serero A, Pierre M, Boisson B, Meinnel T (2000) Identification of eukaryotic peptide deformylases reveals universality of N-terminal protein processing mechanisms. EMBO J 19: 5916–5929. 37. Serero A, Giglione C, Meinnel T (2001) Distinctive features of the two classes of eukaryotic peptide deformylases. J Mol Biol 314: 695–708. 38. Dardel F, Ragusa S, Lazennec C, Blanquet S, Meinnel T (1998) Solution structure of nickel-peptide deformylase. J Mol Biol 280: 501–513. 39. Meinnel T, Blanquet S, Dardel F (1996) A new subclass of the zinc metalloproteases superfamily revealed by the solution structure of peptide deformylase. J Mol Biol 262: 375–386. 40. Larue V, Seijo B, Tisne C, Dardel F (2009) 1H, 13C and 15N NMR assignments of the E. coli peptide deformylase in complex with a natural inhibitor called actinonin. Biomolecular NMR Assignments 3: 153–155. 41. Amero CD, Byerly DW, McElroy CA, Simmons A, Foster MP (2009) Ligand- induced changes in the structure and dynamics of Escherichia coli peptide deformylase. Biochemistry 48: 7595–7607. 42. Berg AK, Srivastava DK (2009) Delineation of alternative conformational states in Escherichia coli peptide deformylase via thermodynamic studies for the binding of actinonin. Biochemistry 48: 1584–1594. 43. Zhou Z, Song X, Gong W (2005) Novel conformational states of peptide deformylase from pathogenic bacterium Leptospira interrogans: implications for population shift. J Biol Chem 280: 42391–42396. 44. Clements JM, Beckett RP, Brown A, Catlin G, Lobell M, et al. (2001) Antibiotic activity and characterization of BB-3497, a novel peptide deformylase inhibitor. Antimicrob Agents Chemother 45: 563–570. 45. Yoon HJ, Kim HL, Lee SK, Kim HW, Lee JY, et al. (2004) Crystal structure of peptide deformylase from Staphylococcus aureus in complex with actinonin, a naturally occurring antibacterial agent. Proteins 57: 639–642. 46. Moon JH, Park JK, Kim EE (2005) Structure analysis of peptide deformylase from Bacillus cereus. Proteins 61: 217–220. 47. Park J, Fu H, Pei D (2004) Peptidyl aldehydes as slow-binding inhibitors of dual- specificity phosphatases. Bioorg Med Chem Lett 14: 685–687. 48. Velazquez-Campoy A, Ohtaka H, Nezami A, Muzammil S, Freire E (2004) Isothermal titration calorimetry. Curr Protoc Cell Biol Chapter 17: Unit 17 18. 49. Boularot A, Giglione C, Petit S, Duroc Y, Sousa RA, et al. (2007) Discovery and refinement of a new structural class of potent peptide deformylase inhibitors. J Med Chem 50: 10–20. 50. Hackbarth CJ, Chen DZ, Lewis JG, Clark K, Mangold JB, et al. (2002) N-alkyl urea hydroxamic acids as a new class of peptide deformylase inhibitors with antibacterial activity. Antimicrob Agents Chemother 46: 2752–2764. 51. Ragusa S, Mouchet P, Lazennec C, Dive V, Meinnel T (1999) Substrate recognition and selectivity of peptide deformylase. Similarities and differences with metzincins and thermolysin. J Mol Biol 289: 1445–1457. 52. Fraser JS, Clarkson MW, Degnan SC, Erion R, Kern D, et al. (2009) Hidden alternative structures of proline isomerase essential for catalysis. Nature 462: 669–673. 53. Lang PT, Ng HL, Fraser JS, Corn JE, Echols N, et al. (2010) Automated electron-density sampling reveals widespread conformational polymorphism in proteins. Protein Sci 19: 1420–1431. 54. Meinnel T, Patiny L, Ragusa S, Blanquet S (1999) Design and synthesis of substrate analogue inhibitors of peptide deformylase. Biochemistry 38: 4287–4295. 55. Schechter I, Berger A (1967) On the size of the active site in proteases. I. Papain. Bochem Biophys Res Commun 27: 157–162. 56. Fersht AR, Shi JP, Knill-Jones J, Lowe DM, Wilkinson AJ, et al. (1985) Hydrogen bonding and biological specificity analysed by protein engineering. Nature 314: 235–238. 57. Takano K, Yamagata Y, Kubota M, Funahashi J, Fujii S, et al. (1999) Contribution of hydrogen bonds to the conformational stability of human lysozyme: calorimetry and X-ray analysis of six Ser —. Ala mutants. Biochemistry 38: 6623–6629. 58. Yuan Z, Trias J, White RJ (2001) Deformylase as a novel antibacterial target. Drug Discov Today 6: 954–961. 59. Wang Q, Zhang D, Wang J, Cai Z, Xu W (2006) Docking studies of Nickel- Peptide deformylase (PDF) inhibitors: exploring the new binding pockets. Biophys Chem 122: 43–49. 60. Li Y, Chen Z, Gong W (2002) Enzymatic properties of a new peptide deformylase from pathogenic bacterium Leptospira interrogans. Biochem Biophys Res Commun 295: 884–889. 61. Bracchi-Ricard V, Nguyen KT, Zhou Y, Rajagopalan PT, Chakrabarti D, et al. (2001) Characterization of an eukaryotic peptide deformylase from Plasmodium falciparum. Arch Biochem Biophys 396: 162–170. 62. Dirk LM, Williams MA, Houtz RL (2001) Eukaryotic peptide deformylases. Nuclear-encoded and chloroplast-targeted enzymes in Arabidopsis. Plant Physiol 127: 97–107. 63. Dirk LM, Schmidt JJ, Cai Y, Barnes JC, Hanger KM, et al. (2008) Insights into the substrate specificity of plant peptide deformylase, an essential enzyme with potential for the development of novel biotechnology applications in agriculture. Biochem J 413: 417–427. 64. Lee JE, Smith GD, Horvatin C, Huang DJ, Cornell KA, et al. (2005) Structural snapshots of MTA/AdoHcy nucleosidase along the reaction coordinate provide insights into enzyme and nucleoside flexibility during catalysis. J Mol Biol 352: 559–574. 65. Wang Y, Liu L, Wei Z, Cheng Z, Lin Y, et al. (2006) Seeing the process of histidine phosphorylation in human bisphosphoglycerate mutase. J Biol Chem 281: 39642–39648. 66. Parker JB, Bianchet MA, Krosky DJ, Friedman JI, Amzel LM, et al. (2007) Enzymatic capture of an extrahelical thymine in the search for uracil in DNA. Nature 449: 433–437. 67. Towler P, Staker B, Prasad SG, Menon S, Tang J, et al. (2004) ACE2 X-ray structures reveal a large hinge-bending motion important for inhibitor binding and catalysis. J Biol Chem 279: 17996–18007. 68. Teague SJ (2003) Implications of protein flexibility for drug discovery. Nat Rev Drug Discov 2: 527–541. 69. Geremia S, Campagnolo M, Schinzel R, Johnson LN (2002) Enzymatic catalysis in crystals of Escherichia coli maltodextrin phosphorylase. J Mol Biol 322: 413–423. 70. Pargellis C, Tong L, Churchill L, Cirillo PF, Gilmore T, et al. (2002) Inhibition of p38 MAP kinase by utilizing a novel allosteric binding site. Nat Struct Biol 9: 268–272. 71. Lazennec C, Meinnel T (1997) Formate dehydrogenase-coupled spectrophoto- metric assay of peptide deformylase. Anal Biochem 244: 180–182. 72. Henderson PJ (1972) A linear equation that describes the steady-state kinetics of enzymes and subcellular particles interacting with tightly bound inhibitors. Biochem J 127: 321–333. 73. Kabsch W (1993) Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Cryst 26: 795–800. 74. McCoy AJ, Grosse-Kunstleve RW, Storoni LC, Read RJ (2005) Likelihood- enhanced fast translation functions. Acta Crystallogr D Biol Crystallogr 61: 458–464. 75. Brunger AT, Adams PD, Clore GM, DeLano WL, Gros P, et al. (1998) Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr 54(Pt 5): 905–921. 76. Robien MA, Nguyen KT, Kumar A, Hirsh I, Turley S, et al. (2004) An improved crystal form of Plasmodium falciparum peptide deformylase. Protein Sci 13: 1155–1163. 77. Roussel A, Cambillau C (1989) TURBO-FRODO. In: Graphics S, ed. Silicon Graphics geometry partners directory Mountain View, CA. pp 77–78. 78. Murshudov GN, Vagin AA, Dodson EJ (1997) Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystal- logr 53: 240–255. 79. Laskowski RA, Moss DS, Thornton JM (1993) Main-chain bond lengths and bond angles in protein structures. J Mol Biol 231: 1049–1067. The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 16 May 2011 | Volume 9 | Issue 5 | e1001066 80. Gouet P, Courcelle E, Stuart DI, Metoz F (1999) ESPript: analysis of multiple sequence alignments in PostScript. Bioinformatics 15: 305–308. 81. Kabsch W, Sander C (1983) Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features. Biopolymers 22: 2577–2637. 82. Kyte J, Doolittle RF (1982) A simple method for displaying the hydropathic character of a protein. J Mol Biol 157: 105–132. 83. Levitt M (1976) A simplified representation of protein conformations for rapid simulation of protein folding. J Mol Biol 104: 59–107. The Dynamics of Induced Fit at High Resolution PLoS Biology | www.plosbiology.org 17 May 2011 | Volume 9 | Issue 5 | e1001066
3M6U
Multi-site-specific 16S rRNA methyltransferase RsmF from Thermus thermophilus in space group 43
Multi-site-specific 16S rRNA methyltransferase RsmF from Thermus thermophilus HASAN DEMIRCI,1,3 LINE H.G. LARSEN,2,3 TRINE HANSEN,2 ANETTE RASMUSSEN,2 ASHWIN CADAMBI,1 STEVEN T. GREGORY,1 FINN KIRPEKAR,2 and GERWALD JOGL1 1Department of Molecular Biology, Cell Biology and Biochemistry, Brown University, Providence, Rhode Island 02912, USA 2Department of Biochemistry and Molecular Biology, University of Southern Denmark, 5230 Odense M, Denmark ABSTRACT Cells devote a significant effort toward the production of multiple modified nucleotides in rRNAs, which fine tune the ribosome function. Here, we report that two methyltransferases, RsmB and RsmF, are responsible for all four 5-methylcytidine (m5C) modifications in 16S rRNA of Thermus thermophilus. Like Escherichia coli RsmB, T. thermophilus RsmB produces m5C967. In contrast to E. coli RsmF, which introduces a single m5C1407 modification, T. thermophilus RsmF modifies three positions, generating m5C1400 and m5C1404 in addition to m5C1407. These three residues are clustered near the decoding site of the ribosome, but are situated in distinct structural contexts, suggesting a requirement for flexibility in the RsmF active site that is absent from the E. coli enzyme. Two of these residues, C1400 and C1404, are sufficiently buried in the mature ribosome structure so as to require extensive unfolding of the rRNA to be accessible to RsmF. In vitro, T. thermophilus RsmF methylates C1400, C1404, and C1407 in a 30S subunit substrate, but only C1400 and C1404 when naked 16S rRNA is the substrate. The multispecificity of T. thermophilus RsmF is potentially explained by three crystal structures of the enzyme in a complex with cofactor S-adenosyl-methionine at up to 1.3 A˚ resolution. In addition to confirming the overall structural similarity to E. coli RsmF, these structures also reveal that key segments in the active site are likely to be dynamic in solution, thereby expanding substrate recognition by T. thermophilus RsmF. Keywords: rRNA methyltransferase; 5-methylcytidine; RsmB; RsmF; RNA mass spectrometry INTRODUCTION Ribosomal RNAs (rRNAs) are post-transcriptionally mod- ified in all three domains of life, and many modifications are phylogenetically conserved. Most modifications are located in functionally important regions of the ribosome, where they probably act to fine tune protein synthesis (Agris 2004; Gustilo et al. 2008). Complete modification maps of bacterial 16S rRNAs have been determined for only a hand- ful of species, and among these are the enteric bacterium Escherichia coli and the extremely thermophilic bacterium Thermus thermophilus (Guymon et al. 2006). Despite the large phylogenetic divergence of these two organisms, their ribosome modification patterns are quite similar. Of the 11 E. coli and 14 T. thermophilus 16S rRNA modifications, eight are identical. This suggests a set of common functional requirements conserved since divergence from their last common ancestor, and also suggests common recognition mechanisms among their modifying enzymes. For most ribosome modifications, a single enzyme recog- nizes and modifies a single site. However, there exist nota- ble exceptions. Among these are dimethylation of two adja- cent adenosines in 16S rRNA by KsgA (Helser et al. 1972); pseudouridylation of three adjacent residues in tRNAs by TruA (Hur and Stroud 2007); pseudouridylation of several tRNA residues by Pus1 (Motorin et al. 1998), Pus2 (Behm- Ansmant et al. 2007), or Pus7 (Behm-Ansmant et al. 2003); or methylation of four tRNA positions by Saccharomyces cerevisiae Trm4 (Motorin and Grosjean 1999). Even with these multi-site-specific enzymes, however, homologs from various species generally modify the same residues. E. coli 16S rRNA contains two 5-methylcytidine (m5C) residues, located in or near the highly conserved decoding 3These authors contributed equally to this work. Abbreviations: rRNA, ribosomal RNA; AdoMet, S-adenosyl-L-methionine; m5C, 5-methyl-cytidine; MALDI mass spectrometry, matrix-assisted laser desorption ionization mass spectrometry. Reprint requests to: Gerwald Jogl, Department of Molecular Biology, Cell Biology and Biochemistry, Brown University, Box G-E129, Provi- dence, RI 02912, USA; e-mail: Gerwald_Jogl@brown.edu; fax: (401) 863-6114; or Finn Kirpekar, Department of Biochemistry and Molecular Biology, University of Southern Denmark, Campusvej 55, 5230 Odense M, Denmark; e-mail: f.kir@bmb.sdu.dk; fax: (+45) 65502467. Article published online ahead of print. Article and publication date are at http://www.rnajournal.org/cgi/doi/10.1261/rna.2088310. 1584 RNA (2010), 16:1584–1596. Published by Cold Spring Harbor Laboratory Press. Copyright  2010 RNA Society. center of the 30S subunit (Fig. 1). An m5C967 modification is produced by RsmB (also called Fmu), while an m5C1407 modification is produced by RsmF, formerly known as YebU (Andersen and Douthwaite 2006). T. thermophilus 16S rRNA contains m5C967 and m5C1407, as well as two additional m5C nucleotides, m5C1400 and m5C1404 (E. coli rRNA numbering used throughout) (Guymon et al. 2006). While the m5C967 and m5C1407 modifications are pre- sumably produced by RsmB and RsmF homologs, respec- tively, the source of the two additional m5C residues has been unknown. Here we demonstrate that T. thermophilus RsmF is a multi-site-specific methyltransferase and, in contrast to the single-site-specific E. coli RsmF, is respon- sible for the synthesis of three modifications: m5C1407, m5C1400, and m5C1404. We also demonstrate that RsmB is responsible for the synthesis of m5C967 in T. thermophilus as well as is in E. coli, thereby accounting for all four m5C modifications of 16S rRNA. We present crystal structures of T. thermophilus RsmF up to 1.3 A˚ resolution that reveal a dynamic region in the active site that is absent from the E. coli RsmF structure, providing a possible explanation for the expanded recognition capacity of the T. thermophilus methyltransferase. RESULTS Identification of T. thermophilus 16S rRNA m5C-methyltransferases With the E. coli RsmB and RsmF protein sequences as queries, we used conventional BLAST searches (Altschul et al. 1990) to identify potential homologs encoded by the T. thermophilus HB8 genome (data not shown). Both RsmB and RsmF have the highest similarity to the T. thermophilus protein encoded by TTHA1387 (BLAST scores of 106 and 190, respectively) and second-highest similarity to the pro- tein encoded by TTHA0851 (BLAST scores of 93 and 81, respectively). The simplest interpretation of these results is that TTHA1387 encodes RsmF, responsible for methylation of C1407, leaving TTHA0851 as the most likely candidate for the gene encoding RsmB, responsible for methylation of C967. The similarities of the two E. coli enzymes with other T. thermophilus proteins were far too low to reveal potential candidates responsible for methylation of C1400 and C1404. We next constructed T. thermophilus strains in which either TTHA0851 or TTHA1387 was inactivated by the homologous recombination and insertion of a heat stable kanamycin-resistance gene. 16S rRNA was isolated from these null mutants and subfragments of z50 nucleotides (nt) around the regions of interest were further purified, digested with RNase T1, and analyzed by MALDI mass spectrometry (Fig. 2). Comparison of the TTHA1387 null mutant to wild-type T. thermophilus HB8 indicates three clear differences, each corresponding to the disappearance of a methyl group (z14.0 Da). The RNase T1 digestion fragment harboring m5C1407, the nucleotide methylated by RsmF in E. coli, is absent in the null mutant, indicating that TTHA1387 is indeed rsmF. The predicted RNase T1 fragment reduced by 14.0 Da is obscured by another RNase T1 fragment that is present in both the wild-type and TTHA1387 null mutants (Fig. 2B). Unexpectedly, two addi- tional RNase T1 fragments are also reduced by 14.0 Da. One of these contains C1400 while the other contains C1404. This latter RNase T1 fragment from wild-type T. thermophilus contains three methyl groups, two on m4Cm1402 and one on m5C1404 (Guymon et al. 2006), preventing an unambiguous identification of the missing methyl group. We therefore performed tandem mass spec- trometry on the 1402CCCG1405 RNase T1 fragment with two methyl groups from the TTHA1387 null mutant and com- pared it with the triply methylated wild-type RNase T1 fragment (Fig. 2C). The clear w2 ions, as well as the less intense z3 ions, display a 14.0 Da mass difference between the two samples, showing that the methylations on m4Cm1402 were not affected by inactivation of TTHA1387. Tandem mass spectrometry was also performed on the RNase T1 fragments appearing as a consequence of the lack of methylations on C1400 and C1407 (data not shown). As expected, the C1400-containing fragment revealed no FIGURE 1. Secondary structure diagram of the 39 minor domain of 16S rRNA indicating the position of the three RsmF substrate nucleotides. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1585 indications of a methyl group, whereas the RNase T1 fragment with C1407 ex- hibited a fragmentation pattern corre- sponding to the expected mass overlap with an RNase T1 fragment of a different sequence. In summary, our data lead us to conclude that TTHA1387 encodes an RsmF m5C methyltransferase responsible for synthesizing m5C1400, m5C1404, and m5C1407 in 16S rRNA of T. thermophilus. The above results left us with TTHA0851 as the sole candidate for the gene encoding the m5C967 methyl- transferase. An approach conceptually identical to that described above re- vealed that disruption of TTHA0851 reduced the relevant RNase T1 fragment by 14.0 Da (Supplemental Fig. 1A). Since this fragment is methylated at G966 and C967 in the wild-type strain (Guymon et al. 2006), tandem mass spectrometry was again performed (Sup- plemental Fig. 1B), showing that only the methyl group on C967 was absent. In agreement with the suggested nomen- clature for rRNA modifying enzymes (Andersen and Douthwaite 2006), we hereafter refer to TTHA0851 as rsmB due to substrate specificity identical to the originally identified enzyme from E. coli (Gu et al. 1999). Effect of temperature on growth of the rsmF null mutant One possible explanation for methyla- tion of multiple sites by RsmF is that such additional methyl groups improve ribosomal function at elevated temper- ature. To address this possibility, we ex- amined the effect of temperature on growth of the rsmF null mutant. Wild- type T. thermophilus and the rsmF null mutant were cocultured at three tem- peratures and these cocultures were se- rially subcultured for seven cycles of 24 h each. The proportion of wild-type and rsmF null mutant cells in each mixed culture was determined by spreading di- lutions onto TEM plates with or without kanamycin. After seven cycles, no differ- ence in the relative proportions of the wild-type and rsmF mutant was exhib- ited at 70°C. However, at 60°C, the rsmF null mutant constituted only around 5% FIGURE 2. (Legend on next page) Demirci et al. 1586 RNA, Vol. 16, No. 8 of the population, and at 80°C the rsmF null mutant was unable to grow at all. Thus, methylation by RsmF appears to facilitate growth at temperatures outside the optimal growth temperature. RsmF substrate preference The T. thermophilus rsmB and rsmF genes were each cloned into an E. coli expression plasmid in order to produce proteins for X-ray crystallography and in vitro methylation studies. The expression constructs were equipped with C-terminal histidine6 tags to facilitate protein purification. While we achieved a high expression level of RsmF, we were unable to do so with RsmB despite a series of optimization attempts. Consequently, in vitro substrate and structure analyses were performed exclusively with RsmF. E. coli RsmF requires the 30S ribosomal subunit as a substrate when the activity is assayed in vitro (Andersen and Douthwaite 2006). We assayed 70S ribosomes, 30S ribosomal subunits, and 16S rRNA for the ability to serve as substrates for methylation by RsmF in vitro. 16S rRNA subfragments of z50 nt around the target sites were purified after the in vitro assay and analyzed by mass spectrometry as described above. In vitro methylation at 70°C showed an interesting but rather complex substrate pattern. RsmF completely methylates C1400 when either 16S rRNA or 30S subunits are used as a substrate. It methylates C1404 to z35% with 16S rRNA and completely with 30S subunits, and it produces only trace amounts of methylation of C1407 with 16S rRNA and z75% with 30S subunits (Fig. 3). There were no indications of the 70S ribosome being a substrate in vitro. Curiously, T. thermophilus RsmF expressed in an E. coli rsmF null mutant almost completely methylated, in vivo, positions C1400 and C1404, but not C1407 (data not shown). X-ray crystal structures of RsmF We determined the structure of T. thermophilus RsmF (456 amino acids) in three different crystal forms and in a com- plex with cofactor AdoMet to up to 1.3 A˚ resolution (Figs. 4, 5). The structure was solved in space group P43 (data set RsmF1, 1.4 A˚ resolution) by molecular replacement using a search model generated with the program Modeller (Eswar et al. 2008) from the catalytic domain of the RsmF homolog YebU from E. coli (Pdb 2FRX) (Hallberg et al. 2006). The structures of the AdoMet-bound form in space group P2 (RsmF2, 1.82 A˚ resolution), of the AdoMet- bound form (RsmF3, 1.3 A˚ resolution), and of the apo- form (RsmF4, 1.68 A˚ resolution) in space group P21212 were subsequently solved by molecular replacement with the refined RsmF1 model. There are two molecules in the asymmetric unit in space groups P43 and P2 and one mol- ecule in space group P21212. Electron density is generally well defined in all crystal forms. The majority of residues (92.0%, 92.3%, 93.1%, and 92.6%) are in the most favored region of the Ramachandran plot for RsmF1, RsmF2, RsmF3, and RsmF4, respectively, and there are no residues in the disallowed region. The final models consist of residues 5–178, 194–198, and 201–456 and five additional residues from the histidine6 affinity tag in both chains of data set RsmF1; residues 2–456 and five affinity-tag residues in both chains of data set RsmF2; residues 1–456 and six affinity-tag residues in data set RsmF3; and resi- dues 1–456 and seven affinity-tag residues in RsmF4. The N-terminal a-amino group was ordered in data sets RsmF3 and RsmF4 and contained additional electron density, which we interpreted as N-(dihydroxymethyl)-L-methio- nine, the hydrated form of N-formyl-methionine. Data collection and refinement statistics are given in Table 1. The overall structure of RsmF consists of a central canonical class I methyltransferase catalytic domain with additional N-terminal and C-terminal domains (Figs. 4, 5). The catalytic do- main is formed by a central seven- stranded b-sheet that is flanked on both sides by three helices of varying lengths. An inserted region between strand b7 and helix a11 contains additional heli- ces a9 and a10, which interact with the two N-terminal helices a1 and a2. A second inserted region following strand b9 includes the short helices a13 and FIGURE 2. (A) MALDI mass spectra of an RNase T1-digested 16S rRNA subfragment (pos. 1378–1432) from wild-type cells (upper panel) or from the TTHA1387 (putative rsmF) null mutant (lower panel). Expected digestion products are labeled; fragments affected by the null mutation are set in italics. (B) Expansion of the signals affected by the TTHA1387 null mu- tation. The sequence and methylation status of the RNase T1 products are indicated. (C) MALDI tandem mass spectrometry of the methylated RNase T1 fragment of 16S rRNA (pos. 1404–1407); wild-type cells (upper panel), TTHA1387 null mutant (lower panel). Mass spectrometric fragments used to deduce the methylation status are labeled. The position of the backbone fragments (nomenclature according to McLuckey et al. [1992]) in the sequence is shown. MH+, precursor ion selected for fragmentation; C, cytosine; mC, methylated cytosine; C>p, cytidine-2´-39-monophosphate; me, methyl group. FIGURE 3. In vitro methylation with 30S ribosomal subunits or 16S rRNA from the T. thermophilus rsmF null mutant as a substrate. Effect on C1400-, C1404-, and C1407-harboring RNase T1 products. In vitro methylated products are set in italics. *, artifact signal arising from the enzyme preparation. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1587 a14, which interact with helix a12. Furthermore, RsmF contains three additional smaller domains, an N-terminal domain consisting of a three-stranded b-sheet and two flank- ing helices (Fig. 5B, colored in blue), and two C-terminal domains consisting of four-stranded b-sheets and two or one helix (Fig. 5B, colored in magenta and red). Cofactor binding, substrate docking, and conformational flexibility in the active site The coordination of AdoMet in the T. thermophilus RsmF active site is similar to that seen in other class I methyl- transferases. However, the previously published structure of E. coli RsmF did not contain the cofactor AdoMet in the active site, precluding a direct comparison. Both the T. thermophilus and E. coli RsmF cofactor-binding sites reveal a new variation for the methyltransferase signature motif I (Malone et al. 1995), with the highly conserved GxGxG sequence replaced by 109AAAPG113. The combination of three alanines and a proline results in a loop conformation that is very similar to that observed in other methyltrans- ferases with a GxGxG motif (e.g., RsmC) (Demirci et al. 2008a). In RsmF, the amide hydrogen atom of the last glycine residue forms a hydrogen bond with the cofactor carboxy group (Fig. 6). Other key interactions with AdoMet are well conserved in RsmF. The cofactor adenine ring is located in a mainly hydrophobic pocket lined by residues Val134, Pro160, and Leu211. This pocket is open toward the solvent. The adenine amino group is not specifically recog- nized and interacts with solvent water molecules. The ribose hydroxyl groups form hydrogen bonds with Glu133 and Arg138, and the methionine amino group interacts with Asp177. The AdoMet cofactor is bound in a cleft in the RsmF active site, which suggests that substrate cytidine bases are inserted into the active site in an unstacked conforma- tion. Inspection of the electrostatic charge distribution re- veals a large positively charged surface region, which would be consistent with binding to an RNA surface and modifi- cation of the substrate base in an unstacked orientation (Fig. 6C). To evaluate the possible orientation of a substrate base in the active site, we performed computational docking calculations with the program Dock6 (Lang et al. 2009). The resulting positions of cytosine and m5C in the presence of AdoMet in data set RsmF3 are highly similar to each other, with m5C placed into the active site with its phosphate group toward a positively charged pocket at the entrance of the active site cleft (Fig. 6E). The position of the phosphate group is close to a sulfate molecule that we observed in data set RsmF1, providing further support for the results of the docking calculation (Fig. 6F). Interestingly, we observed that three active site segments were disordered in data set RsmF1. These segments include residues 179–193 (including helices a9 and a10 and the catalytic Cys180), residues 199–200, and the N-terminal residues 1–5, which interact with the first two segments (Fig. 6E,F, colored in green). We observed electron density for the intervening residues 194–198, which formed a lattice con- tact with a neighboring molecule. However, the position of FIGURE 4. Structure-based sequence alignment of RsmF from T. thermophilus and E. coli. Secondary structure elements of T. thermophilus RsmF are indicated on top. The color scheme for the secondary structure elements is as in Figure 5A. The position of the variant methyltransferase motif I is marked with a red box. A flexible region observed in the active site is marked with a green box; residues interacting with the cofactor are marked with orange boxes. Demirci et al. 1588 RNA, Vol. 16, No. 8 these five residues was not related to their position in the other three data sets, suggesting that the extended active site region between residues 179 and 201 can reorient in the RsmF structure. This observation suggests that this active site region is dynamic, which may be important for sub- strate binding at 72°C, the optimum growth temperature of T. thermophilus. DISCUSSION Substrate recognition mechanisms We have identified the two enzymes responsible for the synthesis of the four m5C modifications of T. thermophilus 16S rRNA, and characterized the RsmF methyltransferase responsible for synthesizing three of these. rRNA modifying enzymes in bacteria are generally highly specific, with a one-to-one association between the modifying enzyme and the modification. A few cases of multitarget ribosome mod- ifying enzymes have been reported (Helser et al. 1972; Demirci et al. 2008b), but to our knowledge T. thermophi- lus RsmF is the first rRNA methyltransferase found to mod- ify three different nucleotides. Most ribosome modifying enzymes probably recognize assembly intermediates, and the data presented here are consistent with that notion. T. thermophilus RsmF methylates C1400 and C1404 in vitro using either 16S rRNA or 30S subunits as substrates, whereas both E. coli (Andersen and Douthwaite 2006) and T. thermophilus RsmF exclusively utilize 30S subunits as substrates for methylation of C1407. This may reflect that C1400 and C1404 methylations do not rely on the asso- ciation of ribosomal proteins in order to be recognized by T. thermophilus RsmF. C1407 methylation, in contrast, depends on both rRNA and the ribosomal protein for the recognition by RsmF in both T. thermophilus and E. coli. More puzzling is the observation that T. thermophilus RsmF does not methylate E. coli ribosomes in vivo on C1407. It is perhaps worth noting that methylation of C1407 in the T. thermophilus 30S ribosomal subunit in vitro was less efficient than methylation of the other two positions, indicating the need for a particular intermediate assembly structure or for accessory factors. The only clear in vitro FIGURE 5. Overall structure of RsmF. (A) Schematic representation of the position of the substrate bases in the 30S ribosomal subunit (Pdb entry 2WRI) (Gao et al. 2009). C1400, C1404, and C1407 in helix 44 (green) are shown in stick representation with 5-methyl groups as pink spheres. (B) Schematic stereo representation of the overall structure of RsmF. Secondary structure elements are in orange and yellow for the catalytic domain, in salmon and blue for additional N-terminal domains, and in magenta and red for C-terminal domains. The flexible region in the active site including helices a9 and a10 is indicated with an arrow and colored in green. (C) Topology diagram with secondary structure elements colored as in B. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1589 methylation observed at 37°C with T. thermophilus RsmF was on C1400 with 16S rRNA as the substrate (data not shown), which is not evidently related to the methylation pattern in vivo in the heterologous system. It seems unlikely that the aberrant methylation in the heterologous system reflects species-specific differences in the mature 30S ribo- somal subunit, given the extreme sequence and structural conservation of the decoding site. Instead, it may reflect differences in 30S subunit assembly in the two organisms, necessary due to the large difference in growth temperature. The four m5C residues in T. thermophilus 16S rRNA are clustered in and around the functionally critical decoding center at or close to sites of contact with tRNA, mRNA, and EF-G (Ogle et al. 2002; Selmer et al. 2006; Gao et al. 2009). m5C1400, m5C1404, and m5C1407 are located in the subunit body, while m5C967 is located in the subunit head, about 10 A˚ from m5C1400 (Wimberly et al. 2000). Exam- ination of the 30S subunit crystal structure (Wimberly et al. 2000) indicates that the three bases methylated by RsmF are situated in three distinct structural contexts, but provides few clues to a common mode of substrate recognition. C1400 is an unpaired base protruding from a sharply bent segment of rRNA at the junction of helices 43 and 44, while C1404 and C1407 are engaged in Watson–Crick pairs within helix 44 (Fig. 4A). The C5 positions of the latter two bases are not obviously accessible, such that RsmF would need to approach them from the major groove side. While a flipping of C1407 out of the helix via the minor groove could allow access to this base, such a mechanism would be problematic for C1404, whose minor groove side is packed against the rest of the 30S subunit. While m5C1404 and m5C1407 are about 11 A˚ apart (Selmer et al. 2006), m5C1400 is quite distant from both of these bases (about 21 and 30 A˚ , respectively). The pro- truded conformation of m5C1400 in the mature 30S sub- unit is due in part to base-pairing interactions between adjacent bases and the 1500 region of 16S rRNA (C1399– G1504 and G1401–C1501). As the 1500 region is one of the last segments of 16S rRNA to be synthesized, C1400 could potentially be positioned much closer to C1404 and C1407 in an assembly intermediate, prior to the formation of the C1399–G1504 and G1401–C1501 base pairs. RsmF could utilize a single binding mode to then access all three bases, further facilitated by its flexible active site domain. Meth- ylation of C1400 in mature 30S subunits would therefore involve disruption of the adjacent base pairs. A complete understanding of the recognition mechanism of these enzymes will require high-resolution structural data on assembly intermediate-enzyme complexes. Given the large number of potential subunit assembly intermediates, pre- cisely defining the physiological substrate for RsmB and RsmF will be a formidable task. TABLE 1. Data collection and refinement statistics RsmF1 RsmF2 RsmF3 RsmF4 Data collectiona AdoMet AdoMet Space group P43 P2 P21212 P21212 Cell dimensions a, b, c (A˚ ) 71.0, 71.0, 186.7 66.0, 78.3, 108.1 89.7, 109.0, 51.0 89.8, 109.1, 50.8 a, b, g (°) 90, 90, 90 90, 107.1, 90 90, 90, 90 90, 90, 90 Resolution (A˚ )b 30–1.4 (1.55–1.40) 30–1.82 (1.89–1.82) 30–1.30 (1.34–1.30) 30–1.68 (1.74–1.68) Rmerge 0.065 (0.59) 0.08(0.38) 0.058(0.36) 0.15 (0.49) I/sI 29.3(2.15) 12.6 (2.04) 24.3 (2.05) 14.2 (1.73) Completeness (%) 90.1 (72.6) 97.0(86.5) 95.6(66.3) 99.6 (97.8) Redundancy 8.9 (5.4) 2.8(2.0) 5.2(2.1) 6.4 (3.9) Refinement Resolution (A˚ ) 30–1.4 (1.42–1.40) 30–1.82 (1.84–1.82) 30–1.30 (1.32–1.30) 30–1.68 (1.69–1.68) Number of reflections 161,955 (4356) 91,762 (2583) 119,490/2640 109,375 (3244) Rwork/Rfree 0.169/0.189 (0.227/0.263) 0.162/0.194 (0.222/0.259) 0.177/0.191 (0.233/0.234) 0.173/0.192 (0.217/0.266) Number of atoms Protein 6766 7117 3598 3574 Ligand/ion 20 54 27 1 Water 1486 1285 856 665 B-factors Protein 24.1 20.4 15.9 17.2 Ligand/ion 34.5 21.6 17.5 19.1 Water 38.7 36.3 33.6 33.6 RMSDs Bond lengths (A˚ ) 0.009 0.006 0.005 0.004 Bond angles (°) 1.22 1.04 1.14 0.94 aOne crystal used for each data set. bThe highest resolution shell is shown in parentheses. Demirci et al. 1590 RNA, Vol. 16, No. 8 Structural comparison of RsmF with related methyltransferases A database search with Dali (Holm et al. 2008) confirmed the structural similarity of the T. thermophilus and E. coli (PDB 2FRX) (Hallberg et al. 2006) RsmF homologs, which superimpose with a root-mean-square deviation (RMSD) of 1.6 A˚ for 342 Ca atoms (Fig. 7A) and are the only two structures in the Pro- tein Database with this domain organi- zation. Even so, substantial structural differences are observed in most of the loop regions and for the long connect- ing loop between the methyltransferase domain and the first C-terminal domain. While in the active site, the positions of residues in the cofactor-binding site and of the two cysteine residues are conserved (Fig. 7B), there are a number of positively charged residues (Arg30, Arg190, Arg194, His195, and Arg203) in the T. thermophilus structure that are absent from the E. coli enzyme (Fig. 7B). Three of these are located in the flexible region, and the combination of a posi- tive charge and flexibility close to the active site is suggestive of a functional contribution of this region to the mul- tisite specificity of T. thermophilus RsmF (Figs. 6C,D, 7B). Methylation of three rRNA positions may require an increase in the enzyme’s structural dynamics in order to accommodate the 30S subunit in slightly different orientations. Similar observations have been made for other multi-site-specific methyltransferases in- cluding KsgA, which modifies two adja- cent adenosines in the 30S ribosomal subunit (O’Farrell et al. 2004; Demirci et al. 2009), and the PrmA ribosomal protein methyltransferase, which under- goes dramatic interdomain movements to modify multiple lysine residues and the N-terminal a-amino group on the same substrate protein (Demirci et al. 2007, 2008b). The second C-terminal domain in RsmF is related to the RNA-binding PUA (pseudouridine synthase and archaeosine transglycosylase) domains (Perez-Arellano et al. 2007). The RlmI methyltransferase, which produces m5C1962 in 23S rRNA (Purta et al. 2008) also contains a PUA domain (Sunita et al. 2008), although it is N-terminal to the catalytic methyltransferase domain and in a different orientation. PUA domains contain six b-strands, which form a central pseudobarrel closed by a short 310-helix. A comparison of the C-terminal domain in RsmF with a typical PUA domain in archaeosine trans- glycosylase (ArcTGT, Pdb entry 1J2B) reveals that the central fold is similar (53 Ca atoms align with an RMSD FIGURE 6. Substrate docking and conformational flexibility. (A) Cofactor-binding site in RsmF. Bound AdoMet is shown in blue sticks. Hydrogen bonds to coordinating residues are indicated. (B) Final sA-weighted 2mFO-DFC electron density map of the cofactor-binding site (data set RsmF2) contoured at the 1s level. (C,D) Comparison of the electrostatic surface charge distribution between RsmF from T. thermophilus and E. coli. The location of the C-terminal domains and of the flexible region (labeled ‘‘FLEX’’) is indicated with circles. AdoMet and docked m5C are shown as sticks. AdoMet from the T. thermophilus structure is shown with the E. coli structure for comparison. (E) The modeled position of m5C docked into the active site of RsmF. m5C and AdoMet are shown in tan and blue sticks, respectively. Residues in the flexible region are shown in green. (F) The active site region in data set RsmF1. A sulfate ion is observed close to the position of the m5C phosphate group. Residues 194–198 in the flexible region interacting with a neighboring RsmF molecule are indicated with an arrow. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1591 of 1.8 A˚ ), but that several connecting loop regions are substantially shorter and two a-helices and one b-strand of the pseudobarrel are absent. Thus, this PUA-like domain differs considerably from typical PUA domains. However, the similarity to RNA-binding PUA domains and the positive surface charge distribution observed in both RsmF structures are suggestive for a conserved function of the PUA-like domain in RNA recognition. The next most closely related structure is E. coli RsmB (474 residues, PDB 1SQF) (Foster et al. 2003). A total of 264 Ca atoms can be aligned with an RMSD of 1.5 A˚ between RsmB and RsmF. Both enzymes retain the same organization for the N-terminal domains and the core methyltransferase domain. However, an additional 140- residue N-terminal RNA-binding domain provides sub- strate specificity to RsmB, whereas the two C-terminal domains following the core methyltransferase domain (160 residues) are likely to determine substrate recognition by RsmF. Thus, these two enzymes have evolved substrate specificity via acquisition of additional, unrelated RNA recognition domains. While there are as yet no enzyme– substrate complexes for rRNA m5C- methyltransferases, insights into the RsmF catalytic mechanism can be gleaned from a comparison with the RlmD and TrmA m5U methyltransferases in covalent in- termediate complexes with RNA oligonu- cleotides (Lee et al. 2005; Alian et al.2008). DNA and RNA m5C methyltransferases use a thiol from a catalytic cysteine residue to attack the six-position of the pyrimi- dine base to activate the five-position for methyl group transfer (Liu and Santi 2000). Cys180 and Cys230 in RsmF are positioned equivalently to the catalytic Cys324 and the catalytic base Glu358 of TrmA. The substrate nucleotides insert into TrmA and RsmF in unrelated di- rections, consistent with a lack of struc- tural homology outside the methyltrans- ferase domains. Nevertheless, the C5 positions of the pyrimidine rings and of the 5-methyl carbons are quite similar with respect to the catalytic cysteine resi- dues and the AdoMet cofactor. The same structural homology of the active site geometry can be observed in comparison with the RlmD methyltransferase (Sup- plemental Fig. 2; Lee et al. 2005). A possible origin of T. thermophilus RsmF T. thermophilus RsmF shows the highest similarities with proteins from close relatives, namely, Thermus aquaticus and two Meiothermus species of the Thermaceae family. Remarkably, the next highest similarities (BLAST scores between 267 and 359) are with the NOL1/ NOP2/Sun proteins from the Gram-positive Firmicutes phylum (Supplemental Fig. 3). This similarity, together with the fact that the Thermaceae family and most members of the Firmicutes identified in Supplemental Figure 3 are thermophilic, suggest that rsmF has undergone horizontal transfer between the Thermaceae family and members of the Firmicutes phylum. Thus, we speculate that this version of the RsmF protein, which catalyzes methylation of three cytidines, may be adaptive for existence in thermally challenging environments. The effect of the loss of methyl- ation by RsmF on growth at different temperatures is consistent with this notion. Our hypothesis of horizontal transfer of rsmF predicts that RsmF of other members of the Thermaceae family and members of the Firmicutes phylum will also be found to introduce multiple m5C modifications. FIGURE 7. Comparison with other methyltransferases. (A) Differences between the overall structures of RsmF from T. thermophilus (orange) and E. coli (green). AdoMet bound in the T. thermophilus structure is shown as blue sticks. (B) Comparison of the active site region in both enzymes. Residues in the cofactor-binding site and in the flexible region in the T. thermophilus enzyme are shown as sticks. (C) Comparison between RsmF (orange) and a substrate complex structure of the TrmA methyltransferase (cyan/blue). (D) Comparison of the active site region illustrating the differences in the insertion direction of the substrate base in RsmF (m5C in light orange) and in TrmA (m5U in cyan). Demirci et al. 1592 RNA, Vol. 16, No. 8 MATERIALS AND METHODS Cloning of the T. thermophilus rsmB and rsmF genes The T. thermophilus HB8 loci TTHA0851 (GenBank accession number BAD70674) and TTHA1387 (GenBank accession number BAD71210) were PCR amplified from genomic DNA and purified via the High Pure PCR Template Preparation Kit (Roche). The 100 mL PCRs contained 150 ng DNA, 10 mM of each primer, 10 mM dNTP, 1 unit Phusion DNA polymerase (Finnzymes), and 1x Phusion HF buffer. Primers for rsmB amplification were 59-CC CTGGACATATGAGGGCCGG-39 and 59-GGCCAAGATCTTGCC TGAGAG-39, and the temperature cycling was as follows: 98°C/30 sec; 30X (98°C/10 sec, 59°C/20 sec, and 72°C/36 sec); and 72°C/420s. Primers for rsmF amplification were 59-GCTAGGGTACACATA TGCTGCCC-39 and 59-GCACGGGGGTGAGATCTAAGCCC-39, and the temperature cycling was as follows: 98°C/30 sec; 30X (98°C/10 sec, 62°C /20 sec, 72°C/42 sec); and 72°C/420 sec. The desired PCR products were purified from agarose gels using the GFX PCR purification kit (GE Healthcare). The PCR fragments were digested with NdeI and BglII and inserted into the expression vector pLJ102 (Andersen and Douthwaite 2006), generating isopropyl-1-thio-b-D-galactopyranoside (IPTG)-inducible genes for the recombinant proteins with a C-terminal histidine6 tag. The constructs (designated pLJ102-RsmB and pLJ102-RsmF) were used to transform an rsmF-deletion derivative of E. coli CP79 (Andersen and Douthwaite 2006). Deletion of the T. thermophilus rsmB and rsmF genes Constructs for inactivation of the T. thermophilus rsmB and rsmF genes were made by inserting the gene for a heat tolerant kanamycin (HTK) nucleotidyltransferase (Hoseki et al. 1999) into the methyltransferase parts of either pLJ102-RsmB or pLJ102- RsmF. The htk part of pUC18-htk (Hashimoto et al. 2001) was amplified by PCR with primers that introduced an upstream AvrII site and a downstream SacI site into the product for later disrup- tion of rsmB. For rsmF disruption, the PCR primers introduced SacI restriction enzyme sites both upstream of and downstream from the htk gene. These sites were used to insert the PCR prod- ucts into pLJ102-RsmB and pLJ102-RsmF to form the plasmids pLJ102-RsmBThtk and pLJ102-RsmFThtk, which were propa- gated in the E. coli strain Top10 (Invitrogen). T. thermophilus HB8 was transformed with pLJ102-RsmBThtk or pLJ102- RsmFThtk selecting for kanamycin resistance as described by others (Hashimoto et al. 2001; Cameron et al. 2004). Kanamycin- resistant transformants were restreaked twice. Gene disruptions were verified by PCR with primers distal to the interrupted rsmB or rsmF genes on genomic DNA; resulting PCR products were characterized by sequencing. Growth competition assays Wild-type and rsmF null mutant liquid cultures were grown at 70°C to saturation, then equal numbers of cells from each were mixed and incubated in 5 mL TEM medium at 60°C, 70°C, or 80°C. After growth for 24 h, 100 mL of the 60°C culture, 10 mL of the 70°C culture, and 1000 mL of the 80°C culture were trans- ferred to a fresh 5-mL medium and incubated at the respective temperatures for another 24 h. This was repeated in independent triplicates for seven cycles. Samples of 1 mL were collected at each dilution and half was plated on TEM plates without antibiotic and the other half was plated on TEM plates with 30 mg/mL kanamycin. The plates were incubated at 70°C. Purification of T. thermophilus ribosomal subunits and ribosomes T. thermophilus culture (1 L) was grown in TEM media (contain- ing 30 mg/mL of kanamycin when appropriate) with shaking at 70°C to an OD600 = 0.6. Cells were harvested and washed once with 100 mL of buffer A (10 mM NH4Cl, 20 mM MgCl2, 100 mM KCl, and 10 mM Tris-HCl [pH 7.5]), then suspended in 10 mL of buffer A, and disrupted by sonication. The lysate was cleared by centrifugation twice in a Beckman JA20 rotor at 16,000 rpm for 10 min at 4°C. Crude ribosomes were collected by centrifugation in a Beckman Ti50 rotor at 19,000 rpm for 19 h at 4°C, and dissolved in buffer A. 70S ribosomes were obtained by centrifu- gation of 100 A260 units of crude ribosomes through a 10%–40% sucrose gradient (200 mM NH4Cl, 20 mM MgCl2, 20 mM Tris- HCl [pH 7.5]) in a Beckman SW28 rotor at 20,000 rpm for 18 h at 4°C. Fractions containing intact 70S ribosomes were pooled and concentrated by centrifugation in a Beckman Ti50 rotor at 40,000 rpm for 22 h at 4°C, washed, and dissolved in buffer A, and stored at 80°C. 50S and 30S ribosomal subunits were obtained by adjusting 100 A260 units of crude ribosomes (10 mM NH4Cl, 2 mM MgCl2, 100 mM KCl, and 10 mM Tris-HCl [pH 7.5]) and passing through a 5%–20% sucrose gradient (200 mM NH4Cl, 1 mM MgCl2 and 20 mM Tris-HCl [pH 7.5]) in a Sorvall AH-629 rotor at 20,000 rpm for 18 h at 4°C. After pooling of the relevant fractions, the subunits were adjusted to 10 mM MgCl2 and pelleted by centrifugation in a Beckman Ti50 rotor at 40,000 rpm for 22 h at 4°C, washed with and dissolved in 10 mM NH4Cl, 10 mM MgCl2, 100 mM KCl, and 10 mM Tris-HCl (pH 7.5), and stored at 80°C. Isolation of 16S rRNA and subfragments from T. thermophilus and E. coli Water (400 mL) was added to 100 mL of 30S ribosomal subunits and the rRNA was extracted with 500 mL phenol, phenol/ chloroform, and chloroform. rRNA was ethanol precipitated and dissolved in water. Purification of 16S rRNA subfragments was performed as previously described (Andersen et al. 2004). Briefly, 16S rRNA was hybridized to an excess of oligodeoxynu- cleotide complementary to either the region 944–990 or the region 1378–1432. Single-stranded nucleic acids were digested with Mung Bean Nuclease and RNase A. The resulting mixture was separated on a polyacrylamide gel. Bands were visualized by ethid- ium bromide staining, excised, and eluted. E. coli CP79 with the endogenous rsmF inactivated, but com- plemented with the T. thermophilus homolog on the plasmid pLJ102-RsmF, were grown at 37°C to an OD450 = 0.45 in 200 mL of LB medium containing 100 mg/L of ampicillin. RsmF expres- sion was induced by addition of IPTG to 1 mM, and incubation for another 3 h. Cells were harvested by centrifugation at 4°C, washed in 100 mL TMN buffer (50 mM Tris-HCl [pH 7.8], 10 mM magnesium acetate, 100 mM NH4Cl), and resuspended in 2 mL TMN buffer prior to lysis by sonication (7 3 30 sec on ice) T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1593 and removal of debris by centrifugation (10 min/14,000 rpm/4°C/ microcentrifuge). Total RNA was recovered from the supernatant by phenol extraction and ethanol precipitation. A 16S rRNA subfragment was isolated as described above using an oligodeoxy- nucleotide complementary to the region 1378–1432. In vitro methylation Reactions contained 50 pmol of 16S rRNA, 30S subunits, or 70S ribosomes from the T. thermophilus TTHA1387 null mutant as the substrate in a total volume of 100 mL (containing 100 mM NH4Cl, 10 mM MgCl2, 40 mM Hepes [pH 7.5]), 6 mM b-mercaptoetha- nol, and 10% glycerol (prepared as a two times concentrated stock solution), 1.5 mM S-adenosyl methionine, and 2 mg of recombi- nantly expressed RsmF (see below). For the reaction at 70°C, water and stock buffer were mixed and left at room temperature for 15 min. Then a substrate, an enzyme and S-adenosyl methionine were added and incubated at 70°C for 1 h. The 37°C reaction was started by mixing water and buffer followed by 15 min at room temperature; the substrate was added and the mixture transferred to 50°C for 5 min. After cooling to 37°C, S-adenosyl methionine and an enzyme were added and the incubation continued for 1 h. Reactions were stopped by phenol/ chloroform extraction and the rRNA was recovered by ethanol precipitation before purification of 16S rRNA subfragments as described above. Control reactions without enzyme or S-adenosyl methionine were carried out in all instances. RNase T1 digestion and mass spectrometry A purified 16S rRNA subfragment (1–2 pmol) was incubated with 2 units RNase T1 (Roche) and 50 mM 3-hydroxypicollinic acid (3-HPA) in a total volume of 2 mL for 4 h at 37°C. MALDI mass spectrometry was performed either on an ABI voyager STR in- strument or a Waters Q-TOF MALDI instrument; MALDI tan- dem mass spectrometry was done on a Waters Q-TOF MALDI instrument. All spectra were recorded in positive ion mode using 3-HPA as the matrix. Experimental details were as previously de- scribed (Douthwaite and Kirpekar 2007). Protein expression and purification for crystallization E. coli BL21 (DE3) (Invitrogen) containing pLJ102-RsmF was grown to midlog phase in LB media at 37°C in the presence of 200 mg/mL ampicillin. Protein expression was induced at 20°C with 400 mM IPTG. Cells were pelleted after 18 h by centrifuga- tion at 4000 rpm for 20 min at 4°C and lysed by ultrasonication on ice in a buffer containing 20 mM Tris-HCl (pH 8.5), 300 mM NaCl, 5 mM b-mercaptoethanol, 0.1% Triton X-100, and 5% glycerol. Cell debris and membranes were pelleted by centrifuga- tion at 11,000 rpm for 30 min at 4°C. The soluble E. coli proteins were precipitated by heat treatment at 65°C for 30 min and pelleted by centrifugation at 11,000 rpm at 4°C for 30 min. Soluble C-terminally hexahistidine-tagged T. thermophilus RsmF was further purified by affinity chromatography with nickel- nitrilotriacetic acid resin (Qiagen). Untagged proteins were re- moved with buffer containing 20 mM Tris-HCl (pH 8.5), 250 mM NaCl, and 1 mM imidazole (pH 8.5). Recombinant RsmF was then eluted with the same buffer containing 150 mM imidazole. The protein was then purified by cation exchange chromatogra- phy (SP) (GE Healthcare) at pH 8.5, using a linear gradient of 10 mM to 1 M NaCl concentration. RsmF fractions were pooled and concentrated and applied to a size-exclusion S200 column (GE Healthcare) equilibrated with a buffer containing 20 mM Tris-HCl (pH 8.5) and 200 mM NaCl. Purified RsmF was concentrated to 13 mg/mL for crystallization trials. The C-terminal hexahistidine tag was not removed for crystallization. For the production of selenomethionyl proteins, the expression construct was trans- formed into B834 (DE3) cells (Novagen). The bacterial growth was carried out in defined LeMaster medium (Hendrickson et al. 1990), and the protein was purified using the same protocol as for the unmodified protein. To form the RsmF-AdoMet complex, purified RsmF was mixed with 4 mM AdoMet incubated at 60°C for 15 min and slowly cooled to room temperature before performing crystallization experiments. Crystallization of RsmF All crystals were obtained using the microbatch technique under oil at 4°C. To obtain the RsmF1 crystal form, 1 mL of protein solution was mixed with the reservoir solution containing 20% (w/v) PEG3350 and 200 mM sodium sulfate decahydrate (pH 6.6). Initial crystals grew over the course of 1–2 wk with maxi- mum dimensions of 0.3 3 0.3 3 0.2 mm. To obtain the RsmF2 crystal form, 1 mL of the RsmF–AdoMet complex was mixed with the reservoir solution containing 200 mM NaCl, 12% w/v PEG8000 and 100 mM HEPES-KOH (pH7.5). Initial crystals grew over the course of 2–3 wk with maximum dimensions of 0.1 3 0.4 3 0.4 mm. To obtain the RsmF3 crystal form, 1 mL of the RsmF–AdoMet complex was mixed with the reservoir solution containing 10% w/v PEG1000, 200 mM NaCl, and 100 mM Tris- HCl (pH 8.5). The initial crystals grew over the course of 1–2 wk with maximum dimensions of 0.05 3 0.3 3 0.4 mm. To obtain the RsmF4 crystal form, 1 mL of the RsmF–AdoMet complex solution was mixed with a reservoir solution containing 160 mM magnesium chloride hexahydrate, 80 mM Tris-HCl (pH 8.5), and 24% w/v PEG4000. Initial crystals grew over the course of 1–2 wk with maximum dimensions of 0.05 3 0.3 3 0.3 mm. RsmF1 crystals were gradually dehydrated by increasing the PEG3350 to 30% w/v and then cryoprotected in a mother liquor supplemented with 25% v/v glycerol and then flash-frozen by being plunged into liquid nitrogen. RsmF2 crystals were cryoprotected in a mother liquor supplemented with 20% v/v ethylene glycol and then flash- frozen by being plunged into liquid nitrogen. RsmF3 crystals were cryoprotected by gradually increasing the concentration of PEG1000 to 30% and then flash-frozen by being plunged into liquid nitrogen. RsmF4 crystals were cryoprotected in a mother liquor supplemented with 20% glycerol and then flash-frozen by being plunged into liquid nitrogen. Data collection X-ray diffraction data for RsmF1, RsmF2, and RsmF4 crystals were collected on a MAR CCD detector at the X4C beamline of the National Synchrotron Light Source in Brookhaven at a wave- length of 0.979 A˚ and 180°C. Diffraction data for RsmF3 crystals were collected on an ADSC CCD detector at the X4A beamline of the National Synchrotron Light Source in Brookhaven at a wave- length of 0.979 A˚ and 180°C. Diffraction data for RsmF1 in space group P43 were collected to 1.4 A˚ resolution with cell Demirci et al. 1594 RNA, Vol. 16, No. 8 dimensions a = 71.0 A˚ , b = 71.0 A˚ , and c = 186.7 A˚ . Diffraction data to 1.82 A˚ for RsmF2 were collected in space group P2 with cell dimensions a = 66.0 A˚ , b = 78.3 A˚ , and c = 108.1 A˚ . Diffraction data to 1.29 A˚ for RsmF3 were collected in space group P21212 with cell dimensions a = 89.7 A˚ , b = 109.0 A˚ , and c = 51.0 A˚ . Diffraction data to 1.68 A˚ for RsmF4 were collected in space group P21212 with cell dimensions a = 89.8 A˚ , b = 109.1 A˚ , and c = 50.8 A˚ . A single crystal was used for each data set. The diffraction images were processed and scaled with the HKL2000 package (Otwinowski and Minor 1997). The data processing statistics are summarized in Table 1. Structure determination and refinement The RsmF structure was solved by molecular replacement with the program Phaser (McCoy et al. 2007) from the CCP4 program suite (Bailey 1994) in space group P43 to 1.4 A˚ resolution (data set RsmF1). The initial search model was built with the program Modeller (Eswar et al. 2008) from the catalytic domain of E. coli YebU (Pdb code 2FRX). After the placement of two RsmF catalytic domains in the asymmetric unit and the initial re- finement with Refmac (Murshudov et al. 1997), the model was further rebuilt with ARP/wARP (Langer et al. 2008). The resulting model was 90% complete and manually checked and completed with Coot (Emsley and Cowtan 2004). Final crystallographic re- finement was performed with the program Phenix (Adams et al. 2002). The other crystal forms were subsequently solved by molecular replacement. The atomic coordinates from the RsmF4 model were then used for initial refinement of the RsmF–AdoMet complex structure in space group P21212 (RsmF3). There are two molecules in the asymmetric unit in data sets RsmF1 and RsmF2, and one molecule in RsmF3 and RsmF4. The crystallographic R/Rfree factors are 0.17/0.19, 0.16/0.19, 0.18/0.19, and 0.17/0.19 for the four data sets: RsmF1, RsmF2, RsmF3, and RsmF4, respectively. The stereochemical quality of the model was assessed with Procheck (Laskowski et al. 1993). The Ramachandran sta- tistics (most favored/additionally allowed/generously allowed/ disallowed) are 91.9%/8.1%/0.0%/0.0% for RsmF1, 91.9%/8.1%/ 0.0%/0.0% for RsmF2, 93.6%/6.4%/0.0%/0.0% for RsmF3, and 92.5%/7.5%/0.0%/0.0% for RsmF4. The refinement statistics are summarized in Table 1. Figures were generated using Pymol (DeLano 2002). Atomic coordinates Coordinates and structure factors have been deposited in the Protein Data Bank with accession codes 3M6U, 3M6V, 3M6W, and 3M6X for data sets RsmF1, RsmF2, RsmF3, and RsmF4, respectively. SUPPLEMENTAL MATERIAL Supplemental material can be found at http://www.rnajournal.org. ACKNOWLEDGMENTS We thank John Schwanof and Randy Abramowitz for access to the X4A and X4C beamlines at the National Synchrotron Light Source. This work was supported by grants GM19756 and GM19756-37S1 from the National Institutes of Health. Received January 14, 2010; accepted April 26, 2010. REFERENCES Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Terwilliger TC. 2002. PHENIX: Building new software for automated crystallo- graphic structure determination. Acta Crystallogr D Biol Crystallogr 58: 1948–1954. Agris PF. 2004. Decoding the genome: A modified view. Nucleic Acids Res 32: 223–238. Alian A, Lee TT, Griner SL, Stroud RM, Finer-Moore J. 2008. Structure of a TrmA–RNA complex: A consensus RNA fold con- tributes to substrate selectivity and catalysis in m5U methyltrans- ferases. Proc Natl Acad Sci 105: 6876–6881. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic local alignment search tool. J Mol Biol 215: 403–410. Andersen NM, Douthwaite S. 2006. YebU is a m5C methyltransferase specific for 16 S rRNA nucleotide 1407. J Mol Biol 359: 777–786. Andersen TE, Porse BT, Kirpekar F. 2004. A novel partial modifica- tion at C2501 in Escherichia coli 23S ribosomal RNA. RNA 10: 907–913. Bailey S. 1994. The CCP4 Suite: Programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 50: 760–763. Behm-Ansmant I, Urban A, Ma X, Yu YT, Motorin Y, Branlant C. 2003. The Saccharomyces cerevisiae U2 snRNA:pseudouridine- synthase Pus7p is a novel multisite-multisubstrate RNA:C- synthase also acting on tRNAs. RNA 9: 1371–1382. Behm-Ansmant I, Branlant C, Motorin Y. 2007. The Saccharomyces cerevisiae Pus2 protein encoded by YGL063w ORF is a mitochon- drial tRNA:C27/28-synthase. RNA 13: 1641–1647. Cameron DM, Gregory ST, Thompson J, Suh MJ, Limbach PA, Dahlberg AE. 2004. Thermus thermophilus L11 methyltransferase, PrmA, is dispensable for growth and preferentially modifies free ribosomal protein L11 prior to ribosome assembly. J Bacteriol 186: 5819–5825. DeLano WL. 2002. The PyMol molecular graphics system . DeLano Scientific, San Carlos, CA. Demirci H, Gregory ST, Dahlberg AE, Jogl G. 2007. Recognition of ribosomal protein L11 by the protein trimethyltransferase PrmA. EMBO J 26: 567–577. Demirci H, Gregory ST, Dahlberg AE, Jogl G. 2008a. Crystal structure of the Thermus thermophilus 16 S rRNA methyltransferase RsmC in complex with cofactor and substrate guanosine. J Biol Chem 283: 26548–26556. Demirci H, Gregory ST, Dahlberg AE, Jogl G. 2008b. Multiple-site trimethylation of ribosomal protein L11 by the PrmA methyl- transferase. Structure 16: 1059–1066. Demirci H, Belardinelli R, Seri E, Gregory ST, Gualerzi C, Dahlberg AE, Jogl G. 2009. Structural rearrangements in the active site of the Thermus thermophilus 16S rRNA methyltransferase KsgA in a bi- nary complex with 59-methylthioadenosine. J Mol Biol 388: 271– 282. Douthwaite S, Kirpekar F. 2007. Identifying modifications in RNA by MALDI mass spectrometry. Methods Enzymol 425: 1–20. Emsley P, Cowtan K. 2004. Coot: Model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132. Eswar N, Eramian D, Webb B, Shen MY, Sali A. 2008. Protein structure modeling with MODELLER. Methods Mol Biol 426: 145– 159. Foster PG, Nunes CR, Greene P, Moustakas D, Stroud RM. 2003. The first structure of an RNA m5C methyltransferase, Fmu, provides insight into catalytic mechanism and specific binding of RNA substrate. Structure 11: 1609–1620. Gao YG, Selmer M, Dunham CM, Weixlbaumer A, Kelley AC, Ramakrishnan V. 2009. The structure of the ribosome with elongation factor G trapped in the post-translocational state. Science 326: 694–699. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1595 Gustilo EM, Vendeix FA, Agris PF. 2008. tRNA’s modifications bring order to gene expression. Curr Opin Microbiol 11: 134–140. Gu XR, Gustafsson C, Ku J, Yu M, Santi DV. 1999. Identification of the 16S rRNA m5C967 methyltransferase from Escherichia coli. Biochemistry 38: 4053–4057. Guymon R, Pomerantz SC, Crain PF, McCloskey JA. 2006. Influence of phylogeny on posttranscriptional modification of rRNA in thermophilic prokaryotes: The complete modification map of 16S rRNA of Thermus thermophilus. Biochemistry 45: 4888–4899. Hallberg BM, Ericsson UB, Johnson KA, Andersen NM, Douthwaite S, Nordlund P, Beuscher AE 4th, Erlandsen H. 2006. The structure of the RNA m5C methyltransferase YebU from Escherichia coli reveals a C-terminal RNA-recruiting PUA domain. J Mol Biol 360: 774–787. Hashimoto Y, Yano T, Kuramitsua S, Kagamiyama H. 2001. Disrup- tion of Thermus thermophilus genes by homologous recombination using a thermostable kanamycin-resistant marker. FEBS Lett 506: 231–234. Helser TL, Davies JE, Dahlberg JE. 1972. Mechanism of kasugamycin resistance in Escherichia coli. Nat New Biol 235: 6–9. Hendrickson WA, Horton JR, LeMaster DM. 1990. Selenomethionyl proteins produced for analysis by multiwavelength anomalous diffraction (MAD): A vehicle for direct determination of three- dimensional structure. EMBO J 9: 1665–1672. Holm L, Ka¨a¨ria¨inen S, Rosenstro¨m P, Schenkel A. 2008. Searching protein structure databases with DaliLite v.3. Bioinformatics 24: 2780–2781. Hoseki J, Yano T, Koyama Y, Kuramitsu S, Kagamiyama H. 1999. Directed evolution of thermostable kanamycin-resistance gene: A convenient selection marker for Thermus thermophilus. J Biochem 126: 951–956. Hur S, Stroud RM. 2007. How U38, 39, and 40 of many tRNAs become the targets for pseudouridylation by TruA. Mol Cell 26: 189–203. Lang PT, Brozell SR, Mukherjee S, Pettersen EF, Meng EC, Thomas V, Rizzo RC, Case DA, James TL, Kuntz ID. 2009. DOCK 6: Combining techniques to model RNA-small molecule complexes. RNA 15: 1219–1230. Langer G, Cohen SX, Lamzin VS, Perrakis A. 2008. Automated macromolecular model building for X-ray crystallography using ARP/wARP version 7. Nat Protoc 3: 1171–1179. Laskowski RA, MacArthur MW, Moss DS, Thornton JM. 1993. PROCHECK: A program to check the stereochemical quality of protein structures. J Appl Crystallogr 26: 283–291. Lee TT, Agarwalla S, Stroud RM. 2005. A unique RNA fold in the RumA-RNA-cofactor ternary complex contributes to substrate selectivity and enzymatic function. Cell 120: 599–611. Liu Y, Santi DV. 2000. m5C RNA and m5C DNA methyl transferases use different cysteine residues as catalysts. Proc Natl Acad Sci 97: 8263–8265. Malone T, Blumenthal RM, Cheng X. 1995. Structure-guided analysis reveals nine sequence motifs conserved among DNA amino- methyltransferases, and suggests a catalytic mechanism for these enzymes. J Mol Biol 253: 618–632. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. 2007. Phaser crystallographic software. J Appl Crystallogr 40: 658–674. McLuckey SA, Van Berkel GJ, Glish GL. 1992. Tandem mass spectrometry of small multiply charged oligonucleotides. J Am Soc Mass Spectrom 3: 60–70. Motorin Y, Grosjean H. 1999. Multisite-specific tRNA:m5C-methyl- transferase (Trm4) in yeast Saccharomyces cerevisiae: Identification of the gene and substrate specificity of the enzyme. RNA 5: 1105– 1118. Motorin Y, Keith G, Simon C, Foiret D, Simos G, Hurt E, Grosjean H. 1998. The yeast tRNA:pseudouridine synthase Pus1p displays a multisite substrate specificity. RNA 4: 856–869. Murshudov GN, Vagin AA, Dodson EJ. 1997. Refinement of macro- molecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystallogr 53: 240–255. O’Farrell HC, Scarsdale JN, Rife JP. 2004. Crystal structure of KsgA, a universally conserved rRNA adenine dimethyltransferase in Escherichia coli. J Mol Biol 339: 337–353. Ogle JM, Murphy FV, Tarry MJ, Ramakrishnan V. 2002. Selection of tRNA by the ribosome requires a transition from an open to a closed form. Cell 111: 721–732. Otwinowski Z, Minor W. 1997. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol 276: 307–326. Pe´rez-Arellano I, Gallego J, Cervera J. 2007. The PUA domain—a structural and functional overview. FEBS J 274: 4972–4984. Purta E, O’Connor M, Bujnicki JM, Douthwaite S. 2008. YccW is the m5C methyltransferase specific for 23S rRNA nucleotide 1962. J Mol Biol 383: 641–651. Selmer M, Dunham CM, Murphy FV 4th, Weixlbaumer A, Petry S, Kelley AC, Weir JR, Ramakrishnan V. 2006. Structure of the 70S ribosome complexed with mRNA and tRNA. Science 313: 1935– 1942. Sunita S, Tkaczuk KL, Purta E, Kasprzak JM, Douthwaite S, Bujnicki JM, Sivaraman J. 2008. Crystal structure of the Escherichia coli 23S rRNA:m5C methyltransferase RlmI (YccW) reveals evolutionary links between RNA modification enzymes. J Mol Biol 383: 652–666. Wimberly BT, Brodersen DE, Clemons WM Jr, Morgan-Warren RJ, Carter AP, Vonrhein C, Hartsch T, Ramakrishnan V. 2000. Structure of the 30S ribosomal subunit. Nature 407: 327–339. Demirci et al. 1596 RNA, Vol. 16, No. 8
3M6V
Multi-site-specific 16S rRNA methyltransferase RsmF from Thermus thermophilus in space group P2 in complex with S-Adenosyl-L-Methionine
Multi-site-specific 16S rRNA methyltransferase RsmF from Thermus thermophilus HASAN DEMIRCI,1,3 LINE H.G. LARSEN,2,3 TRINE HANSEN,2 ANETTE RASMUSSEN,2 ASHWIN CADAMBI,1 STEVEN T. GREGORY,1 FINN KIRPEKAR,2 and GERWALD JOGL1 1Department of Molecular Biology, Cell Biology and Biochemistry, Brown University, Providence, Rhode Island 02912, USA 2Department of Biochemistry and Molecular Biology, University of Southern Denmark, 5230 Odense M, Denmark ABSTRACT Cells devote a significant effort toward the production of multiple modified nucleotides in rRNAs, which fine tune the ribosome function. Here, we report that two methyltransferases, RsmB and RsmF, are responsible for all four 5-methylcytidine (m5C) modifications in 16S rRNA of Thermus thermophilus. Like Escherichia coli RsmB, T. thermophilus RsmB produces m5C967. In contrast to E. coli RsmF, which introduces a single m5C1407 modification, T. thermophilus RsmF modifies three positions, generating m5C1400 and m5C1404 in addition to m5C1407. These three residues are clustered near the decoding site of the ribosome, but are situated in distinct structural contexts, suggesting a requirement for flexibility in the RsmF active site that is absent from the E. coli enzyme. Two of these residues, C1400 and C1404, are sufficiently buried in the mature ribosome structure so as to require extensive unfolding of the rRNA to be accessible to RsmF. In vitro, T. thermophilus RsmF methylates C1400, C1404, and C1407 in a 30S subunit substrate, but only C1400 and C1404 when naked 16S rRNA is the substrate. The multispecificity of T. thermophilus RsmF is potentially explained by three crystal structures of the enzyme in a complex with cofactor S-adenosyl-methionine at up to 1.3 A˚ resolution. In addition to confirming the overall structural similarity to E. coli RsmF, these structures also reveal that key segments in the active site are likely to be dynamic in solution, thereby expanding substrate recognition by T. thermophilus RsmF. Keywords: rRNA methyltransferase; 5-methylcytidine; RsmB; RsmF; RNA mass spectrometry INTRODUCTION Ribosomal RNAs (rRNAs) are post-transcriptionally mod- ified in all three domains of life, and many modifications are phylogenetically conserved. Most modifications are located in functionally important regions of the ribosome, where they probably act to fine tune protein synthesis (Agris 2004; Gustilo et al. 2008). Complete modification maps of bacterial 16S rRNAs have been determined for only a hand- ful of species, and among these are the enteric bacterium Escherichia coli and the extremely thermophilic bacterium Thermus thermophilus (Guymon et al. 2006). Despite the large phylogenetic divergence of these two organisms, their ribosome modification patterns are quite similar. Of the 11 E. coli and 14 T. thermophilus 16S rRNA modifications, eight are identical. This suggests a set of common functional requirements conserved since divergence from their last common ancestor, and also suggests common recognition mechanisms among their modifying enzymes. For most ribosome modifications, a single enzyme recog- nizes and modifies a single site. However, there exist nota- ble exceptions. Among these are dimethylation of two adja- cent adenosines in 16S rRNA by KsgA (Helser et al. 1972); pseudouridylation of three adjacent residues in tRNAs by TruA (Hur and Stroud 2007); pseudouridylation of several tRNA residues by Pus1 (Motorin et al. 1998), Pus2 (Behm- Ansmant et al. 2007), or Pus7 (Behm-Ansmant et al. 2003); or methylation of four tRNA positions by Saccharomyces cerevisiae Trm4 (Motorin and Grosjean 1999). Even with these multi-site-specific enzymes, however, homologs from various species generally modify the same residues. E. coli 16S rRNA contains two 5-methylcytidine (m5C) residues, located in or near the highly conserved decoding 3These authors contributed equally to this work. Abbreviations: rRNA, ribosomal RNA; AdoMet, S-adenosyl-L-methionine; m5C, 5-methyl-cytidine; MALDI mass spectrometry, matrix-assisted laser desorption ionization mass spectrometry. Reprint requests to: Gerwald Jogl, Department of Molecular Biology, Cell Biology and Biochemistry, Brown University, Box G-E129, Provi- dence, RI 02912, USA; e-mail: Gerwald_Jogl@brown.edu; fax: (401) 863-6114; or Finn Kirpekar, Department of Biochemistry and Molecular Biology, University of Southern Denmark, Campusvej 55, 5230 Odense M, Denmark; e-mail: f.kir@bmb.sdu.dk; fax: (+45) 65502467. Article published online ahead of print. Article and publication date are at http://www.rnajournal.org/cgi/doi/10.1261/rna.2088310. 1584 RNA (2010), 16:1584–1596. Published by Cold Spring Harbor Laboratory Press. Copyright  2010 RNA Society. center of the 30S subunit (Fig. 1). An m5C967 modification is produced by RsmB (also called Fmu), while an m5C1407 modification is produced by RsmF, formerly known as YebU (Andersen and Douthwaite 2006). T. thermophilus 16S rRNA contains m5C967 and m5C1407, as well as two additional m5C nucleotides, m5C1400 and m5C1404 (E. coli rRNA numbering used throughout) (Guymon et al. 2006). While the m5C967 and m5C1407 modifications are pre- sumably produced by RsmB and RsmF homologs, respec- tively, the source of the two additional m5C residues has been unknown. Here we demonstrate that T. thermophilus RsmF is a multi-site-specific methyltransferase and, in contrast to the single-site-specific E. coli RsmF, is respon- sible for the synthesis of three modifications: m5C1407, m5C1400, and m5C1404. We also demonstrate that RsmB is responsible for the synthesis of m5C967 in T. thermophilus as well as is in E. coli, thereby accounting for all four m5C modifications of 16S rRNA. We present crystal structures of T. thermophilus RsmF up to 1.3 A˚ resolution that reveal a dynamic region in the active site that is absent from the E. coli RsmF structure, providing a possible explanation for the expanded recognition capacity of the T. thermophilus methyltransferase. RESULTS Identification of T. thermophilus 16S rRNA m5C-methyltransferases With the E. coli RsmB and RsmF protein sequences as queries, we used conventional BLAST searches (Altschul et al. 1990) to identify potential homologs encoded by the T. thermophilus HB8 genome (data not shown). Both RsmB and RsmF have the highest similarity to the T. thermophilus protein encoded by TTHA1387 (BLAST scores of 106 and 190, respectively) and second-highest similarity to the pro- tein encoded by TTHA0851 (BLAST scores of 93 and 81, respectively). The simplest interpretation of these results is that TTHA1387 encodes RsmF, responsible for methylation of C1407, leaving TTHA0851 as the most likely candidate for the gene encoding RsmB, responsible for methylation of C967. The similarities of the two E. coli enzymes with other T. thermophilus proteins were far too low to reveal potential candidates responsible for methylation of C1400 and C1404. We next constructed T. thermophilus strains in which either TTHA0851 or TTHA1387 was inactivated by the homologous recombination and insertion of a heat stable kanamycin-resistance gene. 16S rRNA was isolated from these null mutants and subfragments of z50 nucleotides (nt) around the regions of interest were further purified, digested with RNase T1, and analyzed by MALDI mass spectrometry (Fig. 2). Comparison of the TTHA1387 null mutant to wild-type T. thermophilus HB8 indicates three clear differences, each corresponding to the disappearance of a methyl group (z14.0 Da). The RNase T1 digestion fragment harboring m5C1407, the nucleotide methylated by RsmF in E. coli, is absent in the null mutant, indicating that TTHA1387 is indeed rsmF. The predicted RNase T1 fragment reduced by 14.0 Da is obscured by another RNase T1 fragment that is present in both the wild-type and TTHA1387 null mutants (Fig. 2B). Unexpectedly, two addi- tional RNase T1 fragments are also reduced by 14.0 Da. One of these contains C1400 while the other contains C1404. This latter RNase T1 fragment from wild-type T. thermophilus contains three methyl groups, two on m4Cm1402 and one on m5C1404 (Guymon et al. 2006), preventing an unambiguous identification of the missing methyl group. We therefore performed tandem mass spec- trometry on the 1402CCCG1405 RNase T1 fragment with two methyl groups from the TTHA1387 null mutant and com- pared it with the triply methylated wild-type RNase T1 fragment (Fig. 2C). The clear w2 ions, as well as the less intense z3 ions, display a 14.0 Da mass difference between the two samples, showing that the methylations on m4Cm1402 were not affected by inactivation of TTHA1387. Tandem mass spectrometry was also performed on the RNase T1 fragments appearing as a consequence of the lack of methylations on C1400 and C1407 (data not shown). As expected, the C1400-containing fragment revealed no FIGURE 1. Secondary structure diagram of the 39 minor domain of 16S rRNA indicating the position of the three RsmF substrate nucleotides. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1585 indications of a methyl group, whereas the RNase T1 fragment with C1407 ex- hibited a fragmentation pattern corre- sponding to the expected mass overlap with an RNase T1 fragment of a different sequence. In summary, our data lead us to conclude that TTHA1387 encodes an RsmF m5C methyltransferase responsible for synthesizing m5C1400, m5C1404, and m5C1407 in 16S rRNA of T. thermophilus. The above results left us with TTHA0851 as the sole candidate for the gene encoding the m5C967 methyl- transferase. An approach conceptually identical to that described above re- vealed that disruption of TTHA0851 reduced the relevant RNase T1 fragment by 14.0 Da (Supplemental Fig. 1A). Since this fragment is methylated at G966 and C967 in the wild-type strain (Guymon et al. 2006), tandem mass spectrometry was again performed (Sup- plemental Fig. 1B), showing that only the methyl group on C967 was absent. In agreement with the suggested nomen- clature for rRNA modifying enzymes (Andersen and Douthwaite 2006), we hereafter refer to TTHA0851 as rsmB due to substrate specificity identical to the originally identified enzyme from E. coli (Gu et al. 1999). Effect of temperature on growth of the rsmF null mutant One possible explanation for methyla- tion of multiple sites by RsmF is that such additional methyl groups improve ribosomal function at elevated temper- ature. To address this possibility, we ex- amined the effect of temperature on growth of the rsmF null mutant. Wild- type T. thermophilus and the rsmF null mutant were cocultured at three tem- peratures and these cocultures were se- rially subcultured for seven cycles of 24 h each. The proportion of wild-type and rsmF null mutant cells in each mixed culture was determined by spreading di- lutions onto TEM plates with or without kanamycin. After seven cycles, no differ- ence in the relative proportions of the wild-type and rsmF mutant was exhib- ited at 70°C. However, at 60°C, the rsmF null mutant constituted only around 5% FIGURE 2. (Legend on next page) Demirci et al. 1586 RNA, Vol. 16, No. 8 of the population, and at 80°C the rsmF null mutant was unable to grow at all. Thus, methylation by RsmF appears to facilitate growth at temperatures outside the optimal growth temperature. RsmF substrate preference The T. thermophilus rsmB and rsmF genes were each cloned into an E. coli expression plasmid in order to produce proteins for X-ray crystallography and in vitro methylation studies. The expression constructs were equipped with C-terminal histidine6 tags to facilitate protein purification. While we achieved a high expression level of RsmF, we were unable to do so with RsmB despite a series of optimization attempts. Consequently, in vitro substrate and structure analyses were performed exclusively with RsmF. E. coli RsmF requires the 30S ribosomal subunit as a substrate when the activity is assayed in vitro (Andersen and Douthwaite 2006). We assayed 70S ribosomes, 30S ribosomal subunits, and 16S rRNA for the ability to serve as substrates for methylation by RsmF in vitro. 16S rRNA subfragments of z50 nt around the target sites were purified after the in vitro assay and analyzed by mass spectrometry as described above. In vitro methylation at 70°C showed an interesting but rather complex substrate pattern. RsmF completely methylates C1400 when either 16S rRNA or 30S subunits are used as a substrate. It methylates C1404 to z35% with 16S rRNA and completely with 30S subunits, and it produces only trace amounts of methylation of C1407 with 16S rRNA and z75% with 30S subunits (Fig. 3). There were no indications of the 70S ribosome being a substrate in vitro. Curiously, T. thermophilus RsmF expressed in an E. coli rsmF null mutant almost completely methylated, in vivo, positions C1400 and C1404, but not C1407 (data not shown). X-ray crystal structures of RsmF We determined the structure of T. thermophilus RsmF (456 amino acids) in three different crystal forms and in a com- plex with cofactor AdoMet to up to 1.3 A˚ resolution (Figs. 4, 5). The structure was solved in space group P43 (data set RsmF1, 1.4 A˚ resolution) by molecular replacement using a search model generated with the program Modeller (Eswar et al. 2008) from the catalytic domain of the RsmF homolog YebU from E. coli (Pdb 2FRX) (Hallberg et al. 2006). The structures of the AdoMet-bound form in space group P2 (RsmF2, 1.82 A˚ resolution), of the AdoMet- bound form (RsmF3, 1.3 A˚ resolution), and of the apo- form (RsmF4, 1.68 A˚ resolution) in space group P21212 were subsequently solved by molecular replacement with the refined RsmF1 model. There are two molecules in the asymmetric unit in space groups P43 and P2 and one mol- ecule in space group P21212. Electron density is generally well defined in all crystal forms. The majority of residues (92.0%, 92.3%, 93.1%, and 92.6%) are in the most favored region of the Ramachandran plot for RsmF1, RsmF2, RsmF3, and RsmF4, respectively, and there are no residues in the disallowed region. The final models consist of residues 5–178, 194–198, and 201–456 and five additional residues from the histidine6 affinity tag in both chains of data set RsmF1; residues 2–456 and five affinity-tag residues in both chains of data set RsmF2; residues 1–456 and six affinity-tag residues in data set RsmF3; and resi- dues 1–456 and seven affinity-tag residues in RsmF4. The N-terminal a-amino group was ordered in data sets RsmF3 and RsmF4 and contained additional electron density, which we interpreted as N-(dihydroxymethyl)-L-methio- nine, the hydrated form of N-formyl-methionine. Data collection and refinement statistics are given in Table 1. The overall structure of RsmF consists of a central canonical class I methyltransferase catalytic domain with additional N-terminal and C-terminal domains (Figs. 4, 5). The catalytic do- main is formed by a central seven- stranded b-sheet that is flanked on both sides by three helices of varying lengths. An inserted region between strand b7 and helix a11 contains additional heli- ces a9 and a10, which interact with the two N-terminal helices a1 and a2. A second inserted region following strand b9 includes the short helices a13 and FIGURE 2. (A) MALDI mass spectra of an RNase T1-digested 16S rRNA subfragment (pos. 1378–1432) from wild-type cells (upper panel) or from the TTHA1387 (putative rsmF) null mutant (lower panel). Expected digestion products are labeled; fragments affected by the null mutation are set in italics. (B) Expansion of the signals affected by the TTHA1387 null mu- tation. The sequence and methylation status of the RNase T1 products are indicated. (C) MALDI tandem mass spectrometry of the methylated RNase T1 fragment of 16S rRNA (pos. 1404–1407); wild-type cells (upper panel), TTHA1387 null mutant (lower panel). Mass spectrometric fragments used to deduce the methylation status are labeled. The position of the backbone fragments (nomenclature according to McLuckey et al. [1992]) in the sequence is shown. MH+, precursor ion selected for fragmentation; C, cytosine; mC, methylated cytosine; C>p, cytidine-2´-39-monophosphate; me, methyl group. FIGURE 3. In vitro methylation with 30S ribosomal subunits or 16S rRNA from the T. thermophilus rsmF null mutant as a substrate. Effect on C1400-, C1404-, and C1407-harboring RNase T1 products. In vitro methylated products are set in italics. *, artifact signal arising from the enzyme preparation. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1587 a14, which interact with helix a12. Furthermore, RsmF contains three additional smaller domains, an N-terminal domain consisting of a three-stranded b-sheet and two flank- ing helices (Fig. 5B, colored in blue), and two C-terminal domains consisting of four-stranded b-sheets and two or one helix (Fig. 5B, colored in magenta and red). Cofactor binding, substrate docking, and conformational flexibility in the active site The coordination of AdoMet in the T. thermophilus RsmF active site is similar to that seen in other class I methyl- transferases. However, the previously published structure of E. coli RsmF did not contain the cofactor AdoMet in the active site, precluding a direct comparison. Both the T. thermophilus and E. coli RsmF cofactor-binding sites reveal a new variation for the methyltransferase signature motif I (Malone et al. 1995), with the highly conserved GxGxG sequence replaced by 109AAAPG113. The combination of three alanines and a proline results in a loop conformation that is very similar to that observed in other methyltrans- ferases with a GxGxG motif (e.g., RsmC) (Demirci et al. 2008a). In RsmF, the amide hydrogen atom of the last glycine residue forms a hydrogen bond with the cofactor carboxy group (Fig. 6). Other key interactions with AdoMet are well conserved in RsmF. The cofactor adenine ring is located in a mainly hydrophobic pocket lined by residues Val134, Pro160, and Leu211. This pocket is open toward the solvent. The adenine amino group is not specifically recog- nized and interacts with solvent water molecules. The ribose hydroxyl groups form hydrogen bonds with Glu133 and Arg138, and the methionine amino group interacts with Asp177. The AdoMet cofactor is bound in a cleft in the RsmF active site, which suggests that substrate cytidine bases are inserted into the active site in an unstacked conforma- tion. Inspection of the electrostatic charge distribution re- veals a large positively charged surface region, which would be consistent with binding to an RNA surface and modifi- cation of the substrate base in an unstacked orientation (Fig. 6C). To evaluate the possible orientation of a substrate base in the active site, we performed computational docking calculations with the program Dock6 (Lang et al. 2009). The resulting positions of cytosine and m5C in the presence of AdoMet in data set RsmF3 are highly similar to each other, with m5C placed into the active site with its phosphate group toward a positively charged pocket at the entrance of the active site cleft (Fig. 6E). The position of the phosphate group is close to a sulfate molecule that we observed in data set RsmF1, providing further support for the results of the docking calculation (Fig. 6F). Interestingly, we observed that three active site segments were disordered in data set RsmF1. These segments include residues 179–193 (including helices a9 and a10 and the catalytic Cys180), residues 199–200, and the N-terminal residues 1–5, which interact with the first two segments (Fig. 6E,F, colored in green). We observed electron density for the intervening residues 194–198, which formed a lattice con- tact with a neighboring molecule. However, the position of FIGURE 4. Structure-based sequence alignment of RsmF from T. thermophilus and E. coli. Secondary structure elements of T. thermophilus RsmF are indicated on top. The color scheme for the secondary structure elements is as in Figure 5A. The position of the variant methyltransferase motif I is marked with a red box. A flexible region observed in the active site is marked with a green box; residues interacting with the cofactor are marked with orange boxes. Demirci et al. 1588 RNA, Vol. 16, No. 8 these five residues was not related to their position in the other three data sets, suggesting that the extended active site region between residues 179 and 201 can reorient in the RsmF structure. This observation suggests that this active site region is dynamic, which may be important for sub- strate binding at 72°C, the optimum growth temperature of T. thermophilus. DISCUSSION Substrate recognition mechanisms We have identified the two enzymes responsible for the synthesis of the four m5C modifications of T. thermophilus 16S rRNA, and characterized the RsmF methyltransferase responsible for synthesizing three of these. rRNA modifying enzymes in bacteria are generally highly specific, with a one-to-one association between the modifying enzyme and the modification. A few cases of multitarget ribosome mod- ifying enzymes have been reported (Helser et al. 1972; Demirci et al. 2008b), but to our knowledge T. thermophi- lus RsmF is the first rRNA methyltransferase found to mod- ify three different nucleotides. Most ribosome modifying enzymes probably recognize assembly intermediates, and the data presented here are consistent with that notion. T. thermophilus RsmF methylates C1400 and C1404 in vitro using either 16S rRNA or 30S subunits as substrates, whereas both E. coli (Andersen and Douthwaite 2006) and T. thermophilus RsmF exclusively utilize 30S subunits as substrates for methylation of C1407. This may reflect that C1400 and C1404 methylations do not rely on the asso- ciation of ribosomal proteins in order to be recognized by T. thermophilus RsmF. C1407 methylation, in contrast, depends on both rRNA and the ribosomal protein for the recognition by RsmF in both T. thermophilus and E. coli. More puzzling is the observation that T. thermophilus RsmF does not methylate E. coli ribosomes in vivo on C1407. It is perhaps worth noting that methylation of C1407 in the T. thermophilus 30S ribosomal subunit in vitro was less efficient than methylation of the other two positions, indicating the need for a particular intermediate assembly structure or for accessory factors. The only clear in vitro FIGURE 5. Overall structure of RsmF. (A) Schematic representation of the position of the substrate bases in the 30S ribosomal subunit (Pdb entry 2WRI) (Gao et al. 2009). C1400, C1404, and C1407 in helix 44 (green) are shown in stick representation with 5-methyl groups as pink spheres. (B) Schematic stereo representation of the overall structure of RsmF. Secondary structure elements are in orange and yellow for the catalytic domain, in salmon and blue for additional N-terminal domains, and in magenta and red for C-terminal domains. The flexible region in the active site including helices a9 and a10 is indicated with an arrow and colored in green. (C) Topology diagram with secondary structure elements colored as in B. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1589 methylation observed at 37°C with T. thermophilus RsmF was on C1400 with 16S rRNA as the substrate (data not shown), which is not evidently related to the methylation pattern in vivo in the heterologous system. It seems unlikely that the aberrant methylation in the heterologous system reflects species-specific differences in the mature 30S ribo- somal subunit, given the extreme sequence and structural conservation of the decoding site. Instead, it may reflect differences in 30S subunit assembly in the two organisms, necessary due to the large difference in growth temperature. The four m5C residues in T. thermophilus 16S rRNA are clustered in and around the functionally critical decoding center at or close to sites of contact with tRNA, mRNA, and EF-G (Ogle et al. 2002; Selmer et al. 2006; Gao et al. 2009). m5C1400, m5C1404, and m5C1407 are located in the subunit body, while m5C967 is located in the subunit head, about 10 A˚ from m5C1400 (Wimberly et al. 2000). Exam- ination of the 30S subunit crystal structure (Wimberly et al. 2000) indicates that the three bases methylated by RsmF are situated in three distinct structural contexts, but provides few clues to a common mode of substrate recognition. C1400 is an unpaired base protruding from a sharply bent segment of rRNA at the junction of helices 43 and 44, while C1404 and C1407 are engaged in Watson–Crick pairs within helix 44 (Fig. 4A). The C5 positions of the latter two bases are not obviously accessible, such that RsmF would need to approach them from the major groove side. While a flipping of C1407 out of the helix via the minor groove could allow access to this base, such a mechanism would be problematic for C1404, whose minor groove side is packed against the rest of the 30S subunit. While m5C1404 and m5C1407 are about 11 A˚ apart (Selmer et al. 2006), m5C1400 is quite distant from both of these bases (about 21 and 30 A˚ , respectively). The pro- truded conformation of m5C1400 in the mature 30S sub- unit is due in part to base-pairing interactions between adjacent bases and the 1500 region of 16S rRNA (C1399– G1504 and G1401–C1501). As the 1500 region is one of the last segments of 16S rRNA to be synthesized, C1400 could potentially be positioned much closer to C1404 and C1407 in an assembly intermediate, prior to the formation of the C1399–G1504 and G1401–C1501 base pairs. RsmF could utilize a single binding mode to then access all three bases, further facilitated by its flexible active site domain. Meth- ylation of C1400 in mature 30S subunits would therefore involve disruption of the adjacent base pairs. A complete understanding of the recognition mechanism of these enzymes will require high-resolution structural data on assembly intermediate-enzyme complexes. Given the large number of potential subunit assembly intermediates, pre- cisely defining the physiological substrate for RsmB and RsmF will be a formidable task. TABLE 1. Data collection and refinement statistics RsmF1 RsmF2 RsmF3 RsmF4 Data collectiona AdoMet AdoMet Space group P43 P2 P21212 P21212 Cell dimensions a, b, c (A˚ ) 71.0, 71.0, 186.7 66.0, 78.3, 108.1 89.7, 109.0, 51.0 89.8, 109.1, 50.8 a, b, g (°) 90, 90, 90 90, 107.1, 90 90, 90, 90 90, 90, 90 Resolution (A˚ )b 30–1.4 (1.55–1.40) 30–1.82 (1.89–1.82) 30–1.30 (1.34–1.30) 30–1.68 (1.74–1.68) Rmerge 0.065 (0.59) 0.08(0.38) 0.058(0.36) 0.15 (0.49) I/sI 29.3(2.15) 12.6 (2.04) 24.3 (2.05) 14.2 (1.73) Completeness (%) 90.1 (72.6) 97.0(86.5) 95.6(66.3) 99.6 (97.8) Redundancy 8.9 (5.4) 2.8(2.0) 5.2(2.1) 6.4 (3.9) Refinement Resolution (A˚ ) 30–1.4 (1.42–1.40) 30–1.82 (1.84–1.82) 30–1.30 (1.32–1.30) 30–1.68 (1.69–1.68) Number of reflections 161,955 (4356) 91,762 (2583) 119,490/2640 109,375 (3244) Rwork/Rfree 0.169/0.189 (0.227/0.263) 0.162/0.194 (0.222/0.259) 0.177/0.191 (0.233/0.234) 0.173/0.192 (0.217/0.266) Number of atoms Protein 6766 7117 3598 3574 Ligand/ion 20 54 27 1 Water 1486 1285 856 665 B-factors Protein 24.1 20.4 15.9 17.2 Ligand/ion 34.5 21.6 17.5 19.1 Water 38.7 36.3 33.6 33.6 RMSDs Bond lengths (A˚ ) 0.009 0.006 0.005 0.004 Bond angles (°) 1.22 1.04 1.14 0.94 aOne crystal used for each data set. bThe highest resolution shell is shown in parentheses. Demirci et al. 1590 RNA, Vol. 16, No. 8 Structural comparison of RsmF with related methyltransferases A database search with Dali (Holm et al. 2008) confirmed the structural similarity of the T. thermophilus and E. coli (PDB 2FRX) (Hallberg et al. 2006) RsmF homologs, which superimpose with a root-mean-square deviation (RMSD) of 1.6 A˚ for 342 Ca atoms (Fig. 7A) and are the only two structures in the Pro- tein Database with this domain organi- zation. Even so, substantial structural differences are observed in most of the loop regions and for the long connect- ing loop between the methyltransferase domain and the first C-terminal domain. While in the active site, the positions of residues in the cofactor-binding site and of the two cysteine residues are conserved (Fig. 7B), there are a number of positively charged residues (Arg30, Arg190, Arg194, His195, and Arg203) in the T. thermophilus structure that are absent from the E. coli enzyme (Fig. 7B). Three of these are located in the flexible region, and the combination of a posi- tive charge and flexibility close to the active site is suggestive of a functional contribution of this region to the mul- tisite specificity of T. thermophilus RsmF (Figs. 6C,D, 7B). Methylation of three rRNA positions may require an increase in the enzyme’s structural dynamics in order to accommodate the 30S subunit in slightly different orientations. Similar observations have been made for other multi-site-specific methyltransferases in- cluding KsgA, which modifies two adja- cent adenosines in the 30S ribosomal subunit (O’Farrell et al. 2004; Demirci et al. 2009), and the PrmA ribosomal protein methyltransferase, which under- goes dramatic interdomain movements to modify multiple lysine residues and the N-terminal a-amino group on the same substrate protein (Demirci et al. 2007, 2008b). The second C-terminal domain in RsmF is related to the RNA-binding PUA (pseudouridine synthase and archaeosine transglycosylase) domains (Perez-Arellano et al. 2007). The RlmI methyltransferase, which produces m5C1962 in 23S rRNA (Purta et al. 2008) also contains a PUA domain (Sunita et al. 2008), although it is N-terminal to the catalytic methyltransferase domain and in a different orientation. PUA domains contain six b-strands, which form a central pseudobarrel closed by a short 310-helix. A comparison of the C-terminal domain in RsmF with a typical PUA domain in archaeosine trans- glycosylase (ArcTGT, Pdb entry 1J2B) reveals that the central fold is similar (53 Ca atoms align with an RMSD FIGURE 6. Substrate docking and conformational flexibility. (A) Cofactor-binding site in RsmF. Bound AdoMet is shown in blue sticks. Hydrogen bonds to coordinating residues are indicated. (B) Final sA-weighted 2mFO-DFC electron density map of the cofactor-binding site (data set RsmF2) contoured at the 1s level. (C,D) Comparison of the electrostatic surface charge distribution between RsmF from T. thermophilus and E. coli. The location of the C-terminal domains and of the flexible region (labeled ‘‘FLEX’’) is indicated with circles. AdoMet and docked m5C are shown as sticks. AdoMet from the T. thermophilus structure is shown with the E. coli structure for comparison. (E) The modeled position of m5C docked into the active site of RsmF. m5C and AdoMet are shown in tan and blue sticks, respectively. Residues in the flexible region are shown in green. (F) The active site region in data set RsmF1. A sulfate ion is observed close to the position of the m5C phosphate group. Residues 194–198 in the flexible region interacting with a neighboring RsmF molecule are indicated with an arrow. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1591 of 1.8 A˚ ), but that several connecting loop regions are substantially shorter and two a-helices and one b-strand of the pseudobarrel are absent. Thus, this PUA-like domain differs considerably from typical PUA domains. However, the similarity to RNA-binding PUA domains and the positive surface charge distribution observed in both RsmF structures are suggestive for a conserved function of the PUA-like domain in RNA recognition. The next most closely related structure is E. coli RsmB (474 residues, PDB 1SQF) (Foster et al. 2003). A total of 264 Ca atoms can be aligned with an RMSD of 1.5 A˚ between RsmB and RsmF. Both enzymes retain the same organization for the N-terminal domains and the core methyltransferase domain. However, an additional 140- residue N-terminal RNA-binding domain provides sub- strate specificity to RsmB, whereas the two C-terminal domains following the core methyltransferase domain (160 residues) are likely to determine substrate recognition by RsmF. Thus, these two enzymes have evolved substrate specificity via acquisition of additional, unrelated RNA recognition domains. While there are as yet no enzyme– substrate complexes for rRNA m5C- methyltransferases, insights into the RsmF catalytic mechanism can be gleaned from a comparison with the RlmD and TrmA m5U methyltransferases in covalent in- termediate complexes with RNA oligonu- cleotides (Lee et al. 2005; Alian et al.2008). DNA and RNA m5C methyltransferases use a thiol from a catalytic cysteine residue to attack the six-position of the pyrimi- dine base to activate the five-position for methyl group transfer (Liu and Santi 2000). Cys180 and Cys230 in RsmF are positioned equivalently to the catalytic Cys324 and the catalytic base Glu358 of TrmA. The substrate nucleotides insert into TrmA and RsmF in unrelated di- rections, consistent with a lack of struc- tural homology outside the methyltrans- ferase domains. Nevertheless, the C5 positions of the pyrimidine rings and of the 5-methyl carbons are quite similar with respect to the catalytic cysteine resi- dues and the AdoMet cofactor. The same structural homology of the active site geometry can be observed in comparison with the RlmD methyltransferase (Sup- plemental Fig. 2; Lee et al. 2005). A possible origin of T. thermophilus RsmF T. thermophilus RsmF shows the highest similarities with proteins from close relatives, namely, Thermus aquaticus and two Meiothermus species of the Thermaceae family. Remarkably, the next highest similarities (BLAST scores between 267 and 359) are with the NOL1/ NOP2/Sun proteins from the Gram-positive Firmicutes phylum (Supplemental Fig. 3). This similarity, together with the fact that the Thermaceae family and most members of the Firmicutes identified in Supplemental Figure 3 are thermophilic, suggest that rsmF has undergone horizontal transfer between the Thermaceae family and members of the Firmicutes phylum. Thus, we speculate that this version of the RsmF protein, which catalyzes methylation of three cytidines, may be adaptive for existence in thermally challenging environments. The effect of the loss of methyl- ation by RsmF on growth at different temperatures is consistent with this notion. Our hypothesis of horizontal transfer of rsmF predicts that RsmF of other members of the Thermaceae family and members of the Firmicutes phylum will also be found to introduce multiple m5C modifications. FIGURE 7. Comparison with other methyltransferases. (A) Differences between the overall structures of RsmF from T. thermophilus (orange) and E. coli (green). AdoMet bound in the T. thermophilus structure is shown as blue sticks. (B) Comparison of the active site region in both enzymes. Residues in the cofactor-binding site and in the flexible region in the T. thermophilus enzyme are shown as sticks. (C) Comparison between RsmF (orange) and a substrate complex structure of the TrmA methyltransferase (cyan/blue). (D) Comparison of the active site region illustrating the differences in the insertion direction of the substrate base in RsmF (m5C in light orange) and in TrmA (m5U in cyan). Demirci et al. 1592 RNA, Vol. 16, No. 8 MATERIALS AND METHODS Cloning of the T. thermophilus rsmB and rsmF genes The T. thermophilus HB8 loci TTHA0851 (GenBank accession number BAD70674) and TTHA1387 (GenBank accession number BAD71210) were PCR amplified from genomic DNA and purified via the High Pure PCR Template Preparation Kit (Roche). The 100 mL PCRs contained 150 ng DNA, 10 mM of each primer, 10 mM dNTP, 1 unit Phusion DNA polymerase (Finnzymes), and 1x Phusion HF buffer. Primers for rsmB amplification were 59-CC CTGGACATATGAGGGCCGG-39 and 59-GGCCAAGATCTTGCC TGAGAG-39, and the temperature cycling was as follows: 98°C/30 sec; 30X (98°C/10 sec, 59°C/20 sec, and 72°C/36 sec); and 72°C/420s. Primers for rsmF amplification were 59-GCTAGGGTACACATA TGCTGCCC-39 and 59-GCACGGGGGTGAGATCTAAGCCC-39, and the temperature cycling was as follows: 98°C/30 sec; 30X (98°C/10 sec, 62°C /20 sec, 72°C/42 sec); and 72°C/420 sec. The desired PCR products were purified from agarose gels using the GFX PCR purification kit (GE Healthcare). The PCR fragments were digested with NdeI and BglII and inserted into the expression vector pLJ102 (Andersen and Douthwaite 2006), generating isopropyl-1-thio-b-D-galactopyranoside (IPTG)-inducible genes for the recombinant proteins with a C-terminal histidine6 tag. The constructs (designated pLJ102-RsmB and pLJ102-RsmF) were used to transform an rsmF-deletion derivative of E. coli CP79 (Andersen and Douthwaite 2006). Deletion of the T. thermophilus rsmB and rsmF genes Constructs for inactivation of the T. thermophilus rsmB and rsmF genes were made by inserting the gene for a heat tolerant kanamycin (HTK) nucleotidyltransferase (Hoseki et al. 1999) into the methyltransferase parts of either pLJ102-RsmB or pLJ102- RsmF. The htk part of pUC18-htk (Hashimoto et al. 2001) was amplified by PCR with primers that introduced an upstream AvrII site and a downstream SacI site into the product for later disrup- tion of rsmB. For rsmF disruption, the PCR primers introduced SacI restriction enzyme sites both upstream of and downstream from the htk gene. These sites were used to insert the PCR prod- ucts into pLJ102-RsmB and pLJ102-RsmF to form the plasmids pLJ102-RsmBThtk and pLJ102-RsmFThtk, which were propa- gated in the E. coli strain Top10 (Invitrogen). T. thermophilus HB8 was transformed with pLJ102-RsmBThtk or pLJ102- RsmFThtk selecting for kanamycin resistance as described by others (Hashimoto et al. 2001; Cameron et al. 2004). Kanamycin- resistant transformants were restreaked twice. Gene disruptions were verified by PCR with primers distal to the interrupted rsmB or rsmF genes on genomic DNA; resulting PCR products were characterized by sequencing. Growth competition assays Wild-type and rsmF null mutant liquid cultures were grown at 70°C to saturation, then equal numbers of cells from each were mixed and incubated in 5 mL TEM medium at 60°C, 70°C, or 80°C. After growth for 24 h, 100 mL of the 60°C culture, 10 mL of the 70°C culture, and 1000 mL of the 80°C culture were trans- ferred to a fresh 5-mL medium and incubated at the respective temperatures for another 24 h. This was repeated in independent triplicates for seven cycles. Samples of 1 mL were collected at each dilution and half was plated on TEM plates without antibiotic and the other half was plated on TEM plates with 30 mg/mL kanamycin. The plates were incubated at 70°C. Purification of T. thermophilus ribosomal subunits and ribosomes T. thermophilus culture (1 L) was grown in TEM media (contain- ing 30 mg/mL of kanamycin when appropriate) with shaking at 70°C to an OD600 = 0.6. Cells were harvested and washed once with 100 mL of buffer A (10 mM NH4Cl, 20 mM MgCl2, 100 mM KCl, and 10 mM Tris-HCl [pH 7.5]), then suspended in 10 mL of buffer A, and disrupted by sonication. The lysate was cleared by centrifugation twice in a Beckman JA20 rotor at 16,000 rpm for 10 min at 4°C. Crude ribosomes were collected by centrifugation in a Beckman Ti50 rotor at 19,000 rpm for 19 h at 4°C, and dissolved in buffer A. 70S ribosomes were obtained by centrifu- gation of 100 A260 units of crude ribosomes through a 10%–40% sucrose gradient (200 mM NH4Cl, 20 mM MgCl2, 20 mM Tris- HCl [pH 7.5]) in a Beckman SW28 rotor at 20,000 rpm for 18 h at 4°C. Fractions containing intact 70S ribosomes were pooled and concentrated by centrifugation in a Beckman Ti50 rotor at 40,000 rpm for 22 h at 4°C, washed, and dissolved in buffer A, and stored at 80°C. 50S and 30S ribosomal subunits were obtained by adjusting 100 A260 units of crude ribosomes (10 mM NH4Cl, 2 mM MgCl2, 100 mM KCl, and 10 mM Tris-HCl [pH 7.5]) and passing through a 5%–20% sucrose gradient (200 mM NH4Cl, 1 mM MgCl2 and 20 mM Tris-HCl [pH 7.5]) in a Sorvall AH-629 rotor at 20,000 rpm for 18 h at 4°C. After pooling of the relevant fractions, the subunits were adjusted to 10 mM MgCl2 and pelleted by centrifugation in a Beckman Ti50 rotor at 40,000 rpm for 22 h at 4°C, washed with and dissolved in 10 mM NH4Cl, 10 mM MgCl2, 100 mM KCl, and 10 mM Tris-HCl (pH 7.5), and stored at 80°C. Isolation of 16S rRNA and subfragments from T. thermophilus and E. coli Water (400 mL) was added to 100 mL of 30S ribosomal subunits and the rRNA was extracted with 500 mL phenol, phenol/ chloroform, and chloroform. rRNA was ethanol precipitated and dissolved in water. Purification of 16S rRNA subfragments was performed as previously described (Andersen et al. 2004). Briefly, 16S rRNA was hybridized to an excess of oligodeoxynu- cleotide complementary to either the region 944–990 or the region 1378–1432. Single-stranded nucleic acids were digested with Mung Bean Nuclease and RNase A. The resulting mixture was separated on a polyacrylamide gel. Bands were visualized by ethid- ium bromide staining, excised, and eluted. E. coli CP79 with the endogenous rsmF inactivated, but com- plemented with the T. thermophilus homolog on the plasmid pLJ102-RsmF, were grown at 37°C to an OD450 = 0.45 in 200 mL of LB medium containing 100 mg/L of ampicillin. RsmF expres- sion was induced by addition of IPTG to 1 mM, and incubation for another 3 h. Cells were harvested by centrifugation at 4°C, washed in 100 mL TMN buffer (50 mM Tris-HCl [pH 7.8], 10 mM magnesium acetate, 100 mM NH4Cl), and resuspended in 2 mL TMN buffer prior to lysis by sonication (7 3 30 sec on ice) T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1593 and removal of debris by centrifugation (10 min/14,000 rpm/4°C/ microcentrifuge). Total RNA was recovered from the supernatant by phenol extraction and ethanol precipitation. A 16S rRNA subfragment was isolated as described above using an oligodeoxy- nucleotide complementary to the region 1378–1432. In vitro methylation Reactions contained 50 pmol of 16S rRNA, 30S subunits, or 70S ribosomes from the T. thermophilus TTHA1387 null mutant as the substrate in a total volume of 100 mL (containing 100 mM NH4Cl, 10 mM MgCl2, 40 mM Hepes [pH 7.5]), 6 mM b-mercaptoetha- nol, and 10% glycerol (prepared as a two times concentrated stock solution), 1.5 mM S-adenosyl methionine, and 2 mg of recombi- nantly expressed RsmF (see below). For the reaction at 70°C, water and stock buffer were mixed and left at room temperature for 15 min. Then a substrate, an enzyme and S-adenosyl methionine were added and incubated at 70°C for 1 h. The 37°C reaction was started by mixing water and buffer followed by 15 min at room temperature; the substrate was added and the mixture transferred to 50°C for 5 min. After cooling to 37°C, S-adenosyl methionine and an enzyme were added and the incubation continued for 1 h. Reactions were stopped by phenol/ chloroform extraction and the rRNA was recovered by ethanol precipitation before purification of 16S rRNA subfragments as described above. Control reactions without enzyme or S-adenosyl methionine were carried out in all instances. RNase T1 digestion and mass spectrometry A purified 16S rRNA subfragment (1–2 pmol) was incubated with 2 units RNase T1 (Roche) and 50 mM 3-hydroxypicollinic acid (3-HPA) in a total volume of 2 mL for 4 h at 37°C. MALDI mass spectrometry was performed either on an ABI voyager STR in- strument or a Waters Q-TOF MALDI instrument; MALDI tan- dem mass spectrometry was done on a Waters Q-TOF MALDI instrument. All spectra were recorded in positive ion mode using 3-HPA as the matrix. Experimental details were as previously de- scribed (Douthwaite and Kirpekar 2007). Protein expression and purification for crystallization E. coli BL21 (DE3) (Invitrogen) containing pLJ102-RsmF was grown to midlog phase in LB media at 37°C in the presence of 200 mg/mL ampicillin. Protein expression was induced at 20°C with 400 mM IPTG. Cells were pelleted after 18 h by centrifuga- tion at 4000 rpm for 20 min at 4°C and lysed by ultrasonication on ice in a buffer containing 20 mM Tris-HCl (pH 8.5), 300 mM NaCl, 5 mM b-mercaptoethanol, 0.1% Triton X-100, and 5% glycerol. Cell debris and membranes were pelleted by centrifuga- tion at 11,000 rpm for 30 min at 4°C. The soluble E. coli proteins were precipitated by heat treatment at 65°C for 30 min and pelleted by centrifugation at 11,000 rpm at 4°C for 30 min. Soluble C-terminally hexahistidine-tagged T. thermophilus RsmF was further purified by affinity chromatography with nickel- nitrilotriacetic acid resin (Qiagen). Untagged proteins were re- moved with buffer containing 20 mM Tris-HCl (pH 8.5), 250 mM NaCl, and 1 mM imidazole (pH 8.5). Recombinant RsmF was then eluted with the same buffer containing 150 mM imidazole. The protein was then purified by cation exchange chromatogra- phy (SP) (GE Healthcare) at pH 8.5, using a linear gradient of 10 mM to 1 M NaCl concentration. RsmF fractions were pooled and concentrated and applied to a size-exclusion S200 column (GE Healthcare) equilibrated with a buffer containing 20 mM Tris-HCl (pH 8.5) and 200 mM NaCl. Purified RsmF was concentrated to 13 mg/mL for crystallization trials. The C-terminal hexahistidine tag was not removed for crystallization. For the production of selenomethionyl proteins, the expression construct was trans- formed into B834 (DE3) cells (Novagen). The bacterial growth was carried out in defined LeMaster medium (Hendrickson et al. 1990), and the protein was purified using the same protocol as for the unmodified protein. To form the RsmF-AdoMet complex, purified RsmF was mixed with 4 mM AdoMet incubated at 60°C for 15 min and slowly cooled to room temperature before performing crystallization experiments. Crystallization of RsmF All crystals were obtained using the microbatch technique under oil at 4°C. To obtain the RsmF1 crystal form, 1 mL of protein solution was mixed with the reservoir solution containing 20% (w/v) PEG3350 and 200 mM sodium sulfate decahydrate (pH 6.6). Initial crystals grew over the course of 1–2 wk with maxi- mum dimensions of 0.3 3 0.3 3 0.2 mm. To obtain the RsmF2 crystal form, 1 mL of the RsmF–AdoMet complex was mixed with the reservoir solution containing 200 mM NaCl, 12% w/v PEG8000 and 100 mM HEPES-KOH (pH7.5). Initial crystals grew over the course of 2–3 wk with maximum dimensions of 0.1 3 0.4 3 0.4 mm. To obtain the RsmF3 crystal form, 1 mL of the RsmF–AdoMet complex was mixed with the reservoir solution containing 10% w/v PEG1000, 200 mM NaCl, and 100 mM Tris- HCl (pH 8.5). The initial crystals grew over the course of 1–2 wk with maximum dimensions of 0.05 3 0.3 3 0.4 mm. To obtain the RsmF4 crystal form, 1 mL of the RsmF–AdoMet complex solution was mixed with a reservoir solution containing 160 mM magnesium chloride hexahydrate, 80 mM Tris-HCl (pH 8.5), and 24% w/v PEG4000. Initial crystals grew over the course of 1–2 wk with maximum dimensions of 0.05 3 0.3 3 0.3 mm. RsmF1 crystals were gradually dehydrated by increasing the PEG3350 to 30% w/v and then cryoprotected in a mother liquor supplemented with 25% v/v glycerol and then flash-frozen by being plunged into liquid nitrogen. RsmF2 crystals were cryoprotected in a mother liquor supplemented with 20% v/v ethylene glycol and then flash- frozen by being plunged into liquid nitrogen. RsmF3 crystals were cryoprotected by gradually increasing the concentration of PEG1000 to 30% and then flash-frozen by being plunged into liquid nitrogen. RsmF4 crystals were cryoprotected in a mother liquor supplemented with 20% glycerol and then flash-frozen by being plunged into liquid nitrogen. Data collection X-ray diffraction data for RsmF1, RsmF2, and RsmF4 crystals were collected on a MAR CCD detector at the X4C beamline of the National Synchrotron Light Source in Brookhaven at a wave- length of 0.979 A˚ and 180°C. Diffraction data for RsmF3 crystals were collected on an ADSC CCD detector at the X4A beamline of the National Synchrotron Light Source in Brookhaven at a wave- length of 0.979 A˚ and 180°C. Diffraction data for RsmF1 in space group P43 were collected to 1.4 A˚ resolution with cell Demirci et al. 1594 RNA, Vol. 16, No. 8 dimensions a = 71.0 A˚ , b = 71.0 A˚ , and c = 186.7 A˚ . Diffraction data to 1.82 A˚ for RsmF2 were collected in space group P2 with cell dimensions a = 66.0 A˚ , b = 78.3 A˚ , and c = 108.1 A˚ . Diffraction data to 1.29 A˚ for RsmF3 were collected in space group P21212 with cell dimensions a = 89.7 A˚ , b = 109.0 A˚ , and c = 51.0 A˚ . Diffraction data to 1.68 A˚ for RsmF4 were collected in space group P21212 with cell dimensions a = 89.8 A˚ , b = 109.1 A˚ , and c = 50.8 A˚ . A single crystal was used for each data set. The diffraction images were processed and scaled with the HKL2000 package (Otwinowski and Minor 1997). The data processing statistics are summarized in Table 1. Structure determination and refinement The RsmF structure was solved by molecular replacement with the program Phaser (McCoy et al. 2007) from the CCP4 program suite (Bailey 1994) in space group P43 to 1.4 A˚ resolution (data set RsmF1). The initial search model was built with the program Modeller (Eswar et al. 2008) from the catalytic domain of E. coli YebU (Pdb code 2FRX). After the placement of two RsmF catalytic domains in the asymmetric unit and the initial re- finement with Refmac (Murshudov et al. 1997), the model was further rebuilt with ARP/wARP (Langer et al. 2008). The resulting model was 90% complete and manually checked and completed with Coot (Emsley and Cowtan 2004). Final crystallographic re- finement was performed with the program Phenix (Adams et al. 2002). The other crystal forms were subsequently solved by molecular replacement. The atomic coordinates from the RsmF4 model were then used for initial refinement of the RsmF–AdoMet complex structure in space group P21212 (RsmF3). There are two molecules in the asymmetric unit in data sets RsmF1 and RsmF2, and one molecule in RsmF3 and RsmF4. The crystallographic R/Rfree factors are 0.17/0.19, 0.16/0.19, 0.18/0.19, and 0.17/0.19 for the four data sets: RsmF1, RsmF2, RsmF3, and RsmF4, respectively. The stereochemical quality of the model was assessed with Procheck (Laskowski et al. 1993). The Ramachandran sta- tistics (most favored/additionally allowed/generously allowed/ disallowed) are 91.9%/8.1%/0.0%/0.0% for RsmF1, 91.9%/8.1%/ 0.0%/0.0% for RsmF2, 93.6%/6.4%/0.0%/0.0% for RsmF3, and 92.5%/7.5%/0.0%/0.0% for RsmF4. The refinement statistics are summarized in Table 1. Figures were generated using Pymol (DeLano 2002). Atomic coordinates Coordinates and structure factors have been deposited in the Protein Data Bank with accession codes 3M6U, 3M6V, 3M6W, and 3M6X for data sets RsmF1, RsmF2, RsmF3, and RsmF4, respectively. SUPPLEMENTAL MATERIAL Supplemental material can be found at http://www.rnajournal.org. ACKNOWLEDGMENTS We thank John Schwanof and Randy Abramowitz for access to the X4A and X4C beamlines at the National Synchrotron Light Source. This work was supported by grants GM19756 and GM19756-37S1 from the National Institutes of Health. Received January 14, 2010; accepted April 26, 2010. REFERENCES Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Terwilliger TC. 2002. PHENIX: Building new software for automated crystallo- graphic structure determination. Acta Crystallogr D Biol Crystallogr 58: 1948–1954. Agris PF. 2004. Decoding the genome: A modified view. Nucleic Acids Res 32: 223–238. Alian A, Lee TT, Griner SL, Stroud RM, Finer-Moore J. 2008. Structure of a TrmA–RNA complex: A consensus RNA fold con- tributes to substrate selectivity and catalysis in m5U methyltrans- ferases. Proc Natl Acad Sci 105: 6876–6881. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic local alignment search tool. J Mol Biol 215: 403–410. Andersen NM, Douthwaite S. 2006. YebU is a m5C methyltransferase specific for 16 S rRNA nucleotide 1407. J Mol Biol 359: 777–786. Andersen TE, Porse BT, Kirpekar F. 2004. A novel partial modifica- tion at C2501 in Escherichia coli 23S ribosomal RNA. RNA 10: 907–913. Bailey S. 1994. The CCP4 Suite: Programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 50: 760–763. Behm-Ansmant I, Urban A, Ma X, Yu YT, Motorin Y, Branlant C. 2003. The Saccharomyces cerevisiae U2 snRNA:pseudouridine- synthase Pus7p is a novel multisite-multisubstrate RNA:C- synthase also acting on tRNAs. RNA 9: 1371–1382. Behm-Ansmant I, Branlant C, Motorin Y. 2007. The Saccharomyces cerevisiae Pus2 protein encoded by YGL063w ORF is a mitochon- drial tRNA:C27/28-synthase. RNA 13: 1641–1647. Cameron DM, Gregory ST, Thompson J, Suh MJ, Limbach PA, Dahlberg AE. 2004. Thermus thermophilus L11 methyltransferase, PrmA, is dispensable for growth and preferentially modifies free ribosomal protein L11 prior to ribosome assembly. J Bacteriol 186: 5819–5825. DeLano WL. 2002. The PyMol molecular graphics system . DeLano Scientific, San Carlos, CA. Demirci H, Gregory ST, Dahlberg AE, Jogl G. 2007. Recognition of ribosomal protein L11 by the protein trimethyltransferase PrmA. EMBO J 26: 567–577. Demirci H, Gregory ST, Dahlberg AE, Jogl G. 2008a. Crystal structure of the Thermus thermophilus 16 S rRNA methyltransferase RsmC in complex with cofactor and substrate guanosine. J Biol Chem 283: 26548–26556. Demirci H, Gregory ST, Dahlberg AE, Jogl G. 2008b. Multiple-site trimethylation of ribosomal protein L11 by the PrmA methyl- transferase. Structure 16: 1059–1066. Demirci H, Belardinelli R, Seri E, Gregory ST, Gualerzi C, Dahlberg AE, Jogl G. 2009. Structural rearrangements in the active site of the Thermus thermophilus 16S rRNA methyltransferase KsgA in a bi- nary complex with 59-methylthioadenosine. J Mol Biol 388: 271– 282. Douthwaite S, Kirpekar F. 2007. Identifying modifications in RNA by MALDI mass spectrometry. Methods Enzymol 425: 1–20. Emsley P, Cowtan K. 2004. Coot: Model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132. Eswar N, Eramian D, Webb B, Shen MY, Sali A. 2008. Protein structure modeling with MODELLER. Methods Mol Biol 426: 145– 159. Foster PG, Nunes CR, Greene P, Moustakas D, Stroud RM. 2003. The first structure of an RNA m5C methyltransferase, Fmu, provides insight into catalytic mechanism and specific binding of RNA substrate. Structure 11: 1609–1620. Gao YG, Selmer M, Dunham CM, Weixlbaumer A, Kelley AC, Ramakrishnan V. 2009. The structure of the ribosome with elongation factor G trapped in the post-translocational state. Science 326: 694–699. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1595 Gustilo EM, Vendeix FA, Agris PF. 2008. tRNA’s modifications bring order to gene expression. Curr Opin Microbiol 11: 134–140. Gu XR, Gustafsson C, Ku J, Yu M, Santi DV. 1999. Identification of the 16S rRNA m5C967 methyltransferase from Escherichia coli. Biochemistry 38: 4053–4057. Guymon R, Pomerantz SC, Crain PF, McCloskey JA. 2006. Influence of phylogeny on posttranscriptional modification of rRNA in thermophilic prokaryotes: The complete modification map of 16S rRNA of Thermus thermophilus. Biochemistry 45: 4888–4899. Hallberg BM, Ericsson UB, Johnson KA, Andersen NM, Douthwaite S, Nordlund P, Beuscher AE 4th, Erlandsen H. 2006. The structure of the RNA m5C methyltransferase YebU from Escherichia coli reveals a C-terminal RNA-recruiting PUA domain. J Mol Biol 360: 774–787. Hashimoto Y, Yano T, Kuramitsua S, Kagamiyama H. 2001. Disrup- tion of Thermus thermophilus genes by homologous recombination using a thermostable kanamycin-resistant marker. FEBS Lett 506: 231–234. Helser TL, Davies JE, Dahlberg JE. 1972. Mechanism of kasugamycin resistance in Escherichia coli. Nat New Biol 235: 6–9. Hendrickson WA, Horton JR, LeMaster DM. 1990. Selenomethionyl proteins produced for analysis by multiwavelength anomalous diffraction (MAD): A vehicle for direct determination of three- dimensional structure. EMBO J 9: 1665–1672. Holm L, Ka¨a¨ria¨inen S, Rosenstro¨m P, Schenkel A. 2008. Searching protein structure databases with DaliLite v.3. Bioinformatics 24: 2780–2781. Hoseki J, Yano T, Koyama Y, Kuramitsu S, Kagamiyama H. 1999. Directed evolution of thermostable kanamycin-resistance gene: A convenient selection marker for Thermus thermophilus. J Biochem 126: 951–956. Hur S, Stroud RM. 2007. How U38, 39, and 40 of many tRNAs become the targets for pseudouridylation by TruA. Mol Cell 26: 189–203. Lang PT, Brozell SR, Mukherjee S, Pettersen EF, Meng EC, Thomas V, Rizzo RC, Case DA, James TL, Kuntz ID. 2009. DOCK 6: Combining techniques to model RNA-small molecule complexes. RNA 15: 1219–1230. Langer G, Cohen SX, Lamzin VS, Perrakis A. 2008. Automated macromolecular model building for X-ray crystallography using ARP/wARP version 7. Nat Protoc 3: 1171–1179. Laskowski RA, MacArthur MW, Moss DS, Thornton JM. 1993. PROCHECK: A program to check the stereochemical quality of protein structures. J Appl Crystallogr 26: 283–291. Lee TT, Agarwalla S, Stroud RM. 2005. A unique RNA fold in the RumA-RNA-cofactor ternary complex contributes to substrate selectivity and enzymatic function. Cell 120: 599–611. Liu Y, Santi DV. 2000. m5C RNA and m5C DNA methyl transferases use different cysteine residues as catalysts. Proc Natl Acad Sci 97: 8263–8265. Malone T, Blumenthal RM, Cheng X. 1995. Structure-guided analysis reveals nine sequence motifs conserved among DNA amino- methyltransferases, and suggests a catalytic mechanism for these enzymes. J Mol Biol 253: 618–632. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. 2007. Phaser crystallographic software. J Appl Crystallogr 40: 658–674. McLuckey SA, Van Berkel GJ, Glish GL. 1992. Tandem mass spectrometry of small multiply charged oligonucleotides. J Am Soc Mass Spectrom 3: 60–70. Motorin Y, Grosjean H. 1999. Multisite-specific tRNA:m5C-methyl- transferase (Trm4) in yeast Saccharomyces cerevisiae: Identification of the gene and substrate specificity of the enzyme. RNA 5: 1105– 1118. Motorin Y, Keith G, Simon C, Foiret D, Simos G, Hurt E, Grosjean H. 1998. The yeast tRNA:pseudouridine synthase Pus1p displays a multisite substrate specificity. RNA 4: 856–869. Murshudov GN, Vagin AA, Dodson EJ. 1997. Refinement of macro- molecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystallogr 53: 240–255. O’Farrell HC, Scarsdale JN, Rife JP. 2004. Crystal structure of KsgA, a universally conserved rRNA adenine dimethyltransferase in Escherichia coli. J Mol Biol 339: 337–353. Ogle JM, Murphy FV, Tarry MJ, Ramakrishnan V. 2002. Selection of tRNA by the ribosome requires a transition from an open to a closed form. Cell 111: 721–732. Otwinowski Z, Minor W. 1997. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol 276: 307–326. Pe´rez-Arellano I, Gallego J, Cervera J. 2007. The PUA domain—a structural and functional overview. FEBS J 274: 4972–4984. Purta E, O’Connor M, Bujnicki JM, Douthwaite S. 2008. YccW is the m5C methyltransferase specific for 23S rRNA nucleotide 1962. J Mol Biol 383: 641–651. Selmer M, Dunham CM, Murphy FV 4th, Weixlbaumer A, Petry S, Kelley AC, Weir JR, Ramakrishnan V. 2006. Structure of the 70S ribosome complexed with mRNA and tRNA. Science 313: 1935– 1942. Sunita S, Tkaczuk KL, Purta E, Kasprzak JM, Douthwaite S, Bujnicki JM, Sivaraman J. 2008. Crystal structure of the Escherichia coli 23S rRNA:m5C methyltransferase RlmI (YccW) reveals evolutionary links between RNA modification enzymes. J Mol Biol 383: 652–666. Wimberly BT, Brodersen DE, Clemons WM Jr, Morgan-Warren RJ, Carter AP, Vonrhein C, Hartsch T, Ramakrishnan V. 2000. Structure of the 30S ribosomal subunit. Nature 407: 327–339. Demirci et al. 1596 RNA, Vol. 16, No. 8
3M6W
Multi-site-specific 16S rRNA methyltransferase RsmF from Thermus thermophilus in space group P21212 in complex with S-Adenosyl-L-Methionine
Multi-site-specific 16S rRNA methyltransferase RsmF from Thermus thermophilus HASAN DEMIRCI,1,3 LINE H.G. LARSEN,2,3 TRINE HANSEN,2 ANETTE RASMUSSEN,2 ASHWIN CADAMBI,1 STEVEN T. GREGORY,1 FINN KIRPEKAR,2 and GERWALD JOGL1 1Department of Molecular Biology, Cell Biology and Biochemistry, Brown University, Providence, Rhode Island 02912, USA 2Department of Biochemistry and Molecular Biology, University of Southern Denmark, 5230 Odense M, Denmark ABSTRACT Cells devote a significant effort toward the production of multiple modified nucleotides in rRNAs, which fine tune the ribosome function. Here, we report that two methyltransferases, RsmB and RsmF, are responsible for all four 5-methylcytidine (m5C) modifications in 16S rRNA of Thermus thermophilus. Like Escherichia coli RsmB, T. thermophilus RsmB produces m5C967. In contrast to E. coli RsmF, which introduces a single m5C1407 modification, T. thermophilus RsmF modifies three positions, generating m5C1400 and m5C1404 in addition to m5C1407. These three residues are clustered near the decoding site of the ribosome, but are situated in distinct structural contexts, suggesting a requirement for flexibility in the RsmF active site that is absent from the E. coli enzyme. Two of these residues, C1400 and C1404, are sufficiently buried in the mature ribosome structure so as to require extensive unfolding of the rRNA to be accessible to RsmF. In vitro, T. thermophilus RsmF methylates C1400, C1404, and C1407 in a 30S subunit substrate, but only C1400 and C1404 when naked 16S rRNA is the substrate. The multispecificity of T. thermophilus RsmF is potentially explained by three crystal structures of the enzyme in a complex with cofactor S-adenosyl-methionine at up to 1.3 A˚ resolution. In addition to confirming the overall structural similarity to E. coli RsmF, these structures also reveal that key segments in the active site are likely to be dynamic in solution, thereby expanding substrate recognition by T. thermophilus RsmF. Keywords: rRNA methyltransferase; 5-methylcytidine; RsmB; RsmF; RNA mass spectrometry INTRODUCTION Ribosomal RNAs (rRNAs) are post-transcriptionally mod- ified in all three domains of life, and many modifications are phylogenetically conserved. Most modifications are located in functionally important regions of the ribosome, where they probably act to fine tune protein synthesis (Agris 2004; Gustilo et al. 2008). Complete modification maps of bacterial 16S rRNAs have been determined for only a hand- ful of species, and among these are the enteric bacterium Escherichia coli and the extremely thermophilic bacterium Thermus thermophilus (Guymon et al. 2006). Despite the large phylogenetic divergence of these two organisms, their ribosome modification patterns are quite similar. Of the 11 E. coli and 14 T. thermophilus 16S rRNA modifications, eight are identical. This suggests a set of common functional requirements conserved since divergence from their last common ancestor, and also suggests common recognition mechanisms among their modifying enzymes. For most ribosome modifications, a single enzyme recog- nizes and modifies a single site. However, there exist nota- ble exceptions. Among these are dimethylation of two adja- cent adenosines in 16S rRNA by KsgA (Helser et al. 1972); pseudouridylation of three adjacent residues in tRNAs by TruA (Hur and Stroud 2007); pseudouridylation of several tRNA residues by Pus1 (Motorin et al. 1998), Pus2 (Behm- Ansmant et al. 2007), or Pus7 (Behm-Ansmant et al. 2003); or methylation of four tRNA positions by Saccharomyces cerevisiae Trm4 (Motorin and Grosjean 1999). Even with these multi-site-specific enzymes, however, homologs from various species generally modify the same residues. E. coli 16S rRNA contains two 5-methylcytidine (m5C) residues, located in or near the highly conserved decoding 3These authors contributed equally to this work. Abbreviations: rRNA, ribosomal RNA; AdoMet, S-adenosyl-L-methionine; m5C, 5-methyl-cytidine; MALDI mass spectrometry, matrix-assisted laser desorption ionization mass spectrometry. Reprint requests to: Gerwald Jogl, Department of Molecular Biology, Cell Biology and Biochemistry, Brown University, Box G-E129, Provi- dence, RI 02912, USA; e-mail: Gerwald_Jogl@brown.edu; fax: (401) 863-6114; or Finn Kirpekar, Department of Biochemistry and Molecular Biology, University of Southern Denmark, Campusvej 55, 5230 Odense M, Denmark; e-mail: f.kir@bmb.sdu.dk; fax: (+45) 65502467. Article published online ahead of print. Article and publication date are at http://www.rnajournal.org/cgi/doi/10.1261/rna.2088310. 1584 RNA (2010), 16:1584–1596. Published by Cold Spring Harbor Laboratory Press. Copyright  2010 RNA Society. center of the 30S subunit (Fig. 1). An m5C967 modification is produced by RsmB (also called Fmu), while an m5C1407 modification is produced by RsmF, formerly known as YebU (Andersen and Douthwaite 2006). T. thermophilus 16S rRNA contains m5C967 and m5C1407, as well as two additional m5C nucleotides, m5C1400 and m5C1404 (E. coli rRNA numbering used throughout) (Guymon et al. 2006). While the m5C967 and m5C1407 modifications are pre- sumably produced by RsmB and RsmF homologs, respec- tively, the source of the two additional m5C residues has been unknown. Here we demonstrate that T. thermophilus RsmF is a multi-site-specific methyltransferase and, in contrast to the single-site-specific E. coli RsmF, is respon- sible for the synthesis of three modifications: m5C1407, m5C1400, and m5C1404. We also demonstrate that RsmB is responsible for the synthesis of m5C967 in T. thermophilus as well as is in E. coli, thereby accounting for all four m5C modifications of 16S rRNA. We present crystal structures of T. thermophilus RsmF up to 1.3 A˚ resolution that reveal a dynamic region in the active site that is absent from the E. coli RsmF structure, providing a possible explanation for the expanded recognition capacity of the T. thermophilus methyltransferase. RESULTS Identification of T. thermophilus 16S rRNA m5C-methyltransferases With the E. coli RsmB and RsmF protein sequences as queries, we used conventional BLAST searches (Altschul et al. 1990) to identify potential homologs encoded by the T. thermophilus HB8 genome (data not shown). Both RsmB and RsmF have the highest similarity to the T. thermophilus protein encoded by TTHA1387 (BLAST scores of 106 and 190, respectively) and second-highest similarity to the pro- tein encoded by TTHA0851 (BLAST scores of 93 and 81, respectively). The simplest interpretation of these results is that TTHA1387 encodes RsmF, responsible for methylation of C1407, leaving TTHA0851 as the most likely candidate for the gene encoding RsmB, responsible for methylation of C967. The similarities of the two E. coli enzymes with other T. thermophilus proteins were far too low to reveal potential candidates responsible for methylation of C1400 and C1404. We next constructed T. thermophilus strains in which either TTHA0851 or TTHA1387 was inactivated by the homologous recombination and insertion of a heat stable kanamycin-resistance gene. 16S rRNA was isolated from these null mutants and subfragments of z50 nucleotides (nt) around the regions of interest were further purified, digested with RNase T1, and analyzed by MALDI mass spectrometry (Fig. 2). Comparison of the TTHA1387 null mutant to wild-type T. thermophilus HB8 indicates three clear differences, each corresponding to the disappearance of a methyl group (z14.0 Da). The RNase T1 digestion fragment harboring m5C1407, the nucleotide methylated by RsmF in E. coli, is absent in the null mutant, indicating that TTHA1387 is indeed rsmF. The predicted RNase T1 fragment reduced by 14.0 Da is obscured by another RNase T1 fragment that is present in both the wild-type and TTHA1387 null mutants (Fig. 2B). Unexpectedly, two addi- tional RNase T1 fragments are also reduced by 14.0 Da. One of these contains C1400 while the other contains C1404. This latter RNase T1 fragment from wild-type T. thermophilus contains three methyl groups, two on m4Cm1402 and one on m5C1404 (Guymon et al. 2006), preventing an unambiguous identification of the missing methyl group. We therefore performed tandem mass spec- trometry on the 1402CCCG1405 RNase T1 fragment with two methyl groups from the TTHA1387 null mutant and com- pared it with the triply methylated wild-type RNase T1 fragment (Fig. 2C). The clear w2 ions, as well as the less intense z3 ions, display a 14.0 Da mass difference between the two samples, showing that the methylations on m4Cm1402 were not affected by inactivation of TTHA1387. Tandem mass spectrometry was also performed on the RNase T1 fragments appearing as a consequence of the lack of methylations on C1400 and C1407 (data not shown). As expected, the C1400-containing fragment revealed no FIGURE 1. Secondary structure diagram of the 39 minor domain of 16S rRNA indicating the position of the three RsmF substrate nucleotides. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1585 indications of a methyl group, whereas the RNase T1 fragment with C1407 ex- hibited a fragmentation pattern corre- sponding to the expected mass overlap with an RNase T1 fragment of a different sequence. In summary, our data lead us to conclude that TTHA1387 encodes an RsmF m5C methyltransferase responsible for synthesizing m5C1400, m5C1404, and m5C1407 in 16S rRNA of T. thermophilus. The above results left us with TTHA0851 as the sole candidate for the gene encoding the m5C967 methyl- transferase. An approach conceptually identical to that described above re- vealed that disruption of TTHA0851 reduced the relevant RNase T1 fragment by 14.0 Da (Supplemental Fig. 1A). Since this fragment is methylated at G966 and C967 in the wild-type strain (Guymon et al. 2006), tandem mass spectrometry was again performed (Sup- plemental Fig. 1B), showing that only the methyl group on C967 was absent. In agreement with the suggested nomen- clature for rRNA modifying enzymes (Andersen and Douthwaite 2006), we hereafter refer to TTHA0851 as rsmB due to substrate specificity identical to the originally identified enzyme from E. coli (Gu et al. 1999). Effect of temperature on growth of the rsmF null mutant One possible explanation for methyla- tion of multiple sites by RsmF is that such additional methyl groups improve ribosomal function at elevated temper- ature. To address this possibility, we ex- amined the effect of temperature on growth of the rsmF null mutant. Wild- type T. thermophilus and the rsmF null mutant were cocultured at three tem- peratures and these cocultures were se- rially subcultured for seven cycles of 24 h each. The proportion of wild-type and rsmF null mutant cells in each mixed culture was determined by spreading di- lutions onto TEM plates with or without kanamycin. After seven cycles, no differ- ence in the relative proportions of the wild-type and rsmF mutant was exhib- ited at 70°C. However, at 60°C, the rsmF null mutant constituted only around 5% FIGURE 2. (Legend on next page) Demirci et al. 1586 RNA, Vol. 16, No. 8 of the population, and at 80°C the rsmF null mutant was unable to grow at all. Thus, methylation by RsmF appears to facilitate growth at temperatures outside the optimal growth temperature. RsmF substrate preference The T. thermophilus rsmB and rsmF genes were each cloned into an E. coli expression plasmid in order to produce proteins for X-ray crystallography and in vitro methylation studies. The expression constructs were equipped with C-terminal histidine6 tags to facilitate protein purification. While we achieved a high expression level of RsmF, we were unable to do so with RsmB despite a series of optimization attempts. Consequently, in vitro substrate and structure analyses were performed exclusively with RsmF. E. coli RsmF requires the 30S ribosomal subunit as a substrate when the activity is assayed in vitro (Andersen and Douthwaite 2006). We assayed 70S ribosomes, 30S ribosomal subunits, and 16S rRNA for the ability to serve as substrates for methylation by RsmF in vitro. 16S rRNA subfragments of z50 nt around the target sites were purified after the in vitro assay and analyzed by mass spectrometry as described above. In vitro methylation at 70°C showed an interesting but rather complex substrate pattern. RsmF completely methylates C1400 when either 16S rRNA or 30S subunits are used as a substrate. It methylates C1404 to z35% with 16S rRNA and completely with 30S subunits, and it produces only trace amounts of methylation of C1407 with 16S rRNA and z75% with 30S subunits (Fig. 3). There were no indications of the 70S ribosome being a substrate in vitro. Curiously, T. thermophilus RsmF expressed in an E. coli rsmF null mutant almost completely methylated, in vivo, positions C1400 and C1404, but not C1407 (data not shown). X-ray crystal structures of RsmF We determined the structure of T. thermophilus RsmF (456 amino acids) in three different crystal forms and in a com- plex with cofactor AdoMet to up to 1.3 A˚ resolution (Figs. 4, 5). The structure was solved in space group P43 (data set RsmF1, 1.4 A˚ resolution) by molecular replacement using a search model generated with the program Modeller (Eswar et al. 2008) from the catalytic domain of the RsmF homolog YebU from E. coli (Pdb 2FRX) (Hallberg et al. 2006). The structures of the AdoMet-bound form in space group P2 (RsmF2, 1.82 A˚ resolution), of the AdoMet- bound form (RsmF3, 1.3 A˚ resolution), and of the apo- form (RsmF4, 1.68 A˚ resolution) in space group P21212 were subsequently solved by molecular replacement with the refined RsmF1 model. There are two molecules in the asymmetric unit in space groups P43 and P2 and one mol- ecule in space group P21212. Electron density is generally well defined in all crystal forms. The majority of residues (92.0%, 92.3%, 93.1%, and 92.6%) are in the most favored region of the Ramachandran plot for RsmF1, RsmF2, RsmF3, and RsmF4, respectively, and there are no residues in the disallowed region. The final models consist of residues 5–178, 194–198, and 201–456 and five additional residues from the histidine6 affinity tag in both chains of data set RsmF1; residues 2–456 and five affinity-tag residues in both chains of data set RsmF2; residues 1–456 and six affinity-tag residues in data set RsmF3; and resi- dues 1–456 and seven affinity-tag residues in RsmF4. The N-terminal a-amino group was ordered in data sets RsmF3 and RsmF4 and contained additional electron density, which we interpreted as N-(dihydroxymethyl)-L-methio- nine, the hydrated form of N-formyl-methionine. Data collection and refinement statistics are given in Table 1. The overall structure of RsmF consists of a central canonical class I methyltransferase catalytic domain with additional N-terminal and C-terminal domains (Figs. 4, 5). The catalytic do- main is formed by a central seven- stranded b-sheet that is flanked on both sides by three helices of varying lengths. An inserted region between strand b7 and helix a11 contains additional heli- ces a9 and a10, which interact with the two N-terminal helices a1 and a2. A second inserted region following strand b9 includes the short helices a13 and FIGURE 2. (A) MALDI mass spectra of an RNase T1-digested 16S rRNA subfragment (pos. 1378–1432) from wild-type cells (upper panel) or from the TTHA1387 (putative rsmF) null mutant (lower panel). Expected digestion products are labeled; fragments affected by the null mutation are set in italics. (B) Expansion of the signals affected by the TTHA1387 null mu- tation. The sequence and methylation status of the RNase T1 products are indicated. (C) MALDI tandem mass spectrometry of the methylated RNase T1 fragment of 16S rRNA (pos. 1404–1407); wild-type cells (upper panel), TTHA1387 null mutant (lower panel). Mass spectrometric fragments used to deduce the methylation status are labeled. The position of the backbone fragments (nomenclature according to McLuckey et al. [1992]) in the sequence is shown. MH+, precursor ion selected for fragmentation; C, cytosine; mC, methylated cytosine; C>p, cytidine-2´-39-monophosphate; me, methyl group. FIGURE 3. In vitro methylation with 30S ribosomal subunits or 16S rRNA from the T. thermophilus rsmF null mutant as a substrate. Effect on C1400-, C1404-, and C1407-harboring RNase T1 products. In vitro methylated products are set in italics. *, artifact signal arising from the enzyme preparation. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1587 a14, which interact with helix a12. Furthermore, RsmF contains three additional smaller domains, an N-terminal domain consisting of a three-stranded b-sheet and two flank- ing helices (Fig. 5B, colored in blue), and two C-terminal domains consisting of four-stranded b-sheets and two or one helix (Fig. 5B, colored in magenta and red). Cofactor binding, substrate docking, and conformational flexibility in the active site The coordination of AdoMet in the T. thermophilus RsmF active site is similar to that seen in other class I methyl- transferases. However, the previously published structure of E. coli RsmF did not contain the cofactor AdoMet in the active site, precluding a direct comparison. Both the T. thermophilus and E. coli RsmF cofactor-binding sites reveal a new variation for the methyltransferase signature motif I (Malone et al. 1995), with the highly conserved GxGxG sequence replaced by 109AAAPG113. The combination of three alanines and a proline results in a loop conformation that is very similar to that observed in other methyltrans- ferases with a GxGxG motif (e.g., RsmC) (Demirci et al. 2008a). In RsmF, the amide hydrogen atom of the last glycine residue forms a hydrogen bond with the cofactor carboxy group (Fig. 6). Other key interactions with AdoMet are well conserved in RsmF. The cofactor adenine ring is located in a mainly hydrophobic pocket lined by residues Val134, Pro160, and Leu211. This pocket is open toward the solvent. The adenine amino group is not specifically recog- nized and interacts with solvent water molecules. The ribose hydroxyl groups form hydrogen bonds with Glu133 and Arg138, and the methionine amino group interacts with Asp177. The AdoMet cofactor is bound in a cleft in the RsmF active site, which suggests that substrate cytidine bases are inserted into the active site in an unstacked conforma- tion. Inspection of the electrostatic charge distribution re- veals a large positively charged surface region, which would be consistent with binding to an RNA surface and modifi- cation of the substrate base in an unstacked orientation (Fig. 6C). To evaluate the possible orientation of a substrate base in the active site, we performed computational docking calculations with the program Dock6 (Lang et al. 2009). The resulting positions of cytosine and m5C in the presence of AdoMet in data set RsmF3 are highly similar to each other, with m5C placed into the active site with its phosphate group toward a positively charged pocket at the entrance of the active site cleft (Fig. 6E). The position of the phosphate group is close to a sulfate molecule that we observed in data set RsmF1, providing further support for the results of the docking calculation (Fig. 6F). Interestingly, we observed that three active site segments were disordered in data set RsmF1. These segments include residues 179–193 (including helices a9 and a10 and the catalytic Cys180), residues 199–200, and the N-terminal residues 1–5, which interact with the first two segments (Fig. 6E,F, colored in green). We observed electron density for the intervening residues 194–198, which formed a lattice con- tact with a neighboring molecule. However, the position of FIGURE 4. Structure-based sequence alignment of RsmF from T. thermophilus and E. coli. Secondary structure elements of T. thermophilus RsmF are indicated on top. The color scheme for the secondary structure elements is as in Figure 5A. The position of the variant methyltransferase motif I is marked with a red box. A flexible region observed in the active site is marked with a green box; residues interacting with the cofactor are marked with orange boxes. Demirci et al. 1588 RNA, Vol. 16, No. 8 these five residues was not related to their position in the other three data sets, suggesting that the extended active site region between residues 179 and 201 can reorient in the RsmF structure. This observation suggests that this active site region is dynamic, which may be important for sub- strate binding at 72°C, the optimum growth temperature of T. thermophilus. DISCUSSION Substrate recognition mechanisms We have identified the two enzymes responsible for the synthesis of the four m5C modifications of T. thermophilus 16S rRNA, and characterized the RsmF methyltransferase responsible for synthesizing three of these. rRNA modifying enzymes in bacteria are generally highly specific, with a one-to-one association between the modifying enzyme and the modification. A few cases of multitarget ribosome mod- ifying enzymes have been reported (Helser et al. 1972; Demirci et al. 2008b), but to our knowledge T. thermophi- lus RsmF is the first rRNA methyltransferase found to mod- ify three different nucleotides. Most ribosome modifying enzymes probably recognize assembly intermediates, and the data presented here are consistent with that notion. T. thermophilus RsmF methylates C1400 and C1404 in vitro using either 16S rRNA or 30S subunits as substrates, whereas both E. coli (Andersen and Douthwaite 2006) and T. thermophilus RsmF exclusively utilize 30S subunits as substrates for methylation of C1407. This may reflect that C1400 and C1404 methylations do not rely on the asso- ciation of ribosomal proteins in order to be recognized by T. thermophilus RsmF. C1407 methylation, in contrast, depends on both rRNA and the ribosomal protein for the recognition by RsmF in both T. thermophilus and E. coli. More puzzling is the observation that T. thermophilus RsmF does not methylate E. coli ribosomes in vivo on C1407. It is perhaps worth noting that methylation of C1407 in the T. thermophilus 30S ribosomal subunit in vitro was less efficient than methylation of the other two positions, indicating the need for a particular intermediate assembly structure or for accessory factors. The only clear in vitro FIGURE 5. Overall structure of RsmF. (A) Schematic representation of the position of the substrate bases in the 30S ribosomal subunit (Pdb entry 2WRI) (Gao et al. 2009). C1400, C1404, and C1407 in helix 44 (green) are shown in stick representation with 5-methyl groups as pink spheres. (B) Schematic stereo representation of the overall structure of RsmF. Secondary structure elements are in orange and yellow for the catalytic domain, in salmon and blue for additional N-terminal domains, and in magenta and red for C-terminal domains. The flexible region in the active site including helices a9 and a10 is indicated with an arrow and colored in green. (C) Topology diagram with secondary structure elements colored as in B. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1589 methylation observed at 37°C with T. thermophilus RsmF was on C1400 with 16S rRNA as the substrate (data not shown), which is not evidently related to the methylation pattern in vivo in the heterologous system. It seems unlikely that the aberrant methylation in the heterologous system reflects species-specific differences in the mature 30S ribo- somal subunit, given the extreme sequence and structural conservation of the decoding site. Instead, it may reflect differences in 30S subunit assembly in the two organisms, necessary due to the large difference in growth temperature. The four m5C residues in T. thermophilus 16S rRNA are clustered in and around the functionally critical decoding center at or close to sites of contact with tRNA, mRNA, and EF-G (Ogle et al. 2002; Selmer et al. 2006; Gao et al. 2009). m5C1400, m5C1404, and m5C1407 are located in the subunit body, while m5C967 is located in the subunit head, about 10 A˚ from m5C1400 (Wimberly et al. 2000). Exam- ination of the 30S subunit crystal structure (Wimberly et al. 2000) indicates that the three bases methylated by RsmF are situated in three distinct structural contexts, but provides few clues to a common mode of substrate recognition. C1400 is an unpaired base protruding from a sharply bent segment of rRNA at the junction of helices 43 and 44, while C1404 and C1407 are engaged in Watson–Crick pairs within helix 44 (Fig. 4A). The C5 positions of the latter two bases are not obviously accessible, such that RsmF would need to approach them from the major groove side. While a flipping of C1407 out of the helix via the minor groove could allow access to this base, such a mechanism would be problematic for C1404, whose minor groove side is packed against the rest of the 30S subunit. While m5C1404 and m5C1407 are about 11 A˚ apart (Selmer et al. 2006), m5C1400 is quite distant from both of these bases (about 21 and 30 A˚ , respectively). The pro- truded conformation of m5C1400 in the mature 30S sub- unit is due in part to base-pairing interactions between adjacent bases and the 1500 region of 16S rRNA (C1399– G1504 and G1401–C1501). As the 1500 region is one of the last segments of 16S rRNA to be synthesized, C1400 could potentially be positioned much closer to C1404 and C1407 in an assembly intermediate, prior to the formation of the C1399–G1504 and G1401–C1501 base pairs. RsmF could utilize a single binding mode to then access all three bases, further facilitated by its flexible active site domain. Meth- ylation of C1400 in mature 30S subunits would therefore involve disruption of the adjacent base pairs. A complete understanding of the recognition mechanism of these enzymes will require high-resolution structural data on assembly intermediate-enzyme complexes. Given the large number of potential subunit assembly intermediates, pre- cisely defining the physiological substrate for RsmB and RsmF will be a formidable task. TABLE 1. Data collection and refinement statistics RsmF1 RsmF2 RsmF3 RsmF4 Data collectiona AdoMet AdoMet Space group P43 P2 P21212 P21212 Cell dimensions a, b, c (A˚ ) 71.0, 71.0, 186.7 66.0, 78.3, 108.1 89.7, 109.0, 51.0 89.8, 109.1, 50.8 a, b, g (°) 90, 90, 90 90, 107.1, 90 90, 90, 90 90, 90, 90 Resolution (A˚ )b 30–1.4 (1.55–1.40) 30–1.82 (1.89–1.82) 30–1.30 (1.34–1.30) 30–1.68 (1.74–1.68) Rmerge 0.065 (0.59) 0.08(0.38) 0.058(0.36) 0.15 (0.49) I/sI 29.3(2.15) 12.6 (2.04) 24.3 (2.05) 14.2 (1.73) Completeness (%) 90.1 (72.6) 97.0(86.5) 95.6(66.3) 99.6 (97.8) Redundancy 8.9 (5.4) 2.8(2.0) 5.2(2.1) 6.4 (3.9) Refinement Resolution (A˚ ) 30–1.4 (1.42–1.40) 30–1.82 (1.84–1.82) 30–1.30 (1.32–1.30) 30–1.68 (1.69–1.68) Number of reflections 161,955 (4356) 91,762 (2583) 119,490/2640 109,375 (3244) Rwork/Rfree 0.169/0.189 (0.227/0.263) 0.162/0.194 (0.222/0.259) 0.177/0.191 (0.233/0.234) 0.173/0.192 (0.217/0.266) Number of atoms Protein 6766 7117 3598 3574 Ligand/ion 20 54 27 1 Water 1486 1285 856 665 B-factors Protein 24.1 20.4 15.9 17.2 Ligand/ion 34.5 21.6 17.5 19.1 Water 38.7 36.3 33.6 33.6 RMSDs Bond lengths (A˚ ) 0.009 0.006 0.005 0.004 Bond angles (°) 1.22 1.04 1.14 0.94 aOne crystal used for each data set. bThe highest resolution shell is shown in parentheses. Demirci et al. 1590 RNA, Vol. 16, No. 8 Structural comparison of RsmF with related methyltransferases A database search with Dali (Holm et al. 2008) confirmed the structural similarity of the T. thermophilus and E. coli (PDB 2FRX) (Hallberg et al. 2006) RsmF homologs, which superimpose with a root-mean-square deviation (RMSD) of 1.6 A˚ for 342 Ca atoms (Fig. 7A) and are the only two structures in the Pro- tein Database with this domain organi- zation. Even so, substantial structural differences are observed in most of the loop regions and for the long connect- ing loop between the methyltransferase domain and the first C-terminal domain. While in the active site, the positions of residues in the cofactor-binding site and of the two cysteine residues are conserved (Fig. 7B), there are a number of positively charged residues (Arg30, Arg190, Arg194, His195, and Arg203) in the T. thermophilus structure that are absent from the E. coli enzyme (Fig. 7B). Three of these are located in the flexible region, and the combination of a posi- tive charge and flexibility close to the active site is suggestive of a functional contribution of this region to the mul- tisite specificity of T. thermophilus RsmF (Figs. 6C,D, 7B). Methylation of three rRNA positions may require an increase in the enzyme’s structural dynamics in order to accommodate the 30S subunit in slightly different orientations. Similar observations have been made for other multi-site-specific methyltransferases in- cluding KsgA, which modifies two adja- cent adenosines in the 30S ribosomal subunit (O’Farrell et al. 2004; Demirci et al. 2009), and the PrmA ribosomal protein methyltransferase, which under- goes dramatic interdomain movements to modify multiple lysine residues and the N-terminal a-amino group on the same substrate protein (Demirci et al. 2007, 2008b). The second C-terminal domain in RsmF is related to the RNA-binding PUA (pseudouridine synthase and archaeosine transglycosylase) domains (Perez-Arellano et al. 2007). The RlmI methyltransferase, which produces m5C1962 in 23S rRNA (Purta et al. 2008) also contains a PUA domain (Sunita et al. 2008), although it is N-terminal to the catalytic methyltransferase domain and in a different orientation. PUA domains contain six b-strands, which form a central pseudobarrel closed by a short 310-helix. A comparison of the C-terminal domain in RsmF with a typical PUA domain in archaeosine trans- glycosylase (ArcTGT, Pdb entry 1J2B) reveals that the central fold is similar (53 Ca atoms align with an RMSD FIGURE 6. Substrate docking and conformational flexibility. (A) Cofactor-binding site in RsmF. Bound AdoMet is shown in blue sticks. Hydrogen bonds to coordinating residues are indicated. (B) Final sA-weighted 2mFO-DFC electron density map of the cofactor-binding site (data set RsmF2) contoured at the 1s level. (C,D) Comparison of the electrostatic surface charge distribution between RsmF from T. thermophilus and E. coli. The location of the C-terminal domains and of the flexible region (labeled ‘‘FLEX’’) is indicated with circles. AdoMet and docked m5C are shown as sticks. AdoMet from the T. thermophilus structure is shown with the E. coli structure for comparison. (E) The modeled position of m5C docked into the active site of RsmF. m5C and AdoMet are shown in tan and blue sticks, respectively. Residues in the flexible region are shown in green. (F) The active site region in data set RsmF1. A sulfate ion is observed close to the position of the m5C phosphate group. Residues 194–198 in the flexible region interacting with a neighboring RsmF molecule are indicated with an arrow. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1591 of 1.8 A˚ ), but that several connecting loop regions are substantially shorter and two a-helices and one b-strand of the pseudobarrel are absent. Thus, this PUA-like domain differs considerably from typical PUA domains. However, the similarity to RNA-binding PUA domains and the positive surface charge distribution observed in both RsmF structures are suggestive for a conserved function of the PUA-like domain in RNA recognition. The next most closely related structure is E. coli RsmB (474 residues, PDB 1SQF) (Foster et al. 2003). A total of 264 Ca atoms can be aligned with an RMSD of 1.5 A˚ between RsmB and RsmF. Both enzymes retain the same organization for the N-terminal domains and the core methyltransferase domain. However, an additional 140- residue N-terminal RNA-binding domain provides sub- strate specificity to RsmB, whereas the two C-terminal domains following the core methyltransferase domain (160 residues) are likely to determine substrate recognition by RsmF. Thus, these two enzymes have evolved substrate specificity via acquisition of additional, unrelated RNA recognition domains. While there are as yet no enzyme– substrate complexes for rRNA m5C- methyltransferases, insights into the RsmF catalytic mechanism can be gleaned from a comparison with the RlmD and TrmA m5U methyltransferases in covalent in- termediate complexes with RNA oligonu- cleotides (Lee et al. 2005; Alian et al.2008). DNA and RNA m5C methyltransferases use a thiol from a catalytic cysteine residue to attack the six-position of the pyrimi- dine base to activate the five-position for methyl group transfer (Liu and Santi 2000). Cys180 and Cys230 in RsmF are positioned equivalently to the catalytic Cys324 and the catalytic base Glu358 of TrmA. The substrate nucleotides insert into TrmA and RsmF in unrelated di- rections, consistent with a lack of struc- tural homology outside the methyltrans- ferase domains. Nevertheless, the C5 positions of the pyrimidine rings and of the 5-methyl carbons are quite similar with respect to the catalytic cysteine resi- dues and the AdoMet cofactor. The same structural homology of the active site geometry can be observed in comparison with the RlmD methyltransferase (Sup- plemental Fig. 2; Lee et al. 2005). A possible origin of T. thermophilus RsmF T. thermophilus RsmF shows the highest similarities with proteins from close relatives, namely, Thermus aquaticus and two Meiothermus species of the Thermaceae family. Remarkably, the next highest similarities (BLAST scores between 267 and 359) are with the NOL1/ NOP2/Sun proteins from the Gram-positive Firmicutes phylum (Supplemental Fig. 3). This similarity, together with the fact that the Thermaceae family and most members of the Firmicutes identified in Supplemental Figure 3 are thermophilic, suggest that rsmF has undergone horizontal transfer between the Thermaceae family and members of the Firmicutes phylum. Thus, we speculate that this version of the RsmF protein, which catalyzes methylation of three cytidines, may be adaptive for existence in thermally challenging environments. The effect of the loss of methyl- ation by RsmF on growth at different temperatures is consistent with this notion. Our hypothesis of horizontal transfer of rsmF predicts that RsmF of other members of the Thermaceae family and members of the Firmicutes phylum will also be found to introduce multiple m5C modifications. FIGURE 7. Comparison with other methyltransferases. (A) Differences between the overall structures of RsmF from T. thermophilus (orange) and E. coli (green). AdoMet bound in the T. thermophilus structure is shown as blue sticks. (B) Comparison of the active site region in both enzymes. Residues in the cofactor-binding site and in the flexible region in the T. thermophilus enzyme are shown as sticks. (C) Comparison between RsmF (orange) and a substrate complex structure of the TrmA methyltransferase (cyan/blue). (D) Comparison of the active site region illustrating the differences in the insertion direction of the substrate base in RsmF (m5C in light orange) and in TrmA (m5U in cyan). Demirci et al. 1592 RNA, Vol. 16, No. 8 MATERIALS AND METHODS Cloning of the T. thermophilus rsmB and rsmF genes The T. thermophilus HB8 loci TTHA0851 (GenBank accession number BAD70674) and TTHA1387 (GenBank accession number BAD71210) were PCR amplified from genomic DNA and purified via the High Pure PCR Template Preparation Kit (Roche). The 100 mL PCRs contained 150 ng DNA, 10 mM of each primer, 10 mM dNTP, 1 unit Phusion DNA polymerase (Finnzymes), and 1x Phusion HF buffer. Primers for rsmB amplification were 59-CC CTGGACATATGAGGGCCGG-39 and 59-GGCCAAGATCTTGCC TGAGAG-39, and the temperature cycling was as follows: 98°C/30 sec; 30X (98°C/10 sec, 59°C/20 sec, and 72°C/36 sec); and 72°C/420s. Primers for rsmF amplification were 59-GCTAGGGTACACATA TGCTGCCC-39 and 59-GCACGGGGGTGAGATCTAAGCCC-39, and the temperature cycling was as follows: 98°C/30 sec; 30X (98°C/10 sec, 62°C /20 sec, 72°C/42 sec); and 72°C/420 sec. The desired PCR products were purified from agarose gels using the GFX PCR purification kit (GE Healthcare). The PCR fragments were digested with NdeI and BglII and inserted into the expression vector pLJ102 (Andersen and Douthwaite 2006), generating isopropyl-1-thio-b-D-galactopyranoside (IPTG)-inducible genes for the recombinant proteins with a C-terminal histidine6 tag. The constructs (designated pLJ102-RsmB and pLJ102-RsmF) were used to transform an rsmF-deletion derivative of E. coli CP79 (Andersen and Douthwaite 2006). Deletion of the T. thermophilus rsmB and rsmF genes Constructs for inactivation of the T. thermophilus rsmB and rsmF genes were made by inserting the gene for a heat tolerant kanamycin (HTK) nucleotidyltransferase (Hoseki et al. 1999) into the methyltransferase parts of either pLJ102-RsmB or pLJ102- RsmF. The htk part of pUC18-htk (Hashimoto et al. 2001) was amplified by PCR with primers that introduced an upstream AvrII site and a downstream SacI site into the product for later disrup- tion of rsmB. For rsmF disruption, the PCR primers introduced SacI restriction enzyme sites both upstream of and downstream from the htk gene. These sites were used to insert the PCR prod- ucts into pLJ102-RsmB and pLJ102-RsmF to form the plasmids pLJ102-RsmBThtk and pLJ102-RsmFThtk, which were propa- gated in the E. coli strain Top10 (Invitrogen). T. thermophilus HB8 was transformed with pLJ102-RsmBThtk or pLJ102- RsmFThtk selecting for kanamycin resistance as described by others (Hashimoto et al. 2001; Cameron et al. 2004). Kanamycin- resistant transformants were restreaked twice. Gene disruptions were verified by PCR with primers distal to the interrupted rsmB or rsmF genes on genomic DNA; resulting PCR products were characterized by sequencing. Growth competition assays Wild-type and rsmF null mutant liquid cultures were grown at 70°C to saturation, then equal numbers of cells from each were mixed and incubated in 5 mL TEM medium at 60°C, 70°C, or 80°C. After growth for 24 h, 100 mL of the 60°C culture, 10 mL of the 70°C culture, and 1000 mL of the 80°C culture were trans- ferred to a fresh 5-mL medium and incubated at the respective temperatures for another 24 h. This was repeated in independent triplicates for seven cycles. Samples of 1 mL were collected at each dilution and half was plated on TEM plates without antibiotic and the other half was plated on TEM plates with 30 mg/mL kanamycin. The plates were incubated at 70°C. Purification of T. thermophilus ribosomal subunits and ribosomes T. thermophilus culture (1 L) was grown in TEM media (contain- ing 30 mg/mL of kanamycin when appropriate) with shaking at 70°C to an OD600 = 0.6. Cells were harvested and washed once with 100 mL of buffer A (10 mM NH4Cl, 20 mM MgCl2, 100 mM KCl, and 10 mM Tris-HCl [pH 7.5]), then suspended in 10 mL of buffer A, and disrupted by sonication. The lysate was cleared by centrifugation twice in a Beckman JA20 rotor at 16,000 rpm for 10 min at 4°C. Crude ribosomes were collected by centrifugation in a Beckman Ti50 rotor at 19,000 rpm for 19 h at 4°C, and dissolved in buffer A. 70S ribosomes were obtained by centrifu- gation of 100 A260 units of crude ribosomes through a 10%–40% sucrose gradient (200 mM NH4Cl, 20 mM MgCl2, 20 mM Tris- HCl [pH 7.5]) in a Beckman SW28 rotor at 20,000 rpm for 18 h at 4°C. Fractions containing intact 70S ribosomes were pooled and concentrated by centrifugation in a Beckman Ti50 rotor at 40,000 rpm for 22 h at 4°C, washed, and dissolved in buffer A, and stored at 80°C. 50S and 30S ribosomal subunits were obtained by adjusting 100 A260 units of crude ribosomes (10 mM NH4Cl, 2 mM MgCl2, 100 mM KCl, and 10 mM Tris-HCl [pH 7.5]) and passing through a 5%–20% sucrose gradient (200 mM NH4Cl, 1 mM MgCl2 and 20 mM Tris-HCl [pH 7.5]) in a Sorvall AH-629 rotor at 20,000 rpm for 18 h at 4°C. After pooling of the relevant fractions, the subunits were adjusted to 10 mM MgCl2 and pelleted by centrifugation in a Beckman Ti50 rotor at 40,000 rpm for 22 h at 4°C, washed with and dissolved in 10 mM NH4Cl, 10 mM MgCl2, 100 mM KCl, and 10 mM Tris-HCl (pH 7.5), and stored at 80°C. Isolation of 16S rRNA and subfragments from T. thermophilus and E. coli Water (400 mL) was added to 100 mL of 30S ribosomal subunits and the rRNA was extracted with 500 mL phenol, phenol/ chloroform, and chloroform. rRNA was ethanol precipitated and dissolved in water. Purification of 16S rRNA subfragments was performed as previously described (Andersen et al. 2004). Briefly, 16S rRNA was hybridized to an excess of oligodeoxynu- cleotide complementary to either the region 944–990 or the region 1378–1432. Single-stranded nucleic acids were digested with Mung Bean Nuclease and RNase A. The resulting mixture was separated on a polyacrylamide gel. Bands were visualized by ethid- ium bromide staining, excised, and eluted. E. coli CP79 with the endogenous rsmF inactivated, but com- plemented with the T. thermophilus homolog on the plasmid pLJ102-RsmF, were grown at 37°C to an OD450 = 0.45 in 200 mL of LB medium containing 100 mg/L of ampicillin. RsmF expres- sion was induced by addition of IPTG to 1 mM, and incubation for another 3 h. Cells were harvested by centrifugation at 4°C, washed in 100 mL TMN buffer (50 mM Tris-HCl [pH 7.8], 10 mM magnesium acetate, 100 mM NH4Cl), and resuspended in 2 mL TMN buffer prior to lysis by sonication (7 3 30 sec on ice) T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1593 and removal of debris by centrifugation (10 min/14,000 rpm/4°C/ microcentrifuge). Total RNA was recovered from the supernatant by phenol extraction and ethanol precipitation. A 16S rRNA subfragment was isolated as described above using an oligodeoxy- nucleotide complementary to the region 1378–1432. In vitro methylation Reactions contained 50 pmol of 16S rRNA, 30S subunits, or 70S ribosomes from the T. thermophilus TTHA1387 null mutant as the substrate in a total volume of 100 mL (containing 100 mM NH4Cl, 10 mM MgCl2, 40 mM Hepes [pH 7.5]), 6 mM b-mercaptoetha- nol, and 10% glycerol (prepared as a two times concentrated stock solution), 1.5 mM S-adenosyl methionine, and 2 mg of recombi- nantly expressed RsmF (see below). For the reaction at 70°C, water and stock buffer were mixed and left at room temperature for 15 min. Then a substrate, an enzyme and S-adenosyl methionine were added and incubated at 70°C for 1 h. The 37°C reaction was started by mixing water and buffer followed by 15 min at room temperature; the substrate was added and the mixture transferred to 50°C for 5 min. After cooling to 37°C, S-adenosyl methionine and an enzyme were added and the incubation continued for 1 h. Reactions were stopped by phenol/ chloroform extraction and the rRNA was recovered by ethanol precipitation before purification of 16S rRNA subfragments as described above. Control reactions without enzyme or S-adenosyl methionine were carried out in all instances. RNase T1 digestion and mass spectrometry A purified 16S rRNA subfragment (1–2 pmol) was incubated with 2 units RNase T1 (Roche) and 50 mM 3-hydroxypicollinic acid (3-HPA) in a total volume of 2 mL for 4 h at 37°C. MALDI mass spectrometry was performed either on an ABI voyager STR in- strument or a Waters Q-TOF MALDI instrument; MALDI tan- dem mass spectrometry was done on a Waters Q-TOF MALDI instrument. All spectra were recorded in positive ion mode using 3-HPA as the matrix. Experimental details were as previously de- scribed (Douthwaite and Kirpekar 2007). Protein expression and purification for crystallization E. coli BL21 (DE3) (Invitrogen) containing pLJ102-RsmF was grown to midlog phase in LB media at 37°C in the presence of 200 mg/mL ampicillin. Protein expression was induced at 20°C with 400 mM IPTG. Cells were pelleted after 18 h by centrifuga- tion at 4000 rpm for 20 min at 4°C and lysed by ultrasonication on ice in a buffer containing 20 mM Tris-HCl (pH 8.5), 300 mM NaCl, 5 mM b-mercaptoethanol, 0.1% Triton X-100, and 5% glycerol. Cell debris and membranes were pelleted by centrifuga- tion at 11,000 rpm for 30 min at 4°C. The soluble E. coli proteins were precipitated by heat treatment at 65°C for 30 min and pelleted by centrifugation at 11,000 rpm at 4°C for 30 min. Soluble C-terminally hexahistidine-tagged T. thermophilus RsmF was further purified by affinity chromatography with nickel- nitrilotriacetic acid resin (Qiagen). Untagged proteins were re- moved with buffer containing 20 mM Tris-HCl (pH 8.5), 250 mM NaCl, and 1 mM imidazole (pH 8.5). Recombinant RsmF was then eluted with the same buffer containing 150 mM imidazole. The protein was then purified by cation exchange chromatogra- phy (SP) (GE Healthcare) at pH 8.5, using a linear gradient of 10 mM to 1 M NaCl concentration. RsmF fractions were pooled and concentrated and applied to a size-exclusion S200 column (GE Healthcare) equilibrated with a buffer containing 20 mM Tris-HCl (pH 8.5) and 200 mM NaCl. Purified RsmF was concentrated to 13 mg/mL for crystallization trials. The C-terminal hexahistidine tag was not removed for crystallization. For the production of selenomethionyl proteins, the expression construct was trans- formed into B834 (DE3) cells (Novagen). The bacterial growth was carried out in defined LeMaster medium (Hendrickson et al. 1990), and the protein was purified using the same protocol as for the unmodified protein. To form the RsmF-AdoMet complex, purified RsmF was mixed with 4 mM AdoMet incubated at 60°C for 15 min and slowly cooled to room temperature before performing crystallization experiments. Crystallization of RsmF All crystals were obtained using the microbatch technique under oil at 4°C. To obtain the RsmF1 crystal form, 1 mL of protein solution was mixed with the reservoir solution containing 20% (w/v) PEG3350 and 200 mM sodium sulfate decahydrate (pH 6.6). Initial crystals grew over the course of 1–2 wk with maxi- mum dimensions of 0.3 3 0.3 3 0.2 mm. To obtain the RsmF2 crystal form, 1 mL of the RsmF–AdoMet complex was mixed with the reservoir solution containing 200 mM NaCl, 12% w/v PEG8000 and 100 mM HEPES-KOH (pH7.5). Initial crystals grew over the course of 2–3 wk with maximum dimensions of 0.1 3 0.4 3 0.4 mm. To obtain the RsmF3 crystal form, 1 mL of the RsmF–AdoMet complex was mixed with the reservoir solution containing 10% w/v PEG1000, 200 mM NaCl, and 100 mM Tris- HCl (pH 8.5). The initial crystals grew over the course of 1–2 wk with maximum dimensions of 0.05 3 0.3 3 0.4 mm. To obtain the RsmF4 crystal form, 1 mL of the RsmF–AdoMet complex solution was mixed with a reservoir solution containing 160 mM magnesium chloride hexahydrate, 80 mM Tris-HCl (pH 8.5), and 24% w/v PEG4000. Initial crystals grew over the course of 1–2 wk with maximum dimensions of 0.05 3 0.3 3 0.3 mm. RsmF1 crystals were gradually dehydrated by increasing the PEG3350 to 30% w/v and then cryoprotected in a mother liquor supplemented with 25% v/v glycerol and then flash-frozen by being plunged into liquid nitrogen. RsmF2 crystals were cryoprotected in a mother liquor supplemented with 20% v/v ethylene glycol and then flash- frozen by being plunged into liquid nitrogen. RsmF3 crystals were cryoprotected by gradually increasing the concentration of PEG1000 to 30% and then flash-frozen by being plunged into liquid nitrogen. RsmF4 crystals were cryoprotected in a mother liquor supplemented with 20% glycerol and then flash-frozen by being plunged into liquid nitrogen. Data collection X-ray diffraction data for RsmF1, RsmF2, and RsmF4 crystals were collected on a MAR CCD detector at the X4C beamline of the National Synchrotron Light Source in Brookhaven at a wave- length of 0.979 A˚ and 180°C. Diffraction data for RsmF3 crystals were collected on an ADSC CCD detector at the X4A beamline of the National Synchrotron Light Source in Brookhaven at a wave- length of 0.979 A˚ and 180°C. Diffraction data for RsmF1 in space group P43 were collected to 1.4 A˚ resolution with cell Demirci et al. 1594 RNA, Vol. 16, No. 8 dimensions a = 71.0 A˚ , b = 71.0 A˚ , and c = 186.7 A˚ . Diffraction data to 1.82 A˚ for RsmF2 were collected in space group P2 with cell dimensions a = 66.0 A˚ , b = 78.3 A˚ , and c = 108.1 A˚ . Diffraction data to 1.29 A˚ for RsmF3 were collected in space group P21212 with cell dimensions a = 89.7 A˚ , b = 109.0 A˚ , and c = 51.0 A˚ . Diffraction data to 1.68 A˚ for RsmF4 were collected in space group P21212 with cell dimensions a = 89.8 A˚ , b = 109.1 A˚ , and c = 50.8 A˚ . A single crystal was used for each data set. The diffraction images were processed and scaled with the HKL2000 package (Otwinowski and Minor 1997). The data processing statistics are summarized in Table 1. Structure determination and refinement The RsmF structure was solved by molecular replacement with the program Phaser (McCoy et al. 2007) from the CCP4 program suite (Bailey 1994) in space group P43 to 1.4 A˚ resolution (data set RsmF1). The initial search model was built with the program Modeller (Eswar et al. 2008) from the catalytic domain of E. coli YebU (Pdb code 2FRX). After the placement of two RsmF catalytic domains in the asymmetric unit and the initial re- finement with Refmac (Murshudov et al. 1997), the model was further rebuilt with ARP/wARP (Langer et al. 2008). The resulting model was 90% complete and manually checked and completed with Coot (Emsley and Cowtan 2004). Final crystallographic re- finement was performed with the program Phenix (Adams et al. 2002). The other crystal forms were subsequently solved by molecular replacement. The atomic coordinates from the RsmF4 model were then used for initial refinement of the RsmF–AdoMet complex structure in space group P21212 (RsmF3). There are two molecules in the asymmetric unit in data sets RsmF1 and RsmF2, and one molecule in RsmF3 and RsmF4. The crystallographic R/Rfree factors are 0.17/0.19, 0.16/0.19, 0.18/0.19, and 0.17/0.19 for the four data sets: RsmF1, RsmF2, RsmF3, and RsmF4, respectively. The stereochemical quality of the model was assessed with Procheck (Laskowski et al. 1993). The Ramachandran sta- tistics (most favored/additionally allowed/generously allowed/ disallowed) are 91.9%/8.1%/0.0%/0.0% for RsmF1, 91.9%/8.1%/ 0.0%/0.0% for RsmF2, 93.6%/6.4%/0.0%/0.0% for RsmF3, and 92.5%/7.5%/0.0%/0.0% for RsmF4. The refinement statistics are summarized in Table 1. Figures were generated using Pymol (DeLano 2002). Atomic coordinates Coordinates and structure factors have been deposited in the Protein Data Bank with accession codes 3M6U, 3M6V, 3M6W, and 3M6X for data sets RsmF1, RsmF2, RsmF3, and RsmF4, respectively. SUPPLEMENTAL MATERIAL Supplemental material can be found at http://www.rnajournal.org. ACKNOWLEDGMENTS We thank John Schwanof and Randy Abramowitz for access to the X4A and X4C beamlines at the National Synchrotron Light Source. This work was supported by grants GM19756 and GM19756-37S1 from the National Institutes of Health. Received January 14, 2010; accepted April 26, 2010. REFERENCES Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Terwilliger TC. 2002. PHENIX: Building new software for automated crystallo- graphic structure determination. Acta Crystallogr D Biol Crystallogr 58: 1948–1954. Agris PF. 2004. Decoding the genome: A modified view. Nucleic Acids Res 32: 223–238. Alian A, Lee TT, Griner SL, Stroud RM, Finer-Moore J. 2008. Structure of a TrmA–RNA complex: A consensus RNA fold con- tributes to substrate selectivity and catalysis in m5U methyltrans- ferases. Proc Natl Acad Sci 105: 6876–6881. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic local alignment search tool. J Mol Biol 215: 403–410. Andersen NM, Douthwaite S. 2006. YebU is a m5C methyltransferase specific for 16 S rRNA nucleotide 1407. J Mol Biol 359: 777–786. Andersen TE, Porse BT, Kirpekar F. 2004. A novel partial modifica- tion at C2501 in Escherichia coli 23S ribosomal RNA. RNA 10: 907–913. Bailey S. 1994. The CCP4 Suite: Programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 50: 760–763. Behm-Ansmant I, Urban A, Ma X, Yu YT, Motorin Y, Branlant C. 2003. The Saccharomyces cerevisiae U2 snRNA:pseudouridine- synthase Pus7p is a novel multisite-multisubstrate RNA:C- synthase also acting on tRNAs. RNA 9: 1371–1382. Behm-Ansmant I, Branlant C, Motorin Y. 2007. The Saccharomyces cerevisiae Pus2 protein encoded by YGL063w ORF is a mitochon- drial tRNA:C27/28-synthase. RNA 13: 1641–1647. Cameron DM, Gregory ST, Thompson J, Suh MJ, Limbach PA, Dahlberg AE. 2004. Thermus thermophilus L11 methyltransferase, PrmA, is dispensable for growth and preferentially modifies free ribosomal protein L11 prior to ribosome assembly. J Bacteriol 186: 5819–5825. DeLano WL. 2002. The PyMol molecular graphics system . DeLano Scientific, San Carlos, CA. Demirci H, Gregory ST, Dahlberg AE, Jogl G. 2007. Recognition of ribosomal protein L11 by the protein trimethyltransferase PrmA. EMBO J 26: 567–577. Demirci H, Gregory ST, Dahlberg AE, Jogl G. 2008a. Crystal structure of the Thermus thermophilus 16 S rRNA methyltransferase RsmC in complex with cofactor and substrate guanosine. J Biol Chem 283: 26548–26556. Demirci H, Gregory ST, Dahlberg AE, Jogl G. 2008b. Multiple-site trimethylation of ribosomal protein L11 by the PrmA methyl- transferase. Structure 16: 1059–1066. Demirci H, Belardinelli R, Seri E, Gregory ST, Gualerzi C, Dahlberg AE, Jogl G. 2009. Structural rearrangements in the active site of the Thermus thermophilus 16S rRNA methyltransferase KsgA in a bi- nary complex with 59-methylthioadenosine. J Mol Biol 388: 271– 282. Douthwaite S, Kirpekar F. 2007. Identifying modifications in RNA by MALDI mass spectrometry. Methods Enzymol 425: 1–20. Emsley P, Cowtan K. 2004. Coot: Model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132. Eswar N, Eramian D, Webb B, Shen MY, Sali A. 2008. Protein structure modeling with MODELLER. Methods Mol Biol 426: 145– 159. Foster PG, Nunes CR, Greene P, Moustakas D, Stroud RM. 2003. The first structure of an RNA m5C methyltransferase, Fmu, provides insight into catalytic mechanism and specific binding of RNA substrate. Structure 11: 1609–1620. Gao YG, Selmer M, Dunham CM, Weixlbaumer A, Kelley AC, Ramakrishnan V. 2009. The structure of the ribosome with elongation factor G trapped in the post-translocational state. Science 326: 694–699. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1595 Gustilo EM, Vendeix FA, Agris PF. 2008. tRNA’s modifications bring order to gene expression. Curr Opin Microbiol 11: 134–140. Gu XR, Gustafsson C, Ku J, Yu M, Santi DV. 1999. Identification of the 16S rRNA m5C967 methyltransferase from Escherichia coli. Biochemistry 38: 4053–4057. Guymon R, Pomerantz SC, Crain PF, McCloskey JA. 2006. Influence of phylogeny on posttranscriptional modification of rRNA in thermophilic prokaryotes: The complete modification map of 16S rRNA of Thermus thermophilus. Biochemistry 45: 4888–4899. Hallberg BM, Ericsson UB, Johnson KA, Andersen NM, Douthwaite S, Nordlund P, Beuscher AE 4th, Erlandsen H. 2006. The structure of the RNA m5C methyltransferase YebU from Escherichia coli reveals a C-terminal RNA-recruiting PUA domain. J Mol Biol 360: 774–787. Hashimoto Y, Yano T, Kuramitsua S, Kagamiyama H. 2001. Disrup- tion of Thermus thermophilus genes by homologous recombination using a thermostable kanamycin-resistant marker. FEBS Lett 506: 231–234. Helser TL, Davies JE, Dahlberg JE. 1972. Mechanism of kasugamycin resistance in Escherichia coli. Nat New Biol 235: 6–9. Hendrickson WA, Horton JR, LeMaster DM. 1990. Selenomethionyl proteins produced for analysis by multiwavelength anomalous diffraction (MAD): A vehicle for direct determination of three- dimensional structure. EMBO J 9: 1665–1672. Holm L, Ka¨a¨ria¨inen S, Rosenstro¨m P, Schenkel A. 2008. Searching protein structure databases with DaliLite v.3. Bioinformatics 24: 2780–2781. Hoseki J, Yano T, Koyama Y, Kuramitsu S, Kagamiyama H. 1999. Directed evolution of thermostable kanamycin-resistance gene: A convenient selection marker for Thermus thermophilus. J Biochem 126: 951–956. Hur S, Stroud RM. 2007. How U38, 39, and 40 of many tRNAs become the targets for pseudouridylation by TruA. Mol Cell 26: 189–203. Lang PT, Brozell SR, Mukherjee S, Pettersen EF, Meng EC, Thomas V, Rizzo RC, Case DA, James TL, Kuntz ID. 2009. DOCK 6: Combining techniques to model RNA-small molecule complexes. RNA 15: 1219–1230. Langer G, Cohen SX, Lamzin VS, Perrakis A. 2008. Automated macromolecular model building for X-ray crystallography using ARP/wARP version 7. Nat Protoc 3: 1171–1179. Laskowski RA, MacArthur MW, Moss DS, Thornton JM. 1993. PROCHECK: A program to check the stereochemical quality of protein structures. J Appl Crystallogr 26: 283–291. Lee TT, Agarwalla S, Stroud RM. 2005. A unique RNA fold in the RumA-RNA-cofactor ternary complex contributes to substrate selectivity and enzymatic function. Cell 120: 599–611. Liu Y, Santi DV. 2000. m5C RNA and m5C DNA methyl transferases use different cysteine residues as catalysts. Proc Natl Acad Sci 97: 8263–8265. Malone T, Blumenthal RM, Cheng X. 1995. Structure-guided analysis reveals nine sequence motifs conserved among DNA amino- methyltransferases, and suggests a catalytic mechanism for these enzymes. J Mol Biol 253: 618–632. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. 2007. Phaser crystallographic software. J Appl Crystallogr 40: 658–674. McLuckey SA, Van Berkel GJ, Glish GL. 1992. Tandem mass spectrometry of small multiply charged oligonucleotides. J Am Soc Mass Spectrom 3: 60–70. Motorin Y, Grosjean H. 1999. Multisite-specific tRNA:m5C-methyl- transferase (Trm4) in yeast Saccharomyces cerevisiae: Identification of the gene and substrate specificity of the enzyme. RNA 5: 1105– 1118. Motorin Y, Keith G, Simon C, Foiret D, Simos G, Hurt E, Grosjean H. 1998. The yeast tRNA:pseudouridine synthase Pus1p displays a multisite substrate specificity. RNA 4: 856–869. Murshudov GN, Vagin AA, Dodson EJ. 1997. Refinement of macro- molecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystallogr 53: 240–255. O’Farrell HC, Scarsdale JN, Rife JP. 2004. Crystal structure of KsgA, a universally conserved rRNA adenine dimethyltransferase in Escherichia coli. J Mol Biol 339: 337–353. Ogle JM, Murphy FV, Tarry MJ, Ramakrishnan V. 2002. Selection of tRNA by the ribosome requires a transition from an open to a closed form. Cell 111: 721–732. Otwinowski Z, Minor W. 1997. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol 276: 307–326. Pe´rez-Arellano I, Gallego J, Cervera J. 2007. The PUA domain—a structural and functional overview. FEBS J 274: 4972–4984. Purta E, O’Connor M, Bujnicki JM, Douthwaite S. 2008. YccW is the m5C methyltransferase specific for 23S rRNA nucleotide 1962. J Mol Biol 383: 641–651. Selmer M, Dunham CM, Murphy FV 4th, Weixlbaumer A, Petry S, Kelley AC, Weir JR, Ramakrishnan V. 2006. Structure of the 70S ribosome complexed with mRNA and tRNA. Science 313: 1935– 1942. Sunita S, Tkaczuk KL, Purta E, Kasprzak JM, Douthwaite S, Bujnicki JM, Sivaraman J. 2008. Crystal structure of the Escherichia coli 23S rRNA:m5C methyltransferase RlmI (YccW) reveals evolutionary links between RNA modification enzymes. J Mol Biol 383: 652–666. Wimberly BT, Brodersen DE, Clemons WM Jr, Morgan-Warren RJ, Carter AP, Vonrhein C, Hartsch T, Ramakrishnan V. 2000. Structure of the 30S ribosomal subunit. Nature 407: 327–339. Demirci et al. 1596 RNA, Vol. 16, No. 8
3M6X
Multi-site-specific 16S rRNA methyltransferase RsmF from Thermus thermophilus in space group P21212
Multi-site-specific 16S rRNA methyltransferase RsmF from Thermus thermophilus HASAN DEMIRCI,1,3 LINE H.G. LARSEN,2,3 TRINE HANSEN,2 ANETTE RASMUSSEN,2 ASHWIN CADAMBI,1 STEVEN T. GREGORY,1 FINN KIRPEKAR,2 and GERWALD JOGL1 1Department of Molecular Biology, Cell Biology and Biochemistry, Brown University, Providence, Rhode Island 02912, USA 2Department of Biochemistry and Molecular Biology, University of Southern Denmark, 5230 Odense M, Denmark ABSTRACT Cells devote a significant effort toward the production of multiple modified nucleotides in rRNAs, which fine tune the ribosome function. Here, we report that two methyltransferases, RsmB and RsmF, are responsible for all four 5-methylcytidine (m5C) modifications in 16S rRNA of Thermus thermophilus. Like Escherichia coli RsmB, T. thermophilus RsmB produces m5C967. In contrast to E. coli RsmF, which introduces a single m5C1407 modification, T. thermophilus RsmF modifies three positions, generating m5C1400 and m5C1404 in addition to m5C1407. These three residues are clustered near the decoding site of the ribosome, but are situated in distinct structural contexts, suggesting a requirement for flexibility in the RsmF active site that is absent from the E. coli enzyme. Two of these residues, C1400 and C1404, are sufficiently buried in the mature ribosome structure so as to require extensive unfolding of the rRNA to be accessible to RsmF. In vitro, T. thermophilus RsmF methylates C1400, C1404, and C1407 in a 30S subunit substrate, but only C1400 and C1404 when naked 16S rRNA is the substrate. The multispecificity of T. thermophilus RsmF is potentially explained by three crystal structures of the enzyme in a complex with cofactor S-adenosyl-methionine at up to 1.3 A˚ resolution. In addition to confirming the overall structural similarity to E. coli RsmF, these structures also reveal that key segments in the active site are likely to be dynamic in solution, thereby expanding substrate recognition by T. thermophilus RsmF. Keywords: rRNA methyltransferase; 5-methylcytidine; RsmB; RsmF; RNA mass spectrometry INTRODUCTION Ribosomal RNAs (rRNAs) are post-transcriptionally mod- ified in all three domains of life, and many modifications are phylogenetically conserved. Most modifications are located in functionally important regions of the ribosome, where they probably act to fine tune protein synthesis (Agris 2004; Gustilo et al. 2008). Complete modification maps of bacterial 16S rRNAs have been determined for only a hand- ful of species, and among these are the enteric bacterium Escherichia coli and the extremely thermophilic bacterium Thermus thermophilus (Guymon et al. 2006). Despite the large phylogenetic divergence of these two organisms, their ribosome modification patterns are quite similar. Of the 11 E. coli and 14 T. thermophilus 16S rRNA modifications, eight are identical. This suggests a set of common functional requirements conserved since divergence from their last common ancestor, and also suggests common recognition mechanisms among their modifying enzymes. For most ribosome modifications, a single enzyme recog- nizes and modifies a single site. However, there exist nota- ble exceptions. Among these are dimethylation of two adja- cent adenosines in 16S rRNA by KsgA (Helser et al. 1972); pseudouridylation of three adjacent residues in tRNAs by TruA (Hur and Stroud 2007); pseudouridylation of several tRNA residues by Pus1 (Motorin et al. 1998), Pus2 (Behm- Ansmant et al. 2007), or Pus7 (Behm-Ansmant et al. 2003); or methylation of four tRNA positions by Saccharomyces cerevisiae Trm4 (Motorin and Grosjean 1999). Even with these multi-site-specific enzymes, however, homologs from various species generally modify the same residues. E. coli 16S rRNA contains two 5-methylcytidine (m5C) residues, located in or near the highly conserved decoding 3These authors contributed equally to this work. Abbreviations: rRNA, ribosomal RNA; AdoMet, S-adenosyl-L-methionine; m5C, 5-methyl-cytidine; MALDI mass spectrometry, matrix-assisted laser desorption ionization mass spectrometry. Reprint requests to: Gerwald Jogl, Department of Molecular Biology, Cell Biology and Biochemistry, Brown University, Box G-E129, Provi- dence, RI 02912, USA; e-mail: Gerwald_Jogl@brown.edu; fax: (401) 863-6114; or Finn Kirpekar, Department of Biochemistry and Molecular Biology, University of Southern Denmark, Campusvej 55, 5230 Odense M, Denmark; e-mail: f.kir@bmb.sdu.dk; fax: (+45) 65502467. Article published online ahead of print. Article and publication date are at http://www.rnajournal.org/cgi/doi/10.1261/rna.2088310. 1584 RNA (2010), 16:1584–1596. Published by Cold Spring Harbor Laboratory Press. Copyright  2010 RNA Society. center of the 30S subunit (Fig. 1). An m5C967 modification is produced by RsmB (also called Fmu), while an m5C1407 modification is produced by RsmF, formerly known as YebU (Andersen and Douthwaite 2006). T. thermophilus 16S rRNA contains m5C967 and m5C1407, as well as two additional m5C nucleotides, m5C1400 and m5C1404 (E. coli rRNA numbering used throughout) (Guymon et al. 2006). While the m5C967 and m5C1407 modifications are pre- sumably produced by RsmB and RsmF homologs, respec- tively, the source of the two additional m5C residues has been unknown. Here we demonstrate that T. thermophilus RsmF is a multi-site-specific methyltransferase and, in contrast to the single-site-specific E. coli RsmF, is respon- sible for the synthesis of three modifications: m5C1407, m5C1400, and m5C1404. We also demonstrate that RsmB is responsible for the synthesis of m5C967 in T. thermophilus as well as is in E. coli, thereby accounting for all four m5C modifications of 16S rRNA. We present crystal structures of T. thermophilus RsmF up to 1.3 A˚ resolution that reveal a dynamic region in the active site that is absent from the E. coli RsmF structure, providing a possible explanation for the expanded recognition capacity of the T. thermophilus methyltransferase. RESULTS Identification of T. thermophilus 16S rRNA m5C-methyltransferases With the E. coli RsmB and RsmF protein sequences as queries, we used conventional BLAST searches (Altschul et al. 1990) to identify potential homologs encoded by the T. thermophilus HB8 genome (data not shown). Both RsmB and RsmF have the highest similarity to the T. thermophilus protein encoded by TTHA1387 (BLAST scores of 106 and 190, respectively) and second-highest similarity to the pro- tein encoded by TTHA0851 (BLAST scores of 93 and 81, respectively). The simplest interpretation of these results is that TTHA1387 encodes RsmF, responsible for methylation of C1407, leaving TTHA0851 as the most likely candidate for the gene encoding RsmB, responsible for methylation of C967. The similarities of the two E. coli enzymes with other T. thermophilus proteins were far too low to reveal potential candidates responsible for methylation of C1400 and C1404. We next constructed T. thermophilus strains in which either TTHA0851 or TTHA1387 was inactivated by the homologous recombination and insertion of a heat stable kanamycin-resistance gene. 16S rRNA was isolated from these null mutants and subfragments of z50 nucleotides (nt) around the regions of interest were further purified, digested with RNase T1, and analyzed by MALDI mass spectrometry (Fig. 2). Comparison of the TTHA1387 null mutant to wild-type T. thermophilus HB8 indicates three clear differences, each corresponding to the disappearance of a methyl group (z14.0 Da). The RNase T1 digestion fragment harboring m5C1407, the nucleotide methylated by RsmF in E. coli, is absent in the null mutant, indicating that TTHA1387 is indeed rsmF. The predicted RNase T1 fragment reduced by 14.0 Da is obscured by another RNase T1 fragment that is present in both the wild-type and TTHA1387 null mutants (Fig. 2B). Unexpectedly, two addi- tional RNase T1 fragments are also reduced by 14.0 Da. One of these contains C1400 while the other contains C1404. This latter RNase T1 fragment from wild-type T. thermophilus contains three methyl groups, two on m4Cm1402 and one on m5C1404 (Guymon et al. 2006), preventing an unambiguous identification of the missing methyl group. We therefore performed tandem mass spec- trometry on the 1402CCCG1405 RNase T1 fragment with two methyl groups from the TTHA1387 null mutant and com- pared it with the triply methylated wild-type RNase T1 fragment (Fig. 2C). The clear w2 ions, as well as the less intense z3 ions, display a 14.0 Da mass difference between the two samples, showing that the methylations on m4Cm1402 were not affected by inactivation of TTHA1387. Tandem mass spectrometry was also performed on the RNase T1 fragments appearing as a consequence of the lack of methylations on C1400 and C1407 (data not shown). As expected, the C1400-containing fragment revealed no FIGURE 1. Secondary structure diagram of the 39 minor domain of 16S rRNA indicating the position of the three RsmF substrate nucleotides. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1585 indications of a methyl group, whereas the RNase T1 fragment with C1407 ex- hibited a fragmentation pattern corre- sponding to the expected mass overlap with an RNase T1 fragment of a different sequence. In summary, our data lead us to conclude that TTHA1387 encodes an RsmF m5C methyltransferase responsible for synthesizing m5C1400, m5C1404, and m5C1407 in 16S rRNA of T. thermophilus. The above results left us with TTHA0851 as the sole candidate for the gene encoding the m5C967 methyl- transferase. An approach conceptually identical to that described above re- vealed that disruption of TTHA0851 reduced the relevant RNase T1 fragment by 14.0 Da (Supplemental Fig. 1A). Since this fragment is methylated at G966 and C967 in the wild-type strain (Guymon et al. 2006), tandem mass spectrometry was again performed (Sup- plemental Fig. 1B), showing that only the methyl group on C967 was absent. In agreement with the suggested nomen- clature for rRNA modifying enzymes (Andersen and Douthwaite 2006), we hereafter refer to TTHA0851 as rsmB due to substrate specificity identical to the originally identified enzyme from E. coli (Gu et al. 1999). Effect of temperature on growth of the rsmF null mutant One possible explanation for methyla- tion of multiple sites by RsmF is that such additional methyl groups improve ribosomal function at elevated temper- ature. To address this possibility, we ex- amined the effect of temperature on growth of the rsmF null mutant. Wild- type T. thermophilus and the rsmF null mutant were cocultured at three tem- peratures and these cocultures were se- rially subcultured for seven cycles of 24 h each. The proportion of wild-type and rsmF null mutant cells in each mixed culture was determined by spreading di- lutions onto TEM plates with or without kanamycin. After seven cycles, no differ- ence in the relative proportions of the wild-type and rsmF mutant was exhib- ited at 70°C. However, at 60°C, the rsmF null mutant constituted only around 5% FIGURE 2. (Legend on next page) Demirci et al. 1586 RNA, Vol. 16, No. 8 of the population, and at 80°C the rsmF null mutant was unable to grow at all. Thus, methylation by RsmF appears to facilitate growth at temperatures outside the optimal growth temperature. RsmF substrate preference The T. thermophilus rsmB and rsmF genes were each cloned into an E. coli expression plasmid in order to produce proteins for X-ray crystallography and in vitro methylation studies. The expression constructs were equipped with C-terminal histidine6 tags to facilitate protein purification. While we achieved a high expression level of RsmF, we were unable to do so with RsmB despite a series of optimization attempts. Consequently, in vitro substrate and structure analyses were performed exclusively with RsmF. E. coli RsmF requires the 30S ribosomal subunit as a substrate when the activity is assayed in vitro (Andersen and Douthwaite 2006). We assayed 70S ribosomes, 30S ribosomal subunits, and 16S rRNA for the ability to serve as substrates for methylation by RsmF in vitro. 16S rRNA subfragments of z50 nt around the target sites were purified after the in vitro assay and analyzed by mass spectrometry as described above. In vitro methylation at 70°C showed an interesting but rather complex substrate pattern. RsmF completely methylates C1400 when either 16S rRNA or 30S subunits are used as a substrate. It methylates C1404 to z35% with 16S rRNA and completely with 30S subunits, and it produces only trace amounts of methylation of C1407 with 16S rRNA and z75% with 30S subunits (Fig. 3). There were no indications of the 70S ribosome being a substrate in vitro. Curiously, T. thermophilus RsmF expressed in an E. coli rsmF null mutant almost completely methylated, in vivo, positions C1400 and C1404, but not C1407 (data not shown). X-ray crystal structures of RsmF We determined the structure of T. thermophilus RsmF (456 amino acids) in three different crystal forms and in a com- plex with cofactor AdoMet to up to 1.3 A˚ resolution (Figs. 4, 5). The structure was solved in space group P43 (data set RsmF1, 1.4 A˚ resolution) by molecular replacement using a search model generated with the program Modeller (Eswar et al. 2008) from the catalytic domain of the RsmF homolog YebU from E. coli (Pdb 2FRX) (Hallberg et al. 2006). The structures of the AdoMet-bound form in space group P2 (RsmF2, 1.82 A˚ resolution), of the AdoMet- bound form (RsmF3, 1.3 A˚ resolution), and of the apo- form (RsmF4, 1.68 A˚ resolution) in space group P21212 were subsequently solved by molecular replacement with the refined RsmF1 model. There are two molecules in the asymmetric unit in space groups P43 and P2 and one mol- ecule in space group P21212. Electron density is generally well defined in all crystal forms. The majority of residues (92.0%, 92.3%, 93.1%, and 92.6%) are in the most favored region of the Ramachandran plot for RsmF1, RsmF2, RsmF3, and RsmF4, respectively, and there are no residues in the disallowed region. The final models consist of residues 5–178, 194–198, and 201–456 and five additional residues from the histidine6 affinity tag in both chains of data set RsmF1; residues 2–456 and five affinity-tag residues in both chains of data set RsmF2; residues 1–456 and six affinity-tag residues in data set RsmF3; and resi- dues 1–456 and seven affinity-tag residues in RsmF4. The N-terminal a-amino group was ordered in data sets RsmF3 and RsmF4 and contained additional electron density, which we interpreted as N-(dihydroxymethyl)-L-methio- nine, the hydrated form of N-formyl-methionine. Data collection and refinement statistics are given in Table 1. The overall structure of RsmF consists of a central canonical class I methyltransferase catalytic domain with additional N-terminal and C-terminal domains (Figs. 4, 5). The catalytic do- main is formed by a central seven- stranded b-sheet that is flanked on both sides by three helices of varying lengths. An inserted region between strand b7 and helix a11 contains additional heli- ces a9 and a10, which interact with the two N-terminal helices a1 and a2. A second inserted region following strand b9 includes the short helices a13 and FIGURE 2. (A) MALDI mass spectra of an RNase T1-digested 16S rRNA subfragment (pos. 1378–1432) from wild-type cells (upper panel) or from the TTHA1387 (putative rsmF) null mutant (lower panel). Expected digestion products are labeled; fragments affected by the null mutation are set in italics. (B) Expansion of the signals affected by the TTHA1387 null mu- tation. The sequence and methylation status of the RNase T1 products are indicated. (C) MALDI tandem mass spectrometry of the methylated RNase T1 fragment of 16S rRNA (pos. 1404–1407); wild-type cells (upper panel), TTHA1387 null mutant (lower panel). Mass spectrometric fragments used to deduce the methylation status are labeled. The position of the backbone fragments (nomenclature according to McLuckey et al. [1992]) in the sequence is shown. MH+, precursor ion selected for fragmentation; C, cytosine; mC, methylated cytosine; C>p, cytidine-2´-39-monophosphate; me, methyl group. FIGURE 3. In vitro methylation with 30S ribosomal subunits or 16S rRNA from the T. thermophilus rsmF null mutant as a substrate. Effect on C1400-, C1404-, and C1407-harboring RNase T1 products. In vitro methylated products are set in italics. *, artifact signal arising from the enzyme preparation. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1587 a14, which interact with helix a12. Furthermore, RsmF contains three additional smaller domains, an N-terminal domain consisting of a three-stranded b-sheet and two flank- ing helices (Fig. 5B, colored in blue), and two C-terminal domains consisting of four-stranded b-sheets and two or one helix (Fig. 5B, colored in magenta and red). Cofactor binding, substrate docking, and conformational flexibility in the active site The coordination of AdoMet in the T. thermophilus RsmF active site is similar to that seen in other class I methyl- transferases. However, the previously published structure of E. coli RsmF did not contain the cofactor AdoMet in the active site, precluding a direct comparison. Both the T. thermophilus and E. coli RsmF cofactor-binding sites reveal a new variation for the methyltransferase signature motif I (Malone et al. 1995), with the highly conserved GxGxG sequence replaced by 109AAAPG113. The combination of three alanines and a proline results in a loop conformation that is very similar to that observed in other methyltrans- ferases with a GxGxG motif (e.g., RsmC) (Demirci et al. 2008a). In RsmF, the amide hydrogen atom of the last glycine residue forms a hydrogen bond with the cofactor carboxy group (Fig. 6). Other key interactions with AdoMet are well conserved in RsmF. The cofactor adenine ring is located in a mainly hydrophobic pocket lined by residues Val134, Pro160, and Leu211. This pocket is open toward the solvent. The adenine amino group is not specifically recog- nized and interacts with solvent water molecules. The ribose hydroxyl groups form hydrogen bonds with Glu133 and Arg138, and the methionine amino group interacts with Asp177. The AdoMet cofactor is bound in a cleft in the RsmF active site, which suggests that substrate cytidine bases are inserted into the active site in an unstacked conforma- tion. Inspection of the electrostatic charge distribution re- veals a large positively charged surface region, which would be consistent with binding to an RNA surface and modifi- cation of the substrate base in an unstacked orientation (Fig. 6C). To evaluate the possible orientation of a substrate base in the active site, we performed computational docking calculations with the program Dock6 (Lang et al. 2009). The resulting positions of cytosine and m5C in the presence of AdoMet in data set RsmF3 are highly similar to each other, with m5C placed into the active site with its phosphate group toward a positively charged pocket at the entrance of the active site cleft (Fig. 6E). The position of the phosphate group is close to a sulfate molecule that we observed in data set RsmF1, providing further support for the results of the docking calculation (Fig. 6F). Interestingly, we observed that three active site segments were disordered in data set RsmF1. These segments include residues 179–193 (including helices a9 and a10 and the catalytic Cys180), residues 199–200, and the N-terminal residues 1–5, which interact with the first two segments (Fig. 6E,F, colored in green). We observed electron density for the intervening residues 194–198, which formed a lattice con- tact with a neighboring molecule. However, the position of FIGURE 4. Structure-based sequence alignment of RsmF from T. thermophilus and E. coli. Secondary structure elements of T. thermophilus RsmF are indicated on top. The color scheme for the secondary structure elements is as in Figure 5A. The position of the variant methyltransferase motif I is marked with a red box. A flexible region observed in the active site is marked with a green box; residues interacting with the cofactor are marked with orange boxes. Demirci et al. 1588 RNA, Vol. 16, No. 8 these five residues was not related to their position in the other three data sets, suggesting that the extended active site region between residues 179 and 201 can reorient in the RsmF structure. This observation suggests that this active site region is dynamic, which may be important for sub- strate binding at 72°C, the optimum growth temperature of T. thermophilus. DISCUSSION Substrate recognition mechanisms We have identified the two enzymes responsible for the synthesis of the four m5C modifications of T. thermophilus 16S rRNA, and characterized the RsmF methyltransferase responsible for synthesizing three of these. rRNA modifying enzymes in bacteria are generally highly specific, with a one-to-one association between the modifying enzyme and the modification. A few cases of multitarget ribosome mod- ifying enzymes have been reported (Helser et al. 1972; Demirci et al. 2008b), but to our knowledge T. thermophi- lus RsmF is the first rRNA methyltransferase found to mod- ify three different nucleotides. Most ribosome modifying enzymes probably recognize assembly intermediates, and the data presented here are consistent with that notion. T. thermophilus RsmF methylates C1400 and C1404 in vitro using either 16S rRNA or 30S subunits as substrates, whereas both E. coli (Andersen and Douthwaite 2006) and T. thermophilus RsmF exclusively utilize 30S subunits as substrates for methylation of C1407. This may reflect that C1400 and C1404 methylations do not rely on the asso- ciation of ribosomal proteins in order to be recognized by T. thermophilus RsmF. C1407 methylation, in contrast, depends on both rRNA and the ribosomal protein for the recognition by RsmF in both T. thermophilus and E. coli. More puzzling is the observation that T. thermophilus RsmF does not methylate E. coli ribosomes in vivo on C1407. It is perhaps worth noting that methylation of C1407 in the T. thermophilus 30S ribosomal subunit in vitro was less efficient than methylation of the other two positions, indicating the need for a particular intermediate assembly structure or for accessory factors. The only clear in vitro FIGURE 5. Overall structure of RsmF. (A) Schematic representation of the position of the substrate bases in the 30S ribosomal subunit (Pdb entry 2WRI) (Gao et al. 2009). C1400, C1404, and C1407 in helix 44 (green) are shown in stick representation with 5-methyl groups as pink spheres. (B) Schematic stereo representation of the overall structure of RsmF. Secondary structure elements are in orange and yellow for the catalytic domain, in salmon and blue for additional N-terminal domains, and in magenta and red for C-terminal domains. The flexible region in the active site including helices a9 and a10 is indicated with an arrow and colored in green. (C) Topology diagram with secondary structure elements colored as in B. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1589 methylation observed at 37°C with T. thermophilus RsmF was on C1400 with 16S rRNA as the substrate (data not shown), which is not evidently related to the methylation pattern in vivo in the heterologous system. It seems unlikely that the aberrant methylation in the heterologous system reflects species-specific differences in the mature 30S ribo- somal subunit, given the extreme sequence and structural conservation of the decoding site. Instead, it may reflect differences in 30S subunit assembly in the two organisms, necessary due to the large difference in growth temperature. The four m5C residues in T. thermophilus 16S rRNA are clustered in and around the functionally critical decoding center at or close to sites of contact with tRNA, mRNA, and EF-G (Ogle et al. 2002; Selmer et al. 2006; Gao et al. 2009). m5C1400, m5C1404, and m5C1407 are located in the subunit body, while m5C967 is located in the subunit head, about 10 A˚ from m5C1400 (Wimberly et al. 2000). Exam- ination of the 30S subunit crystal structure (Wimberly et al. 2000) indicates that the three bases methylated by RsmF are situated in three distinct structural contexts, but provides few clues to a common mode of substrate recognition. C1400 is an unpaired base protruding from a sharply bent segment of rRNA at the junction of helices 43 and 44, while C1404 and C1407 are engaged in Watson–Crick pairs within helix 44 (Fig. 4A). The C5 positions of the latter two bases are not obviously accessible, such that RsmF would need to approach them from the major groove side. While a flipping of C1407 out of the helix via the minor groove could allow access to this base, such a mechanism would be problematic for C1404, whose minor groove side is packed against the rest of the 30S subunit. While m5C1404 and m5C1407 are about 11 A˚ apart (Selmer et al. 2006), m5C1400 is quite distant from both of these bases (about 21 and 30 A˚ , respectively). The pro- truded conformation of m5C1400 in the mature 30S sub- unit is due in part to base-pairing interactions between adjacent bases and the 1500 region of 16S rRNA (C1399– G1504 and G1401–C1501). As the 1500 region is one of the last segments of 16S rRNA to be synthesized, C1400 could potentially be positioned much closer to C1404 and C1407 in an assembly intermediate, prior to the formation of the C1399–G1504 and G1401–C1501 base pairs. RsmF could utilize a single binding mode to then access all three bases, further facilitated by its flexible active site domain. Meth- ylation of C1400 in mature 30S subunits would therefore involve disruption of the adjacent base pairs. A complete understanding of the recognition mechanism of these enzymes will require high-resolution structural data on assembly intermediate-enzyme complexes. Given the large number of potential subunit assembly intermediates, pre- cisely defining the physiological substrate for RsmB and RsmF will be a formidable task. TABLE 1. Data collection and refinement statistics RsmF1 RsmF2 RsmF3 RsmF4 Data collectiona AdoMet AdoMet Space group P43 P2 P21212 P21212 Cell dimensions a, b, c (A˚ ) 71.0, 71.0, 186.7 66.0, 78.3, 108.1 89.7, 109.0, 51.0 89.8, 109.1, 50.8 a, b, g (°) 90, 90, 90 90, 107.1, 90 90, 90, 90 90, 90, 90 Resolution (A˚ )b 30–1.4 (1.55–1.40) 30–1.82 (1.89–1.82) 30–1.30 (1.34–1.30) 30–1.68 (1.74–1.68) Rmerge 0.065 (0.59) 0.08(0.38) 0.058(0.36) 0.15 (0.49) I/sI 29.3(2.15) 12.6 (2.04) 24.3 (2.05) 14.2 (1.73) Completeness (%) 90.1 (72.6) 97.0(86.5) 95.6(66.3) 99.6 (97.8) Redundancy 8.9 (5.4) 2.8(2.0) 5.2(2.1) 6.4 (3.9) Refinement Resolution (A˚ ) 30–1.4 (1.42–1.40) 30–1.82 (1.84–1.82) 30–1.30 (1.32–1.30) 30–1.68 (1.69–1.68) Number of reflections 161,955 (4356) 91,762 (2583) 119,490/2640 109,375 (3244) Rwork/Rfree 0.169/0.189 (0.227/0.263) 0.162/0.194 (0.222/0.259) 0.177/0.191 (0.233/0.234) 0.173/0.192 (0.217/0.266) Number of atoms Protein 6766 7117 3598 3574 Ligand/ion 20 54 27 1 Water 1486 1285 856 665 B-factors Protein 24.1 20.4 15.9 17.2 Ligand/ion 34.5 21.6 17.5 19.1 Water 38.7 36.3 33.6 33.6 RMSDs Bond lengths (A˚ ) 0.009 0.006 0.005 0.004 Bond angles (°) 1.22 1.04 1.14 0.94 aOne crystal used for each data set. bThe highest resolution shell is shown in parentheses. Demirci et al. 1590 RNA, Vol. 16, No. 8 Structural comparison of RsmF with related methyltransferases A database search with Dali (Holm et al. 2008) confirmed the structural similarity of the T. thermophilus and E. coli (PDB 2FRX) (Hallberg et al. 2006) RsmF homologs, which superimpose with a root-mean-square deviation (RMSD) of 1.6 A˚ for 342 Ca atoms (Fig. 7A) and are the only two structures in the Pro- tein Database with this domain organi- zation. Even so, substantial structural differences are observed in most of the loop regions and for the long connect- ing loop between the methyltransferase domain and the first C-terminal domain. While in the active site, the positions of residues in the cofactor-binding site and of the two cysteine residues are conserved (Fig. 7B), there are a number of positively charged residues (Arg30, Arg190, Arg194, His195, and Arg203) in the T. thermophilus structure that are absent from the E. coli enzyme (Fig. 7B). Three of these are located in the flexible region, and the combination of a posi- tive charge and flexibility close to the active site is suggestive of a functional contribution of this region to the mul- tisite specificity of T. thermophilus RsmF (Figs. 6C,D, 7B). Methylation of three rRNA positions may require an increase in the enzyme’s structural dynamics in order to accommodate the 30S subunit in slightly different orientations. Similar observations have been made for other multi-site-specific methyltransferases in- cluding KsgA, which modifies two adja- cent adenosines in the 30S ribosomal subunit (O’Farrell et al. 2004; Demirci et al. 2009), and the PrmA ribosomal protein methyltransferase, which under- goes dramatic interdomain movements to modify multiple lysine residues and the N-terminal a-amino group on the same substrate protein (Demirci et al. 2007, 2008b). The second C-terminal domain in RsmF is related to the RNA-binding PUA (pseudouridine synthase and archaeosine transglycosylase) domains (Perez-Arellano et al. 2007). The RlmI methyltransferase, which produces m5C1962 in 23S rRNA (Purta et al. 2008) also contains a PUA domain (Sunita et al. 2008), although it is N-terminal to the catalytic methyltransferase domain and in a different orientation. PUA domains contain six b-strands, which form a central pseudobarrel closed by a short 310-helix. A comparison of the C-terminal domain in RsmF with a typical PUA domain in archaeosine trans- glycosylase (ArcTGT, Pdb entry 1J2B) reveals that the central fold is similar (53 Ca atoms align with an RMSD FIGURE 6. Substrate docking and conformational flexibility. (A) Cofactor-binding site in RsmF. Bound AdoMet is shown in blue sticks. Hydrogen bonds to coordinating residues are indicated. (B) Final sA-weighted 2mFO-DFC electron density map of the cofactor-binding site (data set RsmF2) contoured at the 1s level. (C,D) Comparison of the electrostatic surface charge distribution between RsmF from T. thermophilus and E. coli. The location of the C-terminal domains and of the flexible region (labeled ‘‘FLEX’’) is indicated with circles. AdoMet and docked m5C are shown as sticks. AdoMet from the T. thermophilus structure is shown with the E. coli structure for comparison. (E) The modeled position of m5C docked into the active site of RsmF. m5C and AdoMet are shown in tan and blue sticks, respectively. Residues in the flexible region are shown in green. (F) The active site region in data set RsmF1. A sulfate ion is observed close to the position of the m5C phosphate group. Residues 194–198 in the flexible region interacting with a neighboring RsmF molecule are indicated with an arrow. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1591 of 1.8 A˚ ), but that several connecting loop regions are substantially shorter and two a-helices and one b-strand of the pseudobarrel are absent. Thus, this PUA-like domain differs considerably from typical PUA domains. However, the similarity to RNA-binding PUA domains and the positive surface charge distribution observed in both RsmF structures are suggestive for a conserved function of the PUA-like domain in RNA recognition. The next most closely related structure is E. coli RsmB (474 residues, PDB 1SQF) (Foster et al. 2003). A total of 264 Ca atoms can be aligned with an RMSD of 1.5 A˚ between RsmB and RsmF. Both enzymes retain the same organization for the N-terminal domains and the core methyltransferase domain. However, an additional 140- residue N-terminal RNA-binding domain provides sub- strate specificity to RsmB, whereas the two C-terminal domains following the core methyltransferase domain (160 residues) are likely to determine substrate recognition by RsmF. Thus, these two enzymes have evolved substrate specificity via acquisition of additional, unrelated RNA recognition domains. While there are as yet no enzyme– substrate complexes for rRNA m5C- methyltransferases, insights into the RsmF catalytic mechanism can be gleaned from a comparison with the RlmD and TrmA m5U methyltransferases in covalent in- termediate complexes with RNA oligonu- cleotides (Lee et al. 2005; Alian et al.2008). DNA and RNA m5C methyltransferases use a thiol from a catalytic cysteine residue to attack the six-position of the pyrimi- dine base to activate the five-position for methyl group transfer (Liu and Santi 2000). Cys180 and Cys230 in RsmF are positioned equivalently to the catalytic Cys324 and the catalytic base Glu358 of TrmA. The substrate nucleotides insert into TrmA and RsmF in unrelated di- rections, consistent with a lack of struc- tural homology outside the methyltrans- ferase domains. Nevertheless, the C5 positions of the pyrimidine rings and of the 5-methyl carbons are quite similar with respect to the catalytic cysteine resi- dues and the AdoMet cofactor. The same structural homology of the active site geometry can be observed in comparison with the RlmD methyltransferase (Sup- plemental Fig. 2; Lee et al. 2005). A possible origin of T. thermophilus RsmF T. thermophilus RsmF shows the highest similarities with proteins from close relatives, namely, Thermus aquaticus and two Meiothermus species of the Thermaceae family. Remarkably, the next highest similarities (BLAST scores between 267 and 359) are with the NOL1/ NOP2/Sun proteins from the Gram-positive Firmicutes phylum (Supplemental Fig. 3). This similarity, together with the fact that the Thermaceae family and most members of the Firmicutes identified in Supplemental Figure 3 are thermophilic, suggest that rsmF has undergone horizontal transfer between the Thermaceae family and members of the Firmicutes phylum. Thus, we speculate that this version of the RsmF protein, which catalyzes methylation of three cytidines, may be adaptive for existence in thermally challenging environments. The effect of the loss of methyl- ation by RsmF on growth at different temperatures is consistent with this notion. Our hypothesis of horizontal transfer of rsmF predicts that RsmF of other members of the Thermaceae family and members of the Firmicutes phylum will also be found to introduce multiple m5C modifications. FIGURE 7. Comparison with other methyltransferases. (A) Differences between the overall structures of RsmF from T. thermophilus (orange) and E. coli (green). AdoMet bound in the T. thermophilus structure is shown as blue sticks. (B) Comparison of the active site region in both enzymes. Residues in the cofactor-binding site and in the flexible region in the T. thermophilus enzyme are shown as sticks. (C) Comparison between RsmF (orange) and a substrate complex structure of the TrmA methyltransferase (cyan/blue). (D) Comparison of the active site region illustrating the differences in the insertion direction of the substrate base in RsmF (m5C in light orange) and in TrmA (m5U in cyan). Demirci et al. 1592 RNA, Vol. 16, No. 8 MATERIALS AND METHODS Cloning of the T. thermophilus rsmB and rsmF genes The T. thermophilus HB8 loci TTHA0851 (GenBank accession number BAD70674) and TTHA1387 (GenBank accession number BAD71210) were PCR amplified from genomic DNA and purified via the High Pure PCR Template Preparation Kit (Roche). The 100 mL PCRs contained 150 ng DNA, 10 mM of each primer, 10 mM dNTP, 1 unit Phusion DNA polymerase (Finnzymes), and 1x Phusion HF buffer. Primers for rsmB amplification were 59-CC CTGGACATATGAGGGCCGG-39 and 59-GGCCAAGATCTTGCC TGAGAG-39, and the temperature cycling was as follows: 98°C/30 sec; 30X (98°C/10 sec, 59°C/20 sec, and 72°C/36 sec); and 72°C/420s. Primers for rsmF amplification were 59-GCTAGGGTACACATA TGCTGCCC-39 and 59-GCACGGGGGTGAGATCTAAGCCC-39, and the temperature cycling was as follows: 98°C/30 sec; 30X (98°C/10 sec, 62°C /20 sec, 72°C/42 sec); and 72°C/420 sec. The desired PCR products were purified from agarose gels using the GFX PCR purification kit (GE Healthcare). The PCR fragments were digested with NdeI and BglII and inserted into the expression vector pLJ102 (Andersen and Douthwaite 2006), generating isopropyl-1-thio-b-D-galactopyranoside (IPTG)-inducible genes for the recombinant proteins with a C-terminal histidine6 tag. The constructs (designated pLJ102-RsmB and pLJ102-RsmF) were used to transform an rsmF-deletion derivative of E. coli CP79 (Andersen and Douthwaite 2006). Deletion of the T. thermophilus rsmB and rsmF genes Constructs for inactivation of the T. thermophilus rsmB and rsmF genes were made by inserting the gene for a heat tolerant kanamycin (HTK) nucleotidyltransferase (Hoseki et al. 1999) into the methyltransferase parts of either pLJ102-RsmB or pLJ102- RsmF. The htk part of pUC18-htk (Hashimoto et al. 2001) was amplified by PCR with primers that introduced an upstream AvrII site and a downstream SacI site into the product for later disrup- tion of rsmB. For rsmF disruption, the PCR primers introduced SacI restriction enzyme sites both upstream of and downstream from the htk gene. These sites were used to insert the PCR prod- ucts into pLJ102-RsmB and pLJ102-RsmF to form the plasmids pLJ102-RsmBThtk and pLJ102-RsmFThtk, which were propa- gated in the E. coli strain Top10 (Invitrogen). T. thermophilus HB8 was transformed with pLJ102-RsmBThtk or pLJ102- RsmFThtk selecting for kanamycin resistance as described by others (Hashimoto et al. 2001; Cameron et al. 2004). Kanamycin- resistant transformants were restreaked twice. Gene disruptions were verified by PCR with primers distal to the interrupted rsmB or rsmF genes on genomic DNA; resulting PCR products were characterized by sequencing. Growth competition assays Wild-type and rsmF null mutant liquid cultures were grown at 70°C to saturation, then equal numbers of cells from each were mixed and incubated in 5 mL TEM medium at 60°C, 70°C, or 80°C. After growth for 24 h, 100 mL of the 60°C culture, 10 mL of the 70°C culture, and 1000 mL of the 80°C culture were trans- ferred to a fresh 5-mL medium and incubated at the respective temperatures for another 24 h. This was repeated in independent triplicates for seven cycles. Samples of 1 mL were collected at each dilution and half was plated on TEM plates without antibiotic and the other half was plated on TEM plates with 30 mg/mL kanamycin. The plates were incubated at 70°C. Purification of T. thermophilus ribosomal subunits and ribosomes T. thermophilus culture (1 L) was grown in TEM media (contain- ing 30 mg/mL of kanamycin when appropriate) with shaking at 70°C to an OD600 = 0.6. Cells were harvested and washed once with 100 mL of buffer A (10 mM NH4Cl, 20 mM MgCl2, 100 mM KCl, and 10 mM Tris-HCl [pH 7.5]), then suspended in 10 mL of buffer A, and disrupted by sonication. The lysate was cleared by centrifugation twice in a Beckman JA20 rotor at 16,000 rpm for 10 min at 4°C. Crude ribosomes were collected by centrifugation in a Beckman Ti50 rotor at 19,000 rpm for 19 h at 4°C, and dissolved in buffer A. 70S ribosomes were obtained by centrifu- gation of 100 A260 units of crude ribosomes through a 10%–40% sucrose gradient (200 mM NH4Cl, 20 mM MgCl2, 20 mM Tris- HCl [pH 7.5]) in a Beckman SW28 rotor at 20,000 rpm for 18 h at 4°C. Fractions containing intact 70S ribosomes were pooled and concentrated by centrifugation in a Beckman Ti50 rotor at 40,000 rpm for 22 h at 4°C, washed, and dissolved in buffer A, and stored at 80°C. 50S and 30S ribosomal subunits were obtained by adjusting 100 A260 units of crude ribosomes (10 mM NH4Cl, 2 mM MgCl2, 100 mM KCl, and 10 mM Tris-HCl [pH 7.5]) and passing through a 5%–20% sucrose gradient (200 mM NH4Cl, 1 mM MgCl2 and 20 mM Tris-HCl [pH 7.5]) in a Sorvall AH-629 rotor at 20,000 rpm for 18 h at 4°C. After pooling of the relevant fractions, the subunits were adjusted to 10 mM MgCl2 and pelleted by centrifugation in a Beckman Ti50 rotor at 40,000 rpm for 22 h at 4°C, washed with and dissolved in 10 mM NH4Cl, 10 mM MgCl2, 100 mM KCl, and 10 mM Tris-HCl (pH 7.5), and stored at 80°C. Isolation of 16S rRNA and subfragments from T. thermophilus and E. coli Water (400 mL) was added to 100 mL of 30S ribosomal subunits and the rRNA was extracted with 500 mL phenol, phenol/ chloroform, and chloroform. rRNA was ethanol precipitated and dissolved in water. Purification of 16S rRNA subfragments was performed as previously described (Andersen et al. 2004). Briefly, 16S rRNA was hybridized to an excess of oligodeoxynu- cleotide complementary to either the region 944–990 or the region 1378–1432. Single-stranded nucleic acids were digested with Mung Bean Nuclease and RNase A. The resulting mixture was separated on a polyacrylamide gel. Bands were visualized by ethid- ium bromide staining, excised, and eluted. E. coli CP79 with the endogenous rsmF inactivated, but com- plemented with the T. thermophilus homolog on the plasmid pLJ102-RsmF, were grown at 37°C to an OD450 = 0.45 in 200 mL of LB medium containing 100 mg/L of ampicillin. RsmF expres- sion was induced by addition of IPTG to 1 mM, and incubation for another 3 h. Cells were harvested by centrifugation at 4°C, washed in 100 mL TMN buffer (50 mM Tris-HCl [pH 7.8], 10 mM magnesium acetate, 100 mM NH4Cl), and resuspended in 2 mL TMN buffer prior to lysis by sonication (7 3 30 sec on ice) T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1593 and removal of debris by centrifugation (10 min/14,000 rpm/4°C/ microcentrifuge). Total RNA was recovered from the supernatant by phenol extraction and ethanol precipitation. A 16S rRNA subfragment was isolated as described above using an oligodeoxy- nucleotide complementary to the region 1378–1432. In vitro methylation Reactions contained 50 pmol of 16S rRNA, 30S subunits, or 70S ribosomes from the T. thermophilus TTHA1387 null mutant as the substrate in a total volume of 100 mL (containing 100 mM NH4Cl, 10 mM MgCl2, 40 mM Hepes [pH 7.5]), 6 mM b-mercaptoetha- nol, and 10% glycerol (prepared as a two times concentrated stock solution), 1.5 mM S-adenosyl methionine, and 2 mg of recombi- nantly expressed RsmF (see below). For the reaction at 70°C, water and stock buffer were mixed and left at room temperature for 15 min. Then a substrate, an enzyme and S-adenosyl methionine were added and incubated at 70°C for 1 h. The 37°C reaction was started by mixing water and buffer followed by 15 min at room temperature; the substrate was added and the mixture transferred to 50°C for 5 min. After cooling to 37°C, S-adenosyl methionine and an enzyme were added and the incubation continued for 1 h. Reactions were stopped by phenol/ chloroform extraction and the rRNA was recovered by ethanol precipitation before purification of 16S rRNA subfragments as described above. Control reactions without enzyme or S-adenosyl methionine were carried out in all instances. RNase T1 digestion and mass spectrometry A purified 16S rRNA subfragment (1–2 pmol) was incubated with 2 units RNase T1 (Roche) and 50 mM 3-hydroxypicollinic acid (3-HPA) in a total volume of 2 mL for 4 h at 37°C. MALDI mass spectrometry was performed either on an ABI voyager STR in- strument or a Waters Q-TOF MALDI instrument; MALDI tan- dem mass spectrometry was done on a Waters Q-TOF MALDI instrument. All spectra were recorded in positive ion mode using 3-HPA as the matrix. Experimental details were as previously de- scribed (Douthwaite and Kirpekar 2007). Protein expression and purification for crystallization E. coli BL21 (DE3) (Invitrogen) containing pLJ102-RsmF was grown to midlog phase in LB media at 37°C in the presence of 200 mg/mL ampicillin. Protein expression was induced at 20°C with 400 mM IPTG. Cells were pelleted after 18 h by centrifuga- tion at 4000 rpm for 20 min at 4°C and lysed by ultrasonication on ice in a buffer containing 20 mM Tris-HCl (pH 8.5), 300 mM NaCl, 5 mM b-mercaptoethanol, 0.1% Triton X-100, and 5% glycerol. Cell debris and membranes were pelleted by centrifuga- tion at 11,000 rpm for 30 min at 4°C. The soluble E. coli proteins were precipitated by heat treatment at 65°C for 30 min and pelleted by centrifugation at 11,000 rpm at 4°C for 30 min. Soluble C-terminally hexahistidine-tagged T. thermophilus RsmF was further purified by affinity chromatography with nickel- nitrilotriacetic acid resin (Qiagen). Untagged proteins were re- moved with buffer containing 20 mM Tris-HCl (pH 8.5), 250 mM NaCl, and 1 mM imidazole (pH 8.5). Recombinant RsmF was then eluted with the same buffer containing 150 mM imidazole. The protein was then purified by cation exchange chromatogra- phy (SP) (GE Healthcare) at pH 8.5, using a linear gradient of 10 mM to 1 M NaCl concentration. RsmF fractions were pooled and concentrated and applied to a size-exclusion S200 column (GE Healthcare) equilibrated with a buffer containing 20 mM Tris-HCl (pH 8.5) and 200 mM NaCl. Purified RsmF was concentrated to 13 mg/mL for crystallization trials. The C-terminal hexahistidine tag was not removed for crystallization. For the production of selenomethionyl proteins, the expression construct was trans- formed into B834 (DE3) cells (Novagen). The bacterial growth was carried out in defined LeMaster medium (Hendrickson et al. 1990), and the protein was purified using the same protocol as for the unmodified protein. To form the RsmF-AdoMet complex, purified RsmF was mixed with 4 mM AdoMet incubated at 60°C for 15 min and slowly cooled to room temperature before performing crystallization experiments. Crystallization of RsmF All crystals were obtained using the microbatch technique under oil at 4°C. To obtain the RsmF1 crystal form, 1 mL of protein solution was mixed with the reservoir solution containing 20% (w/v) PEG3350 and 200 mM sodium sulfate decahydrate (pH 6.6). Initial crystals grew over the course of 1–2 wk with maxi- mum dimensions of 0.3 3 0.3 3 0.2 mm. To obtain the RsmF2 crystal form, 1 mL of the RsmF–AdoMet complex was mixed with the reservoir solution containing 200 mM NaCl, 12% w/v PEG8000 and 100 mM HEPES-KOH (pH7.5). Initial crystals grew over the course of 2–3 wk with maximum dimensions of 0.1 3 0.4 3 0.4 mm. To obtain the RsmF3 crystal form, 1 mL of the RsmF–AdoMet complex was mixed with the reservoir solution containing 10% w/v PEG1000, 200 mM NaCl, and 100 mM Tris- HCl (pH 8.5). The initial crystals grew over the course of 1–2 wk with maximum dimensions of 0.05 3 0.3 3 0.4 mm. To obtain the RsmF4 crystal form, 1 mL of the RsmF–AdoMet complex solution was mixed with a reservoir solution containing 160 mM magnesium chloride hexahydrate, 80 mM Tris-HCl (pH 8.5), and 24% w/v PEG4000. Initial crystals grew over the course of 1–2 wk with maximum dimensions of 0.05 3 0.3 3 0.3 mm. RsmF1 crystals were gradually dehydrated by increasing the PEG3350 to 30% w/v and then cryoprotected in a mother liquor supplemented with 25% v/v glycerol and then flash-frozen by being plunged into liquid nitrogen. RsmF2 crystals were cryoprotected in a mother liquor supplemented with 20% v/v ethylene glycol and then flash- frozen by being plunged into liquid nitrogen. RsmF3 crystals were cryoprotected by gradually increasing the concentration of PEG1000 to 30% and then flash-frozen by being plunged into liquid nitrogen. RsmF4 crystals were cryoprotected in a mother liquor supplemented with 20% glycerol and then flash-frozen by being plunged into liquid nitrogen. Data collection X-ray diffraction data for RsmF1, RsmF2, and RsmF4 crystals were collected on a MAR CCD detector at the X4C beamline of the National Synchrotron Light Source in Brookhaven at a wave- length of 0.979 A˚ and 180°C. Diffraction data for RsmF3 crystals were collected on an ADSC CCD detector at the X4A beamline of the National Synchrotron Light Source in Brookhaven at a wave- length of 0.979 A˚ and 180°C. Diffraction data for RsmF1 in space group P43 were collected to 1.4 A˚ resolution with cell Demirci et al. 1594 RNA, Vol. 16, No. 8 dimensions a = 71.0 A˚ , b = 71.0 A˚ , and c = 186.7 A˚ . Diffraction data to 1.82 A˚ for RsmF2 were collected in space group P2 with cell dimensions a = 66.0 A˚ , b = 78.3 A˚ , and c = 108.1 A˚ . Diffraction data to 1.29 A˚ for RsmF3 were collected in space group P21212 with cell dimensions a = 89.7 A˚ , b = 109.0 A˚ , and c = 51.0 A˚ . Diffraction data to 1.68 A˚ for RsmF4 were collected in space group P21212 with cell dimensions a = 89.8 A˚ , b = 109.1 A˚ , and c = 50.8 A˚ . A single crystal was used for each data set. The diffraction images were processed and scaled with the HKL2000 package (Otwinowski and Minor 1997). The data processing statistics are summarized in Table 1. Structure determination and refinement The RsmF structure was solved by molecular replacement with the program Phaser (McCoy et al. 2007) from the CCP4 program suite (Bailey 1994) in space group P43 to 1.4 A˚ resolution (data set RsmF1). The initial search model was built with the program Modeller (Eswar et al. 2008) from the catalytic domain of E. coli YebU (Pdb code 2FRX). After the placement of two RsmF catalytic domains in the asymmetric unit and the initial re- finement with Refmac (Murshudov et al. 1997), the model was further rebuilt with ARP/wARP (Langer et al. 2008). The resulting model was 90% complete and manually checked and completed with Coot (Emsley and Cowtan 2004). Final crystallographic re- finement was performed with the program Phenix (Adams et al. 2002). The other crystal forms were subsequently solved by molecular replacement. The atomic coordinates from the RsmF4 model were then used for initial refinement of the RsmF–AdoMet complex structure in space group P21212 (RsmF3). There are two molecules in the asymmetric unit in data sets RsmF1 and RsmF2, and one molecule in RsmF3 and RsmF4. The crystallographic R/Rfree factors are 0.17/0.19, 0.16/0.19, 0.18/0.19, and 0.17/0.19 for the four data sets: RsmF1, RsmF2, RsmF3, and RsmF4, respectively. The stereochemical quality of the model was assessed with Procheck (Laskowski et al. 1993). The Ramachandran sta- tistics (most favored/additionally allowed/generously allowed/ disallowed) are 91.9%/8.1%/0.0%/0.0% for RsmF1, 91.9%/8.1%/ 0.0%/0.0% for RsmF2, 93.6%/6.4%/0.0%/0.0% for RsmF3, and 92.5%/7.5%/0.0%/0.0% for RsmF4. The refinement statistics are summarized in Table 1. Figures were generated using Pymol (DeLano 2002). Atomic coordinates Coordinates and structure factors have been deposited in the Protein Data Bank with accession codes 3M6U, 3M6V, 3M6W, and 3M6X for data sets RsmF1, RsmF2, RsmF3, and RsmF4, respectively. SUPPLEMENTAL MATERIAL Supplemental material can be found at http://www.rnajournal.org. ACKNOWLEDGMENTS We thank John Schwanof and Randy Abramowitz for access to the X4A and X4C beamlines at the National Synchrotron Light Source. This work was supported by grants GM19756 and GM19756-37S1 from the National Institutes of Health. Received January 14, 2010; accepted April 26, 2010. REFERENCES Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Terwilliger TC. 2002. PHENIX: Building new software for automated crystallo- graphic structure determination. Acta Crystallogr D Biol Crystallogr 58: 1948–1954. Agris PF. 2004. Decoding the genome: A modified view. Nucleic Acids Res 32: 223–238. Alian A, Lee TT, Griner SL, Stroud RM, Finer-Moore J. 2008. Structure of a TrmA–RNA complex: A consensus RNA fold con- tributes to substrate selectivity and catalysis in m5U methyltrans- ferases. Proc Natl Acad Sci 105: 6876–6881. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic local alignment search tool. J Mol Biol 215: 403–410. Andersen NM, Douthwaite S. 2006. YebU is a m5C methyltransferase specific for 16 S rRNA nucleotide 1407. J Mol Biol 359: 777–786. Andersen TE, Porse BT, Kirpekar F. 2004. A novel partial modifica- tion at C2501 in Escherichia coli 23S ribosomal RNA. RNA 10: 907–913. Bailey S. 1994. The CCP4 Suite: Programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 50: 760–763. Behm-Ansmant I, Urban A, Ma X, Yu YT, Motorin Y, Branlant C. 2003. The Saccharomyces cerevisiae U2 snRNA:pseudouridine- synthase Pus7p is a novel multisite-multisubstrate RNA:C- synthase also acting on tRNAs. RNA 9: 1371–1382. Behm-Ansmant I, Branlant C, Motorin Y. 2007. The Saccharomyces cerevisiae Pus2 protein encoded by YGL063w ORF is a mitochon- drial tRNA:C27/28-synthase. RNA 13: 1641–1647. Cameron DM, Gregory ST, Thompson J, Suh MJ, Limbach PA, Dahlberg AE. 2004. Thermus thermophilus L11 methyltransferase, PrmA, is dispensable for growth and preferentially modifies free ribosomal protein L11 prior to ribosome assembly. J Bacteriol 186: 5819–5825. DeLano WL. 2002. The PyMol molecular graphics system . DeLano Scientific, San Carlos, CA. Demirci H, Gregory ST, Dahlberg AE, Jogl G. 2007. Recognition of ribosomal protein L11 by the protein trimethyltransferase PrmA. EMBO J 26: 567–577. Demirci H, Gregory ST, Dahlberg AE, Jogl G. 2008a. Crystal structure of the Thermus thermophilus 16 S rRNA methyltransferase RsmC in complex with cofactor and substrate guanosine. J Biol Chem 283: 26548–26556. Demirci H, Gregory ST, Dahlberg AE, Jogl G. 2008b. Multiple-site trimethylation of ribosomal protein L11 by the PrmA methyl- transferase. Structure 16: 1059–1066. Demirci H, Belardinelli R, Seri E, Gregory ST, Gualerzi C, Dahlberg AE, Jogl G. 2009. Structural rearrangements in the active site of the Thermus thermophilus 16S rRNA methyltransferase KsgA in a bi- nary complex with 59-methylthioadenosine. J Mol Biol 388: 271– 282. Douthwaite S, Kirpekar F. 2007. Identifying modifications in RNA by MALDI mass spectrometry. Methods Enzymol 425: 1–20. Emsley P, Cowtan K. 2004. Coot: Model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132. Eswar N, Eramian D, Webb B, Shen MY, Sali A. 2008. Protein structure modeling with MODELLER. Methods Mol Biol 426: 145– 159. Foster PG, Nunes CR, Greene P, Moustakas D, Stroud RM. 2003. The first structure of an RNA m5C methyltransferase, Fmu, provides insight into catalytic mechanism and specific binding of RNA substrate. Structure 11: 1609–1620. Gao YG, Selmer M, Dunham CM, Weixlbaumer A, Kelley AC, Ramakrishnan V. 2009. The structure of the ribosome with elongation factor G trapped in the post-translocational state. Science 326: 694–699. T. thermophilus 16S rRNA methyltransferase RsmF www.rnajournal.org 1595 Gustilo EM, Vendeix FA, Agris PF. 2008. tRNA’s modifications bring order to gene expression. Curr Opin Microbiol 11: 134–140. Gu XR, Gustafsson C, Ku J, Yu M, Santi DV. 1999. Identification of the 16S rRNA m5C967 methyltransferase from Escherichia coli. Biochemistry 38: 4053–4057. Guymon R, Pomerantz SC, Crain PF, McCloskey JA. 2006. Influence of phylogeny on posttranscriptional modification of rRNA in thermophilic prokaryotes: The complete modification map of 16S rRNA of Thermus thermophilus. Biochemistry 45: 4888–4899. Hallberg BM, Ericsson UB, Johnson KA, Andersen NM, Douthwaite S, Nordlund P, Beuscher AE 4th, Erlandsen H. 2006. The structure of the RNA m5C methyltransferase YebU from Escherichia coli reveals a C-terminal RNA-recruiting PUA domain. J Mol Biol 360: 774–787. Hashimoto Y, Yano T, Kuramitsua S, Kagamiyama H. 2001. Disrup- tion of Thermus thermophilus genes by homologous recombination using a thermostable kanamycin-resistant marker. FEBS Lett 506: 231–234. Helser TL, Davies JE, Dahlberg JE. 1972. Mechanism of kasugamycin resistance in Escherichia coli. Nat New Biol 235: 6–9. Hendrickson WA, Horton JR, LeMaster DM. 1990. Selenomethionyl proteins produced for analysis by multiwavelength anomalous diffraction (MAD): A vehicle for direct determination of three- dimensional structure. EMBO J 9: 1665–1672. Holm L, Ka¨a¨ria¨inen S, Rosenstro¨m P, Schenkel A. 2008. Searching protein structure databases with DaliLite v.3. Bioinformatics 24: 2780–2781. Hoseki J, Yano T, Koyama Y, Kuramitsu S, Kagamiyama H. 1999. Directed evolution of thermostable kanamycin-resistance gene: A convenient selection marker for Thermus thermophilus. J Biochem 126: 951–956. Hur S, Stroud RM. 2007. How U38, 39, and 40 of many tRNAs become the targets for pseudouridylation by TruA. Mol Cell 26: 189–203. Lang PT, Brozell SR, Mukherjee S, Pettersen EF, Meng EC, Thomas V, Rizzo RC, Case DA, James TL, Kuntz ID. 2009. DOCK 6: Combining techniques to model RNA-small molecule complexes. RNA 15: 1219–1230. Langer G, Cohen SX, Lamzin VS, Perrakis A. 2008. Automated macromolecular model building for X-ray crystallography using ARP/wARP version 7. Nat Protoc 3: 1171–1179. Laskowski RA, MacArthur MW, Moss DS, Thornton JM. 1993. PROCHECK: A program to check the stereochemical quality of protein structures. J Appl Crystallogr 26: 283–291. Lee TT, Agarwalla S, Stroud RM. 2005. A unique RNA fold in the RumA-RNA-cofactor ternary complex contributes to substrate selectivity and enzymatic function. Cell 120: 599–611. Liu Y, Santi DV. 2000. m5C RNA and m5C DNA methyl transferases use different cysteine residues as catalysts. Proc Natl Acad Sci 97: 8263–8265. Malone T, Blumenthal RM, Cheng X. 1995. Structure-guided analysis reveals nine sequence motifs conserved among DNA amino- methyltransferases, and suggests a catalytic mechanism for these enzymes. J Mol Biol 253: 618–632. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. 2007. Phaser crystallographic software. J Appl Crystallogr 40: 658–674. McLuckey SA, Van Berkel GJ, Glish GL. 1992. Tandem mass spectrometry of small multiply charged oligonucleotides. J Am Soc Mass Spectrom 3: 60–70. Motorin Y, Grosjean H. 1999. Multisite-specific tRNA:m5C-methyl- transferase (Trm4) in yeast Saccharomyces cerevisiae: Identification of the gene and substrate specificity of the enzyme. RNA 5: 1105– 1118. Motorin Y, Keith G, Simon C, Foiret D, Simos G, Hurt E, Grosjean H. 1998. The yeast tRNA:pseudouridine synthase Pus1p displays a multisite substrate specificity. RNA 4: 856–869. Murshudov GN, Vagin AA, Dodson EJ. 1997. Refinement of macro- molecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystallogr 53: 240–255. O’Farrell HC, Scarsdale JN, Rife JP. 2004. Crystal structure of KsgA, a universally conserved rRNA adenine dimethyltransferase in Escherichia coli. J Mol Biol 339: 337–353. Ogle JM, Murphy FV, Tarry MJ, Ramakrishnan V. 2002. Selection of tRNA by the ribosome requires a transition from an open to a closed form. Cell 111: 721–732. Otwinowski Z, Minor W. 1997. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol 276: 307–326. Pe´rez-Arellano I, Gallego J, Cervera J. 2007. The PUA domain—a structural and functional overview. FEBS J 274: 4972–4984. Purta E, O’Connor M, Bujnicki JM, Douthwaite S. 2008. YccW is the m5C methyltransferase specific for 23S rRNA nucleotide 1962. J Mol Biol 383: 641–651. Selmer M, Dunham CM, Murphy FV 4th, Weixlbaumer A, Petry S, Kelley AC, Weir JR, Ramakrishnan V. 2006. Structure of the 70S ribosome complexed with mRNA and tRNA. Science 313: 1935– 1942. Sunita S, Tkaczuk KL, Purta E, Kasprzak JM, Douthwaite S, Bujnicki JM, Sivaraman J. 2008. Crystal structure of the Escherichia coli 23S rRNA:m5C methyltransferase RlmI (YccW) reveals evolutionary links between RNA modification enzymes. J Mol Biol 383: 652–666. Wimberly BT, Brodersen DE, Clemons WM Jr, Morgan-Warren RJ, Carter AP, Vonrhein C, Hartsch T, Ramakrishnan V. 2000. Structure of the 30S ribosomal subunit. Nature 407: 327–339. Demirci et al. 1596 RNA, Vol. 16, No. 8
3M6Z
Crystal structure of an N-terminal 44 kDa fragment of topoisomerase V in the presence of guanidium hydrochloride
Structures of minimal catalytic fragments of topoisomerase V reveals conformational changes relevant for DNA binding Rakhi Rajan*, Bhupesh Taneja*,†, and Alfonso Mondragón*,‡ * Department of Biochemistry, Molecular Biology and Cell Biology, Northwestern University, 2205 Tech Dr, Evanston, IL 60208 Summary Topoisomerase V is an archaeal type I topoisomerase that is unique among topoisomerases due to presence of both topoisomerase and DNA repair activities in the same protein. It is organized as an N-terminal topoisomerase domain followed by 24 tandem helix hairpin helix (HhH) motifs. Structural studies have shown that the active site is buried by the (HhH) motifs. Here we show that the N-terminal domain can relax DNA in the absence of any HhH motifs and that the HhH motifs are required for stable protein-DNA complex formation. Crystal structures of various topoisomerase V fragments show changes in the relative orientation of the domains mediated by a long bent linker helix, and these movements are essential for the DNA to enter the active site. Phosphate ions bound to the protein near the active site helped model DNA in the topoisomerase domain and shows how topoisomerase V may interact with DNA. Introduction DNA topoisomerases are enzymes found in all forms of life (bacteria, eukarya, and archaea) and they regulate the topological state of DNA inside the cell. They form a transient break in a single or double stranded DNA and allow the passage of another single or double DNA strand through the break, before resealing the break (Champoux, 2001) (Schoeffler and Berger, 2008). As a result of this, topoisomerases can relax supercoiled DNA, help in the segregation of DNA strands following replication, and lead to the formation and resolution of knots and catenates (Gellert, 1981). Topoisomerases participate in many aspects of DNA metabolism, such as replication, recombination, and transcription (Champoux, 2001). In addition, they are targets of various anti-cancerous drugs and anti-bacterial agents (Maxwell, 1999; Pommier, 1998; Rothenberg, 1997; Wang et al., 1997). DNA topoisomerases are broadly classified into two types, type I and type II enzymes. Type I enzymes cleave a single strand of a DNA molecule and pass another single or double stranded DNA through the break before resealing the opening. Type II enzymes cleave both ‡Corresponding author: Phone: 847-491-7726, Fax: 847-467-6489, a-mondragon@northwestern.edu. †Present address: Institute of Genomics and Integrative Biology, CSIR, Delhi, India Protein data bank accession codes The final structure factors and coordinates of Topo-31, Topo-44 Form I, Form II, and Form III have been deposited in the Protein Data Bank with accession codes 3M7G, 3M7D, 3M6K, and 3M6Z respectively. Supplementary data Supplementary data are available at Structure Journal Online. Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. NIH Public Access Author Manuscript Structure. Author manuscript; available in PMC 2011 July 14. Published in final edited form as: Structure. 2010 July 14; 18(7): 829–838. doi:10.1016/j.str.2010.03.006. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript strands of a double stranded DNA in concert and pass another double stranded DNA through the break. Type I enzymes use the torsional energy stored in the supercoiled DNA to drive DNA relaxation and hence they do not require high energy cofactors, such as ATP, for their activity (Baker et al., 2009) Type II enzymes, on the other hand, require ATP and Mg2+ for their activity. Type I topoisomerases are further subdivided into three subtypes: IA, IB, and IC (Forterre et al., 2007). Type IA and IB enzymes have been studied extensively (Baker et al., 2009) and there is ample information available about their general mechanism of DNA relaxation and the mode of DNA binding. Type IC, on the other hand, is a relatively new subtype. Currently topoisomerase V is the only member of this family and it has been identified only in the Methanopyrus genus. Previously, topoisomerase V had been considered as a type IB enzyme based on its biochemical characteristics (Slesarev et al., 1993), but the crystal structure of an N-terminal 61 kDa of topoisomerase V (Topo-61) (Taneja et al., 2006) revealed a completely new fold without similarity to other topoisomerases or any other known protein. Furthermore, the orientation of the putative active site residues is also different from other type I topoisomerases, suggesting a different mechanism of cleavage and religation of DNA. These observations, together with the lack of sequence similarity, indicated that topoisomerase V defines a new subtype of type I enzymes (Forterre, 2006). Topoisomerase V was identified in Methanopyrus kandleri, an extremophile isolated from a deep-water ‘black smoker’ chimney in the Gulf of California (Huber et al., 1989). The enzyme is active at very high temperatures (122°C) and high salt concentrations (0.65 M NaCl and 3.1 M potassium glutamate). The unusual characteristic of topoisomerase V is that it has both topoisomerase and DNA repair activities in the same polypeptide (Belova et al., 2001). Based on the sequence analysis of topoisomerase V, it has been predicted that the protein contains 24 helix-hairpin-helix (HhH) DNA binding motifs arranged as 12 (HhH)2 domains around the N-terminal topoisomerase domain (Belova et al., 2002) (Figure 1A). Some of these (HhH)2 domains are involved in the apurinic/apyrimidinic (AP) site- processing activity, but the exact location of the repair active site is not known yet. Topoisomerase V can relax both positively and negatively supercoiled DNA without the need for metal cations or high energy cofactors. Single molecule experiments have shown that topoisomerase V relaxes DNA by a constrained swiveling mechanism, relaxing around 12 turns of DNA per relaxation cycle (Taneja et al., 2007). Type IB enzymes, which also use a constrained swiveling mechanism for DNA relaxation, relax around 19 turns of DNA per relaxation cycle (Koster et al., 2005). The structure of Topo-61 showed that the topoisomerase domain is mainly alpha helical and that the first four (HhH)2 domains curl around the topoisomerase domain (Taneja et al., 2006) (Figure 1B). The topoisomerase and (HhH)2 domains are joined by a long bent helix, termed the “linker helix”. Three of the five putative active site residues are present in a helix-turn-helix (HTH) domain and the other two are present in an intervening loop and a helix. The active site residues are buried by the first (HhH)2 domain and it has been suggested that large conformational changes will be needed for the DNA to access the active site of topoisomerase V (Taneja et al., 2007). Here we present data that shows that the N- terminal 31 kDa fragment of topoisomerase V (Topo-31) has topoisomerase activity, consistent with previous predictions based on the structure. In addition, we show that the Topo-44 fragment (N-terminal 44 kDa fragment of topoisomerase V) can form a stable protein-DNA complex, emphasizing the need of the (HhH)2 domains for binding DNA. We determined a crystal structure of (Topo-31) fragment, which has only the topoisomerase domain, and three different crystal structures of the Topo-44 fragment, which includes the topoisomerase domain and three tandem HhH motifs. In all structures, the topoisomerase domain is very similar. In contrast, the structures of Topo-44 show conformational changes in the linker helix resulting in variable orientations of the (HhH)2 domains when compared Rajan et al. Page 2 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript to the Topo-61 structure. Phosphate ions are present in the vicinity of the topoisomerase active site in two of the Topo-44 structures. Some of the catalytic residues interact with the phosphate ions and may mimic contacts with DNA. These observations suggest that the movement of the (HhH)2 domains is mediated by the linker helix and helps expose the topoisomerase active site to facilitate DNA binding. In addition, the location of the phosphate ions suggests a possible path for the DNA and the way the active site residues interact with it. Results The topoisomerase domain can relax DNA DNA relaxation assays using different topoisomerase V fragments showed that the topoisomerase domain alone is capable of relaxing DNA. Topoisomerase V fragments with different numbers of (HhH)2 domains, Topo-31, Topo-44, and Topo-78, were studied using relaxation assays. Topo-31 has no (HhH)2 domains, Topo-44 has one full and one partial (HhH)2 domain, while Topo-78 has eight full (HhH)2 domains, including a putative DNA repair domain. In addition to standard conditions, the effect of different pH conditions and presence of magnesium ions were also tested. The experiments show that Topo-31 is capable of relaxing DNA, despite the absence of the (HhH)2 domains (Figure 2B). A pH profile analysis for the DNA relaxation assays showed that Topo-78 relaxes DNA over a wider pH range (pH 5 to 9), while Topo-31 and Topo-44 relax DNA optimally at pH 5 (Figure 2A, 2B, 2C). In addition, magnesium is not required for the reaction, but stimulates it at all pH values (Figure 2B, 2C). Topo-78 can relax DNA to the same extent with lower amounts of protein (0.1 μg/reaction) compared to Topo-44 (~1.5 μg/reaction) and Topo-31 (~9 μg/reaction). This could be due to the enhanced DNA binding facilitated by the (HhH)2 domains. Together, these results suggest that, even though the (HhH)2 domains are dispensable for topoisomerase activity, they enhance DNA relaxation activity. In addition, the pH dependence of the DNA relaxation activity indicates that the reaction is likely to involve side chains with ionizable groups in the low pH range, such as glutamates. Finally, the magnesium independence of the reactions confirms that even the smallest fragments do not require metals for activity, although magnesium has a stimulatory effect. This may be due to favorable interactions of the cations with DNA. The (HhH)2 domains enhance DNA binding affinity EMSA experiments with different fragments of topoisomerase V and DNA showed that (HhH)2 domains could help in the formation of a stable protein-DNA complex. Various topoisomerase V fragments (Topo-31, Topo-44, and Topo-78) and single and double stranded DNA were analyzed by EMSA experiments. Topo-44 and Topo-78 formed stable complexes with a 39mer double stranded DNA (Figure 2D), while no DNA binding was observed for the Topo-31 fragment (data not shown). These observations indicate that (HhH)2 domains are necessary for a stable protein-DNA complex and that as few as one and half (HhH)2 domains are enough for formation of a stable protein-DNA complex. EMSA with single stranded DNA showed that Topo-31 and Topo-44 cannot bind to single stranded DNA, while Topo-78 can bind to single stranded DNA (data not shown). Overall Structures The topoisomerase domain of topoisomerase V is a helical-rich compact domain that has no structural similarity to any other known protein. The only recognizable structural element is a HTH that contains some of the active site residues. Not surprisingly, the topoisomerase domain of the four structures (Topo-31, Topo-44 (Forms I, II, and III)) superimpose very well on each other and also to that from the Topo-61 structure. In the Topo-31 structure, two surface loops, residues 39-49 and 120-124, adopt a different conformation compared to the Rajan et al. Page 3 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Topo-61 and Topo-44 structures. These two loops are not always visible in the Topo-44 structures, suggesting that they are mobile regions. The r.m.s.d. for the superposition of the topoisomerase core domain of all the new structures on to the Topo-61 structure range from 0.2 Å to 0.7 Å if the two mobile surface loops are not included (Figure 3A). In general, the topoisomerase domain remains unchanged and is identical in all structures. The (HhH)2 domains also remain largely unchanged, with r.m.s.d. for the superposition of only the (HhH)2 domains from the three Topo-44 crystal forms and equivalent domains in the Topo-61 structure ranging from 0.31 Å to 0.56 Å. The five crystallographically independent structures of Topo-44 (Form I, Form II A and B monomers, and Form III A and B monomers) were compared with each other and to the two crystallographically independent Topo-61 monomers to understand the conformational changes in the protein. The r.m.s.d. for the superposition of all the Topo-44 structures (residues 3-375) on to the Topo-61 fragment or on each other vary between 0.9 Å and 2.7 Å, with the majority above 1.5 Å, showing that in general the structures have slightly different conformations. As mentioned above, the different domains behave as rigid or almost rigid subunits and the only change in the structure is the relative orientation between the topoisomerase and the (HhH)2 domains. The change in orientation of the domains starts at the linker helix (residues 269-295), which acts as a hinge region, and follows into the (HhH)2 domains. At the start of the linker helix, the structures superimpose very well for all five Topo-44 and two Topo-61 structures. In the middle of the linker helix there is a kink after which the linker helix from all the structures shows different orientations (Figure 3B). The flexibility of the linker helix is also evident by the fact that the linker helix in the B subunit of Form III crystals appears in two alternate conformations. The change in the relative orientation of the (HhH)2 and topoisomerase domains (Figure 3C and 3D), suggests that these domains can adopt different orientations and these movements might be necessary for the DNA to access the active site. The topoisomerase domain has a positively charged groove adjacent to the active site The structure of the Topo-31 as well as the structures of the Topo-44 fragment reveals the presence of a positively charged groove in the protein that encompasses the active site region (shown later in Figure 6C). This charged groove had been observed before in the structure of the Topo-61 fragment, although several (HhH)2 motifs partially obstruct it (Taneja et al., 2006). The structure of the Topo-31 confirms the presence of the groove even in the absence of the (HhH)2 motifs. The groove is long and can be deep in some areas. It includes regions of the HTH motifs and extends all the way to the linker helix. All the residues forming the active site pentad point towards the groove. The active site tyrosine, Tyr226, is found near one of the ends of the groove, a region where it widens. The positively charged character of the groove and its presence by the active site strongly suggest that it may be involved in DNA binding. Phosphate ions bind in the groove near the topoisomerase active site An interesting observation stemming from the Form II and Form III Topo-44 structures is the presence of phosphate ions near the positively charged DNA binding groove. All three Topo-44 crystal forms were crystallized in the presence of phosphate-citrate buffer, but only Form II and Form III structures showed phosphate ions bound to the protein, which were assigned based on electron density consistent with a tetrahedral phosphate ion (Figure 4A). Form II and Form III crystals include 1–1.2 M guanidium hydrochloride in the crystallization solution. The high resolution Form III structure shows clear density for three guanidium ions bound to the protein, two very well ordered and one with weak density. The presence of guanidium hydrochloride in the crystals appears to trigger a conformational change allowing the binding of phosphate ions to the protein. It is interesting to note that Rajan et al. Page 4 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Form I crystals did not show any bound phosphate albeit its presence in the crystallization condition. This could be due to the absence of guanidium hydrochloride to trigger the binding of phosphate ions as observed in Form II and Form III structures. There are three phosphate ions in the B subunit and none in the A subunit of the Form II Topo-44 structure. Two of the phosphates are in the topoisomerase active site and one of them forms close contacts with the putative active site residues in the topoisomerase domain (Figure 4B). Form III crystal has seven phosphate ions, three in each subunit and one between both the subunits. In the Form III structure, the phosphate ion near the active site Tyr226 is absent, but it shows several new locations for phosphate ions, especially in the positively charged groove containing the topoisomerase active site (Figure 5A). An overlay of the A and B subunits of the Topo-44 Form III structure with the B subunit of Topo-44 Form II structure shows eight unique phosphate ions (Figure 5A). It clearly shows that there are more phosphate ions bound in the positively charged groove compared to other regions of the protein. Taking into account all structures, there are five unique phosphate ion binding sites in the putative DNA binding groove and an additional one near its end and close to the start of the linker helix. Several pairs of phosphates in the groove are separated by a distance of around 7 Å (Figure 5B), which would be consistent with the phosphate-phosphate distance in adjacent nucleotides in a DNA double helix. One of the phosphates (P1) is found near the active site tyrosine and is coordinated by Tyr226 and Arg131, two residues that have been implicated in cleavage and religation of the DNA (Taneja et al., 2006), and by Glu215, whose charge may be important for interactions with DNA (R.R. and A.M., unpublished observations). The side chains of the tyrosine and the glutamate residues are in contact with Arg144 and His200, the other putative active site residues, and these interactions may help to orient them for the catalytic reaction. Adjacent to P1, there is a second phosphate (P2) at a distance of 7.5 Å which is trapped between the topoisomerase domain and an HhH motif. P2 is coordinated by Arg131, an active site residue, in addition to Arg108 from the topoisomerase domain and Arg293 and serines 322 and 324 from the second HhH motif (Figure 6C). Three more phosphates are found in the groove (P3, P4, and P5) coordinated mainly by positively charged residues, such as Arg37, Lys47, Arg108, Lys134, and Arg135 from the topoisomerase domain and also residues from the linker helix such as Tyr289 and Arg293. In general, some of the side chains can contact more than one phosphate. The distance between P3 and P4 and P4 and P5 is 6.8 Å and 6.5 Å respectively. A final phosphate (P6) is located at the start of the linker helix and on the edge of the groove (Figure 5A). Discussion Topoisomerase V is active at very high temperatures (122°C) and high salt concentrations. DNA relaxation assays with various topoisomerase V fragments (Topo-44 and Topo-61) show that a temperature above 60° C is required for optimal activity, although longer fragments of topoisomerase V can relax DNA at lower temperatures (Taneja et al., 2007). Topo-44 was first identified by limited proteolytic digestion of the full length topoisomerase V protein (Belova et al., 2002) at 80°C. In contrast, Topo-61 is the shortest fragment showing topoisomerase activity when the proteolytic reaction is performed at 37°C (Belova et al., 2002). The N-terminal Topo-31 fragment, which contains neither HhH motifs nor the linker helix, was identified as the smallest region spanning the topoisomerase domain from the crystal structure of Topo-61 fragment (Taneja et al., 2006) and it was suggested that it could represent the minimal domain capable of relaxing DNA. Relaxation experiments with this minimal domain show that this is indeed the case, although the activity is not as robust as with longer fragments. As expected, Topo-31 does not require magnesium for activity, but magnesium enhances it, as is the case for type IB topoisomerases, which also uses a Rajan et al. Page 5 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript swiveling mechanism for DNA relaxation (Stewart et al., 1996). The optimal pH for activity for the Topo-31 and the Topo-44 fragments is around 5. This pH dependence is not observed for the Topo-78 fragment. The DNA relaxation by shorter fragments of topoisomerase V at pH 5 could point to the involvement of some ionizable side chains in the relaxation activity. It could also be simply due to the effects of various side chains on DNA binding. Further experiments with different active site mutations in both longer and shorter fragments of topoisomerase V will be required to probe the pH dependence of the relaxation reaction by shorter topoisomerase V fragments. Gel shift experiments show that Topo-44 and also longer fragments (Topo-78) can bind double stranded DNA. Surprisingly, Topo-31 does not show DNA binding activity in these assays even though it is still capable of relaxing DNA. It appears that the presence of the (HhH)2 domains stabilizes the DNA/protein complex. One possibility is that the (HhH)2 domains could play a similar role to the cap domain present in type IB enzymes, which helps to encircle the DNA during the swiveling reaction (Redinbo et al., 1998). In addition, both short fragments of topoisomerase V do not bind single stranded DNA, whereas Topo-78 can form a stable complex with single stranded DNA (data not shown). (HhH)2 domains binding to single stranded DNA has been observed before. For instance, the N-terminal 8 kDa of mammalian polymerase β, which contains a single HhH motif, binds to single stranded DNA through both helices (Kumar et al., 1990; Liu et al., 1994). The exact mode of single stranded DNA binding by Topo-78 or the possible role in relaxation or repair activities is not yet clear. The structure of Topo-61 showed that the topoisomerase active site of topoisomerase V is buried by one of the (HhH)2 domains suggesting that conformational changes are essential for the protein to bind DNA. The present structures of Topo-44 reinforce this observation and show that the (HhH)2 domains can change their position relative to the topoisomerase domain and that this change is mediated by the movement of the linker helix. The (HhH)2 domains act as rigid individual units, as evidenced by the fact that in different structures they show the same structure and relative orientation of the two HhH motifs. The topoisomerase domain also appears to be rigid showing the same structure even in the total absence of the rest of the protein. The linker helix (residues 269-295), which is a long bent helix, serves as a hinge for the movement of the (HhH)2 domains away from the rigid topoisomerase domain, possibly by responding to interactions with double stranded DNA. This movement has to be quite large. The Topo-44 structures in the absence of DNA capture the regions that move, but do not show the full extent of the movement or indicate the way the HhH motifs interact with DNA. As mentioned before, topoisomerase V binds double stranded DNA and has a groove wide enough to accommodate double strand DNA (Figure 6C). The presence of an HTH domain normally associated with DNA binding, the positively charged nature of it, and several phosphates bound along it suggest that this groove could be involved in DNA binding. In addition, the active site is found in this groove and some residues form part of the HTH domain. Previously, DNA was modeled bound to the topoisomerase domain (Taneja et al., 2006) based on the structures of HTH domains in complex with DNA but there was no evidence to support it. Using the phosphates present in the groove in the current structures, it is possible to refine this model. A superposition of the B subunit of Form II and the A and B subunits of Form III Topo-44 structures shows five different phosphate ions in the positively charged groove which are separated by a distance of around 7 Å, consistent with the distance of consecutive phosphates in B DNA of ~6.4 Å. A sixth phosphate ion is found outside the groove near the linker helix. A double stranded DNA molecule was modeled into the groove based on the positions of the phosphate ions (Figure 6). Only five out of the six phosphates could be placed on the DNA molecule, as one of them was inconsistent with a Rajan et al. Page 6 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript double stranded molecule. Phosphate ions P3, P4, and P5 would correspond to three adjacent phosphates in one DNA strand, while P1, located near the active site, would belong to the opposite strand. A final phosphate (P6) is away from the groove and near the linker helix (Figure 6A, 6B). The fit to the four inner phosphates is excellent and the DNA can be accommodated in the groove of the Topo-31 structure without the need for any major rearrangements of the protein backbone. The fifth phosphate (P6) does not fit as well and a better fit would require movement of either the protein or the DNA, but the change would be relatively modest. Several side chains would need to move, but these changes would also be minor. The major change needed to accommodate the DNA in the structures with the (HhH)2 domains present is the movement of the (HhH)2 domains away from the topoisomerase domain (Figure 6B). The movement of (HhH)2 domains should be feasible as is evident from the Topo-44 structures showing different orientations of the (HhH)2 domains. The location of the (HhH)2 domains after DNA binding is not evident, but one possibility is that they would help enclose the DNA to form a clamp around it, similar to the arrangement in type IB enzymes. In the model of the topoisomerase domain in complex with DNA, the active site residues are in close contact with the backbone of DNA. The catalytic Tyr226 is pointing towards the phosphate of the DNA backbone; Arg131 and Arg144 are positioned to stabilize the protein- DNA covalent complex. Surprisingly Glu215 also appears to interact directly with the DNA phosphate backbone. The other active site residues like His200 and Lys 218 are also near the DNA. The active site is located near the end of the groove, where it widens. At this end, the DNA fits loosely in the groove, which is spacious to accommodate the movement of the strands. The ‘constrained swiveling’ mechanism employed by type IB and IC enzymes necessitates rotation of one strand about the other after forming the covalent protein-DNA intermediate. The position of the active site at the wider end of the putative DNA binding groove would facilitate the rotation of the DNA strand at this end, while holding the rest of the DNA in place through extensive interactions along the groove. Even though type IB and IC enzymes have a similar overall mechanism of action, the structures of fragments of topoisomerase V suggest many differences. Type IB enzymes have two domains which come together to form a C-shaped clamp around the DNA (Perry et al., 2006; Redinbo et al., 1998; Stewart et al., 1998) The protein has an open stage where these domains are separate and this helps in the entry and release of the DNA from the protein active site. A wide DNA binding cavity is not observed in the topoisomerase V structures. Instead, the structures show a positively charged groove which is always present in the protein and does not require domain rearrangements to form. DNA can access this groove after a conformational change involving the movement of the (HhH)2 domains exposing the active site. The (HhH)2 domains could help enclose DNA during the swiveling of the DNA, forming a similar enclosure to the one observed for type IB enzymes. It is not known whether all HhH motifs contact DNA simultaneously, but this appears unlikely without a major rearrangement of the motifs. It is likely that only some of the HhH motifs contact DNA at any given time or that some of the motifs do not have the capacity to bind DNA. Finally, similar to type IB enzymes (Cheng and Shuman, 1998), the putative domain enclosing the DNA is dispensable for activity, although it enhances the relaxation activity markedly. Thus, it is likely that type IB and IC enzymes have several overall similarities in the way that they interact with DNA, but the atomic details are markedly different. There are still many details of the atomic mechanism of type IC topoisomerases that need to be understood. The present functional and structural studies provide new information about topoisomerase V including the observations that the Topo-31 is the minimal fragment capable of DNA relaxation, the (HhH)2 domains enhance binding of the protein to DNA, the changes in relative orientation of the domains is mediated by the linker helix, and several Rajan et al. Page 7 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript phosphate ions bind in a positively charged groove. Furthermore, the position of the phosphate ions in the groove helped in the placement of DNA in the topoisomerase domain and this provides an initial model of how topoisomerase V interacts with DNA. Thus the present study helps to establish the role of different domains more clearly, to illustrate a mechanism to drive the conformational changes needed for activity, and to suggest a possible manner of binding DNA. Additional work on structures of protein/DNA complexes and intermediates in the swiveling reaction are needed to understand the way this new type of topoisomerases interacts with DNA to perform a complex reaction. Experimental Procedures Protein purification The N-terminal 31 kDa (Topo-31: residues 1-269), and 44 kDa (Topo-44: residues 1 to 380) fragments of topoisomerase V protein were cloned into the pET15b plasmid and transformed into Escherichia coli BL21 Rosetta (DE3) cells. The N-terminal 78 kDa (Topo-78: residues 1 to 685) fragment of topoisomerase V protein was cloned into the pET14b plasmid (Belova et al., 2002) and transformed into Escherichia coli BL21(DE3) cells. For protein production, cells were grown at 37° C in LB medium containing 100 μg/ml ampicillin and 100 μg/ml chloramphenicol for Rosetta cells and LB medium with 100 μg/ml ampicillin for BL21(DE3) cells to an optical density (OD600) of 0.6. The cells were then cooled down on ice, followed by the addition of isopropyl β-D-1-thiogalactopyranoside (IPTG) to a final 0.5 mM concentration, and grown overnight at 16° C. Cells were harvested and resuspended in 50 mM Tris pH 8, 500 mM NaCl, 0.5 mM EDTA, 1 mM DTT, flash frozen in liquid nitrogen and stored at −80° C. After thawing the pellet, pepstatin (1μg/ml), benzamidine (1mM), PMSF (1mM), and Brij 58 (0.1%) were added to the cells and the protein was purified as described earlier (Taneja et al., 2006) The protein was further purified by anion exchange and gel filtration chromatography. Pure protein was concentrated and stored in 50 mM Tris pH 8, 250 mM NaCl, and 1 mM DTT. The seleno- methionine substituted Topo-44 was prepared from cells grown in a minimal medium supplemented with nutrients and salts (Doublie, 1997); protein purification followed the same procedure as for the native protein except that 5mM DTT was used in all the purification steps and for storage. Relaxation assays Relaxation assays with the different topoisomerase V fragments were carried out at pH values ranging from 4 to 10. The pH of the buffers was adjusted at 65 °C to account for the change in pH at higher temperature. The different buffers used were: sodium acetate for pH 4 and 5, MES for pH 6, HEPES for pH 7, TRIS for pH 8, CHES for pH 9, and CAPS for pH 10. Topoisomerase activity assays were performed by incubating varying amounts of protein (Topo-31, Topo-44 or Topo-78) with 0.2 μg negatively supercoiled pUC19 DNA in 50 mM of the required buffer, 30 mM NaCl, 0.2 mM or 5 mM EDTA or 1 mM MgCl2. The reactions were carried out at 65 °C for 15 min and terminated by cooling and addition of SDS to a final 1% concentration. The products were resolved on a 1% agarose gel and visualized by ethidium bromide staining. Electrophoretic Mobility Shift Assay For Electrophoretic Mobility Shift Assay (EMSA), 4 μM of a 39mer double stranded DNA oligonucleotide (5′ GCGACGCGAGGCTGGATGGCCTTCCCCATTATGATTCTT3′) was incubated with different concentrations of topoisomerase V fragments in 50 mM sodium acetate pH 5, 30 mM NaCl, 1 mM MgCl2 at 65 °C for 30 minutes. Glycerol was added to the reaction mixture to a final concentration of 8% and the products were separated on a 4 % acrylamide native gel. The gel was stained with ethidium bromide to detect the DNA. When Rajan et al. Page 8 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript a stable protein-DNA complex was formed, there was an upward shift in the band indicating a higher molecular weight complex. Crystallization Topo-31 crystals were grown using the sitting drop vapor diffusion method equilibrated against, 23% PEG 6000, 0.1 M Na citrate pH 5.5, at 22°C. For data collection, the Topo-31 crystals were cryo-protected by adding glycerol to the mother liquor to a final 20% concentration. Topo-44 was crystallized by the hanging drop vapor diffusion method under three different crystallization conditions (Forms I, II, and III). Crystal Form I grew under 0.1 M phosphate citrate pH 5, 0.2 M NaCl, 15% PEG 3350 and 8% dioxane. The crystals were cryo-protected by increasing the PEG concentration to 30%. Form II crystals grew under 0.1 M phosphate citrate pH 5, 0.2 M NaCl, 16% PEG 8000 and 1M guanidium hydrochloride. For cryo-protection, they were transferred to a solution with 1.5X reservoir solution and 20% 2,3 butanediol or 20% DMSO for 10 seconds and immediately flash frozen under liquid nitrogen. Form III crystals grew under 0.1 M phosphate citrate pH 5.5, 0.15 M sodium sulfate, 0.01 M MgCl2, 1 M guanidium hydrochloride, and 28 % PEG 3350. The crystals were grown at 30°C and were cryo-protected by increasing the PEG concentration to 40%. Further details of crystallization are presented in the Supplementary Information. Data collection and structure determination Diffraction data were collected at the Dupont Northwestern Dow and Life Science Collaborative Access Team stations (DND and LS CAT) at the Advanced Photon Source in Argonne National Laboratory. Data collection and refinement statistics are shown in Table I. All data were processed and integrated using XDS (Kabsch, 1993) and scaled with SCALA (Collaborative-Computational-Project-4, 1994). Data on the Topo-31 crystals were collected to 2.4 Å resolution. The structure was solved by Molecular Replacement (McCoy et al., 2007) using the topoisomerase domain from the Topo-61 structure (residues 1-266) (Taneja et al., 2006) as the search model. It was refined with refmac5 (Murshudov et al., 1997) and Phenix (Afonine et al., 2005) to a final Rwork of 20.0 % and Rfree of 24.8 %. Topo-44 Form I crystals diffract to 1.8 Å. The structure of Form I crystals was solved by Molecular Replacement (McCoy et al., 2007) using the topoisomerase domain from the Topo-61 structure as the search model. Model rebuilding was performed using coot (Emsley and Cowtan, 2004), and refinement using refmac5 (Murshudov et al., 1997). The final Rwork and Rfree are 17.5 % and 22.0 % respectively. For Topo-44 Form II and Form III crystals, seleno-methionine derivatized crystals were used for single-wavelength anomalous dispersion (SAD) phasing. AutoSharp (Vonrhein et al., 2007) was used for locating the selenium atoms; model building was done using coot (Emsley and Cowtan, 2004), and refinement was carried out using refmac5 (Murshudov et al., 1997) Three phosphate ions were noticed in the Form II structure; two of which present in the topoisomerase active site and are separated by a distance of ~7.5 Å. The structure was refined to a final Rwork of 24.1 % and Rfree of 28.9 %. Topo-44 Form III crystals diffracted to 1.4 Å. The final Rwork and Rfree are 16.5 % and 18.4%, respectively. An interesting observation is the presence of both phosphate and guanidium ions in the Form III Topo-44 structure. The linker helix and part of the first HhH motif of the B monomer show alternate conformations and were built as two separate chains with occupancy of 0.5 each. Further details on data collection and structure determination are given in the Supplementary Information. All figures were made with Pymol (DeLano, 2002) and the electrostatic surfaces were calculated with APBS (Baker et al., 2001). Rajan et al. Page 9 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We acknowledge staff and instrumentation support from the Keck Biophysics Facility and the Center for Structural Biology at Northwestern University, and DND and LS-CAT at the Advanced Photon Source (APS) at Argonne National Laboratory. Support from the R.H. Lurie Comprehensive Cancer Center of Northwestern University to the Structural Biology Facility is also acknowledged. DND-CAT is supported by Dupont, DOW and the NSF. LS-CAT was supported by the Michigan Economic Development Corporation and the Michigan Technology Tri-Corridor. Use of the APS is supported by the Department of Energy (DOE). Research was supported by NIH grant GM51350 (to AM). References Afonine PV, Grosse-Kunstleve RW, Adams PD. A robust bulk-solvent correction and anisotropic scaling procedure. Acta Crystallogr D Biol Crystallogr. 2005; 61:850–855. [PubMed: 15983406] Baker NA, Sept D, Joseph S, Holst MJ, McCammon JA. Electrostatics of nanosystems: application to microtubules and the ribosome. Proc Natl Acad Sci U S A. 2001; 98:10037–10041. [PubMed: 11517324] Baker NM, Rajan R, Mondragon A. Structural studies of type I topoisomerases. Nucleic Acids Res. 2009; 37:693–701. [PubMed: 19106140] Belova GI, Prasad R, Kozyavkin SA, Lake JA, Wilson SH, Slesarev AI. A type IB topoisomerase with DNA repair activities. Proc Natl Acad Sci U S A. 2001; 98:6015–6020. [PubMed: 11353838] Belova GI, Prasad R, Nazimov IV, Wilson SH, Slesarev AI. The domain organization and properties of individual domains of DNA topoisomerase V, a type 1B topoisomerase with DNA repair activities. J Biol Chem. 2002; 277:4959–4965. [PubMed: 11733530] Champoux JJ. DNA Topoisomerases: Structure, Function, and Mechanism. Annu Rev Biochem. 2001; 70:369–413. [PubMed: 11395412] Cheng C, Shuman S. A catalytic domain of eukaryotic DNA topoisomerase I. J Biol Chem. 1998; 273:11589–11595. [PubMed: 9565576] Collaborative-Computational-Project-4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D. 1994; 50:760–763. [PubMed: 15299374] Davis IW, Murray LW, Richardson JS, Richardson DC. MOLPROBITY: structure validation and all- atom contact analysis for nucleic acids and their complexes. Nucleic Acids Res. 2004; 32:W615– 619. [PubMed: 15215462] DeLano, WL. The PyMol Molecular Graphics System. San Carlos, CA: DeLano Scientific; 2002. Diederichs K, Karplus PA. Improved R-factors for diffraction data analysis in macromolecular crystallography. Nat Struct Biol. 1997; 4:269–275. [PubMed: 9095194] Doublie S. Preparation of selenomethionyl proteins for phase determination. Methods Enzymol. 1997; 276:523–530. [PubMed: 9048379] Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] Forterre P. DNA topoisomerase V: a new fold of mysterious origin. Trends Biotechnol. 2006; 24:245– 247. [PubMed: 16650908] Forterre P, Gribaldo S, Gadelle D, Serre MC. Origin and evolution of DNA topoisomerases. Biochimie. 2007; 89:427–446. [PubMed: 17293019] Gellert M. DNA Topoisomerases. Annu Rev Biochem. 1981; 50:879–910. [PubMed: 6267993] Huber R, Kurr M, Jannasch HW, Stetter KO. A novel group of abyssal methanogenic archaebacteria (Methanopyrus) growing at 110 °C. Nature. 1989; 342:833–834. Kabsch W. Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Crystallogr. 1993; 26:795–800. Rajan et al. Page 10 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Koster DA, Croquette V, Dekker C, Shuman S, Dekker NH. Friction and torque govern the relaxation of DNA supercoils by eukaryotic topoisomerase IB. Nature. 2005; 434:671–674. [PubMed: 15800630] Kumar A, Widen SG, Williams KR, Kedar P, Karpel RL, Wilson SH. Studies of the domain structure of mammalian DNA polymerase beta. Identification of a discrete template binding domain. J Biol Chem. 1990; 265:2124–2131. [PubMed: 2404980] Liu D, DeRose EF, Prasad R, Wilson SH, Mullen GP. Assignments of 1H, 15N, and 13C resonances for the backbone and side chains of the N-terminal domain of DNA polymerase beta. Determination of the secondary structure and tertiary contacts. Biochemistry. 1994; 33:9537– 9545. [PubMed: 8068628] Maxwell A. DNA gyrase as a drug target. Biochem Soc Trans. 1999; 27:48–53. [PubMed: 10093705] McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software. J Appl Crystallogr. 2007; 40:658–674. [PubMed: 19461840] Murshudov GN, Vagin AA, Dodson EJ. Refinement of macromolecular structures by the maximum- likelihood method. Acta Crystallogr D. 1997; 53:240–255. [PubMed: 15299926] Perry K, Hwang Y, Bushman FD, Van Duyne GD. Structural basis for specificity in the poxvirus topoisomerase. Mol Cell. 2006; 23:343–354. [PubMed: 16885024] Pommier Y. Diversity of DNA topoisomerases I and inhibitors. Biochimie. 1998; 80:255–270. [PubMed: 9615865] Redinbo MR, Stewart L, Kuhn P, Champoux JJ, Hol WG. Crystal structures of human topoisomerase I in covalent and noncovalent complexes with DNA. Science. 1998; 279:1504–1513. [PubMed: 9488644] Rothenberg ML. Topoisomerase I inhibitors: review and update. Ann Oncol. 1997; 8:837–855. [PubMed: 9358934] Schoeffler AJ, Berger JM. DNA topoisomerases: harnessing and constraining energy to govern chromosome topology. Q Rev Biophys. 2008; 41:41–101. [PubMed: 18755053] Slesarev AI, Stetter KO, Lake JA, Gellert M, Krah R, Kozyavkin SA. DNA topoisomerase V is a relative of eukaryotic topoisomerase I from a hyperthermophilic prokaryote. Nature. 1993; 364:735–737. [PubMed: 8395022] Stewart L, Ireton GC, Parker LH, Madden KR, Champoux JJ. Biochemical and biophysical analyses of recombinant forms of human topoisomerase I. J Biol Chem. 1996; 271:7593–7601. [PubMed: 8631793] Stewart L, Redinbo MR, Qiu X, Hol WG, Champoux JJ. A model for the mechanism of human topoisomerase I. Science. 1998; 279:1534–1541. [PubMed: 9488652] Taneja B, Patel A, Slesarev A, Mondragon A. Structure of the N-terminal fragment of topoisomerase V reveals a new family of topoisomerases. EMBO J. 2006; 25:398–408. [PubMed: 16395333] Taneja B, Schnurr B, Slesarev A, Marko JF, Mondragon A. Topoisomerase V relaxes supercoiled DNA by a constrained swiveling mechanism. Proc Natl Acad Sci U S A. 2007; 104:14670–14675. [PubMed: 17804808] Vonrhein C, Blanc E, Roversi P, Bricogne G. Automated structure solution with autoSHARP. Methods Mol Biol. 2007; 364:215–230. [PubMed: 17172768] Wang HK, Morris-Natschke SL, Lee KH. Recent advances in the discovery and development of topoisomerase inhibitors as antitumor agents. Med Res Rev. 1997; 17:367–425. [PubMed: 9211397] Rajan et al. Page 11 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Organization of topoisomerase V Topoisomerase V is a multi-domain protein consisting of 24 helix-hairpin-helix (HhH) DNA binding motifs arranged as 12 (HhH)2 domains following the N-terminal topoisomerase domain. A) Schematic diagram of various topoisomerase V fragments. The topoisomerase domain is shown in red, the (HhH)2 domains are shown in alternating colors of cyan and yellow. The (HhH)2 domains with repair activity are shown in green. All fragments shown have topoisomerase activity, but only the full length protein and the Topo78 fragment have repair activity. B) Crystal structure of Topo-61 fragment (Taneja et al., 2006). The coloring scheme is the same as in Figure 1A, except that the linker helix is shown in grey. Rajan et al. Page 12 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. DNA relaxation activity and EMSA for Topo-31, Topo-44 and Topo-78 fragments of topoisomerase V A) pH profile of the DNA relaxation activity of Topo-78 and Topo-44 fragments. 0.2 μg of pUC19 DNA were incubated with 0.1 μg of Topo-78 or 1.5 μg of Topo-44 proteins at 65°C for 15 minutes for the relaxation reaction. The reaction buffer contained 50 mM of the appropriate buffer, 30 mM NaCl and 0.2 mM EDTA. Topo-78 relaxes DNA at a wider pH range (5 to 9) than Topo-44, which relaxes DNA efficiently only at pH 5. DNA relaxation activity of Topo-31 (B) and Topo-44 (C) fragments in the absence and presence of MgCl2. 0.2 μg of pUC19 DNA were incubated with 9 μg of Topo-31 or 1.5 μg of Topo-44 proteins at 65°C for 15 minutes for the relaxation reaction. The reaction buffer contained 50 mM of the appropriate buffer, 30 mM NaCl and 5 mM EDTA or 1 mM MgCl2. Both Topo-31 and Topo-44 fragments can relax DNA in the absence of MgCl2, but MgCl2 enhances the DNA relaxation activity of the topoisomerase V fragments. The black triangle in panels A, B and C represents increasing pH from 4 to 10 by one pH unit. D) EMSA of Topo-44 and Topo-78 fragments with a 39mer double stranded DNA. Both Topo-44 and Topo-78 form stable complexes with DNA, although Topo-78 seems to saturate DNA binding while Topo-44 does not. In addition, Topo-44 shows some cleavage of the DNA (bottom free DNA band), while the cleavage is not apparent in Topo-78. The numbers at the bottom represent the molar ratio of protein to DNA. Rajan et al. Page 13 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Structure of Topo-44 fragments A) Overlay of Form I (green), Form II (B subunit: blue), and Form III (B subunit: magenta) structures onto the Topo-61 structure (B subunit: orange). The topoisomerase domains superimpose very well for all the structures, while the linker helix and (HhH)2 domains show differences in orientation. B) Overlay of the linker helices of Form I, II, and III structures with that of Topo-61. The color scheme is same for all the figures unless mentioned otherwise. Note that the linker helices have the same orientation at the start and they change as they move further down the helix. C) Superposition of Form I, II, and III Topo-44 structures with that of Topo-61. Only the (HhH)2 domains are colored while the remaining parts are shown in gray for clarity. The active site residues are shown as orange sticks. Note that the (HhH)2 domains adopt different orientations in all the structures. D) Orientation of the (HhH)2 domains of Form I, II and Topo-61 structures. In Form I and II structures, the (HhH)2 domains are moved away from the topoisomerase domain. For clarity, the (HhH)2 domains of Form III are not shown. In panels C and D, the topoisomerase domains were superposed to emphasize the different orientation of the other domains. Rajan et al. Page 14 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. Phosphate ions present near the active site of the Topo-44 structure A) Stereo view of a Form III difference electron density map calculated with a model not including the phosphates. The electron density is contoured at 3.7σ and shows the tetrahedral shape of the phosphate ions. The active site residues are shown in stick. B) Stereo view of the interaction of the phosphate ions with the putative active site residues. The B subunit of Form II structure was superimposed onto the B subunit of Form III structure and the phosphates ions from both structures are shown together with the Form II B subunit protein backbone. The interactions made by the phosphate ion with the active site residues and the corresponding distances in Å are represented as black dotted lines. Rajan et al. Page 15 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Representation of the unique phosphate ions from Form II and Form III Topo-44 structures A) An overlay of the A (magenta) and B (brown) subunits of the Form III structure and B (blue) subunit of Form II Topo-44 structures. The positions of eight unique phosphate ions (orange spheres) are shown. Note that most phosphate ions are found along the DNA binding groove of the topoisomerase domain. B) The phosphate ions in the DNA binding groove are separated by distances of around 7 Å. The protein backbone is that of the B subunit of Form III structure. The active site residues are represented as sticks and distances in Å between adjacent phosphate ions are shown as black dotted lines. Rajan et al. Page 16 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. Model showing DNA bound to the topoisomerase domain A) Model of a 17-mer double stranded DNA bound to the Topo-31 structure (teal). The DNA is represented as green sticks, where as phosphate ions are represented as orange sticks. DNA binds along the DNA binding groove and five of the eight phosphate ions noted in the Topo-44 structures coincide with the DNA backbone. B) Model of Topo-44 (Form II, B subunit: blue) binding to 17-mer double stranded DNA. Note that the linker helix and the (HhH)2 domains interfere with DNA binding to the topoisomerase domain and are likely to move away to allow binding. C) Electrostatic surface representation of the Topo-31 structure. The positively charged DNA binding groove is clearly visible and the phosphate ions are bound in this groove. The orientation corresponds to a 90° rotation of the one shown in Figure 6A in the direction of the arrow. Note that the DNA binding groove goes from one end of the molecule to the other and it is narrower at one end (start of the linker helix) and wider at the other end. The putative active site residues (green sticks) are located at the wider end of the groove. Other residues lining the groove and interacting with the phosphate ions are shown as cyan sticks. D) Electrostatic surface representation of Topo-31 with phosphate ions (orange) and DNA (green). Three phosphate ions (P3, P4, and P5) coincide with the phosphates of one of the DNA strands, where as P1 coincides with a phosphate of the opposite DNA strand. The model shows that the DNA binding groove of topoisomerase V is wide enough to bind DNA and that the movement of linker helix and (HhH)2 domains Rajan et al. Page 17 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript are required to accommodate the DNA. The electrostatic potential was calculated with a dielectric constant of 80 for solvent and 2 for protein. The surface is colored with a blue to red gradient from +10 to −10 KbT/ec. Rajan et al. Page 18 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Rajan et al. Page 19 Table 1 Data collection and refinement statistics Topo-31 Topo-44 Form I Topo-44 Form II Topo-44 Form III Data Collection Space group C2221 C121 P41212 P212121 Cell dimensions a=106.7 Å, b=119.4 Å, c=63.7 Å a=104.2 Å, b=47.7 Å, c=81.2 Å (β=112.48) a=b=70.1 Å, c=349.6 Å a=63.6 Å, b=80.1 Å, c=137.2 Å Resolution (Å)a 79.56 – 2.4 (2.53 – 2.4) 75.05 – 1.82 (1.91 – 1.82) 29.5- 2.6 (2.72-2.6) 28.9-1.4 (1.46-1.4) Number of observed reflections 78,729 (11,538 134,411 (13,220) 227,408 (19,917) 1,157,917 (126,319) Number of unique reflections 16,259 (2,346) 32,998 (4,301) 28,151 (3,331) 136,662 (15,986) Completeness (%) 99.8 (99.8) 98.3 (88.6) 99.9 (100.0) 98.8 (95.5) Multiplicity 4.8 (4.9) 4.1 (3.1) 8.1 (6.0) 8.5 (7.9) Rmerge (%)b 4.7 (71.1) 4.0 (16.3) 7.4 (52.2) 4.5 (37.9) Rmeas (%)c 5.3 (79.6) 4.6 (19.4) 7.9 (57.2) 4.8 (40.5) ≪I>/σ(<I>)>d 20.5 (2.5) 23.0 (6.8) 19 (3.2) 27.5 (5.3) Refinement Resolution (Å) 79.56 - 2.4 (2.46 - 2.4) 28.06 -1.82 (1.87 – 1.82) 29.14 – 2.6 (2.67 – 2.6) 28.9 - 1.4 (1.44 - 1.4) Number of reflections working/test 15,419/821 31,317/1,673 26,710/1,438 129,802/6,859 Rwork (%)e 20.0(24.3) 17.5 (17.9) 24.1(36.6) 16.5 (19.3) Rfree(%)f 24.8 (31.1) 22.0 (24.8) 28.9 (45.1) 18.4 (22.1) Protein residues/atomsg 269/2,203 376/3212 727/5,970 738/7,511 Atoms in alternate conformations 0 258 (20 protein residues) 8 (1 protein residue) 2846 (157 protein residues) Water molecules 29 238 30 573 Other atoms - - 3 PO4 7 PO4, 3 Gmh, 3 Mg++, 2 Cl− B-factor (Å2) Protein atoms (chain) 68.4 22.8 A:53.8; B:58.2 A:13.4; B:14.9 Water molecules 59.1 29.3 40.0 23.7 r.m.s. deviations bond lengths (Å) 0.015 0.006 0.01 0.009 bond angles (°) 1.42 0.920 1.2 1.2 Ramachandran ploti Favored regions (%) 94.3 98.9 96.2 98.5 Outliers (%) 0.0 0.0 0.3 0 aNumbers in parenthesis correspond to highest resolution shell. bRmerge= Σ|I − <I>|/ΣI, where I is the observed intensity and <I> the average intensity obtained from multiple measurements. Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Rajan et al. Page 20 cRmeas as described in Diederichs and Karplus (Diederichs and Karplus, 1997). d≪I>/σ(<I>)> = Mean Ih over the standard deviation of the mean Ih averaged over all reflections in a resolution shell. eRwork= Σ ||Fo| − |Fc||/Σ|Fo|, where |Fo| is the observed structure factor amplitude and |Fc| the calculated structure factor amplitude. fRfree: Rfactor based on 5% of the data excluded from refinement. gTotal number of protein atoms, including those in alternate conformations. hGm: guanidinum ion. iAs reported by Molprobity (Davis et al., 2004). Structure. Author manuscript; available in PMC 2011 July 14.
3M71
Crystal Structure of Plant SLAC1 homolog TehA
Homolog Structure of the SLAC1 Anion Channel for Closing Stomata in Leaves Yuhang Chen1,6, Lei Hu2, Marco Punta6,7, Renato Bruni6, Brandan Hillerich6, Brian Kloss6, Burkhard Rost1,6,1, James Love6, Steven A. Siegelbaum2,3,5, and Wayne A. Hendrickson1,4,5,6 1Department of Biochemistry and Molecular Biophysics, Columbia University, New York, NY 10032, USA 2Department of Neuroscience, Columbia University, New York, NY 10032, USA 3Department of Pharmacology, Columbia University, New York, NY 10032, USA 4Department of Physiology and Cellular Biophysics, Columbia University, New York, NY 10032, USA 5Howard Hughes Medical Institute, Columbia University, New York, NY 10032, USA 6NYCOMPS, New York Structural Biology Center 89 Convent Avenue, New York, NY 10027 USA 7Department of Computer Science and Institute for Advanced Study Technical University of Munich D-85748 Munich, Germany Summary The plant SLAC1 anion channel controls turgor pressure in the aperture-defining guard cells of plant stomata, thereby regulating exchange of water vapor and photosynthetic gases in response to environmental signals such as drought or high levels of carbon dioxide. We determined the crystal structure of a bacterial homolog of SLAC1 at 1.20Å resolution, and we have used structure- inspired mutagenesis to analyze the conductance properties of SLAC1 channels. SLAC1 is a symmetric trimer composed from quasi-symmetric subunits, each having ten transmembrane helices arranged from helical hairpin pairs to form a central five-helix transmembrane pore that is gated by an extremely conserved phenylalanine residue. Conformational features suggest a mechanism for control of gating by kinase activation, and electrostatic features of the pore coupled with electrophysiological characteristics suggest that selectivity among different anions is largely a function of the energetic cost of ion dehydration. Users may view, print, copy, download and text and data- mine the content in such documents, for the purposes of academic research, subject always to the full Conditions of use: http://www.nature.com/authors/editorial_policies/license.html#terms Correspondence and request for materials should be addressed to WAH (wayne@convex.hhmi.columbia.edu).. Supplementary Information is linked to the online version of the paper at www.nature.com/nature. Author Contributions YC, LH, MP, RB, BH, BK and JL performed experiments; YC, LH, BR, SAS, and WAH analyzed data; YC, LH, SAS, and WAH prepared the manuscript. Author Information Atomic coordinates and diffraction data are deposited in the Protein Data Bank with accession codes 3M71, 3M73, 3M74, 3M75, 3M76, and 3M7L. See Table S4 for identifications. Reprints and permissions information is available at www.Nature.com/reprints. HHS Public Access Author manuscript Nature. Author manuscript; available in PMC 2013 January 18. Published in final edited form as: Nature. 2010 October 28; 467(7319): 1074–1080. doi:10.1038/nature09487. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Stomatal pores in the leaves of plants permit the influx of atmospheric carbon dioxide in exchange for transpirational evaporation of water1,2. A pair of kidney-shaped guard cells define each pore aperture, and turgor pressure variation in these cells determines the degree of stomatal pore openness. Depending on diverse environmental factors, the stomata close to prevent H2O losses and open to admit CO2 for photosynthesis. Environmental stimuli that lead to stomatal closure include darkness, high CO2 levels, ozone, low air humidity and drought. The plant hormone abscisic acid (ABA) is critical for signal transduction from these stimuli. Mutational screens in Arabidopsis thaliana for CO2 and ozone sensitivity identified a protein with ten predicted transmembrane (TM) helices, now called slow anion channel 1 (SLAC1), as having a central role in the control of stomatal closure3-5. Recent studies proved that SLAC1 is indeed an anion channel6-7, with characteristics like those of slow anion channels found in guard cells8, and that it is activated by phosphorylation from the OST1 kinase9. OST1 activity is negatively regulated by the ABI1 phosphatase10,11, which is in turn inhibited by the stomatal ABA receptors PYR/RCAR12,13 when in the ternary hormone-receptor-phosphatase complex14-18. Thereby, ABA stimulates SLAC1 channel activity. Resulting Cl− efflux through SLAC1 causes membrane depolarization, which activates outward-rectifying K+ channels, leading to KCl and water efflux to further reduce turgor and cause stomatal closure. SLAC1 expression in Arabidopsis is confined to the guard cells of leaves, but other Arabidopsis tissues do have SLAC1 homologs3, named SLAH1-SLAH4. The identifying mutations slac1-14 and slac1-23 are, respectively, in predicted transmembrane segments 9 (S456F) and 1 (G194D) of a protein that includes substantial N- and C-terminal extensions outside a 10-helix transmembrane (TM) domain. SLAH1, which is absent from leaves and lacks the terminal extensions of SLAC1, fully complements the mutant phenotype in slac1-2 guard cell protoplasts3. SLAC1 and homologs are also present in other plant genomes, including nine in rice (Oryza sativa) and five in grape vines (Vitis vinifera). SLAC1 relatives, some quite remote, also occur in bacteria, archaea and fungi. Known prokaryotic homologs contain only the predicted transmembrane domain of SLAC1, but some fungal homologs do have N- and C-terminal extensions. One homolog, Mae1 from the yeast Schizosaccharomyces pombe, functions as a malate uptake transporter19; another, Ssu1 from Saccharomyces cerevisiae and other fungi including Aspergillus fumigatus, is characterized as a sulfite efflux pump20,21; and TehA from Escherichia coli is identified as a tellurite resistance protein by virtue of its association in the tehA/tehB operon22,23. Despite a lack of further biochemical characterization, many homologs are annotated as tellurite resistance/ dicarboxylate transporter (TDT) proteins. We have undertaken structural and functional characterizations of the SLAC1 anion channel. We first solved an atomic-resolution crystal structure of the TehA homolog from Haemophilus influenzae, and we then developed a homology model for Arabidopsis SLAC1. This model allowed us to conduct mutagenesis for functional testing of structure-inspired hypotheses on gating and selectivity. We expressed Haemophilus TehA and Arabidopsis SLAC1 in Xenopus oocytes to characterize channel properties of these proteins and mutant variants. We also determined crystal structures for several mutant variants, including the homolog of slac1-2. Chen et al. Page 2 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Structure of SLAC1 bacterial homolog TehA We performed a bioinformatic analysis of SLAC1-related proteins, first clustering nearly 900 non-redundant sequences into a superfamily at the PSI-BLAST level E≤10−3, then into three distinct families at an initial threshold of E≤10−30, and finally into subfamilies at a typical initial threshold of E≤10−55. Since previous annotation is not well founded in experiment and SLAC1 is now the best characterized member, we adopt a nomenclature defining a SLAC superfamily divided into families identified as SF1-SF3 and subfamilies SF1A, SF1B etc. Family SF1 comprises the plant SLAC proteins and close bacterial homologs, family SF2 comprises a distinct set of bacterial proteins often annotated as exfoliative toxins, and family SF3 comprises the fungal Mae1 and Ssu1 proteins and their archaeal or bacterial homologs, respectively. SLAC family SF1 has three large subfamilies: the plant SLAC and SLAH proteins are in subfamily SF1A, closest bacterial homologs are in SF1B, and the TehA homologs are in SF1C (Fig. 1a). The other families also divide into subfamilies as detailed in Table S1, and family SF1 is divided into sub-subfamilies (Table S2). Two pertinent SF1 sequences are aligned in Fig. 1b. We used a structural genomics approach to obtain structural information, testing expression and purification for 43 bacterial and archaeal likely homologs, assaying for detergent choice and stability on 8 of these, finding two with appropriate profiles by size exclusion chromatography, and obtaining suitable crystals for one. This protein, TehA from H. influenzae (HiTehA), was found to be trimeric both by size-exclusion multi-angle light scattering (SEC-MALS) measurements and by chemical cross-linking. When solubilized in β-octylglucoside, HiTehA crystallized in space group R3 with a=b=96.01Å and c=136.27Å. Each asymmetric unit contains one subunit and 65% solvent. The structure was solved by selenomethionyl (SeMet) SAD phasing, ultimately at 1.50Å resolution (Table S3, Fig. S1), and then refined at 1.20Å resolution (Fig. 2a) to R/Rfree values of 14.1%/15.7% for a model that includes ordered residues 6-313, 213 water molecules and four detergent molecules (Table S4). The crystal structure has TehA trimers aligned with three-fold axes of the lattice (Fig. 2b). Subunits are tightly associated, burying 8947 Å2 of total surface area within trimer interfaces. The electrostatic potential surface is largely negative on the extracellular surface (Fig. 2c) and largely positive on the cytoplasmic surface (Fig. 2d). The membrane orientation is specified experimentally from GFP tagging of E. coli TehA24. Each TehA protomer has ten TM helices, as predicted; however, the fold is novel. Tandemly repeated helical hairpins are arranged with quasi-five-fold symmetry (Figs. 2e,S2). Extracellular inter-helix loops are short (2-5 residues) whereas intracellular inter-helix connections are longer, including a nine-residue helix H2,3 between TM2 and TM3 (Fig. 2f). An inner pentad of outwardly directed, TModd, helices creates an apparent pore through each protomer perpendicular to the putative membrane plane. TMeven helices from the five hairpins surround the inner pore and make an outer layer. Chen et al. Page 3 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Homology model for plant SLAC1 Arabidopsis SLAC1 (AtSLAC1) is substantially similar to bacterial homologs, notably HiTehA (Fig. 1b). All HiTehA TM helices are fully aligned to predicted SLAC1 TM helices, but there are short inter-helical gaps (1-5 residues) in all five extracellular loops and in two of the intracellular loops. The TM domain of AtSLAC1 (residues 188-504) aligns to HiTehA with 19% sequence identity and with a PSI-BLAST E-value of 3×10−22. For comparison, and in keeping with the family tree (Fig. 1a), the TM domain of AtSLAC1 shares sequence identities of 76% with rice SLAC1, 41% with Arabidopsis SLAH1, 25% with an SL1B homolog from Halorhodospira halophila, 11% with S. cerevisiae Ssu1, and 9% with S. pombe Mae1. A conceptual model with the AtSLAC1 sequence transposed onto the HiTehA helices sufficed to guide most of our mutational tests, but a detailed AtSLAC1 homology model helped to refine our ideas. Surface variability and electrostatic potential are plotted onto the surface of this model (Fig. 2g,2h). The most remarkable feature of the TehA structure and corresponding SLAC1 model is the central pore through each protomer. As for acetylcholine receptors25, the SLAC1 pore is formed by five helices; but the SLAC1 helices come from one protein molecule rather than five. The SLAC1 pore has a relatively uniform diameter of approximately 5Å across nearly five helical turns (Fig. S3), except for a pronounced constriction in the middle of the membrane (Fig. 3a) where the pore is occluded by the side chain of Phe 450 (Phe 262 in HiTehA). This residue is the only absolutely conserved amino acid residue of the SLAC1 family. The pore is lined with highly conserved (86% identity among five SLAC1 orthologs; 32% identity between AtSLAC1 and HiTehA) and generally hydrophobic residues (Figs. 1b, 3b,3c,S4). Despite this hydrophobicity, the electrostatic potential on the pore surface is polarized (Fig. 3a), presumably due to an invaginated shape adjacent to charged residues outside the membrane. The generally electropositive character of the cytoplasmic surface likely contributes to anion efflux. Kinks in the pore helices contribute to formation of a relatively constant pore diameter across the membrane. Four of the five HiTehA inner helices have centrally located proline residues, which necessarily generate kinks, and TM9 is kinked at a backbone-coordinated water molecule (Fig. 3d). Proline replaces Gly263 in all SF1A and SF1B relatives, including Pro451 AtSLAC1. This water-displacing change is isostructural26. Centrally located prolines also prevail in TM3, TM5 and TM7 across the SF1 proteins. The outer helices are longer and straighter, but more inclined. The only outer-helix proline kink of HiTehA is in TM6 at the trimer three-fold axis. Two of the gene-identifying mutations in Arabadopsis SLAC1 are selected point mutations, others are disruptive transfer DNA (T-DNA) insertions4. In the AtSLAC1 homology model, the slac1-2 (G194D) mutation3 points into the pore from TM1 and can be accommodated structurally whereas the slac1-1 (S456F) mutation4 points away from the pore six residues after pore-blocking Phe450 on TM9 and would be expected to disruptive. Ser456 interacts with outer helix TM10 in the homology model, and the phenyl bulk from S456F would not fit. Position 456 has alanine in HiTehA, 66% of all 204 SF1 homologs and another 27% have threonine or serine (as in all SLAC1 channels); phenylalanine never occurs among all Chen et al. Page 4 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 814 SLAC superfamily members. Position 194 has glycine in 58% of SF1 homologs and alanine in another 22%. Asp194, which never occurs naturally, would block the pore and be expected to repel anions. Mutational tests of channel function Mutational studies corroborate the hypothesis that a TehA-based SLAC1 model is appropriate. First, as discussed above, the slac1-1 mutant (S456F) is expected to be structurally disruptive, and indeed it is inactive in guard cells4, and the slac1-2 mutation (G194D) is expected to block the pore, and we show below that this variant is also inactive. We have also shown that the introduction of SLAC1-conserved proline residues into HiTehA (A208P/G263P) is accommodated isomorphously26. Moreover, as shown below, channel conductance properties of several mutants are similar for AtSLAC1 and HiTehA. To examine characteristics of the SLAC1 channel in light of the structural model, we performed electrophysiological tests of membrane currents from voltage-clamped Xenopus oocytes following injection of wild-type or mutant AtSLAC1 or HiTehA cRNAs. We observed modest-sized chloride currents with wild-type AtSLAC1 cRNA, as found previously6,7, but did not detect any chloride current following injection of wild-type HiTehA cRNA. We found that SLAC1 Cl− conductance was enhanced when the OST1 kinase cRNA was co-injected with SLAC1 cRNA, but only to the levels found by Lee et al.6 and not to the much higher levels found by Geiger et al.7 with OST1 physically connected to SLAC1 by split YFP linkage. In keeping with structural evidence that Phe262 blocks the HiTehA pore, removal of the phenyl group in HiTehA F262A or in the homologous AtSLAC1 F450A resulted in very large Cl− currents relative to wild-type levels, and the SLAC1 currents were now less enhanced by presence of OST1 (Fig. 4a). The tempting interpretation of an opened gate will require validation with appropriately analyzed single- channel recordings27. In keeping with the slac1-2 phenotype3, neither the functionally impaired AtSLAC1 G194D nor its HiTehA G15D homolog showed any substantial conductance; moreover, further consistent with pore blockage by Asp194 in slac1-2, the large conductances of HiTehA F262A and AtSLAC1 F450A were stopped in the double mutants AtSLAC1 G194D/F450A and HiTehA G15D/F262A (Fig. 4a). Here again, the effects in SLAC1 were independent of OST1. We also tested conductance characteristics for a series of AtSLAC1 F450X substitution mutants – F450A,G,T,V,L– and for the corresponding HiTehA F262X series – F262A,G,T,V,L (Figs. 4b,S5, Table S5). Findings from the two series are roughly parallel; in particular, the alanine and glycine substitutions lead to large currents for both and in comparison to the others. There are distinctions, of course, including generally higher conductances for AtSLAC1 over HiTehA and less conductance of F262T TehA compared to F450T SLAC1. It is also noteworthy that OST1 activation is very muted for the F450A,G,L mutants, which is consistent with SLAC1 gating at Phe450. Crystal structures were also determined for several of the HiTehA mutant variants (Table S4). The structures of F262A (1.15Å), F262V (1.60Å), F262L (1.65Å) and G15D (1.50Å) are all essentially isomorphous with the wild-type TehA structure with changes localized Chen et al. Page 5 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript primarily at the sites of mutation; the same is true for the double mutations of F262A/G15D, F262G/G15D and A208P/G263P26. The F262A structure has a wide open pore (Fig. 5a) with a relatively uniform pore diameter of ~5Å through ~30Å across the membrane (Fig. S3) whereas G15D has a doubly occluded pore (Fig. 5b). The pores of other mutant variants are consistent with the sizes of constrictive residues and with the observed conductances. Gating and activation The crystal structures of TehA and its mutant variants when taken together with the functional studies in Xenopus oocytes point to a crucial role for Phe450 in gating of the SLAC1 anion channel. Conservation of this residue across the SLAC1 family implies functional importance. The occlusion of the pore by the presence of F262 in the structure of wild-type TehA and the openness of the pore upon its substitution by alanine in the structure of the F262A mutant provides physical evidence for a gating role of this residue. This interpretation is supported by the correlated conductance characteristics from variants of the AtSLAC1 and HiTehA channels (Figs. 4b,S5). While these observations may suffice for placing the gate within the channel pore, they do not by themselves suggest a mechanism for gating in response to physiological stimuli. Some insight does come from conformational details defined at high resolution. One important structural clue is that the side chain of Phe262 is in a high energy conformation in the HiTehA structure, with χ1/χ2 at −160°/-4°. Although χ1 is in a preferred trans conformation, the phenyl ring is restricted by contacts with Val210 and Leu18 to a χ2 value near 0° rather than near to the preferred 90° orientation (Fig. 5c). Further evidence that Phe262 is restrained from local equilibrium comes from shifts observed in crystal structures of the F262A, F262V, F262L and F262G/G15D variants, which all show consistent backbone movements that displace Cβ(262) by 0.47-0.78Å (Fig. 5d). By contrast, Leu262 in F262L is in a preferred trans/gauche+ conformation at χ1/χ2 = 177°/63° as is Val262 in F262V at χ1 = −176°. What might control activation of HiTehA is unclear, but for AtSLAC1 activation is by OST1 phosphorylation6,7. The molecular consequences of OST1 phosphorylation of SLAC1 remain unknown, but it is plausible that associated shifts in pore- helix orientations would unlatch Phe450 in SLAC1 from a TehA-like restrained orientation. By analogy with Leu262 in F262L, we expect a preferred rotameric state for Leu450 in AtSLAC1 F450L. Thus, the lack of appreciable OST1 activation of conductance in AtSLAC1 F450L (Fig. 5e) might be explained by the lack of a restraining latch, whereby the channel remains closed despite OST1 activation. Puzzles certainly remain since OST1 does substantially activate AtSLAC1 F450T or F450V, which like HiTehA F262V should also be unrestrained; presumably activating adjustments widen the pore enough for ion permeation past threonine and valine but not leucine. Phosphorylation sites have been discovered in the N- and C-terminal tails of AtSLAC16,7,28 (179 and 51 residues long, respectively), but these alone cannot explain OST1 activation of SLAC1. First SLAH1, which fully complements the slac1-1 mutation, does not have these cytoplasmic tails. Secondly, although OST1 phosphorylation of Ser120 in the N-terminal tail is necessary for SLAC1 activation, it is not sufficient7. Thus, we surmise that direct phosphorylation of the SLAC1 TM domain must be critical, and SLAC1 has four conserved Chen et al. Page 6 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Ser/Thr candidates in its cytoplasmic loops. Moreover, SLAC1 proteins have proline- mediated kinks at the putative Phe450 gate in helix TM9, and also in adjacent helix TM7; these features may play a role in phosphorylation-driven unlatching of the Phe450 gate in SLAC1. Ion selectivity and discrimination Our studies of SLAC1 channel relative ion permeabilities, based on measurements of current reversal potential, are consistent with earlier work demonstrating that AtSLAC1 conducts anions but not cations and is selective among anions, with greater permeability for nitrate than for chloride (as in Vicia faba guard cell protoplasts27) and much reduced permeability for malate, bicarbonate or sulfate6,7. We also find that SLAC1 has little permeability for sulfite. Additionally, we find that wild-type SLAC1, F450A and F450T all have similar relative permeabilities to chloride, sulfite and malate, despite having widely different conductance levels, but the gating mutants do show small but significant decreases in nitrate permeability (Fig. 4c, Table S6). The relative insensitivity of anion permeability to gating residue changes suggests that selectivity for these anions may occur away from the central constriction at the channel gate. To some extent, ionic discrimination must depend on pore geometry; thus, an organic anion such as malate may be simply too large to pass through the 5-Å wide pore. Although the SLAC1 pore is lined largely with hydrophobic sidechains (Fig. 3c), it is sprinkled with hydroxyl groups from serine and threonine residues (16%); their electropositive hydrogen atoms may facilitate conductance. Most strikingly, the electrostatic potential within the AtSLAC1 pore is electropositive throughout (Fig. 3a). This polarization, promoted by charges on extra-membranous loops, no doubt contributes significantly in discrimination against cations. The relative anion permeability sequence of SLAC1 determined by us and others, I− > NO3− > Br− > Cl− 6,7, corresponds to selectivity sequence 1 compiled by Wright and Diamond29 for a range of anion-selective proteins. This sequence correlates inversely with the hydration energies of monovalent anions – anions with a lower hydration energy have a greater channel permeability. It is thought to be generated in proteins with weak, low field-strength, anion binding sites, where selectivity is largely determined by the energetic cost of anion dehydration. These selectivity results are thus consistent with the SLAC1 structure, where the pore lacks any obvious anion binding site. Distinctiveness of the SLAC1 channel SLAC1 anion channels are entirely novel in structure and, apparently, in the mechanism for ion conductance. The best characterized of anion channels belong to the CLC family of Cl− channels and transporters30-32. CLC channels have an altogether different architecture from the SLAC1 channel, and the mechanism for selectivity is also very different. Bacterial CLC transporters bind halide ions at three sites in a highly constricted pore30. By contrast, the SLAC1 pore has a relatively uniform diameter across the membrane, except where closed by the gating phenyl group, and we do not find discrete ion binding sites. CLC selectivity is governed by specific residues surrounding these binding sites30,32. The anion selectivity Chen et al. Page 7 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript sequence for the CLC channels of Cl−>Br−>NO −3>I−, opposite of that in SLAC1, is consistent with the high field-strength anion binding sites in CLC channels29. Interestingly, as for AtSLAC1, an Arabidopsis CLCa channel also preferentially transports nitrate ions33, and an E. coli CLC channel is converted to preference of nitrate when a generally conserved serine at the central site is substituted with proline as in AtCLCa32. SLAC1 also differs radically from other structurally characterized anion channels and transporters. These include the VDAC1 voltage-gated anion channel from mitochondrial outer membranes, which has a porin-like beta-barrel structure34-36, and a light-driven halorhodopsin chloride pump, which has a transmembrane conductance pathway similar to that of the proton-pumping pore of bacteriorhodopsin37. Although its channel structure is still only known by homology to other ABC transporters, CFTR is another obviously distinct chloride channel38. Cys-loop receptors also include anion channels39, and these are similar to SLAC1 in having five-helix pores25, but here selectivity is governed by charged groups at the entrance to the pore, which distinguish the anion-selective GABAA and glycine receptors from the cation-selective acetylcholine and serotonin 5HT3 receptors39. Finally, the recently identified TMEM16A gene for calcium-activated chloride channels40-42 appears to encode an 8-TM protein that is again distinct from SLAC1. Stomatal guard cells show both rapidly activated (R-type) and slow (S-type) anion channel activity43. Although slac1 guard cells have very defective S-type activity, their R-type currents are normal4. Guard cell protoplasts from the slac1-2 mutant abnormally accumulate Cl−, K+, malate and fumarate3, whereas SLAC1 shows negligible malate conductance7. As for SLAC1-associated K+ movements, other channels or transporters must be responsible for SLAC1-associated malate movements. Recent studies indicate that AtALMT12, an aluminium-activated malate transporter (ALMT) family member, is a malate-dependent R- type anion channel44 needed for stomatal closure45. Conclusions We find that many functional properties of the plant SLAC1 anion channel are explained well by the structure of an uncharacterized bacterial TehA protein that has been associated with tellurite resistance. SLAC1 and TehA belong to distinct subfamilies within one branch of a larger SLAC1 superfamily, but AtSLAC1 and HiTehA are sufficiently similar (19% sequence identity) that the SLAC1 homology model is predictive for function, including a verified placement of the identifying slac1-2 mutation G194D and a phenylalanine gate. One remaining puzzle concerns the structural change that activating phosphorylation elicits in SLAC1, and another puzzle is the biochemical role of the TehA homologs in bacteria. In a companion paper26, we examine functional and structural properties of TehA in bacteria, showing that it is anion channel, although actually not conferring tellurite resistance, and identifying a mutant variant with properties suggestive of an activated state. Thus, SLAC1 and TehA likely represent a large family of selective anion channels controlled by environmental stimuli. Chen et al. Page 8 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript METHODS Selection of target sequences TehA from E. coli (EcTehA) was a centrally selected sequence for expansion as a NYCOMPS targeted family. Target selection criteria utilized at NYCOMPS are described in details elsewhere46. Briefly, the EcTehA sequence was run against a dataset of about 40,000 predicted alpha helical integral membrane protein sequences from prokaryotic genomes (NYCOMPS98 dataset46) using PSI-BLAST50. Sequences that matched EcTehA with an E- value lower than 10−3 in an alignment extending over at least 50% of both predicted TM regions and passing our post-seed-expansion filtering criteria46 were passed to the protein production pipeline. Protein expression screening Full-length homologs from the following 38 species, including 2 sequences each from 5 of these, were amplified from genomic DNA by PCR: Thermoplasma acidophilum, Lactococcus lactis subsp. lactis Il1403, Streptococcus pyogenes M1 GAS, Streptococcus pneumoniae TIGR4, Haemophilus influenzae Rd KW20, Methanocaldococcus jannaschii DSM 2661 (2), Escherichia coli K12, Salmonella typhimurium LT2, Clostridium perfringens ATCC 13124, Xanthomonas campestris pv. campestris str. ATCC 33913, Streptococcus agalactiae 2603V/R, Shewanella oneidensis MR-1, Streptococcus mutans UA159, Archaeoglobus fulgidus DSM 4304, Vibrio parahaemolyticus RIMD 2210633 (2), Pseudomonas syringae pv. tomato str. DC3000 (2), Enterococcus faecalis V583, Pyrococcus horikoshii OT3, Bordetella bronchiseptica RB50, Streptomyces coelicolor A3, Corynebacterium glutamicum ATCC 13032, Picrophilus torridus DSM 9790, Acinetobacter sp. ADP1, Vibrio fischeri ES114, Pseudomonas fluorescens Pf-5, Sulfolobus acidocaldarius DSM 639, Colwellia psychrerythraea 34H, Neisseria meningitidis MC58, Rhodobacter sphaeroides 2.4.1, Vibrio cholerae O1 biovar eltor str. N16961 (2), Marinobacter aquaeolei VT8, Acinetobacter baumannii ATCC 17978, Klebsiella pneumoniae subsp. pneumoniae MGH 78578, Streptomyces avermitili MA-4680, Bordetella parapertussis 12822 (2), Streptococcus thermophilus LMG 18311, Salmonella enterica subsp. enterica serovar Paratyphi A str. ATCC 9150, and Anaeromyxobacter dehalogenans 2CP-C. Selected cDNAs were cloned into a modified pET vector (Novagen, Inc) that fuses a FLAG and deca-histidine tag at the C-terminus, which are cleavable by TEV protease. Proteins were expressed in E. coli BL21(DE3) plysS by a high throughput format (0.6ml in a deep well block) and purified after lysis by sonication using metal affinity purification in a buffer containing N-Dodecyl-β-D-Maltopyranoside. Samples were passed over an analytical size exclusion column in 12 different detergent-containing mobile phases, which included N- dodecyl-β-D-maltopyranoside (DDM), N-decyl-β-D-altopyranoside (DM), N-nonyl-β-D- altopyranoside (NM), N-octyl-β-D-altopyranoside (OM), N-octyl-β-D-Glucopyranoside (OG), N-nonyl-β-D-Glucopyranoside (NG) and lauryl dimethyl amine oxide (LDAO). Multi-angle light scattering with refractive index detection was used to analyze the oligomeric state51. The E. coli and H. influenzae proteins were judged to be monodisperse and stable and were passed to scale up. Chen et al. Page 9 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Scaled-up production and purification For scale up, in brief, transformed BL21 plysS cells were grown at 37°C in 2× YT media to OD 0.6-0.8 after being inoculated with 1% of the overnight culture. The culture was induced with 0.4mM IPTG and continued to grow at 37 °C for another 4 hours. The cells were harvested by centrifugation and stored at −80°C before use. Selenomethionyl (SeMet) TehA was expressed in a similar way, but using containing SeMet in place of methionine in defined minimal media. Cells were resuspended in a buffer containing 50 mM Tris-HCl (pH 8.0) and 200 mM NaCl and lysed using a French Press with two passes at 15-20,000 psi. Cell debris were removed by centrifugation at 10,000g for 20min, and the membrane fraction was isolated from that supernatant by ultra-centrifugation at 150,000g for 1 hr. The membrane fraction was homogenized in a solubilization buffer containing 50 mM Tris (pH 8.0) and 200 mM NaCl, and incubated with final concentration of 1% (w/v) dodecyl-β- D-maltopyranoside (DDM, Anatrace, Inc) for 1 hr at 4°C. The non-dissolved matter was remove by ultracentrifugation at 150,000g for 30 min, and the supernatant was loaded to a 5ml Hitrap Ni2+-NTA affinity column (GE Healthcare), pre-equilibrated with the same solubilization buffer supplemented with 0.05% DDM. After 20 column volume buffer wash, the protein was eluted with 250mM imidazole in the solublization buffer. The Flag and 10- His tags were removed by adding super TEV at 1:100 mass ratio and incubating at 4°C overnight. Tag removal was confirmed by SDS-PAGE, and the resulting sample was concentrated to around 10mg/ml. Preparative size exclusion chromatography was carried out on a Superdex-200 column for further purification, removal of TEV protease and the cleaved tag, and for buffer and detergent exchange. The gel-filtration buffer contained 10 mM Tris (pH 8.0), 200 mM NaCl, 1mM EDTA, 0.5 mM Tris [2-carboxyethyl] phosphine (TCEP), and 2×CMC of detergent. In the case of HiTehA, the protein was well behaved and stable in nearly all tested detergents, and we have purified it from DDM, DM, NM, OM, OG and LDAO. Protein characterization We performed N-terminal amino-acid sequencing of purified HiTehA and EcTehA prior to TEV protease treatment. Results from these analyses proved that true initiating methionine residue is located 14 residues after one annotated as N-terminal. The intervening nucleotide sequence contains a Shine-Delgarno sequence. For cross-linking experiments, purified HiTehA was incubated with 10 mM disuccinimidyl glutarate at room temperate for 30 min, and 100mM Tris-HCl pH 7.5 was added to stop the reaction. The incubated sample when run on a 8-25% gradient SDS-PAGE gel showed a ladder consistent with a trimeric structure. Crystallization and data collection Purified protein was concentrated to ~10mg/ml for initial crystal trials in a Mosquito robot with commercial screens from Hampton research, Emerald Biosystems and Molecular Dimension. We obtained crystals of HiTehA from protein in detergent DDM, DM, NM, OM, OG and LDAO, but only those from LDAO and OG gave diffraction to beyond 4Å spacings. Crystals that proved useful were all grown at 4°C using the sitting-drop vapor Chen et al. Page 10 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript diffusion method. After extensive optimization we reached conditions supporting very high resolution. Wild-type HiTehA and most of its variants crystallized from 1mM ZnSO4, 50mM Hepes-Na pH 7.8 and 28% PEG600. HiTehA F262A crystallized from 200mM Li2SO4, 100mM glycine pH9.3 and 33% PEG400). Addition of 10mM spermidine as an additive helped to obtain slightly better diffracting crystals. Cryoprotection was achieved by adding 5% ethylene glycol or PEG400 to the crystallization solution. Structure determination and refinement Native and SeMet SAD data set were collected at NSLS beamline X4A and processed using the software HKL200052. Crystals of HiTehA grew in space group R3 with a=b=c = 85.0Å and α=β=γ = 93.5°. All subsequent manipulations were done in the hexagonal setting of this space group with a=b= 96.0Å and c = 136.7Å. The asymmetric unit contains one TehA protomer and 65% solvent volume. The structure was determined at 2.0Å resolution from single-wavelength anomalous diffraction (SAD) using selenomethionine-substituted protein crystals. Assessment of data quality for phasing, location of heavy atom sites and initial phases were calculated using the HKL2MAP interface to SHELX programs53. All the secondary structure elements were clearly visible in the experimental electron density map. Automatic model building was done in Arp/wArp54 and completed manually in the program COOT55. The model was refined against native data at 1.20Å resolution using the program Refmac5.5 in CCP456, with anisotropic B factor restrained refinement applied. Subsequent structural analyses of mutant variants were refined as isomorphous structures. Site-directed mutagenesis Site-directed mutants were constructed using the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA 92037) and expressed from pET vectors in E. coli BL21(DE3) plysS cells as for the wild-type protein. Electrophysiology All constructs were cloned into plasmid pGHME2, linearized, and transcribed into cRNA using T7 polymerase (mMessage mMachine, Ambion). Oocytes were injected with 50 nl of cRNA solution each, at a constant concentration of 0.5 mg/ml for HiTehA, or 0.5 mg/ml for AtSLAC1 constructs, with or without 0.5 mg/ml of AtOST1 constructs and recorded in voltage-clamp experiments 2 days after injection. For mixed expression experiments, 25nl of cRNA solution was injected for each AtSLAC1 component. Two-electrode voltage clamp recordings were performed to measure HiTehA or AtSLAC1 currents as described6,7. The pipette solutions contained 3 M KCl. For voltage-clamp current recordings, the bath solution contained 1 mM MgCl2, 1 mM CaCl2, 10 mM Mes/Tris (pH 5.6) and 30 mM CsCl; For anion selectivity measurements, the bath solutions contained 50 mM Cl−, NO −3, malate, or 30 mM SO 2-3 sodium salts, 45 mM Na-gluconate, 1 mM Ca-gluconate2, 1 mM Mg- gluconate2, 1 mM K-gluconate, as well as 10 mM Tris/Mes (pH 5.6). Osmolality was adjusted with D-mannitol to 220 mOsmol/kg. The bath electrode was a 3M KCl agar bridge. The testing potentials were +50 to −110 or −130 mV in 20 mV decrements with 7.5 s duration. The holding and return potentials were 0 mV with a 1.45 s or 2 s, respectively. I-V Chen et al. Page 11 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript relations of HiTehA or AtSLAC1 were generated from the current reached after 0.5 s in testing potentials. The Goldman-Hodgkin-Katz equation was applied to estimate permeability ratios for monovalent ions as described6. For divalent anions, the permeability ratios were derived according to Fatt and Ginsborg57. Bioinformatic analysis of SLAC-related proteins Sequences related to SLAC1 were analyzed comprehensively by PSI-BLAST50. Searches at E<10−3 starting from five disparate homologs each identified a common pool of over 900 proteins, which when pooled were used for sub-classification into families and subfamilies. Details of these analyses are reported in footnotes to Table S1. Molecular figures were produced in PyMOL58. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We thank Filippo Mancia, Larry Shapiro and Ming Zhou for helpful discussions; Patricia Rodriguez and Arianne Morrison for help with cloning and expression; Jonah Cheung, Mark Collins, Chris Min, Sheng Ye and Zhen Zhang for advice in protein chemistry and crystallography; John Schwanof and Randy Abramowitz for help with synchrotron experiments; John Riley for help with Xenopus oocyte injections; and Min Su for advice on the PSI- BLAST analysis of SLAC relatives. This project was supported in part by an award to the New York Consortium on Membrane Protein Structure (NYCOMPS) from the NIGMS Protein Structure Initiative. Beamline X4A at the National Synchrotron Light Source (NSLS), a DOE facility at Brookhaven National Laboratory, is supported by the New York Structural Biology Center. References 1. Hetherington AM, Woodward FI. The role of stomata in sensing and driving environmental change. Nature. 2003; 424:901–908. [PubMed: 12931178] 2. Sirichandra C, Wasilewska A, Vlad F, Valon C, Leung J. The guard cell as a single-cell model towards understanding drought tolerance and abscisic acid action. J. Exp. Bot. 2009; 60:1439–1463. [PubMed: 19181866] 3. Negi J, et al. CO2 regulator SLAC1 and its homologues are essential for anion homeostatsis in plant cells. Nature. 2008; 452:483–486. [PubMed: 18305482] 4. Vahisalu T, et al. SLAC1 is required for plant guard cell S-type anion channel function in stomatal signalling. Nature. 2008; 452:487–491. [PubMed: 18305484] 5. Saji S, et al. Disruption of a gene encoding C4-dicarboxylate transporter-like protein increases ozone sensitivity through deregulation of the stomatal response in Arabidopsis thaliana. Plant Cell Physiol. 2008; 49:2–10. [PubMed: 18084014] 6. Lee SC, Lan W, Buchanan BB, Luan S. A protein kinase-phophatase pair interacts with an ion channel to regulate ABA signaling in plant guard cells. Proc. Natl. Acad. Sci. USA. 2009; 106:21419–21424. [PubMed: 19955427] 7. Geiger D, et al. Activity of guard cell anion channel SLAC1 is controlled by drought-stress signaling kinase-phosphatase pair. Proc. Natl. Acad. Sci. USA. 2009; 106:21425–21430. [PubMed: 19955405] 8. Schroeder JI, Hagiwara S. Cytosolic calcium regulates ion channels in the plasma membrane of Vicia faba guard cells. Nature. 1989; 338:427–430. 9. Mustilli A, Merlot S, Vavasseur A, Fenzi F, Giraudat J. Arabidopsis OST1 protein kinase mediates the regulation of stomatal aperture by abscisic acid and acts upstream of reactive oxygen species production. Plant Cell. 2002; 14:3089–3099. [PubMed: 12468729] Chen et al. Page 12 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 10. Leung J, et al. Arabidopsis ABA response gene ABI1: features of a calcium-modulated protein phosphatase. Science. 1994; 264:1448–1452. [PubMed: 7910981] 11. Meyer K, Leube MP, Grill E. A protein phosphatase 2C involved in ABA signal transduction in Arabidopsis thaliana. Science. 1994; 264:1452–1455. [PubMed: 8197457] 12. Ma Y, et al. Regulators of PP2C phosphatase activity function as abscisic acid sensors. Science. 2009; 324:1064–1068. [PubMed: 19407143] 13. Park S, et al. Abscisic acid inhibits type 2C protein phosphatases via the PYR/PYL family of START proteins. Science. 2009; 324:1068–1071. [PubMed: 19407142] 14. Melcher K, et al. A gate-latch-lock mechanism for hormone signalling by abscisic acid receptors. Nature. 2009; 462:602–608. [PubMed: 19898420] 15. Miyazono K, et al. Structural basis of abscisic acid signalling. Nature. 2009; 462:609–614. [PubMed: 19855379] 16. Fujii H, et al. In vitro reconstitution of an abscisic acid signalling pathway. Nature. 2009; 462:660– 664. [PubMed: 19924127] 17. Yin P, et al. Structural insights into the mechanism of abscisic acid signaling by PYL proteins. Nat. Struct. Mol. Biol. 2009; 16:1230–1236. [PubMed: 19893533] 18. Nishimura N, et al. PYR/PYL/RCAR family members are major in-vivo ABI1 protein phosphatase 2C-interacting proteins in Arabidopsis. Plant J. 2010; 61:290–299. [PubMed: 19874541] 19. Grobler J, Bauer F, Subden RE, van Vuuren HJ. The MAE1 gene of Schizosaccharomyces pombe encodes a permease for malate and other C4 dicarboxylic acids. Yeast. 1995; 11:1485–1491. [PubMed: 8750236] 20. Park H, Bakalinsky AT. SSU1 mediates sulphite efflux in Saccharomyces cerevisiae. Yeast. 2000; 16:881–888. [PubMed: 10870099] 21. Léchenne B, et al. Sulphite efflux pumps in Aspergillus fumigatus and dermatophytes. Microbiology. 2007; 153:905–913. [PubMed: 17322211] 22. Walter EG, Weiner JH, Taylor DE. Nucleotide sequence and overexpression of the tellurite- resistance determinant from the IncHII plasmid pHH1508a. Gene. 1991; 101:1–7. [PubMed: 2060788] 23. Taylor DE, Hou Y, Turner RJ, Weiner JH. Location of a potassium tellurite resistance operon (tehA tehB) within the terminus of Escherichia coli K-12. J Bacteriol. 1994; 176:2740–2742. [PubMed: 8169225] 24. Daley DO, et al. Global topology analysis of the Escherichia coli inner membrane proteome. Science. 2005; 308:1321–1323. [PubMed: 15919996] 25. Unwin N. Refined structure of the nicotinic acetylcholine receptor at 4Å resolution. J. Mol. Biol. 2005; 346:967–989. [PubMed: 15701510] 26. Chen Y, Hu L, Siegelbaum SA, Hendrickson WA. Structure-based analysis of anion channel TehA and methyltransferase TehB implicated in bacterial tellurite resistance. Submitted. 27. Schmidt C, Schroeder JI. Anion Selectivity of Slow Anion Channels in the Plasma Membrane of Guard Cells (Large Nitrate Permeability). Plant Physiol. 1994; 106:383–391. [PubMed: 12232336] 28. Vahisalu T, et al. Ozone-triggered rapid stomatal response involves the production of reactive oxygen species, and is controlled by SLAC1 and OST1. Plant J. 2010; 62:442–53. [PubMed: 20128877] 29. Wright EM, Diamond JM. Anion selectivity in biological systems. Physiol. Rev. 1977; 57:109– 156. [PubMed: 834775] 30. Dutzler R, Campbell EB, MacKinnon R. Gating the selectivity filter in ClC chloride channels. Science. 2003; 300:108–112. [PubMed: 12649487] 31. Accardi A, Miller C. Secondary active transport mediated by a prokaryotic homologue of ClC Cl- channels. Nature. 2004; 427:803–807. [PubMed: 14985752] 32. Picollo A, Malvezzi M, Houtman JC, Accardi A. Basis of substrate binding and conservation of selectivity in the CLC family of channels and transporters. Nat. Struct. Mol. Biol. 2009; 16:294– 301. [PubMed: 19219045] 33. De Angeli A, et al. The nitrate/proton antiporter AtCLCa mediates nitrate accumulation in plant vacuoles. Nature. 2006; 442:939–942. [PubMed: 16878138] Chen et al. Page 13 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 34. Hiller S, Garces RG, Malia TJ, Orekhov VY, Colombini M, Wagner G. Solution structure of the integral human membrane protein VDAC-1 in detergent micelles. Science. 2008; 321:1206–1210. [PubMed: 18755977] 35. Bayrhuber M, et al. Structure of the human voltage-dependent anion channel. Proc. Natl. Acad. Sci. USA. 2008; 105:15370–15375. [PubMed: 18832158] 36. Ujwal R, et al. The crystal structure of mouse VDAC1 at 2.3 Å resolution reveals mechanistic insights into metabolite gating. Proc. Natl. Acad. Sci. USA. 2008; 105:17742–17747. [PubMed: 18988731] 37. Kouyama T, et al. Crystal structure of the light-driven chloride pump halorhodopsin from Natronomonas pharaonis. J. Mol. Biol. 2010; 396:564–579. [PubMed: 19961859] 38. Gadsby DC, Vergani P, Csanády L. The ABC protein turned chloride channel whose failure causes cystic fibrosis. Nature. 2006; 440:477–483. [PubMed: 16554808] 39. Miller PS, Smart TG. Binding, activation and modulation of Cys-loop receptors. Trends Pharmacol. Sci. 2010; 31:161–174. [PubMed: 20096941] 40. Yang YD, et al. TMEM16A confers receptor-activated calcium-dependent chloride conductance. Nature. 2008; 455:1210–5. [PubMed: 18724360] 41. Caputo A, et al. TMEM16A, a membrane protein associated with calcium-dependent chloride channel activity. Science. 2008; 322:590–594. [PubMed: 18772398] 42. Schroeder BC, Cheng T, Jan YN, Jan LY. Expression cloning of TMEM16A as a calcium- activated chloride channel subunit. Cell. 2008; 134:1019–1029. [PubMed: 18805094] 43. Schroeder JI, Keller BU. Two types of anion channel currents in guard cells with distinct voltage regulation. Proc. Natl. Acad. Sci. USA. 1992; 89:5025–5029. [PubMed: 1375754] 44. Meyer S, et al. AtALMT12 represents an R-type anion channel required for stomatal movement in Arabidopsis guard cells. Plant J. Jul 12.2010 Epub ahead of print. 45. Sasaki T, et al. Closing plant stomata requires a homolog of an aluminum-activated malate transporter. Plant Cell Physiol. 2010; 51:354–65. [PubMed: 20154005] 46. Punta M, et al. Structural genomics target selection for the New York Consortium on Membrane Protein Structure. J. Struct. Funct. Genomics. 2009; 4:255–268. [PubMed: 19859826] 47. Papadopoulos JS, Agarwala R. COBALT: constraint-based alignment tool for multiple protein sequences. Bioinformatics. 2007; 23:1073–1079. [PubMed: 17332019] 48. Landau M, et al. ConSurf 2005: the projection of evolutionary conservation scores of residues on protein structures. Nucleic Acids Res. 2005; 33:W299–W302. [PubMed: 15980475] 49. Rocchia W, et al. Rapid grid-based construction of the molecular surface for both molecules and geometric objects: applications to the finite difference Poisson-Boltzmann method. J. Comp. Chem. 2002; 23:128–137. [PubMed: 11913378] 50. Altschul SF, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997; 25:3389–3402. [PubMed: 9254694] 51. Kendrick BS, Kerwin BA, Chang BS, Philo JS. Online size-exclusion high-performance liquid chromatography light scattering and differential refractometry methods to determine degree of polymer conjugation to proteins and protein-protein or protein-ligand association states. Anal. Biochem. 2001; 299:136–146. [PubMed: 11730335] 52. Otwinowski Z, Minor W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 1997; 276:307–326. 53. Pape P, Schneider TR. HKL2MAP: a graphical user interface for phasing with SHELX programs. J. Appl. Cryst. 2004; 37:843–844. 54. Perrakis A, Morris R, Lamzin VS. Automated protein model building combined with iterative structure refinement. Nat. Struct. Biol. 1999; 6:458–463. [PubMed: 10331874] 55. Emsley P, Cowtan K. COOT: model-building tools for molecular graphics. Acta Crystallogr. D. 2004; 60:2126–2132. [PubMed: 15572765] 56. CCP4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D. 1994; 50:760– 763. [PubMed: 15299374] 57. Fatt P, Ginsborg BL. The ionic requirements for the production of action potentials in crustacean muscle fibers. J. Physiol. 1958; 142:516–543. [PubMed: 13576452] Chen et al. Page 14 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 58. DeLano, WL. The PyMOL molecular graphics system. DeLano Scientific; San Carlos, CA, USA: 2002. http://www.pymol.org Chen et al. Page 15 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 1. Sequence analysis for the SLAC1 superfamily a, Family tree. The presentation was computed by the program COBALT47 from representative subfamily sequences (Table S1), including Arabidopsis thaliana SLAC1 for SF1Ai, Haemophilus influenzae TehA for SF1Ci, Escherichia coli TehA for SF1Cii, Vibrio parahaemolyticus for SF2A, Staphylococcus aureus for SF2B, Aspergillus fumigatus Ssu1 for SF3A, and Schizosaccharomyces pombe Mae1 for SF3B. b, Structure-based sequence alignment of TehA from Haemophilus influenzae (HiTehA) and SLAC1 from Arabidopsis thaliana (AtSLAC1). The TehA structure has been used to restrict sequence gaps to inter- helical segments. Superior coils define extents of the HiTehA helical segments; red letters mark residue identities; red boxes are drawn for residues that are >95% identical within the plant subfamily SF1A for AtSLAC1 or within the TehA subfamily for HiTehA; red diamonds mark HiTehA residues that line the central pore; and the colored inferior bar encodes ConSurf sequence variability48 for the SF1 family of 204 non-redundant proteins. Chen et al. Page 16 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 2. Crystal structure of HiTehA and homology model of AtSLAC1 a, Electron density distribution from the HiTehA crystal structure at 1.2Å resolution. The map has (2Fo-Fc) coefficients based on the superimposed model. Contours are at 2.5σ. b, Ribbon diagram of the HiTehA trimer. Each protomer is colored spectrally from blue at its N-terminus to red at its C-terminus. c, DelPhi49 electrostatic potential at the extracellular surface. Electronegative and electropositive potential are colored in degrees of red and blue saturation, respectively. d, Electrostatic potential49 at the intracellular surface. e, Ribbon diagram of an HiTehA protomer viewed from outside the membrane. The ribbon is colored spectrally as in 2b. f, Ribbon diagram of an HiTehA protomer viewed from within the membrane from below the view of e. g, Surface of a homology model of AtSLAC1, viewed as in f, and colored by sequence variability48. h, Surface of AtSLAC1 as in g, but colored by electrostatic potential49. Chen et al. Page 17 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 3. Putative structure of the SLAC1 conductance pore a, Cross-section through the homology model of AtSLAC1. The model is viewed as in 2i, with the electrostatic potential49 shown on the external surface of the molecular envelope. The side chain of Phe450 is shown as a stick model (red) on the backbone ribbon colored yellow. b, Cross-section as in a, but colored by surface conservation48 as in 2h. c, Pore- lining residues in the SLAC1 homology model. A cylinder model (left), spectrally colored as in 2b, provides a key for viewing the rolled-open structure (left) with pore-lining residues of AtSLAC1 shown on the TModd helices. d, Ribbon diagrams of HiTehA TM9 (left) and TM7 (right) viewed from within the conductance pore. The side chains of Pro207 and Phe262 are shown as well as a kink-stabilizing water HOH25 that is coordinated by the NH of Gly263 and by C=O groups of Gly202 and Ala259. Density contours are shown for the water molecule. Chen et al. Page 18 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 4. Ionic conductance measurements a, Typical microelectrode voltage-clamp current traces from oocytes injected with various channel cRNAs. Left column, oocytes injected with cRNAs encoding wild-type HiTehA channels (WT), or F262A, G15D/F262A or G15D mutants. Right two columns, oocytes injected with cRNAs for WT AtSLAC1, or F450A, G194D/F450A or G194D mutants, with or without co-injection of AtOST1. Dotted lines represent zero current levels. Extracellular solution contained 30mM CsCl. Schematic keys at the left place phenyl and/or aspartyl gates. b, Effects of gating residue mutations. Mean chloride currents, measured at −90 mV, are shown comparing WT HiTehA and its mutant series F262A,G,T,V,L with WT AtSLAC1 and its corresponding series F450A,G,T,V,L, both alone and co-expressed with AtOST1. Full I-V relations are shown in Fig. S5. c, Effect of gating residues on relative AtSLAC1 anion permeabilities. Relative permeabilities (P[X]/P[Cl]) for chloride, nitrate, sulfite and malate of WT, F450A and F450T SLAC1 channels were measured from the change in current reversal potential with Cl− or anion X− as the sole permeant anion in the bath solution (Methods, Table S6). Chen et al. Page 19 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 5. Structural features at the SLAC1 homolog gate a,b, Cross-sections through the conductance pores of HiTehA F262A and HiTehA G15D. The view and presentations are as in 3a, except that helices are colored purple. c, Molecular basis for conformational strain in gating residue Phe262 of HiTehA. Helices TM1 (left), TM9 (center) and TM7 (right) are viewed from within the pore and presented as ribbon diagrams with selected side chains drawn in stick representation. The local low-energy conformation for the phenyl ring (χ2 = 90°) is shown in thin lines with short contacts indicated in dashed lines: Leu18 Cδ2 – Phe262 Cδ1, d = 2.4 Å; Val210 Cγ2 – Phe262 Cε2, d = 2.8 Å. d, Conformational shifts consequent to release of strain in gating residue Phe262. Cα backbone structures of F262A, F262V and F262L HiTehA were first superimposed onto WT HiTehA, and residues 258-266 were drawn in stereo as superimposed. All backbone atoms are shown for peptides 262±1, but only Cα atoms are shown otherwise. The WT backbone and phenyl group are green; other backbone are all magenta; side chains of Ala262, Val262 and Leu262 are cyan, blue and red, respectively; oxygen-directed bonds are red and nitrogen-directed bonds are blue. e, Typical microelectrode voltage-clamp current traces from oocytes injected with wild-type (WT) or F450L AtSLAC1 cRNA. Experimental conditions and displays are as in 4a. Chen et al. Page 20 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript
3M73
Crystal Structure of Plant SLAC1 homolog TehA
Homolog Structure of the SLAC1 Anion Channel for Closing Stomata in Leaves Yuhang Chen1,6, Lei Hu2, Marco Punta6,7, Renato Bruni6, Brandan Hillerich6, Brian Kloss6, Burkhard Rost1,6,1, James Love6, Steven A. Siegelbaum2,3,5, and Wayne A. Hendrickson1,4,5,6 1Department of Biochemistry and Molecular Biophysics, Columbia University, New York, NY 10032, USA 2Department of Neuroscience, Columbia University, New York, NY 10032, USA 3Department of Pharmacology, Columbia University, New York, NY 10032, USA 4Department of Physiology and Cellular Biophysics, Columbia University, New York, NY 10032, USA 5Howard Hughes Medical Institute, Columbia University, New York, NY 10032, USA 6NYCOMPS, New York Structural Biology Center 89 Convent Avenue, New York, NY 10027 USA 7Department of Computer Science and Institute for Advanced Study Technical University of Munich D-85748 Munich, Germany Summary The plant SLAC1 anion channel controls turgor pressure in the aperture-defining guard cells of plant stomata, thereby regulating exchange of water vapor and photosynthetic gases in response to environmental signals such as drought or high levels of carbon dioxide. We determined the crystal structure of a bacterial homolog of SLAC1 at 1.20Å resolution, and we have used structure- inspired mutagenesis to analyze the conductance properties of SLAC1 channels. SLAC1 is a symmetric trimer composed from quasi-symmetric subunits, each having ten transmembrane helices arranged from helical hairpin pairs to form a central five-helix transmembrane pore that is gated by an extremely conserved phenylalanine residue. Conformational features suggest a mechanism for control of gating by kinase activation, and electrostatic features of the pore coupled with electrophysiological characteristics suggest that selectivity among different anions is largely a function of the energetic cost of ion dehydration. Users may view, print, copy, download and text and data- mine the content in such documents, for the purposes of academic research, subject always to the full Conditions of use: http://www.nature.com/authors/editorial_policies/license.html#terms Correspondence and request for materials should be addressed to WAH (wayne@convex.hhmi.columbia.edu).. Supplementary Information is linked to the online version of the paper at www.nature.com/nature. Author Contributions YC, LH, MP, RB, BH, BK and JL performed experiments; YC, LH, BR, SAS, and WAH analyzed data; YC, LH, SAS, and WAH prepared the manuscript. Author Information Atomic coordinates and diffraction data are deposited in the Protein Data Bank with accession codes 3M71, 3M73, 3M74, 3M75, 3M76, and 3M7L. See Table S4 for identifications. Reprints and permissions information is available at www.Nature.com/reprints. HHS Public Access Author manuscript Nature. Author manuscript; available in PMC 2013 January 18. Published in final edited form as: Nature. 2010 October 28; 467(7319): 1074–1080. doi:10.1038/nature09487. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Stomatal pores in the leaves of plants permit the influx of atmospheric carbon dioxide in exchange for transpirational evaporation of water1,2. A pair of kidney-shaped guard cells define each pore aperture, and turgor pressure variation in these cells determines the degree of stomatal pore openness. Depending on diverse environmental factors, the stomata close to prevent H2O losses and open to admit CO2 for photosynthesis. Environmental stimuli that lead to stomatal closure include darkness, high CO2 levels, ozone, low air humidity and drought. The plant hormone abscisic acid (ABA) is critical for signal transduction from these stimuli. Mutational screens in Arabidopsis thaliana for CO2 and ozone sensitivity identified a protein with ten predicted transmembrane (TM) helices, now called slow anion channel 1 (SLAC1), as having a central role in the control of stomatal closure3-5. Recent studies proved that SLAC1 is indeed an anion channel6-7, with characteristics like those of slow anion channels found in guard cells8, and that it is activated by phosphorylation from the OST1 kinase9. OST1 activity is negatively regulated by the ABI1 phosphatase10,11, which is in turn inhibited by the stomatal ABA receptors PYR/RCAR12,13 when in the ternary hormone-receptor-phosphatase complex14-18. Thereby, ABA stimulates SLAC1 channel activity. Resulting Cl− efflux through SLAC1 causes membrane depolarization, which activates outward-rectifying K+ channels, leading to KCl and water efflux to further reduce turgor and cause stomatal closure. SLAC1 expression in Arabidopsis is confined to the guard cells of leaves, but other Arabidopsis tissues do have SLAC1 homologs3, named SLAH1-SLAH4. The identifying mutations slac1-14 and slac1-23 are, respectively, in predicted transmembrane segments 9 (S456F) and 1 (G194D) of a protein that includes substantial N- and C-terminal extensions outside a 10-helix transmembrane (TM) domain. SLAH1, which is absent from leaves and lacks the terminal extensions of SLAC1, fully complements the mutant phenotype in slac1-2 guard cell protoplasts3. SLAC1 and homologs are also present in other plant genomes, including nine in rice (Oryza sativa) and five in grape vines (Vitis vinifera). SLAC1 relatives, some quite remote, also occur in bacteria, archaea and fungi. Known prokaryotic homologs contain only the predicted transmembrane domain of SLAC1, but some fungal homologs do have N- and C-terminal extensions. One homolog, Mae1 from the yeast Schizosaccharomyces pombe, functions as a malate uptake transporter19; another, Ssu1 from Saccharomyces cerevisiae and other fungi including Aspergillus fumigatus, is characterized as a sulfite efflux pump20,21; and TehA from Escherichia coli is identified as a tellurite resistance protein by virtue of its association in the tehA/tehB operon22,23. Despite a lack of further biochemical characterization, many homologs are annotated as tellurite resistance/ dicarboxylate transporter (TDT) proteins. We have undertaken structural and functional characterizations of the SLAC1 anion channel. We first solved an atomic-resolution crystal structure of the TehA homolog from Haemophilus influenzae, and we then developed a homology model for Arabidopsis SLAC1. This model allowed us to conduct mutagenesis for functional testing of structure-inspired hypotheses on gating and selectivity. We expressed Haemophilus TehA and Arabidopsis SLAC1 in Xenopus oocytes to characterize channel properties of these proteins and mutant variants. We also determined crystal structures for several mutant variants, including the homolog of slac1-2. Chen et al. Page 2 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Structure of SLAC1 bacterial homolog TehA We performed a bioinformatic analysis of SLAC1-related proteins, first clustering nearly 900 non-redundant sequences into a superfamily at the PSI-BLAST level E≤10−3, then into three distinct families at an initial threshold of E≤10−30, and finally into subfamilies at a typical initial threshold of E≤10−55. Since previous annotation is not well founded in experiment and SLAC1 is now the best characterized member, we adopt a nomenclature defining a SLAC superfamily divided into families identified as SF1-SF3 and subfamilies SF1A, SF1B etc. Family SF1 comprises the plant SLAC proteins and close bacterial homologs, family SF2 comprises a distinct set of bacterial proteins often annotated as exfoliative toxins, and family SF3 comprises the fungal Mae1 and Ssu1 proteins and their archaeal or bacterial homologs, respectively. SLAC family SF1 has three large subfamilies: the plant SLAC and SLAH proteins are in subfamily SF1A, closest bacterial homologs are in SF1B, and the TehA homologs are in SF1C (Fig. 1a). The other families also divide into subfamilies as detailed in Table S1, and family SF1 is divided into sub-subfamilies (Table S2). Two pertinent SF1 sequences are aligned in Fig. 1b. We used a structural genomics approach to obtain structural information, testing expression and purification for 43 bacterial and archaeal likely homologs, assaying for detergent choice and stability on 8 of these, finding two with appropriate profiles by size exclusion chromatography, and obtaining suitable crystals for one. This protein, TehA from H. influenzae (HiTehA), was found to be trimeric both by size-exclusion multi-angle light scattering (SEC-MALS) measurements and by chemical cross-linking. When solubilized in β-octylglucoside, HiTehA crystallized in space group R3 with a=b=96.01Å and c=136.27Å. Each asymmetric unit contains one subunit and 65% solvent. The structure was solved by selenomethionyl (SeMet) SAD phasing, ultimately at 1.50Å resolution (Table S3, Fig. S1), and then refined at 1.20Å resolution (Fig. 2a) to R/Rfree values of 14.1%/15.7% for a model that includes ordered residues 6-313, 213 water molecules and four detergent molecules (Table S4). The crystal structure has TehA trimers aligned with three-fold axes of the lattice (Fig. 2b). Subunits are tightly associated, burying 8947 Å2 of total surface area within trimer interfaces. The electrostatic potential surface is largely negative on the extracellular surface (Fig. 2c) and largely positive on the cytoplasmic surface (Fig. 2d). The membrane orientation is specified experimentally from GFP tagging of E. coli TehA24. Each TehA protomer has ten TM helices, as predicted; however, the fold is novel. Tandemly repeated helical hairpins are arranged with quasi-five-fold symmetry (Figs. 2e,S2). Extracellular inter-helix loops are short (2-5 residues) whereas intracellular inter-helix connections are longer, including a nine-residue helix H2,3 between TM2 and TM3 (Fig. 2f). An inner pentad of outwardly directed, TModd, helices creates an apparent pore through each protomer perpendicular to the putative membrane plane. TMeven helices from the five hairpins surround the inner pore and make an outer layer. Chen et al. Page 3 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Homology model for plant SLAC1 Arabidopsis SLAC1 (AtSLAC1) is substantially similar to bacterial homologs, notably HiTehA (Fig. 1b). All HiTehA TM helices are fully aligned to predicted SLAC1 TM helices, but there are short inter-helical gaps (1-5 residues) in all five extracellular loops and in two of the intracellular loops. The TM domain of AtSLAC1 (residues 188-504) aligns to HiTehA with 19% sequence identity and with a PSI-BLAST E-value of 3×10−22. For comparison, and in keeping with the family tree (Fig. 1a), the TM domain of AtSLAC1 shares sequence identities of 76% with rice SLAC1, 41% with Arabidopsis SLAH1, 25% with an SL1B homolog from Halorhodospira halophila, 11% with S. cerevisiae Ssu1, and 9% with S. pombe Mae1. A conceptual model with the AtSLAC1 sequence transposed onto the HiTehA helices sufficed to guide most of our mutational tests, but a detailed AtSLAC1 homology model helped to refine our ideas. Surface variability and electrostatic potential are plotted onto the surface of this model (Fig. 2g,2h). The most remarkable feature of the TehA structure and corresponding SLAC1 model is the central pore through each protomer. As for acetylcholine receptors25, the SLAC1 pore is formed by five helices; but the SLAC1 helices come from one protein molecule rather than five. The SLAC1 pore has a relatively uniform diameter of approximately 5Å across nearly five helical turns (Fig. S3), except for a pronounced constriction in the middle of the membrane (Fig. 3a) where the pore is occluded by the side chain of Phe 450 (Phe 262 in HiTehA). This residue is the only absolutely conserved amino acid residue of the SLAC1 family. The pore is lined with highly conserved (86% identity among five SLAC1 orthologs; 32% identity between AtSLAC1 and HiTehA) and generally hydrophobic residues (Figs. 1b, 3b,3c,S4). Despite this hydrophobicity, the electrostatic potential on the pore surface is polarized (Fig. 3a), presumably due to an invaginated shape adjacent to charged residues outside the membrane. The generally electropositive character of the cytoplasmic surface likely contributes to anion efflux. Kinks in the pore helices contribute to formation of a relatively constant pore diameter across the membrane. Four of the five HiTehA inner helices have centrally located proline residues, which necessarily generate kinks, and TM9 is kinked at a backbone-coordinated water molecule (Fig. 3d). Proline replaces Gly263 in all SF1A and SF1B relatives, including Pro451 AtSLAC1. This water-displacing change is isostructural26. Centrally located prolines also prevail in TM3, TM5 and TM7 across the SF1 proteins. The outer helices are longer and straighter, but more inclined. The only outer-helix proline kink of HiTehA is in TM6 at the trimer three-fold axis. Two of the gene-identifying mutations in Arabadopsis SLAC1 are selected point mutations, others are disruptive transfer DNA (T-DNA) insertions4. In the AtSLAC1 homology model, the slac1-2 (G194D) mutation3 points into the pore from TM1 and can be accommodated structurally whereas the slac1-1 (S456F) mutation4 points away from the pore six residues after pore-blocking Phe450 on TM9 and would be expected to disruptive. Ser456 interacts with outer helix TM10 in the homology model, and the phenyl bulk from S456F would not fit. Position 456 has alanine in HiTehA, 66% of all 204 SF1 homologs and another 27% have threonine or serine (as in all SLAC1 channels); phenylalanine never occurs among all Chen et al. Page 4 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 814 SLAC superfamily members. Position 194 has glycine in 58% of SF1 homologs and alanine in another 22%. Asp194, which never occurs naturally, would block the pore and be expected to repel anions. Mutational tests of channel function Mutational studies corroborate the hypothesis that a TehA-based SLAC1 model is appropriate. First, as discussed above, the slac1-1 mutant (S456F) is expected to be structurally disruptive, and indeed it is inactive in guard cells4, and the slac1-2 mutation (G194D) is expected to block the pore, and we show below that this variant is also inactive. We have also shown that the introduction of SLAC1-conserved proline residues into HiTehA (A208P/G263P) is accommodated isomorphously26. Moreover, as shown below, channel conductance properties of several mutants are similar for AtSLAC1 and HiTehA. To examine characteristics of the SLAC1 channel in light of the structural model, we performed electrophysiological tests of membrane currents from voltage-clamped Xenopus oocytes following injection of wild-type or mutant AtSLAC1 or HiTehA cRNAs. We observed modest-sized chloride currents with wild-type AtSLAC1 cRNA, as found previously6,7, but did not detect any chloride current following injection of wild-type HiTehA cRNA. We found that SLAC1 Cl− conductance was enhanced when the OST1 kinase cRNA was co-injected with SLAC1 cRNA, but only to the levels found by Lee et al.6 and not to the much higher levels found by Geiger et al.7 with OST1 physically connected to SLAC1 by split YFP linkage. In keeping with structural evidence that Phe262 blocks the HiTehA pore, removal of the phenyl group in HiTehA F262A or in the homologous AtSLAC1 F450A resulted in very large Cl− currents relative to wild-type levels, and the SLAC1 currents were now less enhanced by presence of OST1 (Fig. 4a). The tempting interpretation of an opened gate will require validation with appropriately analyzed single- channel recordings27. In keeping with the slac1-2 phenotype3, neither the functionally impaired AtSLAC1 G194D nor its HiTehA G15D homolog showed any substantial conductance; moreover, further consistent with pore blockage by Asp194 in slac1-2, the large conductances of HiTehA F262A and AtSLAC1 F450A were stopped in the double mutants AtSLAC1 G194D/F450A and HiTehA G15D/F262A (Fig. 4a). Here again, the effects in SLAC1 were independent of OST1. We also tested conductance characteristics for a series of AtSLAC1 F450X substitution mutants – F450A,G,T,V,L– and for the corresponding HiTehA F262X series – F262A,G,T,V,L (Figs. 4b,S5, Table S5). Findings from the two series are roughly parallel; in particular, the alanine and glycine substitutions lead to large currents for both and in comparison to the others. There are distinctions, of course, including generally higher conductances for AtSLAC1 over HiTehA and less conductance of F262T TehA compared to F450T SLAC1. It is also noteworthy that OST1 activation is very muted for the F450A,G,L mutants, which is consistent with SLAC1 gating at Phe450. Crystal structures were also determined for several of the HiTehA mutant variants (Table S4). The structures of F262A (1.15Å), F262V (1.60Å), F262L (1.65Å) and G15D (1.50Å) are all essentially isomorphous with the wild-type TehA structure with changes localized Chen et al. Page 5 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript primarily at the sites of mutation; the same is true for the double mutations of F262A/G15D, F262G/G15D and A208P/G263P26. The F262A structure has a wide open pore (Fig. 5a) with a relatively uniform pore diameter of ~5Å through ~30Å across the membrane (Fig. S3) whereas G15D has a doubly occluded pore (Fig. 5b). The pores of other mutant variants are consistent with the sizes of constrictive residues and with the observed conductances. Gating and activation The crystal structures of TehA and its mutant variants when taken together with the functional studies in Xenopus oocytes point to a crucial role for Phe450 in gating of the SLAC1 anion channel. Conservation of this residue across the SLAC1 family implies functional importance. The occlusion of the pore by the presence of F262 in the structure of wild-type TehA and the openness of the pore upon its substitution by alanine in the structure of the F262A mutant provides physical evidence for a gating role of this residue. This interpretation is supported by the correlated conductance characteristics from variants of the AtSLAC1 and HiTehA channels (Figs. 4b,S5). While these observations may suffice for placing the gate within the channel pore, they do not by themselves suggest a mechanism for gating in response to physiological stimuli. Some insight does come from conformational details defined at high resolution. One important structural clue is that the side chain of Phe262 is in a high energy conformation in the HiTehA structure, with χ1/χ2 at −160°/-4°. Although χ1 is in a preferred trans conformation, the phenyl ring is restricted by contacts with Val210 and Leu18 to a χ2 value near 0° rather than near to the preferred 90° orientation (Fig. 5c). Further evidence that Phe262 is restrained from local equilibrium comes from shifts observed in crystal structures of the F262A, F262V, F262L and F262G/G15D variants, which all show consistent backbone movements that displace Cβ(262) by 0.47-0.78Å (Fig. 5d). By contrast, Leu262 in F262L is in a preferred trans/gauche+ conformation at χ1/χ2 = 177°/63° as is Val262 in F262V at χ1 = −176°. What might control activation of HiTehA is unclear, but for AtSLAC1 activation is by OST1 phosphorylation6,7. The molecular consequences of OST1 phosphorylation of SLAC1 remain unknown, but it is plausible that associated shifts in pore- helix orientations would unlatch Phe450 in SLAC1 from a TehA-like restrained orientation. By analogy with Leu262 in F262L, we expect a preferred rotameric state for Leu450 in AtSLAC1 F450L. Thus, the lack of appreciable OST1 activation of conductance in AtSLAC1 F450L (Fig. 5e) might be explained by the lack of a restraining latch, whereby the channel remains closed despite OST1 activation. Puzzles certainly remain since OST1 does substantially activate AtSLAC1 F450T or F450V, which like HiTehA F262V should also be unrestrained; presumably activating adjustments widen the pore enough for ion permeation past threonine and valine but not leucine. Phosphorylation sites have been discovered in the N- and C-terminal tails of AtSLAC16,7,28 (179 and 51 residues long, respectively), but these alone cannot explain OST1 activation of SLAC1. First SLAH1, which fully complements the slac1-1 mutation, does not have these cytoplasmic tails. Secondly, although OST1 phosphorylation of Ser120 in the N-terminal tail is necessary for SLAC1 activation, it is not sufficient7. Thus, we surmise that direct phosphorylation of the SLAC1 TM domain must be critical, and SLAC1 has four conserved Chen et al. Page 6 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Ser/Thr candidates in its cytoplasmic loops. Moreover, SLAC1 proteins have proline- mediated kinks at the putative Phe450 gate in helix TM9, and also in adjacent helix TM7; these features may play a role in phosphorylation-driven unlatching of the Phe450 gate in SLAC1. Ion selectivity and discrimination Our studies of SLAC1 channel relative ion permeabilities, based on measurements of current reversal potential, are consistent with earlier work demonstrating that AtSLAC1 conducts anions but not cations and is selective among anions, with greater permeability for nitrate than for chloride (as in Vicia faba guard cell protoplasts27) and much reduced permeability for malate, bicarbonate or sulfate6,7. We also find that SLAC1 has little permeability for sulfite. Additionally, we find that wild-type SLAC1, F450A and F450T all have similar relative permeabilities to chloride, sulfite and malate, despite having widely different conductance levels, but the gating mutants do show small but significant decreases in nitrate permeability (Fig. 4c, Table S6). The relative insensitivity of anion permeability to gating residue changes suggests that selectivity for these anions may occur away from the central constriction at the channel gate. To some extent, ionic discrimination must depend on pore geometry; thus, an organic anion such as malate may be simply too large to pass through the 5-Å wide pore. Although the SLAC1 pore is lined largely with hydrophobic sidechains (Fig. 3c), it is sprinkled with hydroxyl groups from serine and threonine residues (16%); their electropositive hydrogen atoms may facilitate conductance. Most strikingly, the electrostatic potential within the AtSLAC1 pore is electropositive throughout (Fig. 3a). This polarization, promoted by charges on extra-membranous loops, no doubt contributes significantly in discrimination against cations. The relative anion permeability sequence of SLAC1 determined by us and others, I− > NO3− > Br− > Cl− 6,7, corresponds to selectivity sequence 1 compiled by Wright and Diamond29 for a range of anion-selective proteins. This sequence correlates inversely with the hydration energies of monovalent anions – anions with a lower hydration energy have a greater channel permeability. It is thought to be generated in proteins with weak, low field-strength, anion binding sites, where selectivity is largely determined by the energetic cost of anion dehydration. These selectivity results are thus consistent with the SLAC1 structure, where the pore lacks any obvious anion binding site. Distinctiveness of the SLAC1 channel SLAC1 anion channels are entirely novel in structure and, apparently, in the mechanism for ion conductance. The best characterized of anion channels belong to the CLC family of Cl− channels and transporters30-32. CLC channels have an altogether different architecture from the SLAC1 channel, and the mechanism for selectivity is also very different. Bacterial CLC transporters bind halide ions at three sites in a highly constricted pore30. By contrast, the SLAC1 pore has a relatively uniform diameter across the membrane, except where closed by the gating phenyl group, and we do not find discrete ion binding sites. CLC selectivity is governed by specific residues surrounding these binding sites30,32. The anion selectivity Chen et al. Page 7 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript sequence for the CLC channels of Cl−>Br−>NO −3>I−, opposite of that in SLAC1, is consistent with the high field-strength anion binding sites in CLC channels29. Interestingly, as for AtSLAC1, an Arabidopsis CLCa channel also preferentially transports nitrate ions33, and an E. coli CLC channel is converted to preference of nitrate when a generally conserved serine at the central site is substituted with proline as in AtCLCa32. SLAC1 also differs radically from other structurally characterized anion channels and transporters. These include the VDAC1 voltage-gated anion channel from mitochondrial outer membranes, which has a porin-like beta-barrel structure34-36, and a light-driven halorhodopsin chloride pump, which has a transmembrane conductance pathway similar to that of the proton-pumping pore of bacteriorhodopsin37. Although its channel structure is still only known by homology to other ABC transporters, CFTR is another obviously distinct chloride channel38. Cys-loop receptors also include anion channels39, and these are similar to SLAC1 in having five-helix pores25, but here selectivity is governed by charged groups at the entrance to the pore, which distinguish the anion-selective GABAA and glycine receptors from the cation-selective acetylcholine and serotonin 5HT3 receptors39. Finally, the recently identified TMEM16A gene for calcium-activated chloride channels40-42 appears to encode an 8-TM protein that is again distinct from SLAC1. Stomatal guard cells show both rapidly activated (R-type) and slow (S-type) anion channel activity43. Although slac1 guard cells have very defective S-type activity, their R-type currents are normal4. Guard cell protoplasts from the slac1-2 mutant abnormally accumulate Cl−, K+, malate and fumarate3, whereas SLAC1 shows negligible malate conductance7. As for SLAC1-associated K+ movements, other channels or transporters must be responsible for SLAC1-associated malate movements. Recent studies indicate that AtALMT12, an aluminium-activated malate transporter (ALMT) family member, is a malate-dependent R- type anion channel44 needed for stomatal closure45. Conclusions We find that many functional properties of the plant SLAC1 anion channel are explained well by the structure of an uncharacterized bacterial TehA protein that has been associated with tellurite resistance. SLAC1 and TehA belong to distinct subfamilies within one branch of a larger SLAC1 superfamily, but AtSLAC1 and HiTehA are sufficiently similar (19% sequence identity) that the SLAC1 homology model is predictive for function, including a verified placement of the identifying slac1-2 mutation G194D and a phenylalanine gate. One remaining puzzle concerns the structural change that activating phosphorylation elicits in SLAC1, and another puzzle is the biochemical role of the TehA homologs in bacteria. In a companion paper26, we examine functional and structural properties of TehA in bacteria, showing that it is anion channel, although actually not conferring tellurite resistance, and identifying a mutant variant with properties suggestive of an activated state. Thus, SLAC1 and TehA likely represent a large family of selective anion channels controlled by environmental stimuli. Chen et al. Page 8 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript METHODS Selection of target sequences TehA from E. coli (EcTehA) was a centrally selected sequence for expansion as a NYCOMPS targeted family. Target selection criteria utilized at NYCOMPS are described in details elsewhere46. Briefly, the EcTehA sequence was run against a dataset of about 40,000 predicted alpha helical integral membrane protein sequences from prokaryotic genomes (NYCOMPS98 dataset46) using PSI-BLAST50. Sequences that matched EcTehA with an E- value lower than 10−3 in an alignment extending over at least 50% of both predicted TM regions and passing our post-seed-expansion filtering criteria46 were passed to the protein production pipeline. Protein expression screening Full-length homologs from the following 38 species, including 2 sequences each from 5 of these, were amplified from genomic DNA by PCR: Thermoplasma acidophilum, Lactococcus lactis subsp. lactis Il1403, Streptococcus pyogenes M1 GAS, Streptococcus pneumoniae TIGR4, Haemophilus influenzae Rd KW20, Methanocaldococcus jannaschii DSM 2661 (2), Escherichia coli K12, Salmonella typhimurium LT2, Clostridium perfringens ATCC 13124, Xanthomonas campestris pv. campestris str. ATCC 33913, Streptococcus agalactiae 2603V/R, Shewanella oneidensis MR-1, Streptococcus mutans UA159, Archaeoglobus fulgidus DSM 4304, Vibrio parahaemolyticus RIMD 2210633 (2), Pseudomonas syringae pv. tomato str. DC3000 (2), Enterococcus faecalis V583, Pyrococcus horikoshii OT3, Bordetella bronchiseptica RB50, Streptomyces coelicolor A3, Corynebacterium glutamicum ATCC 13032, Picrophilus torridus DSM 9790, Acinetobacter sp. ADP1, Vibrio fischeri ES114, Pseudomonas fluorescens Pf-5, Sulfolobus acidocaldarius DSM 639, Colwellia psychrerythraea 34H, Neisseria meningitidis MC58, Rhodobacter sphaeroides 2.4.1, Vibrio cholerae O1 biovar eltor str. N16961 (2), Marinobacter aquaeolei VT8, Acinetobacter baumannii ATCC 17978, Klebsiella pneumoniae subsp. pneumoniae MGH 78578, Streptomyces avermitili MA-4680, Bordetella parapertussis 12822 (2), Streptococcus thermophilus LMG 18311, Salmonella enterica subsp. enterica serovar Paratyphi A str. ATCC 9150, and Anaeromyxobacter dehalogenans 2CP-C. Selected cDNAs were cloned into a modified pET vector (Novagen, Inc) that fuses a FLAG and deca-histidine tag at the C-terminus, which are cleavable by TEV protease. Proteins were expressed in E. coli BL21(DE3) plysS by a high throughput format (0.6ml in a deep well block) and purified after lysis by sonication using metal affinity purification in a buffer containing N-Dodecyl-β-D-Maltopyranoside. Samples were passed over an analytical size exclusion column in 12 different detergent-containing mobile phases, which included N- dodecyl-β-D-maltopyranoside (DDM), N-decyl-β-D-altopyranoside (DM), N-nonyl-β-D- altopyranoside (NM), N-octyl-β-D-altopyranoside (OM), N-octyl-β-D-Glucopyranoside (OG), N-nonyl-β-D-Glucopyranoside (NG) and lauryl dimethyl amine oxide (LDAO). Multi-angle light scattering with refractive index detection was used to analyze the oligomeric state51. The E. coli and H. influenzae proteins were judged to be monodisperse and stable and were passed to scale up. Chen et al. Page 9 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Scaled-up production and purification For scale up, in brief, transformed BL21 plysS cells were grown at 37°C in 2× YT media to OD 0.6-0.8 after being inoculated with 1% of the overnight culture. The culture was induced with 0.4mM IPTG and continued to grow at 37 °C for another 4 hours. The cells were harvested by centrifugation and stored at −80°C before use. Selenomethionyl (SeMet) TehA was expressed in a similar way, but using containing SeMet in place of methionine in defined minimal media. Cells were resuspended in a buffer containing 50 mM Tris-HCl (pH 8.0) and 200 mM NaCl and lysed using a French Press with two passes at 15-20,000 psi. Cell debris were removed by centrifugation at 10,000g for 20min, and the membrane fraction was isolated from that supernatant by ultra-centrifugation at 150,000g for 1 hr. The membrane fraction was homogenized in a solubilization buffer containing 50 mM Tris (pH 8.0) and 200 mM NaCl, and incubated with final concentration of 1% (w/v) dodecyl-β- D-maltopyranoside (DDM, Anatrace, Inc) for 1 hr at 4°C. The non-dissolved matter was remove by ultracentrifugation at 150,000g for 30 min, and the supernatant was loaded to a 5ml Hitrap Ni2+-NTA affinity column (GE Healthcare), pre-equilibrated with the same solubilization buffer supplemented with 0.05% DDM. After 20 column volume buffer wash, the protein was eluted with 250mM imidazole in the solublization buffer. The Flag and 10- His tags were removed by adding super TEV at 1:100 mass ratio and incubating at 4°C overnight. Tag removal was confirmed by SDS-PAGE, and the resulting sample was concentrated to around 10mg/ml. Preparative size exclusion chromatography was carried out on a Superdex-200 column for further purification, removal of TEV protease and the cleaved tag, and for buffer and detergent exchange. The gel-filtration buffer contained 10 mM Tris (pH 8.0), 200 mM NaCl, 1mM EDTA, 0.5 mM Tris [2-carboxyethyl] phosphine (TCEP), and 2×CMC of detergent. In the case of HiTehA, the protein was well behaved and stable in nearly all tested detergents, and we have purified it from DDM, DM, NM, OM, OG and LDAO. Protein characterization We performed N-terminal amino-acid sequencing of purified HiTehA and EcTehA prior to TEV protease treatment. Results from these analyses proved that true initiating methionine residue is located 14 residues after one annotated as N-terminal. The intervening nucleotide sequence contains a Shine-Delgarno sequence. For cross-linking experiments, purified HiTehA was incubated with 10 mM disuccinimidyl glutarate at room temperate for 30 min, and 100mM Tris-HCl pH 7.5 was added to stop the reaction. The incubated sample when run on a 8-25% gradient SDS-PAGE gel showed a ladder consistent with a trimeric structure. Crystallization and data collection Purified protein was concentrated to ~10mg/ml for initial crystal trials in a Mosquito robot with commercial screens from Hampton research, Emerald Biosystems and Molecular Dimension. We obtained crystals of HiTehA from protein in detergent DDM, DM, NM, OM, OG and LDAO, but only those from LDAO and OG gave diffraction to beyond 4Å spacings. Crystals that proved useful were all grown at 4°C using the sitting-drop vapor Chen et al. Page 10 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript diffusion method. After extensive optimization we reached conditions supporting very high resolution. Wild-type HiTehA and most of its variants crystallized from 1mM ZnSO4, 50mM Hepes-Na pH 7.8 and 28% PEG600. HiTehA F262A crystallized from 200mM Li2SO4, 100mM glycine pH9.3 and 33% PEG400). Addition of 10mM spermidine as an additive helped to obtain slightly better diffracting crystals. Cryoprotection was achieved by adding 5% ethylene glycol or PEG400 to the crystallization solution. Structure determination and refinement Native and SeMet SAD data set were collected at NSLS beamline X4A and processed using the software HKL200052. Crystals of HiTehA grew in space group R3 with a=b=c = 85.0Å and α=β=γ = 93.5°. All subsequent manipulations were done in the hexagonal setting of this space group with a=b= 96.0Å and c = 136.7Å. The asymmetric unit contains one TehA protomer and 65% solvent volume. The structure was determined at 2.0Å resolution from single-wavelength anomalous diffraction (SAD) using selenomethionine-substituted protein crystals. Assessment of data quality for phasing, location of heavy atom sites and initial phases were calculated using the HKL2MAP interface to SHELX programs53. All the secondary structure elements were clearly visible in the experimental electron density map. Automatic model building was done in Arp/wArp54 and completed manually in the program COOT55. The model was refined against native data at 1.20Å resolution using the program Refmac5.5 in CCP456, with anisotropic B factor restrained refinement applied. Subsequent structural analyses of mutant variants were refined as isomorphous structures. Site-directed mutagenesis Site-directed mutants were constructed using the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA 92037) and expressed from pET vectors in E. coli BL21(DE3) plysS cells as for the wild-type protein. Electrophysiology All constructs were cloned into plasmid pGHME2, linearized, and transcribed into cRNA using T7 polymerase (mMessage mMachine, Ambion). Oocytes were injected with 50 nl of cRNA solution each, at a constant concentration of 0.5 mg/ml for HiTehA, or 0.5 mg/ml for AtSLAC1 constructs, with or without 0.5 mg/ml of AtOST1 constructs and recorded in voltage-clamp experiments 2 days after injection. For mixed expression experiments, 25nl of cRNA solution was injected for each AtSLAC1 component. Two-electrode voltage clamp recordings were performed to measure HiTehA or AtSLAC1 currents as described6,7. The pipette solutions contained 3 M KCl. For voltage-clamp current recordings, the bath solution contained 1 mM MgCl2, 1 mM CaCl2, 10 mM Mes/Tris (pH 5.6) and 30 mM CsCl; For anion selectivity measurements, the bath solutions contained 50 mM Cl−, NO −3, malate, or 30 mM SO 2-3 sodium salts, 45 mM Na-gluconate, 1 mM Ca-gluconate2, 1 mM Mg- gluconate2, 1 mM K-gluconate, as well as 10 mM Tris/Mes (pH 5.6). Osmolality was adjusted with D-mannitol to 220 mOsmol/kg. The bath electrode was a 3M KCl agar bridge. The testing potentials were +50 to −110 or −130 mV in 20 mV decrements with 7.5 s duration. The holding and return potentials were 0 mV with a 1.45 s or 2 s, respectively. I-V Chen et al. Page 11 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript relations of HiTehA or AtSLAC1 were generated from the current reached after 0.5 s in testing potentials. The Goldman-Hodgkin-Katz equation was applied to estimate permeability ratios for monovalent ions as described6. For divalent anions, the permeability ratios were derived according to Fatt and Ginsborg57. Bioinformatic analysis of SLAC-related proteins Sequences related to SLAC1 were analyzed comprehensively by PSI-BLAST50. Searches at E<10−3 starting from five disparate homologs each identified a common pool of over 900 proteins, which when pooled were used for sub-classification into families and subfamilies. Details of these analyses are reported in footnotes to Table S1. Molecular figures were produced in PyMOL58. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We thank Filippo Mancia, Larry Shapiro and Ming Zhou for helpful discussions; Patricia Rodriguez and Arianne Morrison for help with cloning and expression; Jonah Cheung, Mark Collins, Chris Min, Sheng Ye and Zhen Zhang for advice in protein chemistry and crystallography; John Schwanof and Randy Abramowitz for help with synchrotron experiments; John Riley for help with Xenopus oocyte injections; and Min Su for advice on the PSI- BLAST analysis of SLAC relatives. This project was supported in part by an award to the New York Consortium on Membrane Protein Structure (NYCOMPS) from the NIGMS Protein Structure Initiative. Beamline X4A at the National Synchrotron Light Source (NSLS), a DOE facility at Brookhaven National Laboratory, is supported by the New York Structural Biology Center. References 1. Hetherington AM, Woodward FI. The role of stomata in sensing and driving environmental change. Nature. 2003; 424:901–908. [PubMed: 12931178] 2. Sirichandra C, Wasilewska A, Vlad F, Valon C, Leung J. The guard cell as a single-cell model towards understanding drought tolerance and abscisic acid action. J. Exp. Bot. 2009; 60:1439–1463. [PubMed: 19181866] 3. Negi J, et al. CO2 regulator SLAC1 and its homologues are essential for anion homeostatsis in plant cells. Nature. 2008; 452:483–486. [PubMed: 18305482] 4. Vahisalu T, et al. SLAC1 is required for plant guard cell S-type anion channel function in stomatal signalling. Nature. 2008; 452:487–491. [PubMed: 18305484] 5. Saji S, et al. Disruption of a gene encoding C4-dicarboxylate transporter-like protein increases ozone sensitivity through deregulation of the stomatal response in Arabidopsis thaliana. Plant Cell Physiol. 2008; 49:2–10. [PubMed: 18084014] 6. Lee SC, Lan W, Buchanan BB, Luan S. A protein kinase-phophatase pair interacts with an ion channel to regulate ABA signaling in plant guard cells. Proc. Natl. Acad. Sci. USA. 2009; 106:21419–21424. [PubMed: 19955427] 7. Geiger D, et al. Activity of guard cell anion channel SLAC1 is controlled by drought-stress signaling kinase-phosphatase pair. Proc. Natl. Acad. Sci. USA. 2009; 106:21425–21430. [PubMed: 19955405] 8. Schroeder JI, Hagiwara S. Cytosolic calcium regulates ion channels in the plasma membrane of Vicia faba guard cells. Nature. 1989; 338:427–430. 9. Mustilli A, Merlot S, Vavasseur A, Fenzi F, Giraudat J. Arabidopsis OST1 protein kinase mediates the regulation of stomatal aperture by abscisic acid and acts upstream of reactive oxygen species production. Plant Cell. 2002; 14:3089–3099. [PubMed: 12468729] Chen et al. Page 12 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 10. Leung J, et al. Arabidopsis ABA response gene ABI1: features of a calcium-modulated protein phosphatase. Science. 1994; 264:1448–1452. [PubMed: 7910981] 11. Meyer K, Leube MP, Grill E. A protein phosphatase 2C involved in ABA signal transduction in Arabidopsis thaliana. Science. 1994; 264:1452–1455. [PubMed: 8197457] 12. Ma Y, et al. Regulators of PP2C phosphatase activity function as abscisic acid sensors. Science. 2009; 324:1064–1068. [PubMed: 19407143] 13. Park S, et al. Abscisic acid inhibits type 2C protein phosphatases via the PYR/PYL family of START proteins. Science. 2009; 324:1068–1071. [PubMed: 19407142] 14. Melcher K, et al. A gate-latch-lock mechanism for hormone signalling by abscisic acid receptors. Nature. 2009; 462:602–608. [PubMed: 19898420] 15. Miyazono K, et al. Structural basis of abscisic acid signalling. Nature. 2009; 462:609–614. [PubMed: 19855379] 16. Fujii H, et al. In vitro reconstitution of an abscisic acid signalling pathway. Nature. 2009; 462:660– 664. [PubMed: 19924127] 17. Yin P, et al. Structural insights into the mechanism of abscisic acid signaling by PYL proteins. Nat. Struct. Mol. Biol. 2009; 16:1230–1236. [PubMed: 19893533] 18. Nishimura N, et al. PYR/PYL/RCAR family members are major in-vivo ABI1 protein phosphatase 2C-interacting proteins in Arabidopsis. Plant J. 2010; 61:290–299. [PubMed: 19874541] 19. Grobler J, Bauer F, Subden RE, van Vuuren HJ. The MAE1 gene of Schizosaccharomyces pombe encodes a permease for malate and other C4 dicarboxylic acids. Yeast. 1995; 11:1485–1491. [PubMed: 8750236] 20. Park H, Bakalinsky AT. SSU1 mediates sulphite efflux in Saccharomyces cerevisiae. Yeast. 2000; 16:881–888. [PubMed: 10870099] 21. Léchenne B, et al. Sulphite efflux pumps in Aspergillus fumigatus and dermatophytes. Microbiology. 2007; 153:905–913. [PubMed: 17322211] 22. Walter EG, Weiner JH, Taylor DE. Nucleotide sequence and overexpression of the tellurite- resistance determinant from the IncHII plasmid pHH1508a. Gene. 1991; 101:1–7. [PubMed: 2060788] 23. Taylor DE, Hou Y, Turner RJ, Weiner JH. Location of a potassium tellurite resistance operon (tehA tehB) within the terminus of Escherichia coli K-12. J Bacteriol. 1994; 176:2740–2742. [PubMed: 8169225] 24. Daley DO, et al. Global topology analysis of the Escherichia coli inner membrane proteome. Science. 2005; 308:1321–1323. [PubMed: 15919996] 25. Unwin N. Refined structure of the nicotinic acetylcholine receptor at 4Å resolution. J. Mol. Biol. 2005; 346:967–989. [PubMed: 15701510] 26. Chen Y, Hu L, Siegelbaum SA, Hendrickson WA. Structure-based analysis of anion channel TehA and methyltransferase TehB implicated in bacterial tellurite resistance. Submitted. 27. Schmidt C, Schroeder JI. Anion Selectivity of Slow Anion Channels in the Plasma Membrane of Guard Cells (Large Nitrate Permeability). Plant Physiol. 1994; 106:383–391. [PubMed: 12232336] 28. Vahisalu T, et al. Ozone-triggered rapid stomatal response involves the production of reactive oxygen species, and is controlled by SLAC1 and OST1. Plant J. 2010; 62:442–53. [PubMed: 20128877] 29. Wright EM, Diamond JM. Anion selectivity in biological systems. Physiol. Rev. 1977; 57:109– 156. [PubMed: 834775] 30. Dutzler R, Campbell EB, MacKinnon R. Gating the selectivity filter in ClC chloride channels. Science. 2003; 300:108–112. [PubMed: 12649487] 31. Accardi A, Miller C. Secondary active transport mediated by a prokaryotic homologue of ClC Cl- channels. Nature. 2004; 427:803–807. [PubMed: 14985752] 32. Picollo A, Malvezzi M, Houtman JC, Accardi A. Basis of substrate binding and conservation of selectivity in the CLC family of channels and transporters. Nat. Struct. Mol. Biol. 2009; 16:294– 301. [PubMed: 19219045] 33. De Angeli A, et al. The nitrate/proton antiporter AtCLCa mediates nitrate accumulation in plant vacuoles. Nature. 2006; 442:939–942. [PubMed: 16878138] Chen et al. Page 13 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 34. Hiller S, Garces RG, Malia TJ, Orekhov VY, Colombini M, Wagner G. Solution structure of the integral human membrane protein VDAC-1 in detergent micelles. Science. 2008; 321:1206–1210. [PubMed: 18755977] 35. Bayrhuber M, et al. Structure of the human voltage-dependent anion channel. Proc. Natl. Acad. Sci. USA. 2008; 105:15370–15375. [PubMed: 18832158] 36. Ujwal R, et al. The crystal structure of mouse VDAC1 at 2.3 Å resolution reveals mechanistic insights into metabolite gating. Proc. Natl. Acad. Sci. USA. 2008; 105:17742–17747. [PubMed: 18988731] 37. Kouyama T, et al. Crystal structure of the light-driven chloride pump halorhodopsin from Natronomonas pharaonis. J. Mol. Biol. 2010; 396:564–579. [PubMed: 19961859] 38. Gadsby DC, Vergani P, Csanády L. The ABC protein turned chloride channel whose failure causes cystic fibrosis. Nature. 2006; 440:477–483. [PubMed: 16554808] 39. Miller PS, Smart TG. Binding, activation and modulation of Cys-loop receptors. Trends Pharmacol. Sci. 2010; 31:161–174. [PubMed: 20096941] 40. Yang YD, et al. TMEM16A confers receptor-activated calcium-dependent chloride conductance. Nature. 2008; 455:1210–5. [PubMed: 18724360] 41. Caputo A, et al. TMEM16A, a membrane protein associated with calcium-dependent chloride channel activity. Science. 2008; 322:590–594. [PubMed: 18772398] 42. Schroeder BC, Cheng T, Jan YN, Jan LY. Expression cloning of TMEM16A as a calcium- activated chloride channel subunit. Cell. 2008; 134:1019–1029. [PubMed: 18805094] 43. Schroeder JI, Keller BU. Two types of anion channel currents in guard cells with distinct voltage regulation. Proc. Natl. Acad. Sci. USA. 1992; 89:5025–5029. [PubMed: 1375754] 44. Meyer S, et al. AtALMT12 represents an R-type anion channel required for stomatal movement in Arabidopsis guard cells. Plant J. Jul 12.2010 Epub ahead of print. 45. Sasaki T, et al. Closing plant stomata requires a homolog of an aluminum-activated malate transporter. Plant Cell Physiol. 2010; 51:354–65. [PubMed: 20154005] 46. Punta M, et al. Structural genomics target selection for the New York Consortium on Membrane Protein Structure. J. Struct. Funct. Genomics. 2009; 4:255–268. [PubMed: 19859826] 47. Papadopoulos JS, Agarwala R. COBALT: constraint-based alignment tool for multiple protein sequences. Bioinformatics. 2007; 23:1073–1079. [PubMed: 17332019] 48. Landau M, et al. ConSurf 2005: the projection of evolutionary conservation scores of residues on protein structures. Nucleic Acids Res. 2005; 33:W299–W302. [PubMed: 15980475] 49. Rocchia W, et al. Rapid grid-based construction of the molecular surface for both molecules and geometric objects: applications to the finite difference Poisson-Boltzmann method. J. Comp. Chem. 2002; 23:128–137. [PubMed: 11913378] 50. Altschul SF, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997; 25:3389–3402. [PubMed: 9254694] 51. Kendrick BS, Kerwin BA, Chang BS, Philo JS. Online size-exclusion high-performance liquid chromatography light scattering and differential refractometry methods to determine degree of polymer conjugation to proteins and protein-protein or protein-ligand association states. Anal. Biochem. 2001; 299:136–146. [PubMed: 11730335] 52. Otwinowski Z, Minor W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 1997; 276:307–326. 53. Pape P, Schneider TR. HKL2MAP: a graphical user interface for phasing with SHELX programs. J. Appl. Cryst. 2004; 37:843–844. 54. Perrakis A, Morris R, Lamzin VS. Automated protein model building combined with iterative structure refinement. Nat. Struct. Biol. 1999; 6:458–463. [PubMed: 10331874] 55. Emsley P, Cowtan K. COOT: model-building tools for molecular graphics. Acta Crystallogr. D. 2004; 60:2126–2132. [PubMed: 15572765] 56. CCP4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D. 1994; 50:760– 763. [PubMed: 15299374] 57. Fatt P, Ginsborg BL. The ionic requirements for the production of action potentials in crustacean muscle fibers. J. Physiol. 1958; 142:516–543. [PubMed: 13576452] Chen et al. Page 14 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 58. DeLano, WL. The PyMOL molecular graphics system. DeLano Scientific; San Carlos, CA, USA: 2002. http://www.pymol.org Chen et al. Page 15 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 1. Sequence analysis for the SLAC1 superfamily a, Family tree. The presentation was computed by the program COBALT47 from representative subfamily sequences (Table S1), including Arabidopsis thaliana SLAC1 for SF1Ai, Haemophilus influenzae TehA for SF1Ci, Escherichia coli TehA for SF1Cii, Vibrio parahaemolyticus for SF2A, Staphylococcus aureus for SF2B, Aspergillus fumigatus Ssu1 for SF3A, and Schizosaccharomyces pombe Mae1 for SF3B. b, Structure-based sequence alignment of TehA from Haemophilus influenzae (HiTehA) and SLAC1 from Arabidopsis thaliana (AtSLAC1). The TehA structure has been used to restrict sequence gaps to inter- helical segments. Superior coils define extents of the HiTehA helical segments; red letters mark residue identities; red boxes are drawn for residues that are >95% identical within the plant subfamily SF1A for AtSLAC1 or within the TehA subfamily for HiTehA; red diamonds mark HiTehA residues that line the central pore; and the colored inferior bar encodes ConSurf sequence variability48 for the SF1 family of 204 non-redundant proteins. Chen et al. Page 16 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 2. Crystal structure of HiTehA and homology model of AtSLAC1 a, Electron density distribution from the HiTehA crystal structure at 1.2Å resolution. The map has (2Fo-Fc) coefficients based on the superimposed model. Contours are at 2.5σ. b, Ribbon diagram of the HiTehA trimer. Each protomer is colored spectrally from blue at its N-terminus to red at its C-terminus. c, DelPhi49 electrostatic potential at the extracellular surface. Electronegative and electropositive potential are colored in degrees of red and blue saturation, respectively. d, Electrostatic potential49 at the intracellular surface. e, Ribbon diagram of an HiTehA protomer viewed from outside the membrane. The ribbon is colored spectrally as in 2b. f, Ribbon diagram of an HiTehA protomer viewed from within the membrane from below the view of e. g, Surface of a homology model of AtSLAC1, viewed as in f, and colored by sequence variability48. h, Surface of AtSLAC1 as in g, but colored by electrostatic potential49. Chen et al. Page 17 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 3. Putative structure of the SLAC1 conductance pore a, Cross-section through the homology model of AtSLAC1. The model is viewed as in 2i, with the electrostatic potential49 shown on the external surface of the molecular envelope. The side chain of Phe450 is shown as a stick model (red) on the backbone ribbon colored yellow. b, Cross-section as in a, but colored by surface conservation48 as in 2h. c, Pore- lining residues in the SLAC1 homology model. A cylinder model (left), spectrally colored as in 2b, provides a key for viewing the rolled-open structure (left) with pore-lining residues of AtSLAC1 shown on the TModd helices. d, Ribbon diagrams of HiTehA TM9 (left) and TM7 (right) viewed from within the conductance pore. The side chains of Pro207 and Phe262 are shown as well as a kink-stabilizing water HOH25 that is coordinated by the NH of Gly263 and by C=O groups of Gly202 and Ala259. Density contours are shown for the water molecule. Chen et al. Page 18 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 4. Ionic conductance measurements a, Typical microelectrode voltage-clamp current traces from oocytes injected with various channel cRNAs. Left column, oocytes injected with cRNAs encoding wild-type HiTehA channels (WT), or F262A, G15D/F262A or G15D mutants. Right two columns, oocytes injected with cRNAs for WT AtSLAC1, or F450A, G194D/F450A or G194D mutants, with or without co-injection of AtOST1. Dotted lines represent zero current levels. Extracellular solution contained 30mM CsCl. Schematic keys at the left place phenyl and/or aspartyl gates. b, Effects of gating residue mutations. Mean chloride currents, measured at −90 mV, are shown comparing WT HiTehA and its mutant series F262A,G,T,V,L with WT AtSLAC1 and its corresponding series F450A,G,T,V,L, both alone and co-expressed with AtOST1. Full I-V relations are shown in Fig. S5. c, Effect of gating residues on relative AtSLAC1 anion permeabilities. Relative permeabilities (P[X]/P[Cl]) for chloride, nitrate, sulfite and malate of WT, F450A and F450T SLAC1 channels were measured from the change in current reversal potential with Cl− or anion X− as the sole permeant anion in the bath solution (Methods, Table S6). Chen et al. Page 19 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 5. Structural features at the SLAC1 homolog gate a,b, Cross-sections through the conductance pores of HiTehA F262A and HiTehA G15D. The view and presentations are as in 3a, except that helices are colored purple. c, Molecular basis for conformational strain in gating residue Phe262 of HiTehA. Helices TM1 (left), TM9 (center) and TM7 (right) are viewed from within the pore and presented as ribbon diagrams with selected side chains drawn in stick representation. The local low-energy conformation for the phenyl ring (χ2 = 90°) is shown in thin lines with short contacts indicated in dashed lines: Leu18 Cδ2 – Phe262 Cδ1, d = 2.4 Å; Val210 Cγ2 – Phe262 Cε2, d = 2.8 Å. d, Conformational shifts consequent to release of strain in gating residue Phe262. Cα backbone structures of F262A, F262V and F262L HiTehA were first superimposed onto WT HiTehA, and residues 258-266 were drawn in stereo as superimposed. All backbone atoms are shown for peptides 262±1, but only Cα atoms are shown otherwise. The WT backbone and phenyl group are green; other backbone are all magenta; side chains of Ala262, Val262 and Leu262 are cyan, blue and red, respectively; oxygen-directed bonds are red and nitrogen-directed bonds are blue. e, Typical microelectrode voltage-clamp current traces from oocytes injected with wild-type (WT) or F450L AtSLAC1 cRNA. Experimental conditions and displays are as in 4a. Chen et al. Page 20 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript
3M74
Crystal Structure of Plant SLAC1 homolog TehA
Homolog Structure of the SLAC1 Anion Channel for Closing Stomata in Leaves Yuhang Chen1,6, Lei Hu2, Marco Punta6,7, Renato Bruni6, Brandan Hillerich6, Brian Kloss6, Burkhard Rost1,6,1, James Love6, Steven A. Siegelbaum2,3,5, and Wayne A. Hendrickson1,4,5,6 1Department of Biochemistry and Molecular Biophysics, Columbia University, New York, NY 10032, USA 2Department of Neuroscience, Columbia University, New York, NY 10032, USA 3Department of Pharmacology, Columbia University, New York, NY 10032, USA 4Department of Physiology and Cellular Biophysics, Columbia University, New York, NY 10032, USA 5Howard Hughes Medical Institute, Columbia University, New York, NY 10032, USA 6NYCOMPS, New York Structural Biology Center 89 Convent Avenue, New York, NY 10027 USA 7Department of Computer Science and Institute for Advanced Study Technical University of Munich D-85748 Munich, Germany Summary The plant SLAC1 anion channel controls turgor pressure in the aperture-defining guard cells of plant stomata, thereby regulating exchange of water vapor and photosynthetic gases in response to environmental signals such as drought or high levels of carbon dioxide. We determined the crystal structure of a bacterial homolog of SLAC1 at 1.20Å resolution, and we have used structure- inspired mutagenesis to analyze the conductance properties of SLAC1 channels. SLAC1 is a symmetric trimer composed from quasi-symmetric subunits, each having ten transmembrane helices arranged from helical hairpin pairs to form a central five-helix transmembrane pore that is gated by an extremely conserved phenylalanine residue. Conformational features suggest a mechanism for control of gating by kinase activation, and electrostatic features of the pore coupled with electrophysiological characteristics suggest that selectivity among different anions is largely a function of the energetic cost of ion dehydration. Users may view, print, copy, download and text and data- mine the content in such documents, for the purposes of academic research, subject always to the full Conditions of use: http://www.nature.com/authors/editorial_policies/license.html#terms Correspondence and request for materials should be addressed to WAH (wayne@convex.hhmi.columbia.edu).. Supplementary Information is linked to the online version of the paper at www.nature.com/nature. Author Contributions YC, LH, MP, RB, BH, BK and JL performed experiments; YC, LH, BR, SAS, and WAH analyzed data; YC, LH, SAS, and WAH prepared the manuscript. Author Information Atomic coordinates and diffraction data are deposited in the Protein Data Bank with accession codes 3M71, 3M73, 3M74, 3M75, 3M76, and 3M7L. See Table S4 for identifications. Reprints and permissions information is available at www.Nature.com/reprints. HHS Public Access Author manuscript Nature. Author manuscript; available in PMC 2013 January 18. Published in final edited form as: Nature. 2010 October 28; 467(7319): 1074–1080. doi:10.1038/nature09487. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Stomatal pores in the leaves of plants permit the influx of atmospheric carbon dioxide in exchange for transpirational evaporation of water1,2. A pair of kidney-shaped guard cells define each pore aperture, and turgor pressure variation in these cells determines the degree of stomatal pore openness. Depending on diverse environmental factors, the stomata close to prevent H2O losses and open to admit CO2 for photosynthesis. Environmental stimuli that lead to stomatal closure include darkness, high CO2 levels, ozone, low air humidity and drought. The plant hormone abscisic acid (ABA) is critical for signal transduction from these stimuli. Mutational screens in Arabidopsis thaliana for CO2 and ozone sensitivity identified a protein with ten predicted transmembrane (TM) helices, now called slow anion channel 1 (SLAC1), as having a central role in the control of stomatal closure3-5. Recent studies proved that SLAC1 is indeed an anion channel6-7, with characteristics like those of slow anion channels found in guard cells8, and that it is activated by phosphorylation from the OST1 kinase9. OST1 activity is negatively regulated by the ABI1 phosphatase10,11, which is in turn inhibited by the stomatal ABA receptors PYR/RCAR12,13 when in the ternary hormone-receptor-phosphatase complex14-18. Thereby, ABA stimulates SLAC1 channel activity. Resulting Cl− efflux through SLAC1 causes membrane depolarization, which activates outward-rectifying K+ channels, leading to KCl and water efflux to further reduce turgor and cause stomatal closure. SLAC1 expression in Arabidopsis is confined to the guard cells of leaves, but other Arabidopsis tissues do have SLAC1 homologs3, named SLAH1-SLAH4. The identifying mutations slac1-14 and slac1-23 are, respectively, in predicted transmembrane segments 9 (S456F) and 1 (G194D) of a protein that includes substantial N- and C-terminal extensions outside a 10-helix transmembrane (TM) domain. SLAH1, which is absent from leaves and lacks the terminal extensions of SLAC1, fully complements the mutant phenotype in slac1-2 guard cell protoplasts3. SLAC1 and homologs are also present in other plant genomes, including nine in rice (Oryza sativa) and five in grape vines (Vitis vinifera). SLAC1 relatives, some quite remote, also occur in bacteria, archaea and fungi. Known prokaryotic homologs contain only the predicted transmembrane domain of SLAC1, but some fungal homologs do have N- and C-terminal extensions. One homolog, Mae1 from the yeast Schizosaccharomyces pombe, functions as a malate uptake transporter19; another, Ssu1 from Saccharomyces cerevisiae and other fungi including Aspergillus fumigatus, is characterized as a sulfite efflux pump20,21; and TehA from Escherichia coli is identified as a tellurite resistance protein by virtue of its association in the tehA/tehB operon22,23. Despite a lack of further biochemical characterization, many homologs are annotated as tellurite resistance/ dicarboxylate transporter (TDT) proteins. We have undertaken structural and functional characterizations of the SLAC1 anion channel. We first solved an atomic-resolution crystal structure of the TehA homolog from Haemophilus influenzae, and we then developed a homology model for Arabidopsis SLAC1. This model allowed us to conduct mutagenesis for functional testing of structure-inspired hypotheses on gating and selectivity. We expressed Haemophilus TehA and Arabidopsis SLAC1 in Xenopus oocytes to characterize channel properties of these proteins and mutant variants. We also determined crystal structures for several mutant variants, including the homolog of slac1-2. Chen et al. Page 2 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Structure of SLAC1 bacterial homolog TehA We performed a bioinformatic analysis of SLAC1-related proteins, first clustering nearly 900 non-redundant sequences into a superfamily at the PSI-BLAST level E≤10−3, then into three distinct families at an initial threshold of E≤10−30, and finally into subfamilies at a typical initial threshold of E≤10−55. Since previous annotation is not well founded in experiment and SLAC1 is now the best characterized member, we adopt a nomenclature defining a SLAC superfamily divided into families identified as SF1-SF3 and subfamilies SF1A, SF1B etc. Family SF1 comprises the plant SLAC proteins and close bacterial homologs, family SF2 comprises a distinct set of bacterial proteins often annotated as exfoliative toxins, and family SF3 comprises the fungal Mae1 and Ssu1 proteins and their archaeal or bacterial homologs, respectively. SLAC family SF1 has three large subfamilies: the plant SLAC and SLAH proteins are in subfamily SF1A, closest bacterial homologs are in SF1B, and the TehA homologs are in SF1C (Fig. 1a). The other families also divide into subfamilies as detailed in Table S1, and family SF1 is divided into sub-subfamilies (Table S2). Two pertinent SF1 sequences are aligned in Fig. 1b. We used a structural genomics approach to obtain structural information, testing expression and purification for 43 bacterial and archaeal likely homologs, assaying for detergent choice and stability on 8 of these, finding two with appropriate profiles by size exclusion chromatography, and obtaining suitable crystals for one. This protein, TehA from H. influenzae (HiTehA), was found to be trimeric both by size-exclusion multi-angle light scattering (SEC-MALS) measurements and by chemical cross-linking. When solubilized in β-octylglucoside, HiTehA crystallized in space group R3 with a=b=96.01Å and c=136.27Å. Each asymmetric unit contains one subunit and 65% solvent. The structure was solved by selenomethionyl (SeMet) SAD phasing, ultimately at 1.50Å resolution (Table S3, Fig. S1), and then refined at 1.20Å resolution (Fig. 2a) to R/Rfree values of 14.1%/15.7% for a model that includes ordered residues 6-313, 213 water molecules and four detergent molecules (Table S4). The crystal structure has TehA trimers aligned with three-fold axes of the lattice (Fig. 2b). Subunits are tightly associated, burying 8947 Å2 of total surface area within trimer interfaces. The electrostatic potential surface is largely negative on the extracellular surface (Fig. 2c) and largely positive on the cytoplasmic surface (Fig. 2d). The membrane orientation is specified experimentally from GFP tagging of E. coli TehA24. Each TehA protomer has ten TM helices, as predicted; however, the fold is novel. Tandemly repeated helical hairpins are arranged with quasi-five-fold symmetry (Figs. 2e,S2). Extracellular inter-helix loops are short (2-5 residues) whereas intracellular inter-helix connections are longer, including a nine-residue helix H2,3 between TM2 and TM3 (Fig. 2f). An inner pentad of outwardly directed, TModd, helices creates an apparent pore through each protomer perpendicular to the putative membrane plane. TMeven helices from the five hairpins surround the inner pore and make an outer layer. Chen et al. Page 3 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Homology model for plant SLAC1 Arabidopsis SLAC1 (AtSLAC1) is substantially similar to bacterial homologs, notably HiTehA (Fig. 1b). All HiTehA TM helices are fully aligned to predicted SLAC1 TM helices, but there are short inter-helical gaps (1-5 residues) in all five extracellular loops and in two of the intracellular loops. The TM domain of AtSLAC1 (residues 188-504) aligns to HiTehA with 19% sequence identity and with a PSI-BLAST E-value of 3×10−22. For comparison, and in keeping with the family tree (Fig. 1a), the TM domain of AtSLAC1 shares sequence identities of 76% with rice SLAC1, 41% with Arabidopsis SLAH1, 25% with an SL1B homolog from Halorhodospira halophila, 11% with S. cerevisiae Ssu1, and 9% with S. pombe Mae1. A conceptual model with the AtSLAC1 sequence transposed onto the HiTehA helices sufficed to guide most of our mutational tests, but a detailed AtSLAC1 homology model helped to refine our ideas. Surface variability and electrostatic potential are plotted onto the surface of this model (Fig. 2g,2h). The most remarkable feature of the TehA structure and corresponding SLAC1 model is the central pore through each protomer. As for acetylcholine receptors25, the SLAC1 pore is formed by five helices; but the SLAC1 helices come from one protein molecule rather than five. The SLAC1 pore has a relatively uniform diameter of approximately 5Å across nearly five helical turns (Fig. S3), except for a pronounced constriction in the middle of the membrane (Fig. 3a) where the pore is occluded by the side chain of Phe 450 (Phe 262 in HiTehA). This residue is the only absolutely conserved amino acid residue of the SLAC1 family. The pore is lined with highly conserved (86% identity among five SLAC1 orthologs; 32% identity between AtSLAC1 and HiTehA) and generally hydrophobic residues (Figs. 1b, 3b,3c,S4). Despite this hydrophobicity, the electrostatic potential on the pore surface is polarized (Fig. 3a), presumably due to an invaginated shape adjacent to charged residues outside the membrane. The generally electropositive character of the cytoplasmic surface likely contributes to anion efflux. Kinks in the pore helices contribute to formation of a relatively constant pore diameter across the membrane. Four of the five HiTehA inner helices have centrally located proline residues, which necessarily generate kinks, and TM9 is kinked at a backbone-coordinated water molecule (Fig. 3d). Proline replaces Gly263 in all SF1A and SF1B relatives, including Pro451 AtSLAC1. This water-displacing change is isostructural26. Centrally located prolines also prevail in TM3, TM5 and TM7 across the SF1 proteins. The outer helices are longer and straighter, but more inclined. The only outer-helix proline kink of HiTehA is in TM6 at the trimer three-fold axis. Two of the gene-identifying mutations in Arabadopsis SLAC1 are selected point mutations, others are disruptive transfer DNA (T-DNA) insertions4. In the AtSLAC1 homology model, the slac1-2 (G194D) mutation3 points into the pore from TM1 and can be accommodated structurally whereas the slac1-1 (S456F) mutation4 points away from the pore six residues after pore-blocking Phe450 on TM9 and would be expected to disruptive. Ser456 interacts with outer helix TM10 in the homology model, and the phenyl bulk from S456F would not fit. Position 456 has alanine in HiTehA, 66% of all 204 SF1 homologs and another 27% have threonine or serine (as in all SLAC1 channels); phenylalanine never occurs among all Chen et al. Page 4 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 814 SLAC superfamily members. Position 194 has glycine in 58% of SF1 homologs and alanine in another 22%. Asp194, which never occurs naturally, would block the pore and be expected to repel anions. Mutational tests of channel function Mutational studies corroborate the hypothesis that a TehA-based SLAC1 model is appropriate. First, as discussed above, the slac1-1 mutant (S456F) is expected to be structurally disruptive, and indeed it is inactive in guard cells4, and the slac1-2 mutation (G194D) is expected to block the pore, and we show below that this variant is also inactive. We have also shown that the introduction of SLAC1-conserved proline residues into HiTehA (A208P/G263P) is accommodated isomorphously26. Moreover, as shown below, channel conductance properties of several mutants are similar for AtSLAC1 and HiTehA. To examine characteristics of the SLAC1 channel in light of the structural model, we performed electrophysiological tests of membrane currents from voltage-clamped Xenopus oocytes following injection of wild-type or mutant AtSLAC1 or HiTehA cRNAs. We observed modest-sized chloride currents with wild-type AtSLAC1 cRNA, as found previously6,7, but did not detect any chloride current following injection of wild-type HiTehA cRNA. We found that SLAC1 Cl− conductance was enhanced when the OST1 kinase cRNA was co-injected with SLAC1 cRNA, but only to the levels found by Lee et al.6 and not to the much higher levels found by Geiger et al.7 with OST1 physically connected to SLAC1 by split YFP linkage. In keeping with structural evidence that Phe262 blocks the HiTehA pore, removal of the phenyl group in HiTehA F262A or in the homologous AtSLAC1 F450A resulted in very large Cl− currents relative to wild-type levels, and the SLAC1 currents were now less enhanced by presence of OST1 (Fig. 4a). The tempting interpretation of an opened gate will require validation with appropriately analyzed single- channel recordings27. In keeping with the slac1-2 phenotype3, neither the functionally impaired AtSLAC1 G194D nor its HiTehA G15D homolog showed any substantial conductance; moreover, further consistent with pore blockage by Asp194 in slac1-2, the large conductances of HiTehA F262A and AtSLAC1 F450A were stopped in the double mutants AtSLAC1 G194D/F450A and HiTehA G15D/F262A (Fig. 4a). Here again, the effects in SLAC1 were independent of OST1. We also tested conductance characteristics for a series of AtSLAC1 F450X substitution mutants – F450A,G,T,V,L– and for the corresponding HiTehA F262X series – F262A,G,T,V,L (Figs. 4b,S5, Table S5). Findings from the two series are roughly parallel; in particular, the alanine and glycine substitutions lead to large currents for both and in comparison to the others. There are distinctions, of course, including generally higher conductances for AtSLAC1 over HiTehA and less conductance of F262T TehA compared to F450T SLAC1. It is also noteworthy that OST1 activation is very muted for the F450A,G,L mutants, which is consistent with SLAC1 gating at Phe450. Crystal structures were also determined for several of the HiTehA mutant variants (Table S4). The structures of F262A (1.15Å), F262V (1.60Å), F262L (1.65Å) and G15D (1.50Å) are all essentially isomorphous with the wild-type TehA structure with changes localized Chen et al. Page 5 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript primarily at the sites of mutation; the same is true for the double mutations of F262A/G15D, F262G/G15D and A208P/G263P26. The F262A structure has a wide open pore (Fig. 5a) with a relatively uniform pore diameter of ~5Å through ~30Å across the membrane (Fig. S3) whereas G15D has a doubly occluded pore (Fig. 5b). The pores of other mutant variants are consistent with the sizes of constrictive residues and with the observed conductances. Gating and activation The crystal structures of TehA and its mutant variants when taken together with the functional studies in Xenopus oocytes point to a crucial role for Phe450 in gating of the SLAC1 anion channel. Conservation of this residue across the SLAC1 family implies functional importance. The occlusion of the pore by the presence of F262 in the structure of wild-type TehA and the openness of the pore upon its substitution by alanine in the structure of the F262A mutant provides physical evidence for a gating role of this residue. This interpretation is supported by the correlated conductance characteristics from variants of the AtSLAC1 and HiTehA channels (Figs. 4b,S5). While these observations may suffice for placing the gate within the channel pore, they do not by themselves suggest a mechanism for gating in response to physiological stimuli. Some insight does come from conformational details defined at high resolution. One important structural clue is that the side chain of Phe262 is in a high energy conformation in the HiTehA structure, with χ1/χ2 at −160°/-4°. Although χ1 is in a preferred trans conformation, the phenyl ring is restricted by contacts with Val210 and Leu18 to a χ2 value near 0° rather than near to the preferred 90° orientation (Fig. 5c). Further evidence that Phe262 is restrained from local equilibrium comes from shifts observed in crystal structures of the F262A, F262V, F262L and F262G/G15D variants, which all show consistent backbone movements that displace Cβ(262) by 0.47-0.78Å (Fig. 5d). By contrast, Leu262 in F262L is in a preferred trans/gauche+ conformation at χ1/χ2 = 177°/63° as is Val262 in F262V at χ1 = −176°. What might control activation of HiTehA is unclear, but for AtSLAC1 activation is by OST1 phosphorylation6,7. The molecular consequences of OST1 phosphorylation of SLAC1 remain unknown, but it is plausible that associated shifts in pore- helix orientations would unlatch Phe450 in SLAC1 from a TehA-like restrained orientation. By analogy with Leu262 in F262L, we expect a preferred rotameric state for Leu450 in AtSLAC1 F450L. Thus, the lack of appreciable OST1 activation of conductance in AtSLAC1 F450L (Fig. 5e) might be explained by the lack of a restraining latch, whereby the channel remains closed despite OST1 activation. Puzzles certainly remain since OST1 does substantially activate AtSLAC1 F450T or F450V, which like HiTehA F262V should also be unrestrained; presumably activating adjustments widen the pore enough for ion permeation past threonine and valine but not leucine. Phosphorylation sites have been discovered in the N- and C-terminal tails of AtSLAC16,7,28 (179 and 51 residues long, respectively), but these alone cannot explain OST1 activation of SLAC1. First SLAH1, which fully complements the slac1-1 mutation, does not have these cytoplasmic tails. Secondly, although OST1 phosphorylation of Ser120 in the N-terminal tail is necessary for SLAC1 activation, it is not sufficient7. Thus, we surmise that direct phosphorylation of the SLAC1 TM domain must be critical, and SLAC1 has four conserved Chen et al. Page 6 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Ser/Thr candidates in its cytoplasmic loops. Moreover, SLAC1 proteins have proline- mediated kinks at the putative Phe450 gate in helix TM9, and also in adjacent helix TM7; these features may play a role in phosphorylation-driven unlatching of the Phe450 gate in SLAC1. Ion selectivity and discrimination Our studies of SLAC1 channel relative ion permeabilities, based on measurements of current reversal potential, are consistent with earlier work demonstrating that AtSLAC1 conducts anions but not cations and is selective among anions, with greater permeability for nitrate than for chloride (as in Vicia faba guard cell protoplasts27) and much reduced permeability for malate, bicarbonate or sulfate6,7. We also find that SLAC1 has little permeability for sulfite. Additionally, we find that wild-type SLAC1, F450A and F450T all have similar relative permeabilities to chloride, sulfite and malate, despite having widely different conductance levels, but the gating mutants do show small but significant decreases in nitrate permeability (Fig. 4c, Table S6). The relative insensitivity of anion permeability to gating residue changes suggests that selectivity for these anions may occur away from the central constriction at the channel gate. To some extent, ionic discrimination must depend on pore geometry; thus, an organic anion such as malate may be simply too large to pass through the 5-Å wide pore. Although the SLAC1 pore is lined largely with hydrophobic sidechains (Fig. 3c), it is sprinkled with hydroxyl groups from serine and threonine residues (16%); their electropositive hydrogen atoms may facilitate conductance. Most strikingly, the electrostatic potential within the AtSLAC1 pore is electropositive throughout (Fig. 3a). This polarization, promoted by charges on extra-membranous loops, no doubt contributes significantly in discrimination against cations. The relative anion permeability sequence of SLAC1 determined by us and others, I− > NO3− > Br− > Cl− 6,7, corresponds to selectivity sequence 1 compiled by Wright and Diamond29 for a range of anion-selective proteins. This sequence correlates inversely with the hydration energies of monovalent anions – anions with a lower hydration energy have a greater channel permeability. It is thought to be generated in proteins with weak, low field-strength, anion binding sites, where selectivity is largely determined by the energetic cost of anion dehydration. These selectivity results are thus consistent with the SLAC1 structure, where the pore lacks any obvious anion binding site. Distinctiveness of the SLAC1 channel SLAC1 anion channels are entirely novel in structure and, apparently, in the mechanism for ion conductance. The best characterized of anion channels belong to the CLC family of Cl− channels and transporters30-32. CLC channels have an altogether different architecture from the SLAC1 channel, and the mechanism for selectivity is also very different. Bacterial CLC transporters bind halide ions at three sites in a highly constricted pore30. By contrast, the SLAC1 pore has a relatively uniform diameter across the membrane, except where closed by the gating phenyl group, and we do not find discrete ion binding sites. CLC selectivity is governed by specific residues surrounding these binding sites30,32. The anion selectivity Chen et al. Page 7 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript sequence for the CLC channels of Cl−>Br−>NO −3>I−, opposite of that in SLAC1, is consistent with the high field-strength anion binding sites in CLC channels29. Interestingly, as for AtSLAC1, an Arabidopsis CLCa channel also preferentially transports nitrate ions33, and an E. coli CLC channel is converted to preference of nitrate when a generally conserved serine at the central site is substituted with proline as in AtCLCa32. SLAC1 also differs radically from other structurally characterized anion channels and transporters. These include the VDAC1 voltage-gated anion channel from mitochondrial outer membranes, which has a porin-like beta-barrel structure34-36, and a light-driven halorhodopsin chloride pump, which has a transmembrane conductance pathway similar to that of the proton-pumping pore of bacteriorhodopsin37. Although its channel structure is still only known by homology to other ABC transporters, CFTR is another obviously distinct chloride channel38. Cys-loop receptors also include anion channels39, and these are similar to SLAC1 in having five-helix pores25, but here selectivity is governed by charged groups at the entrance to the pore, which distinguish the anion-selective GABAA and glycine receptors from the cation-selective acetylcholine and serotonin 5HT3 receptors39. Finally, the recently identified TMEM16A gene for calcium-activated chloride channels40-42 appears to encode an 8-TM protein that is again distinct from SLAC1. Stomatal guard cells show both rapidly activated (R-type) and slow (S-type) anion channel activity43. Although slac1 guard cells have very defective S-type activity, their R-type currents are normal4. Guard cell protoplasts from the slac1-2 mutant abnormally accumulate Cl−, K+, malate and fumarate3, whereas SLAC1 shows negligible malate conductance7. As for SLAC1-associated K+ movements, other channels or transporters must be responsible for SLAC1-associated malate movements. Recent studies indicate that AtALMT12, an aluminium-activated malate transporter (ALMT) family member, is a malate-dependent R- type anion channel44 needed for stomatal closure45. Conclusions We find that many functional properties of the plant SLAC1 anion channel are explained well by the structure of an uncharacterized bacterial TehA protein that has been associated with tellurite resistance. SLAC1 and TehA belong to distinct subfamilies within one branch of a larger SLAC1 superfamily, but AtSLAC1 and HiTehA are sufficiently similar (19% sequence identity) that the SLAC1 homology model is predictive for function, including a verified placement of the identifying slac1-2 mutation G194D and a phenylalanine gate. One remaining puzzle concerns the structural change that activating phosphorylation elicits in SLAC1, and another puzzle is the biochemical role of the TehA homologs in bacteria. In a companion paper26, we examine functional and structural properties of TehA in bacteria, showing that it is anion channel, although actually not conferring tellurite resistance, and identifying a mutant variant with properties suggestive of an activated state. Thus, SLAC1 and TehA likely represent a large family of selective anion channels controlled by environmental stimuli. Chen et al. Page 8 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript METHODS Selection of target sequences TehA from E. coli (EcTehA) was a centrally selected sequence for expansion as a NYCOMPS targeted family. Target selection criteria utilized at NYCOMPS are described in details elsewhere46. Briefly, the EcTehA sequence was run against a dataset of about 40,000 predicted alpha helical integral membrane protein sequences from prokaryotic genomes (NYCOMPS98 dataset46) using PSI-BLAST50. Sequences that matched EcTehA with an E- value lower than 10−3 in an alignment extending over at least 50% of both predicted TM regions and passing our post-seed-expansion filtering criteria46 were passed to the protein production pipeline. Protein expression screening Full-length homologs from the following 38 species, including 2 sequences each from 5 of these, were amplified from genomic DNA by PCR: Thermoplasma acidophilum, Lactococcus lactis subsp. lactis Il1403, Streptococcus pyogenes M1 GAS, Streptococcus pneumoniae TIGR4, Haemophilus influenzae Rd KW20, Methanocaldococcus jannaschii DSM 2661 (2), Escherichia coli K12, Salmonella typhimurium LT2, Clostridium perfringens ATCC 13124, Xanthomonas campestris pv. campestris str. ATCC 33913, Streptococcus agalactiae 2603V/R, Shewanella oneidensis MR-1, Streptococcus mutans UA159, Archaeoglobus fulgidus DSM 4304, Vibrio parahaemolyticus RIMD 2210633 (2), Pseudomonas syringae pv. tomato str. DC3000 (2), Enterococcus faecalis V583, Pyrococcus horikoshii OT3, Bordetella bronchiseptica RB50, Streptomyces coelicolor A3, Corynebacterium glutamicum ATCC 13032, Picrophilus torridus DSM 9790, Acinetobacter sp. ADP1, Vibrio fischeri ES114, Pseudomonas fluorescens Pf-5, Sulfolobus acidocaldarius DSM 639, Colwellia psychrerythraea 34H, Neisseria meningitidis MC58, Rhodobacter sphaeroides 2.4.1, Vibrio cholerae O1 biovar eltor str. N16961 (2), Marinobacter aquaeolei VT8, Acinetobacter baumannii ATCC 17978, Klebsiella pneumoniae subsp. pneumoniae MGH 78578, Streptomyces avermitili MA-4680, Bordetella parapertussis 12822 (2), Streptococcus thermophilus LMG 18311, Salmonella enterica subsp. enterica serovar Paratyphi A str. ATCC 9150, and Anaeromyxobacter dehalogenans 2CP-C. Selected cDNAs were cloned into a modified pET vector (Novagen, Inc) that fuses a FLAG and deca-histidine tag at the C-terminus, which are cleavable by TEV protease. Proteins were expressed in E. coli BL21(DE3) plysS by a high throughput format (0.6ml in a deep well block) and purified after lysis by sonication using metal affinity purification in a buffer containing N-Dodecyl-β-D-Maltopyranoside. Samples were passed over an analytical size exclusion column in 12 different detergent-containing mobile phases, which included N- dodecyl-β-D-maltopyranoside (DDM), N-decyl-β-D-altopyranoside (DM), N-nonyl-β-D- altopyranoside (NM), N-octyl-β-D-altopyranoside (OM), N-octyl-β-D-Glucopyranoside (OG), N-nonyl-β-D-Glucopyranoside (NG) and lauryl dimethyl amine oxide (LDAO). Multi-angle light scattering with refractive index detection was used to analyze the oligomeric state51. The E. coli and H. influenzae proteins were judged to be monodisperse and stable and were passed to scale up. Chen et al. Page 9 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Scaled-up production and purification For scale up, in brief, transformed BL21 plysS cells were grown at 37°C in 2× YT media to OD 0.6-0.8 after being inoculated with 1% of the overnight culture. The culture was induced with 0.4mM IPTG and continued to grow at 37 °C for another 4 hours. The cells were harvested by centrifugation and stored at −80°C before use. Selenomethionyl (SeMet) TehA was expressed in a similar way, but using containing SeMet in place of methionine in defined minimal media. Cells were resuspended in a buffer containing 50 mM Tris-HCl (pH 8.0) and 200 mM NaCl and lysed using a French Press with two passes at 15-20,000 psi. Cell debris were removed by centrifugation at 10,000g for 20min, and the membrane fraction was isolated from that supernatant by ultra-centrifugation at 150,000g for 1 hr. The membrane fraction was homogenized in a solubilization buffer containing 50 mM Tris (pH 8.0) and 200 mM NaCl, and incubated with final concentration of 1% (w/v) dodecyl-β- D-maltopyranoside (DDM, Anatrace, Inc) for 1 hr at 4°C. The non-dissolved matter was remove by ultracentrifugation at 150,000g for 30 min, and the supernatant was loaded to a 5ml Hitrap Ni2+-NTA affinity column (GE Healthcare), pre-equilibrated with the same solubilization buffer supplemented with 0.05% DDM. After 20 column volume buffer wash, the protein was eluted with 250mM imidazole in the solublization buffer. The Flag and 10- His tags were removed by adding super TEV at 1:100 mass ratio and incubating at 4°C overnight. Tag removal was confirmed by SDS-PAGE, and the resulting sample was concentrated to around 10mg/ml. Preparative size exclusion chromatography was carried out on a Superdex-200 column for further purification, removal of TEV protease and the cleaved tag, and for buffer and detergent exchange. The gel-filtration buffer contained 10 mM Tris (pH 8.0), 200 mM NaCl, 1mM EDTA, 0.5 mM Tris [2-carboxyethyl] phosphine (TCEP), and 2×CMC of detergent. In the case of HiTehA, the protein was well behaved and stable in nearly all tested detergents, and we have purified it from DDM, DM, NM, OM, OG and LDAO. Protein characterization We performed N-terminal amino-acid sequencing of purified HiTehA and EcTehA prior to TEV protease treatment. Results from these analyses proved that true initiating methionine residue is located 14 residues after one annotated as N-terminal. The intervening nucleotide sequence contains a Shine-Delgarno sequence. For cross-linking experiments, purified HiTehA was incubated with 10 mM disuccinimidyl glutarate at room temperate for 30 min, and 100mM Tris-HCl pH 7.5 was added to stop the reaction. The incubated sample when run on a 8-25% gradient SDS-PAGE gel showed a ladder consistent with a trimeric structure. Crystallization and data collection Purified protein was concentrated to ~10mg/ml for initial crystal trials in a Mosquito robot with commercial screens from Hampton research, Emerald Biosystems and Molecular Dimension. We obtained crystals of HiTehA from protein in detergent DDM, DM, NM, OM, OG and LDAO, but only those from LDAO and OG gave diffraction to beyond 4Å spacings. Crystals that proved useful were all grown at 4°C using the sitting-drop vapor Chen et al. Page 10 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript diffusion method. After extensive optimization we reached conditions supporting very high resolution. Wild-type HiTehA and most of its variants crystallized from 1mM ZnSO4, 50mM Hepes-Na pH 7.8 and 28% PEG600. HiTehA F262A crystallized from 200mM Li2SO4, 100mM glycine pH9.3 and 33% PEG400). Addition of 10mM spermidine as an additive helped to obtain slightly better diffracting crystals. Cryoprotection was achieved by adding 5% ethylene glycol or PEG400 to the crystallization solution. Structure determination and refinement Native and SeMet SAD data set were collected at NSLS beamline X4A and processed using the software HKL200052. Crystals of HiTehA grew in space group R3 with a=b=c = 85.0Å and α=β=γ = 93.5°. All subsequent manipulations were done in the hexagonal setting of this space group with a=b= 96.0Å and c = 136.7Å. The asymmetric unit contains one TehA protomer and 65% solvent volume. The structure was determined at 2.0Å resolution from single-wavelength anomalous diffraction (SAD) using selenomethionine-substituted protein crystals. Assessment of data quality for phasing, location of heavy atom sites and initial phases were calculated using the HKL2MAP interface to SHELX programs53. All the secondary structure elements were clearly visible in the experimental electron density map. Automatic model building was done in Arp/wArp54 and completed manually in the program COOT55. The model was refined against native data at 1.20Å resolution using the program Refmac5.5 in CCP456, with anisotropic B factor restrained refinement applied. Subsequent structural analyses of mutant variants were refined as isomorphous structures. Site-directed mutagenesis Site-directed mutants were constructed using the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA 92037) and expressed from pET vectors in E. coli BL21(DE3) plysS cells as for the wild-type protein. Electrophysiology All constructs were cloned into plasmid pGHME2, linearized, and transcribed into cRNA using T7 polymerase (mMessage mMachine, Ambion). Oocytes were injected with 50 nl of cRNA solution each, at a constant concentration of 0.5 mg/ml for HiTehA, or 0.5 mg/ml for AtSLAC1 constructs, with or without 0.5 mg/ml of AtOST1 constructs and recorded in voltage-clamp experiments 2 days after injection. For mixed expression experiments, 25nl of cRNA solution was injected for each AtSLAC1 component. Two-electrode voltage clamp recordings were performed to measure HiTehA or AtSLAC1 currents as described6,7. The pipette solutions contained 3 M KCl. For voltage-clamp current recordings, the bath solution contained 1 mM MgCl2, 1 mM CaCl2, 10 mM Mes/Tris (pH 5.6) and 30 mM CsCl; For anion selectivity measurements, the bath solutions contained 50 mM Cl−, NO −3, malate, or 30 mM SO 2-3 sodium salts, 45 mM Na-gluconate, 1 mM Ca-gluconate2, 1 mM Mg- gluconate2, 1 mM K-gluconate, as well as 10 mM Tris/Mes (pH 5.6). Osmolality was adjusted with D-mannitol to 220 mOsmol/kg. The bath electrode was a 3M KCl agar bridge. The testing potentials were +50 to −110 or −130 mV in 20 mV decrements with 7.5 s duration. The holding and return potentials were 0 mV with a 1.45 s or 2 s, respectively. I-V Chen et al. Page 11 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript relations of HiTehA or AtSLAC1 were generated from the current reached after 0.5 s in testing potentials. The Goldman-Hodgkin-Katz equation was applied to estimate permeability ratios for monovalent ions as described6. For divalent anions, the permeability ratios were derived according to Fatt and Ginsborg57. Bioinformatic analysis of SLAC-related proteins Sequences related to SLAC1 were analyzed comprehensively by PSI-BLAST50. Searches at E<10−3 starting from five disparate homologs each identified a common pool of over 900 proteins, which when pooled were used for sub-classification into families and subfamilies. Details of these analyses are reported in footnotes to Table S1. Molecular figures were produced in PyMOL58. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We thank Filippo Mancia, Larry Shapiro and Ming Zhou for helpful discussions; Patricia Rodriguez and Arianne Morrison for help with cloning and expression; Jonah Cheung, Mark Collins, Chris Min, Sheng Ye and Zhen Zhang for advice in protein chemistry and crystallography; John Schwanof and Randy Abramowitz for help with synchrotron experiments; John Riley for help with Xenopus oocyte injections; and Min Su for advice on the PSI- BLAST analysis of SLAC relatives. This project was supported in part by an award to the New York Consortium on Membrane Protein Structure (NYCOMPS) from the NIGMS Protein Structure Initiative. Beamline X4A at the National Synchrotron Light Source (NSLS), a DOE facility at Brookhaven National Laboratory, is supported by the New York Structural Biology Center. References 1. Hetherington AM, Woodward FI. The role of stomata in sensing and driving environmental change. Nature. 2003; 424:901–908. [PubMed: 12931178] 2. Sirichandra C, Wasilewska A, Vlad F, Valon C, Leung J. The guard cell as a single-cell model towards understanding drought tolerance and abscisic acid action. J. Exp. Bot. 2009; 60:1439–1463. [PubMed: 19181866] 3. Negi J, et al. CO2 regulator SLAC1 and its homologues are essential for anion homeostatsis in plant cells. Nature. 2008; 452:483–486. [PubMed: 18305482] 4. Vahisalu T, et al. SLAC1 is required for plant guard cell S-type anion channel function in stomatal signalling. Nature. 2008; 452:487–491. [PubMed: 18305484] 5. Saji S, et al. Disruption of a gene encoding C4-dicarboxylate transporter-like protein increases ozone sensitivity through deregulation of the stomatal response in Arabidopsis thaliana. Plant Cell Physiol. 2008; 49:2–10. [PubMed: 18084014] 6. Lee SC, Lan W, Buchanan BB, Luan S. A protein kinase-phophatase pair interacts with an ion channel to regulate ABA signaling in plant guard cells. Proc. Natl. Acad. Sci. USA. 2009; 106:21419–21424. [PubMed: 19955427] 7. Geiger D, et al. Activity of guard cell anion channel SLAC1 is controlled by drought-stress signaling kinase-phosphatase pair. Proc. Natl. Acad. Sci. USA. 2009; 106:21425–21430. [PubMed: 19955405] 8. Schroeder JI, Hagiwara S. Cytosolic calcium regulates ion channels in the plasma membrane of Vicia faba guard cells. Nature. 1989; 338:427–430. 9. Mustilli A, Merlot S, Vavasseur A, Fenzi F, Giraudat J. Arabidopsis OST1 protein kinase mediates the regulation of stomatal aperture by abscisic acid and acts upstream of reactive oxygen species production. Plant Cell. 2002; 14:3089–3099. [PubMed: 12468729] Chen et al. Page 12 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 10. Leung J, et al. Arabidopsis ABA response gene ABI1: features of a calcium-modulated protein phosphatase. Science. 1994; 264:1448–1452. [PubMed: 7910981] 11. Meyer K, Leube MP, Grill E. A protein phosphatase 2C involved in ABA signal transduction in Arabidopsis thaliana. Science. 1994; 264:1452–1455. [PubMed: 8197457] 12. Ma Y, et al. Regulators of PP2C phosphatase activity function as abscisic acid sensors. Science. 2009; 324:1064–1068. [PubMed: 19407143] 13. Park S, et al. Abscisic acid inhibits type 2C protein phosphatases via the PYR/PYL family of START proteins. Science. 2009; 324:1068–1071. [PubMed: 19407142] 14. Melcher K, et al. A gate-latch-lock mechanism for hormone signalling by abscisic acid receptors. Nature. 2009; 462:602–608. [PubMed: 19898420] 15. Miyazono K, et al. Structural basis of abscisic acid signalling. Nature. 2009; 462:609–614. [PubMed: 19855379] 16. Fujii H, et al. In vitro reconstitution of an abscisic acid signalling pathway. Nature. 2009; 462:660– 664. [PubMed: 19924127] 17. Yin P, et al. Structural insights into the mechanism of abscisic acid signaling by PYL proteins. Nat. Struct. Mol. Biol. 2009; 16:1230–1236. [PubMed: 19893533] 18. Nishimura N, et al. PYR/PYL/RCAR family members are major in-vivo ABI1 protein phosphatase 2C-interacting proteins in Arabidopsis. Plant J. 2010; 61:290–299. [PubMed: 19874541] 19. Grobler J, Bauer F, Subden RE, van Vuuren HJ. The MAE1 gene of Schizosaccharomyces pombe encodes a permease for malate and other C4 dicarboxylic acids. Yeast. 1995; 11:1485–1491. [PubMed: 8750236] 20. Park H, Bakalinsky AT. SSU1 mediates sulphite efflux in Saccharomyces cerevisiae. Yeast. 2000; 16:881–888. [PubMed: 10870099] 21. Léchenne B, et al. Sulphite efflux pumps in Aspergillus fumigatus and dermatophytes. Microbiology. 2007; 153:905–913. [PubMed: 17322211] 22. Walter EG, Weiner JH, Taylor DE. Nucleotide sequence and overexpression of the tellurite- resistance determinant from the IncHII plasmid pHH1508a. Gene. 1991; 101:1–7. [PubMed: 2060788] 23. Taylor DE, Hou Y, Turner RJ, Weiner JH. Location of a potassium tellurite resistance operon (tehA tehB) within the terminus of Escherichia coli K-12. J Bacteriol. 1994; 176:2740–2742. [PubMed: 8169225] 24. Daley DO, et al. Global topology analysis of the Escherichia coli inner membrane proteome. Science. 2005; 308:1321–1323. [PubMed: 15919996] 25. Unwin N. Refined structure of the nicotinic acetylcholine receptor at 4Å resolution. J. Mol. Biol. 2005; 346:967–989. [PubMed: 15701510] 26. Chen Y, Hu L, Siegelbaum SA, Hendrickson WA. Structure-based analysis of anion channel TehA and methyltransferase TehB implicated in bacterial tellurite resistance. Submitted. 27. Schmidt C, Schroeder JI. Anion Selectivity of Slow Anion Channels in the Plasma Membrane of Guard Cells (Large Nitrate Permeability). Plant Physiol. 1994; 106:383–391. [PubMed: 12232336] 28. Vahisalu T, et al. Ozone-triggered rapid stomatal response involves the production of reactive oxygen species, and is controlled by SLAC1 and OST1. Plant J. 2010; 62:442–53. [PubMed: 20128877] 29. Wright EM, Diamond JM. Anion selectivity in biological systems. Physiol. Rev. 1977; 57:109– 156. [PubMed: 834775] 30. Dutzler R, Campbell EB, MacKinnon R. Gating the selectivity filter in ClC chloride channels. Science. 2003; 300:108–112. [PubMed: 12649487] 31. Accardi A, Miller C. Secondary active transport mediated by a prokaryotic homologue of ClC Cl- channels. Nature. 2004; 427:803–807. [PubMed: 14985752] 32. Picollo A, Malvezzi M, Houtman JC, Accardi A. Basis of substrate binding and conservation of selectivity in the CLC family of channels and transporters. Nat. Struct. Mol. Biol. 2009; 16:294– 301. [PubMed: 19219045] 33. De Angeli A, et al. The nitrate/proton antiporter AtCLCa mediates nitrate accumulation in plant vacuoles. Nature. 2006; 442:939–942. [PubMed: 16878138] Chen et al. Page 13 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 34. Hiller S, Garces RG, Malia TJ, Orekhov VY, Colombini M, Wagner G. Solution structure of the integral human membrane protein VDAC-1 in detergent micelles. Science. 2008; 321:1206–1210. [PubMed: 18755977] 35. Bayrhuber M, et al. Structure of the human voltage-dependent anion channel. Proc. Natl. Acad. Sci. USA. 2008; 105:15370–15375. [PubMed: 18832158] 36. Ujwal R, et al. The crystal structure of mouse VDAC1 at 2.3 Å resolution reveals mechanistic insights into metabolite gating. Proc. Natl. Acad. Sci. USA. 2008; 105:17742–17747. [PubMed: 18988731] 37. Kouyama T, et al. Crystal structure of the light-driven chloride pump halorhodopsin from Natronomonas pharaonis. J. Mol. Biol. 2010; 396:564–579. [PubMed: 19961859] 38. Gadsby DC, Vergani P, Csanády L. The ABC protein turned chloride channel whose failure causes cystic fibrosis. Nature. 2006; 440:477–483. [PubMed: 16554808] 39. Miller PS, Smart TG. Binding, activation and modulation of Cys-loop receptors. Trends Pharmacol. Sci. 2010; 31:161–174. [PubMed: 20096941] 40. Yang YD, et al. TMEM16A confers receptor-activated calcium-dependent chloride conductance. Nature. 2008; 455:1210–5. [PubMed: 18724360] 41. Caputo A, et al. TMEM16A, a membrane protein associated with calcium-dependent chloride channel activity. Science. 2008; 322:590–594. [PubMed: 18772398] 42. Schroeder BC, Cheng T, Jan YN, Jan LY. Expression cloning of TMEM16A as a calcium- activated chloride channel subunit. Cell. 2008; 134:1019–1029. [PubMed: 18805094] 43. Schroeder JI, Keller BU. Two types of anion channel currents in guard cells with distinct voltage regulation. Proc. Natl. Acad. Sci. USA. 1992; 89:5025–5029. [PubMed: 1375754] 44. Meyer S, et al. AtALMT12 represents an R-type anion channel required for stomatal movement in Arabidopsis guard cells. Plant J. Jul 12.2010 Epub ahead of print. 45. Sasaki T, et al. Closing plant stomata requires a homolog of an aluminum-activated malate transporter. Plant Cell Physiol. 2010; 51:354–65. [PubMed: 20154005] 46. Punta M, et al. Structural genomics target selection for the New York Consortium on Membrane Protein Structure. J. Struct. Funct. Genomics. 2009; 4:255–268. [PubMed: 19859826] 47. Papadopoulos JS, Agarwala R. COBALT: constraint-based alignment tool for multiple protein sequences. Bioinformatics. 2007; 23:1073–1079. [PubMed: 17332019] 48. Landau M, et al. ConSurf 2005: the projection of evolutionary conservation scores of residues on protein structures. Nucleic Acids Res. 2005; 33:W299–W302. [PubMed: 15980475] 49. Rocchia W, et al. Rapid grid-based construction of the molecular surface for both molecules and geometric objects: applications to the finite difference Poisson-Boltzmann method. J. Comp. Chem. 2002; 23:128–137. [PubMed: 11913378] 50. Altschul SF, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997; 25:3389–3402. [PubMed: 9254694] 51. Kendrick BS, Kerwin BA, Chang BS, Philo JS. Online size-exclusion high-performance liquid chromatography light scattering and differential refractometry methods to determine degree of polymer conjugation to proteins and protein-protein or protein-ligand association states. Anal. Biochem. 2001; 299:136–146. [PubMed: 11730335] 52. Otwinowski Z, Minor W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 1997; 276:307–326. 53. Pape P, Schneider TR. HKL2MAP: a graphical user interface for phasing with SHELX programs. J. Appl. Cryst. 2004; 37:843–844. 54. Perrakis A, Morris R, Lamzin VS. Automated protein model building combined with iterative structure refinement. Nat. Struct. Biol. 1999; 6:458–463. [PubMed: 10331874] 55. Emsley P, Cowtan K. COOT: model-building tools for molecular graphics. Acta Crystallogr. D. 2004; 60:2126–2132. [PubMed: 15572765] 56. CCP4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D. 1994; 50:760– 763. [PubMed: 15299374] 57. Fatt P, Ginsborg BL. The ionic requirements for the production of action potentials in crustacean muscle fibers. J. Physiol. 1958; 142:516–543. [PubMed: 13576452] Chen et al. Page 14 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 58. DeLano, WL. The PyMOL molecular graphics system. DeLano Scientific; San Carlos, CA, USA: 2002. http://www.pymol.org Chen et al. Page 15 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 1. Sequence analysis for the SLAC1 superfamily a, Family tree. The presentation was computed by the program COBALT47 from representative subfamily sequences (Table S1), including Arabidopsis thaliana SLAC1 for SF1Ai, Haemophilus influenzae TehA for SF1Ci, Escherichia coli TehA for SF1Cii, Vibrio parahaemolyticus for SF2A, Staphylococcus aureus for SF2B, Aspergillus fumigatus Ssu1 for SF3A, and Schizosaccharomyces pombe Mae1 for SF3B. b, Structure-based sequence alignment of TehA from Haemophilus influenzae (HiTehA) and SLAC1 from Arabidopsis thaliana (AtSLAC1). The TehA structure has been used to restrict sequence gaps to inter- helical segments. Superior coils define extents of the HiTehA helical segments; red letters mark residue identities; red boxes are drawn for residues that are >95% identical within the plant subfamily SF1A for AtSLAC1 or within the TehA subfamily for HiTehA; red diamonds mark HiTehA residues that line the central pore; and the colored inferior bar encodes ConSurf sequence variability48 for the SF1 family of 204 non-redundant proteins. Chen et al. Page 16 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 2. Crystal structure of HiTehA and homology model of AtSLAC1 a, Electron density distribution from the HiTehA crystal structure at 1.2Å resolution. The map has (2Fo-Fc) coefficients based on the superimposed model. Contours are at 2.5σ. b, Ribbon diagram of the HiTehA trimer. Each protomer is colored spectrally from blue at its N-terminus to red at its C-terminus. c, DelPhi49 electrostatic potential at the extracellular surface. Electronegative and electropositive potential are colored in degrees of red and blue saturation, respectively. d, Electrostatic potential49 at the intracellular surface. e, Ribbon diagram of an HiTehA protomer viewed from outside the membrane. The ribbon is colored spectrally as in 2b. f, Ribbon diagram of an HiTehA protomer viewed from within the membrane from below the view of e. g, Surface of a homology model of AtSLAC1, viewed as in f, and colored by sequence variability48. h, Surface of AtSLAC1 as in g, but colored by electrostatic potential49. Chen et al. Page 17 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 3. Putative structure of the SLAC1 conductance pore a, Cross-section through the homology model of AtSLAC1. The model is viewed as in 2i, with the electrostatic potential49 shown on the external surface of the molecular envelope. The side chain of Phe450 is shown as a stick model (red) on the backbone ribbon colored yellow. b, Cross-section as in a, but colored by surface conservation48 as in 2h. c, Pore- lining residues in the SLAC1 homology model. A cylinder model (left), spectrally colored as in 2b, provides a key for viewing the rolled-open structure (left) with pore-lining residues of AtSLAC1 shown on the TModd helices. d, Ribbon diagrams of HiTehA TM9 (left) and TM7 (right) viewed from within the conductance pore. The side chains of Pro207 and Phe262 are shown as well as a kink-stabilizing water HOH25 that is coordinated by the NH of Gly263 and by C=O groups of Gly202 and Ala259. Density contours are shown for the water molecule. Chen et al. Page 18 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 4. Ionic conductance measurements a, Typical microelectrode voltage-clamp current traces from oocytes injected with various channel cRNAs. Left column, oocytes injected with cRNAs encoding wild-type HiTehA channels (WT), or F262A, G15D/F262A or G15D mutants. Right two columns, oocytes injected with cRNAs for WT AtSLAC1, or F450A, G194D/F450A or G194D mutants, with or without co-injection of AtOST1. Dotted lines represent zero current levels. Extracellular solution contained 30mM CsCl. Schematic keys at the left place phenyl and/or aspartyl gates. b, Effects of gating residue mutations. Mean chloride currents, measured at −90 mV, are shown comparing WT HiTehA and its mutant series F262A,G,T,V,L with WT AtSLAC1 and its corresponding series F450A,G,T,V,L, both alone and co-expressed with AtOST1. Full I-V relations are shown in Fig. S5. c, Effect of gating residues on relative AtSLAC1 anion permeabilities. Relative permeabilities (P[X]/P[Cl]) for chloride, nitrate, sulfite and malate of WT, F450A and F450T SLAC1 channels were measured from the change in current reversal potential with Cl− or anion X− as the sole permeant anion in the bath solution (Methods, Table S6). Chen et al. Page 19 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 5. Structural features at the SLAC1 homolog gate a,b, Cross-sections through the conductance pores of HiTehA F262A and HiTehA G15D. The view and presentations are as in 3a, except that helices are colored purple. c, Molecular basis for conformational strain in gating residue Phe262 of HiTehA. Helices TM1 (left), TM9 (center) and TM7 (right) are viewed from within the pore and presented as ribbon diagrams with selected side chains drawn in stick representation. The local low-energy conformation for the phenyl ring (χ2 = 90°) is shown in thin lines with short contacts indicated in dashed lines: Leu18 Cδ2 – Phe262 Cδ1, d = 2.4 Å; Val210 Cγ2 – Phe262 Cε2, d = 2.8 Å. d, Conformational shifts consequent to release of strain in gating residue Phe262. Cα backbone structures of F262A, F262V and F262L HiTehA were first superimposed onto WT HiTehA, and residues 258-266 were drawn in stereo as superimposed. All backbone atoms are shown for peptides 262±1, but only Cα atoms are shown otherwise. The WT backbone and phenyl group are green; other backbone are all magenta; side chains of Ala262, Val262 and Leu262 are cyan, blue and red, respectively; oxygen-directed bonds are red and nitrogen-directed bonds are blue. e, Typical microelectrode voltage-clamp current traces from oocytes injected with wild-type (WT) or F450L AtSLAC1 cRNA. Experimental conditions and displays are as in 4a. Chen et al. Page 20 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript
3M75
Crystal Structure of Plant SLAC1 homolog TehA
Homolog Structure of the SLAC1 Anion Channel for Closing Stomata in Leaves Yuhang Chen1,6, Lei Hu2, Marco Punta6,7, Renato Bruni6, Brandan Hillerich6, Brian Kloss6, Burkhard Rost1,6,1, James Love6, Steven A. Siegelbaum2,3,5, and Wayne A. Hendrickson1,4,5,6 1Department of Biochemistry and Molecular Biophysics, Columbia University, New York, NY 10032, USA 2Department of Neuroscience, Columbia University, New York, NY 10032, USA 3Department of Pharmacology, Columbia University, New York, NY 10032, USA 4Department of Physiology and Cellular Biophysics, Columbia University, New York, NY 10032, USA 5Howard Hughes Medical Institute, Columbia University, New York, NY 10032, USA 6NYCOMPS, New York Structural Biology Center 89 Convent Avenue, New York, NY 10027 USA 7Department of Computer Science and Institute for Advanced Study Technical University of Munich D-85748 Munich, Germany Summary The plant SLAC1 anion channel controls turgor pressure in the aperture-defining guard cells of plant stomata, thereby regulating exchange of water vapor and photosynthetic gases in response to environmental signals such as drought or high levels of carbon dioxide. We determined the crystal structure of a bacterial homolog of SLAC1 at 1.20Å resolution, and we have used structure- inspired mutagenesis to analyze the conductance properties of SLAC1 channels. SLAC1 is a symmetric trimer composed from quasi-symmetric subunits, each having ten transmembrane helices arranged from helical hairpin pairs to form a central five-helix transmembrane pore that is gated by an extremely conserved phenylalanine residue. Conformational features suggest a mechanism for control of gating by kinase activation, and electrostatic features of the pore coupled with electrophysiological characteristics suggest that selectivity among different anions is largely a function of the energetic cost of ion dehydration. Users may view, print, copy, download and text and data- mine the content in such documents, for the purposes of academic research, subject always to the full Conditions of use: http://www.nature.com/authors/editorial_policies/license.html#terms Correspondence and request for materials should be addressed to WAH (wayne@convex.hhmi.columbia.edu).. Supplementary Information is linked to the online version of the paper at www.nature.com/nature. Author Contributions YC, LH, MP, RB, BH, BK and JL performed experiments; YC, LH, BR, SAS, and WAH analyzed data; YC, LH, SAS, and WAH prepared the manuscript. Author Information Atomic coordinates and diffraction data are deposited in the Protein Data Bank with accession codes 3M71, 3M73, 3M74, 3M75, 3M76, and 3M7L. See Table S4 for identifications. Reprints and permissions information is available at www.Nature.com/reprints. HHS Public Access Author manuscript Nature. Author manuscript; available in PMC 2013 January 18. Published in final edited form as: Nature. 2010 October 28; 467(7319): 1074–1080. doi:10.1038/nature09487. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Stomatal pores in the leaves of plants permit the influx of atmospheric carbon dioxide in exchange for transpirational evaporation of water1,2. A pair of kidney-shaped guard cells define each pore aperture, and turgor pressure variation in these cells determines the degree of stomatal pore openness. Depending on diverse environmental factors, the stomata close to prevent H2O losses and open to admit CO2 for photosynthesis. Environmental stimuli that lead to stomatal closure include darkness, high CO2 levels, ozone, low air humidity and drought. The plant hormone abscisic acid (ABA) is critical for signal transduction from these stimuli. Mutational screens in Arabidopsis thaliana for CO2 and ozone sensitivity identified a protein with ten predicted transmembrane (TM) helices, now called slow anion channel 1 (SLAC1), as having a central role in the control of stomatal closure3-5. Recent studies proved that SLAC1 is indeed an anion channel6-7, with characteristics like those of slow anion channels found in guard cells8, and that it is activated by phosphorylation from the OST1 kinase9. OST1 activity is negatively regulated by the ABI1 phosphatase10,11, which is in turn inhibited by the stomatal ABA receptors PYR/RCAR12,13 when in the ternary hormone-receptor-phosphatase complex14-18. Thereby, ABA stimulates SLAC1 channel activity. Resulting Cl− efflux through SLAC1 causes membrane depolarization, which activates outward-rectifying K+ channels, leading to KCl and water efflux to further reduce turgor and cause stomatal closure. SLAC1 expression in Arabidopsis is confined to the guard cells of leaves, but other Arabidopsis tissues do have SLAC1 homologs3, named SLAH1-SLAH4. The identifying mutations slac1-14 and slac1-23 are, respectively, in predicted transmembrane segments 9 (S456F) and 1 (G194D) of a protein that includes substantial N- and C-terminal extensions outside a 10-helix transmembrane (TM) domain. SLAH1, which is absent from leaves and lacks the terminal extensions of SLAC1, fully complements the mutant phenotype in slac1-2 guard cell protoplasts3. SLAC1 and homologs are also present in other plant genomes, including nine in rice (Oryza sativa) and five in grape vines (Vitis vinifera). SLAC1 relatives, some quite remote, also occur in bacteria, archaea and fungi. Known prokaryotic homologs contain only the predicted transmembrane domain of SLAC1, but some fungal homologs do have N- and C-terminal extensions. One homolog, Mae1 from the yeast Schizosaccharomyces pombe, functions as a malate uptake transporter19; another, Ssu1 from Saccharomyces cerevisiae and other fungi including Aspergillus fumigatus, is characterized as a sulfite efflux pump20,21; and TehA from Escherichia coli is identified as a tellurite resistance protein by virtue of its association in the tehA/tehB operon22,23. Despite a lack of further biochemical characterization, many homologs are annotated as tellurite resistance/ dicarboxylate transporter (TDT) proteins. We have undertaken structural and functional characterizations of the SLAC1 anion channel. We first solved an atomic-resolution crystal structure of the TehA homolog from Haemophilus influenzae, and we then developed a homology model for Arabidopsis SLAC1. This model allowed us to conduct mutagenesis for functional testing of structure-inspired hypotheses on gating and selectivity. We expressed Haemophilus TehA and Arabidopsis SLAC1 in Xenopus oocytes to characterize channel properties of these proteins and mutant variants. We also determined crystal structures for several mutant variants, including the homolog of slac1-2. Chen et al. Page 2 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Structure of SLAC1 bacterial homolog TehA We performed a bioinformatic analysis of SLAC1-related proteins, first clustering nearly 900 non-redundant sequences into a superfamily at the PSI-BLAST level E≤10−3, then into three distinct families at an initial threshold of E≤10−30, and finally into subfamilies at a typical initial threshold of E≤10−55. Since previous annotation is not well founded in experiment and SLAC1 is now the best characterized member, we adopt a nomenclature defining a SLAC superfamily divided into families identified as SF1-SF3 and subfamilies SF1A, SF1B etc. Family SF1 comprises the plant SLAC proteins and close bacterial homologs, family SF2 comprises a distinct set of bacterial proteins often annotated as exfoliative toxins, and family SF3 comprises the fungal Mae1 and Ssu1 proteins and their archaeal or bacterial homologs, respectively. SLAC family SF1 has three large subfamilies: the plant SLAC and SLAH proteins are in subfamily SF1A, closest bacterial homologs are in SF1B, and the TehA homologs are in SF1C (Fig. 1a). The other families also divide into subfamilies as detailed in Table S1, and family SF1 is divided into sub-subfamilies (Table S2). Two pertinent SF1 sequences are aligned in Fig. 1b. We used a structural genomics approach to obtain structural information, testing expression and purification for 43 bacterial and archaeal likely homologs, assaying for detergent choice and stability on 8 of these, finding two with appropriate profiles by size exclusion chromatography, and obtaining suitable crystals for one. This protein, TehA from H. influenzae (HiTehA), was found to be trimeric both by size-exclusion multi-angle light scattering (SEC-MALS) measurements and by chemical cross-linking. When solubilized in β-octylglucoside, HiTehA crystallized in space group R3 with a=b=96.01Å and c=136.27Å. Each asymmetric unit contains one subunit and 65% solvent. The structure was solved by selenomethionyl (SeMet) SAD phasing, ultimately at 1.50Å resolution (Table S3, Fig. S1), and then refined at 1.20Å resolution (Fig. 2a) to R/Rfree values of 14.1%/15.7% for a model that includes ordered residues 6-313, 213 water molecules and four detergent molecules (Table S4). The crystal structure has TehA trimers aligned with three-fold axes of the lattice (Fig. 2b). Subunits are tightly associated, burying 8947 Å2 of total surface area within trimer interfaces. The electrostatic potential surface is largely negative on the extracellular surface (Fig. 2c) and largely positive on the cytoplasmic surface (Fig. 2d). The membrane orientation is specified experimentally from GFP tagging of E. coli TehA24. Each TehA protomer has ten TM helices, as predicted; however, the fold is novel. Tandemly repeated helical hairpins are arranged with quasi-five-fold symmetry (Figs. 2e,S2). Extracellular inter-helix loops are short (2-5 residues) whereas intracellular inter-helix connections are longer, including a nine-residue helix H2,3 between TM2 and TM3 (Fig. 2f). An inner pentad of outwardly directed, TModd, helices creates an apparent pore through each protomer perpendicular to the putative membrane plane. TMeven helices from the five hairpins surround the inner pore and make an outer layer. Chen et al. Page 3 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Homology model for plant SLAC1 Arabidopsis SLAC1 (AtSLAC1) is substantially similar to bacterial homologs, notably HiTehA (Fig. 1b). All HiTehA TM helices are fully aligned to predicted SLAC1 TM helices, but there are short inter-helical gaps (1-5 residues) in all five extracellular loops and in two of the intracellular loops. The TM domain of AtSLAC1 (residues 188-504) aligns to HiTehA with 19% sequence identity and with a PSI-BLAST E-value of 3×10−22. For comparison, and in keeping with the family tree (Fig. 1a), the TM domain of AtSLAC1 shares sequence identities of 76% with rice SLAC1, 41% with Arabidopsis SLAH1, 25% with an SL1B homolog from Halorhodospira halophila, 11% with S. cerevisiae Ssu1, and 9% with S. pombe Mae1. A conceptual model with the AtSLAC1 sequence transposed onto the HiTehA helices sufficed to guide most of our mutational tests, but a detailed AtSLAC1 homology model helped to refine our ideas. Surface variability and electrostatic potential are plotted onto the surface of this model (Fig. 2g,2h). The most remarkable feature of the TehA structure and corresponding SLAC1 model is the central pore through each protomer. As for acetylcholine receptors25, the SLAC1 pore is formed by five helices; but the SLAC1 helices come from one protein molecule rather than five. The SLAC1 pore has a relatively uniform diameter of approximately 5Å across nearly five helical turns (Fig. S3), except for a pronounced constriction in the middle of the membrane (Fig. 3a) where the pore is occluded by the side chain of Phe 450 (Phe 262 in HiTehA). This residue is the only absolutely conserved amino acid residue of the SLAC1 family. The pore is lined with highly conserved (86% identity among five SLAC1 orthologs; 32% identity between AtSLAC1 and HiTehA) and generally hydrophobic residues (Figs. 1b, 3b,3c,S4). Despite this hydrophobicity, the electrostatic potential on the pore surface is polarized (Fig. 3a), presumably due to an invaginated shape adjacent to charged residues outside the membrane. The generally electropositive character of the cytoplasmic surface likely contributes to anion efflux. Kinks in the pore helices contribute to formation of a relatively constant pore diameter across the membrane. Four of the five HiTehA inner helices have centrally located proline residues, which necessarily generate kinks, and TM9 is kinked at a backbone-coordinated water molecule (Fig. 3d). Proline replaces Gly263 in all SF1A and SF1B relatives, including Pro451 AtSLAC1. This water-displacing change is isostructural26. Centrally located prolines also prevail in TM3, TM5 and TM7 across the SF1 proteins. The outer helices are longer and straighter, but more inclined. The only outer-helix proline kink of HiTehA is in TM6 at the trimer three-fold axis. Two of the gene-identifying mutations in Arabadopsis SLAC1 are selected point mutations, others are disruptive transfer DNA (T-DNA) insertions4. In the AtSLAC1 homology model, the slac1-2 (G194D) mutation3 points into the pore from TM1 and can be accommodated structurally whereas the slac1-1 (S456F) mutation4 points away from the pore six residues after pore-blocking Phe450 on TM9 and would be expected to disruptive. Ser456 interacts with outer helix TM10 in the homology model, and the phenyl bulk from S456F would not fit. Position 456 has alanine in HiTehA, 66% of all 204 SF1 homologs and another 27% have threonine or serine (as in all SLAC1 channels); phenylalanine never occurs among all Chen et al. Page 4 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 814 SLAC superfamily members. Position 194 has glycine in 58% of SF1 homologs and alanine in another 22%. Asp194, which never occurs naturally, would block the pore and be expected to repel anions. Mutational tests of channel function Mutational studies corroborate the hypothesis that a TehA-based SLAC1 model is appropriate. First, as discussed above, the slac1-1 mutant (S456F) is expected to be structurally disruptive, and indeed it is inactive in guard cells4, and the slac1-2 mutation (G194D) is expected to block the pore, and we show below that this variant is also inactive. We have also shown that the introduction of SLAC1-conserved proline residues into HiTehA (A208P/G263P) is accommodated isomorphously26. Moreover, as shown below, channel conductance properties of several mutants are similar for AtSLAC1 and HiTehA. To examine characteristics of the SLAC1 channel in light of the structural model, we performed electrophysiological tests of membrane currents from voltage-clamped Xenopus oocytes following injection of wild-type or mutant AtSLAC1 or HiTehA cRNAs. We observed modest-sized chloride currents with wild-type AtSLAC1 cRNA, as found previously6,7, but did not detect any chloride current following injection of wild-type HiTehA cRNA. We found that SLAC1 Cl− conductance was enhanced when the OST1 kinase cRNA was co-injected with SLAC1 cRNA, but only to the levels found by Lee et al.6 and not to the much higher levels found by Geiger et al.7 with OST1 physically connected to SLAC1 by split YFP linkage. In keeping with structural evidence that Phe262 blocks the HiTehA pore, removal of the phenyl group in HiTehA F262A or in the homologous AtSLAC1 F450A resulted in very large Cl− currents relative to wild-type levels, and the SLAC1 currents were now less enhanced by presence of OST1 (Fig. 4a). The tempting interpretation of an opened gate will require validation with appropriately analyzed single- channel recordings27. In keeping with the slac1-2 phenotype3, neither the functionally impaired AtSLAC1 G194D nor its HiTehA G15D homolog showed any substantial conductance; moreover, further consistent with pore blockage by Asp194 in slac1-2, the large conductances of HiTehA F262A and AtSLAC1 F450A were stopped in the double mutants AtSLAC1 G194D/F450A and HiTehA G15D/F262A (Fig. 4a). Here again, the effects in SLAC1 were independent of OST1. We also tested conductance characteristics for a series of AtSLAC1 F450X substitution mutants – F450A,G,T,V,L– and for the corresponding HiTehA F262X series – F262A,G,T,V,L (Figs. 4b,S5, Table S5). Findings from the two series are roughly parallel; in particular, the alanine and glycine substitutions lead to large currents for both and in comparison to the others. There are distinctions, of course, including generally higher conductances for AtSLAC1 over HiTehA and less conductance of F262T TehA compared to F450T SLAC1. It is also noteworthy that OST1 activation is very muted for the F450A,G,L mutants, which is consistent with SLAC1 gating at Phe450. Crystal structures were also determined for several of the HiTehA mutant variants (Table S4). The structures of F262A (1.15Å), F262V (1.60Å), F262L (1.65Å) and G15D (1.50Å) are all essentially isomorphous with the wild-type TehA structure with changes localized Chen et al. Page 5 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript primarily at the sites of mutation; the same is true for the double mutations of F262A/G15D, F262G/G15D and A208P/G263P26. The F262A structure has a wide open pore (Fig. 5a) with a relatively uniform pore diameter of ~5Å through ~30Å across the membrane (Fig. S3) whereas G15D has a doubly occluded pore (Fig. 5b). The pores of other mutant variants are consistent with the sizes of constrictive residues and with the observed conductances. Gating and activation The crystal structures of TehA and its mutant variants when taken together with the functional studies in Xenopus oocytes point to a crucial role for Phe450 in gating of the SLAC1 anion channel. Conservation of this residue across the SLAC1 family implies functional importance. The occlusion of the pore by the presence of F262 in the structure of wild-type TehA and the openness of the pore upon its substitution by alanine in the structure of the F262A mutant provides physical evidence for a gating role of this residue. This interpretation is supported by the correlated conductance characteristics from variants of the AtSLAC1 and HiTehA channels (Figs. 4b,S5). While these observations may suffice for placing the gate within the channel pore, they do not by themselves suggest a mechanism for gating in response to physiological stimuli. Some insight does come from conformational details defined at high resolution. One important structural clue is that the side chain of Phe262 is in a high energy conformation in the HiTehA structure, with χ1/χ2 at −160°/-4°. Although χ1 is in a preferred trans conformation, the phenyl ring is restricted by contacts with Val210 and Leu18 to a χ2 value near 0° rather than near to the preferred 90° orientation (Fig. 5c). Further evidence that Phe262 is restrained from local equilibrium comes from shifts observed in crystal structures of the F262A, F262V, F262L and F262G/G15D variants, which all show consistent backbone movements that displace Cβ(262) by 0.47-0.78Å (Fig. 5d). By contrast, Leu262 in F262L is in a preferred trans/gauche+ conformation at χ1/χ2 = 177°/63° as is Val262 in F262V at χ1 = −176°. What might control activation of HiTehA is unclear, but for AtSLAC1 activation is by OST1 phosphorylation6,7. The molecular consequences of OST1 phosphorylation of SLAC1 remain unknown, but it is plausible that associated shifts in pore- helix orientations would unlatch Phe450 in SLAC1 from a TehA-like restrained orientation. By analogy with Leu262 in F262L, we expect a preferred rotameric state for Leu450 in AtSLAC1 F450L. Thus, the lack of appreciable OST1 activation of conductance in AtSLAC1 F450L (Fig. 5e) might be explained by the lack of a restraining latch, whereby the channel remains closed despite OST1 activation. Puzzles certainly remain since OST1 does substantially activate AtSLAC1 F450T or F450V, which like HiTehA F262V should also be unrestrained; presumably activating adjustments widen the pore enough for ion permeation past threonine and valine but not leucine. Phosphorylation sites have been discovered in the N- and C-terminal tails of AtSLAC16,7,28 (179 and 51 residues long, respectively), but these alone cannot explain OST1 activation of SLAC1. First SLAH1, which fully complements the slac1-1 mutation, does not have these cytoplasmic tails. Secondly, although OST1 phosphorylation of Ser120 in the N-terminal tail is necessary for SLAC1 activation, it is not sufficient7. Thus, we surmise that direct phosphorylation of the SLAC1 TM domain must be critical, and SLAC1 has four conserved Chen et al. Page 6 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Ser/Thr candidates in its cytoplasmic loops. Moreover, SLAC1 proteins have proline- mediated kinks at the putative Phe450 gate in helix TM9, and also in adjacent helix TM7; these features may play a role in phosphorylation-driven unlatching of the Phe450 gate in SLAC1. Ion selectivity and discrimination Our studies of SLAC1 channel relative ion permeabilities, based on measurements of current reversal potential, are consistent with earlier work demonstrating that AtSLAC1 conducts anions but not cations and is selective among anions, with greater permeability for nitrate than for chloride (as in Vicia faba guard cell protoplasts27) and much reduced permeability for malate, bicarbonate or sulfate6,7. We also find that SLAC1 has little permeability for sulfite. Additionally, we find that wild-type SLAC1, F450A and F450T all have similar relative permeabilities to chloride, sulfite and malate, despite having widely different conductance levels, but the gating mutants do show small but significant decreases in nitrate permeability (Fig. 4c, Table S6). The relative insensitivity of anion permeability to gating residue changes suggests that selectivity for these anions may occur away from the central constriction at the channel gate. To some extent, ionic discrimination must depend on pore geometry; thus, an organic anion such as malate may be simply too large to pass through the 5-Å wide pore. Although the SLAC1 pore is lined largely with hydrophobic sidechains (Fig. 3c), it is sprinkled with hydroxyl groups from serine and threonine residues (16%); their electropositive hydrogen atoms may facilitate conductance. Most strikingly, the electrostatic potential within the AtSLAC1 pore is electropositive throughout (Fig. 3a). This polarization, promoted by charges on extra-membranous loops, no doubt contributes significantly in discrimination against cations. The relative anion permeability sequence of SLAC1 determined by us and others, I− > NO3− > Br− > Cl− 6,7, corresponds to selectivity sequence 1 compiled by Wright and Diamond29 for a range of anion-selective proteins. This sequence correlates inversely with the hydration energies of monovalent anions – anions with a lower hydration energy have a greater channel permeability. It is thought to be generated in proteins with weak, low field-strength, anion binding sites, where selectivity is largely determined by the energetic cost of anion dehydration. These selectivity results are thus consistent with the SLAC1 structure, where the pore lacks any obvious anion binding site. Distinctiveness of the SLAC1 channel SLAC1 anion channels are entirely novel in structure and, apparently, in the mechanism for ion conductance. The best characterized of anion channels belong to the CLC family of Cl− channels and transporters30-32. CLC channels have an altogether different architecture from the SLAC1 channel, and the mechanism for selectivity is also very different. Bacterial CLC transporters bind halide ions at three sites in a highly constricted pore30. By contrast, the SLAC1 pore has a relatively uniform diameter across the membrane, except where closed by the gating phenyl group, and we do not find discrete ion binding sites. CLC selectivity is governed by specific residues surrounding these binding sites30,32. The anion selectivity Chen et al. Page 7 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript sequence for the CLC channels of Cl−>Br−>NO −3>I−, opposite of that in SLAC1, is consistent with the high field-strength anion binding sites in CLC channels29. Interestingly, as for AtSLAC1, an Arabidopsis CLCa channel also preferentially transports nitrate ions33, and an E. coli CLC channel is converted to preference of nitrate when a generally conserved serine at the central site is substituted with proline as in AtCLCa32. SLAC1 also differs radically from other structurally characterized anion channels and transporters. These include the VDAC1 voltage-gated anion channel from mitochondrial outer membranes, which has a porin-like beta-barrel structure34-36, and a light-driven halorhodopsin chloride pump, which has a transmembrane conductance pathway similar to that of the proton-pumping pore of bacteriorhodopsin37. Although its channel structure is still only known by homology to other ABC transporters, CFTR is another obviously distinct chloride channel38. Cys-loop receptors also include anion channels39, and these are similar to SLAC1 in having five-helix pores25, but here selectivity is governed by charged groups at the entrance to the pore, which distinguish the anion-selective GABAA and glycine receptors from the cation-selective acetylcholine and serotonin 5HT3 receptors39. Finally, the recently identified TMEM16A gene for calcium-activated chloride channels40-42 appears to encode an 8-TM protein that is again distinct from SLAC1. Stomatal guard cells show both rapidly activated (R-type) and slow (S-type) anion channel activity43. Although slac1 guard cells have very defective S-type activity, their R-type currents are normal4. Guard cell protoplasts from the slac1-2 mutant abnormally accumulate Cl−, K+, malate and fumarate3, whereas SLAC1 shows negligible malate conductance7. As for SLAC1-associated K+ movements, other channels or transporters must be responsible for SLAC1-associated malate movements. Recent studies indicate that AtALMT12, an aluminium-activated malate transporter (ALMT) family member, is a malate-dependent R- type anion channel44 needed for stomatal closure45. Conclusions We find that many functional properties of the plant SLAC1 anion channel are explained well by the structure of an uncharacterized bacterial TehA protein that has been associated with tellurite resistance. SLAC1 and TehA belong to distinct subfamilies within one branch of a larger SLAC1 superfamily, but AtSLAC1 and HiTehA are sufficiently similar (19% sequence identity) that the SLAC1 homology model is predictive for function, including a verified placement of the identifying slac1-2 mutation G194D and a phenylalanine gate. One remaining puzzle concerns the structural change that activating phosphorylation elicits in SLAC1, and another puzzle is the biochemical role of the TehA homologs in bacteria. In a companion paper26, we examine functional and structural properties of TehA in bacteria, showing that it is anion channel, although actually not conferring tellurite resistance, and identifying a mutant variant with properties suggestive of an activated state. Thus, SLAC1 and TehA likely represent a large family of selective anion channels controlled by environmental stimuli. Chen et al. Page 8 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript METHODS Selection of target sequences TehA from E. coli (EcTehA) was a centrally selected sequence for expansion as a NYCOMPS targeted family. Target selection criteria utilized at NYCOMPS are described in details elsewhere46. Briefly, the EcTehA sequence was run against a dataset of about 40,000 predicted alpha helical integral membrane protein sequences from prokaryotic genomes (NYCOMPS98 dataset46) using PSI-BLAST50. Sequences that matched EcTehA with an E- value lower than 10−3 in an alignment extending over at least 50% of both predicted TM regions and passing our post-seed-expansion filtering criteria46 were passed to the protein production pipeline. Protein expression screening Full-length homologs from the following 38 species, including 2 sequences each from 5 of these, were amplified from genomic DNA by PCR: Thermoplasma acidophilum, Lactococcus lactis subsp. lactis Il1403, Streptococcus pyogenes M1 GAS, Streptococcus pneumoniae TIGR4, Haemophilus influenzae Rd KW20, Methanocaldococcus jannaschii DSM 2661 (2), Escherichia coli K12, Salmonella typhimurium LT2, Clostridium perfringens ATCC 13124, Xanthomonas campestris pv. campestris str. ATCC 33913, Streptococcus agalactiae 2603V/R, Shewanella oneidensis MR-1, Streptococcus mutans UA159, Archaeoglobus fulgidus DSM 4304, Vibrio parahaemolyticus RIMD 2210633 (2), Pseudomonas syringae pv. tomato str. DC3000 (2), Enterococcus faecalis V583, Pyrococcus horikoshii OT3, Bordetella bronchiseptica RB50, Streptomyces coelicolor A3, Corynebacterium glutamicum ATCC 13032, Picrophilus torridus DSM 9790, Acinetobacter sp. ADP1, Vibrio fischeri ES114, Pseudomonas fluorescens Pf-5, Sulfolobus acidocaldarius DSM 639, Colwellia psychrerythraea 34H, Neisseria meningitidis MC58, Rhodobacter sphaeroides 2.4.1, Vibrio cholerae O1 biovar eltor str. N16961 (2), Marinobacter aquaeolei VT8, Acinetobacter baumannii ATCC 17978, Klebsiella pneumoniae subsp. pneumoniae MGH 78578, Streptomyces avermitili MA-4680, Bordetella parapertussis 12822 (2), Streptococcus thermophilus LMG 18311, Salmonella enterica subsp. enterica serovar Paratyphi A str. ATCC 9150, and Anaeromyxobacter dehalogenans 2CP-C. Selected cDNAs were cloned into a modified pET vector (Novagen, Inc) that fuses a FLAG and deca-histidine tag at the C-terminus, which are cleavable by TEV protease. Proteins were expressed in E. coli BL21(DE3) plysS by a high throughput format (0.6ml in a deep well block) and purified after lysis by sonication using metal affinity purification in a buffer containing N-Dodecyl-β-D-Maltopyranoside. Samples were passed over an analytical size exclusion column in 12 different detergent-containing mobile phases, which included N- dodecyl-β-D-maltopyranoside (DDM), N-decyl-β-D-altopyranoside (DM), N-nonyl-β-D- altopyranoside (NM), N-octyl-β-D-altopyranoside (OM), N-octyl-β-D-Glucopyranoside (OG), N-nonyl-β-D-Glucopyranoside (NG) and lauryl dimethyl amine oxide (LDAO). Multi-angle light scattering with refractive index detection was used to analyze the oligomeric state51. The E. coli and H. influenzae proteins were judged to be monodisperse and stable and were passed to scale up. Chen et al. Page 9 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Scaled-up production and purification For scale up, in brief, transformed BL21 plysS cells were grown at 37°C in 2× YT media to OD 0.6-0.8 after being inoculated with 1% of the overnight culture. The culture was induced with 0.4mM IPTG and continued to grow at 37 °C for another 4 hours. The cells were harvested by centrifugation and stored at −80°C before use. Selenomethionyl (SeMet) TehA was expressed in a similar way, but using containing SeMet in place of methionine in defined minimal media. Cells were resuspended in a buffer containing 50 mM Tris-HCl (pH 8.0) and 200 mM NaCl and lysed using a French Press with two passes at 15-20,000 psi. Cell debris were removed by centrifugation at 10,000g for 20min, and the membrane fraction was isolated from that supernatant by ultra-centrifugation at 150,000g for 1 hr. The membrane fraction was homogenized in a solubilization buffer containing 50 mM Tris (pH 8.0) and 200 mM NaCl, and incubated with final concentration of 1% (w/v) dodecyl-β- D-maltopyranoside (DDM, Anatrace, Inc) for 1 hr at 4°C. The non-dissolved matter was remove by ultracentrifugation at 150,000g for 30 min, and the supernatant was loaded to a 5ml Hitrap Ni2+-NTA affinity column (GE Healthcare), pre-equilibrated with the same solubilization buffer supplemented with 0.05% DDM. After 20 column volume buffer wash, the protein was eluted with 250mM imidazole in the solublization buffer. The Flag and 10- His tags were removed by adding super TEV at 1:100 mass ratio and incubating at 4°C overnight. Tag removal was confirmed by SDS-PAGE, and the resulting sample was concentrated to around 10mg/ml. Preparative size exclusion chromatography was carried out on a Superdex-200 column for further purification, removal of TEV protease and the cleaved tag, and for buffer and detergent exchange. The gel-filtration buffer contained 10 mM Tris (pH 8.0), 200 mM NaCl, 1mM EDTA, 0.5 mM Tris [2-carboxyethyl] phosphine (TCEP), and 2×CMC of detergent. In the case of HiTehA, the protein was well behaved and stable in nearly all tested detergents, and we have purified it from DDM, DM, NM, OM, OG and LDAO. Protein characterization We performed N-terminal amino-acid sequencing of purified HiTehA and EcTehA prior to TEV protease treatment. Results from these analyses proved that true initiating methionine residue is located 14 residues after one annotated as N-terminal. The intervening nucleotide sequence contains a Shine-Delgarno sequence. For cross-linking experiments, purified HiTehA was incubated with 10 mM disuccinimidyl glutarate at room temperate for 30 min, and 100mM Tris-HCl pH 7.5 was added to stop the reaction. The incubated sample when run on a 8-25% gradient SDS-PAGE gel showed a ladder consistent with a trimeric structure. Crystallization and data collection Purified protein was concentrated to ~10mg/ml for initial crystal trials in a Mosquito robot with commercial screens from Hampton research, Emerald Biosystems and Molecular Dimension. We obtained crystals of HiTehA from protein in detergent DDM, DM, NM, OM, OG and LDAO, but only those from LDAO and OG gave diffraction to beyond 4Å spacings. Crystals that proved useful were all grown at 4°C using the sitting-drop vapor Chen et al. Page 10 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript diffusion method. After extensive optimization we reached conditions supporting very high resolution. Wild-type HiTehA and most of its variants crystallized from 1mM ZnSO4, 50mM Hepes-Na pH 7.8 and 28% PEG600. HiTehA F262A crystallized from 200mM Li2SO4, 100mM glycine pH9.3 and 33% PEG400). Addition of 10mM spermidine as an additive helped to obtain slightly better diffracting crystals. Cryoprotection was achieved by adding 5% ethylene glycol or PEG400 to the crystallization solution. Structure determination and refinement Native and SeMet SAD data set were collected at NSLS beamline X4A and processed using the software HKL200052. Crystals of HiTehA grew in space group R3 with a=b=c = 85.0Å and α=β=γ = 93.5°. All subsequent manipulations were done in the hexagonal setting of this space group with a=b= 96.0Å and c = 136.7Å. The asymmetric unit contains one TehA protomer and 65% solvent volume. The structure was determined at 2.0Å resolution from single-wavelength anomalous diffraction (SAD) using selenomethionine-substituted protein crystals. Assessment of data quality for phasing, location of heavy atom sites and initial phases were calculated using the HKL2MAP interface to SHELX programs53. All the secondary structure elements were clearly visible in the experimental electron density map. Automatic model building was done in Arp/wArp54 and completed manually in the program COOT55. The model was refined against native data at 1.20Å resolution using the program Refmac5.5 in CCP456, with anisotropic B factor restrained refinement applied. Subsequent structural analyses of mutant variants were refined as isomorphous structures. Site-directed mutagenesis Site-directed mutants were constructed using the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA 92037) and expressed from pET vectors in E. coli BL21(DE3) plysS cells as for the wild-type protein. Electrophysiology All constructs were cloned into plasmid pGHME2, linearized, and transcribed into cRNA using T7 polymerase (mMessage mMachine, Ambion). Oocytes were injected with 50 nl of cRNA solution each, at a constant concentration of 0.5 mg/ml for HiTehA, or 0.5 mg/ml for AtSLAC1 constructs, with or without 0.5 mg/ml of AtOST1 constructs and recorded in voltage-clamp experiments 2 days after injection. For mixed expression experiments, 25nl of cRNA solution was injected for each AtSLAC1 component. Two-electrode voltage clamp recordings were performed to measure HiTehA or AtSLAC1 currents as described6,7. The pipette solutions contained 3 M KCl. For voltage-clamp current recordings, the bath solution contained 1 mM MgCl2, 1 mM CaCl2, 10 mM Mes/Tris (pH 5.6) and 30 mM CsCl; For anion selectivity measurements, the bath solutions contained 50 mM Cl−, NO −3, malate, or 30 mM SO 2-3 sodium salts, 45 mM Na-gluconate, 1 mM Ca-gluconate2, 1 mM Mg- gluconate2, 1 mM K-gluconate, as well as 10 mM Tris/Mes (pH 5.6). Osmolality was adjusted with D-mannitol to 220 mOsmol/kg. The bath electrode was a 3M KCl agar bridge. The testing potentials were +50 to −110 or −130 mV in 20 mV decrements with 7.5 s duration. The holding and return potentials were 0 mV with a 1.45 s or 2 s, respectively. I-V Chen et al. Page 11 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript relations of HiTehA or AtSLAC1 were generated from the current reached after 0.5 s in testing potentials. The Goldman-Hodgkin-Katz equation was applied to estimate permeability ratios for monovalent ions as described6. For divalent anions, the permeability ratios were derived according to Fatt and Ginsborg57. Bioinformatic analysis of SLAC-related proteins Sequences related to SLAC1 were analyzed comprehensively by PSI-BLAST50. Searches at E<10−3 starting from five disparate homologs each identified a common pool of over 900 proteins, which when pooled were used for sub-classification into families and subfamilies. Details of these analyses are reported in footnotes to Table S1. Molecular figures were produced in PyMOL58. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We thank Filippo Mancia, Larry Shapiro and Ming Zhou for helpful discussions; Patricia Rodriguez and Arianne Morrison for help with cloning and expression; Jonah Cheung, Mark Collins, Chris Min, Sheng Ye and Zhen Zhang for advice in protein chemistry and crystallography; John Schwanof and Randy Abramowitz for help with synchrotron experiments; John Riley for help with Xenopus oocyte injections; and Min Su for advice on the PSI- BLAST analysis of SLAC relatives. This project was supported in part by an award to the New York Consortium on Membrane Protein Structure (NYCOMPS) from the NIGMS Protein Structure Initiative. Beamline X4A at the National Synchrotron Light Source (NSLS), a DOE facility at Brookhaven National Laboratory, is supported by the New York Structural Biology Center. References 1. Hetherington AM, Woodward FI. The role of stomata in sensing and driving environmental change. Nature. 2003; 424:901–908. [PubMed: 12931178] 2. Sirichandra C, Wasilewska A, Vlad F, Valon C, Leung J. The guard cell as a single-cell model towards understanding drought tolerance and abscisic acid action. J. Exp. Bot. 2009; 60:1439–1463. [PubMed: 19181866] 3. Negi J, et al. CO2 regulator SLAC1 and its homologues are essential for anion homeostatsis in plant cells. Nature. 2008; 452:483–486. [PubMed: 18305482] 4. Vahisalu T, et al. SLAC1 is required for plant guard cell S-type anion channel function in stomatal signalling. Nature. 2008; 452:487–491. [PubMed: 18305484] 5. Saji S, et al. Disruption of a gene encoding C4-dicarboxylate transporter-like protein increases ozone sensitivity through deregulation of the stomatal response in Arabidopsis thaliana. Plant Cell Physiol. 2008; 49:2–10. [PubMed: 18084014] 6. Lee SC, Lan W, Buchanan BB, Luan S. A protein kinase-phophatase pair interacts with an ion channel to regulate ABA signaling in plant guard cells. Proc. Natl. Acad. Sci. USA. 2009; 106:21419–21424. [PubMed: 19955427] 7. Geiger D, et al. Activity of guard cell anion channel SLAC1 is controlled by drought-stress signaling kinase-phosphatase pair. Proc. Natl. Acad. Sci. USA. 2009; 106:21425–21430. [PubMed: 19955405] 8. Schroeder JI, Hagiwara S. Cytosolic calcium regulates ion channels in the plasma membrane of Vicia faba guard cells. Nature. 1989; 338:427–430. 9. Mustilli A, Merlot S, Vavasseur A, Fenzi F, Giraudat J. Arabidopsis OST1 protein kinase mediates the regulation of stomatal aperture by abscisic acid and acts upstream of reactive oxygen species production. Plant Cell. 2002; 14:3089–3099. [PubMed: 12468729] Chen et al. Page 12 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 10. Leung J, et al. Arabidopsis ABA response gene ABI1: features of a calcium-modulated protein phosphatase. Science. 1994; 264:1448–1452. [PubMed: 7910981] 11. Meyer K, Leube MP, Grill E. A protein phosphatase 2C involved in ABA signal transduction in Arabidopsis thaliana. Science. 1994; 264:1452–1455. [PubMed: 8197457] 12. Ma Y, et al. Regulators of PP2C phosphatase activity function as abscisic acid sensors. Science. 2009; 324:1064–1068. [PubMed: 19407143] 13. Park S, et al. Abscisic acid inhibits type 2C protein phosphatases via the PYR/PYL family of START proteins. Science. 2009; 324:1068–1071. [PubMed: 19407142] 14. Melcher K, et al. A gate-latch-lock mechanism for hormone signalling by abscisic acid receptors. Nature. 2009; 462:602–608. [PubMed: 19898420] 15. Miyazono K, et al. Structural basis of abscisic acid signalling. Nature. 2009; 462:609–614. [PubMed: 19855379] 16. Fujii H, et al. In vitro reconstitution of an abscisic acid signalling pathway. Nature. 2009; 462:660– 664. [PubMed: 19924127] 17. Yin P, et al. Structural insights into the mechanism of abscisic acid signaling by PYL proteins. Nat. Struct. Mol. Biol. 2009; 16:1230–1236. [PubMed: 19893533] 18. Nishimura N, et al. PYR/PYL/RCAR family members are major in-vivo ABI1 protein phosphatase 2C-interacting proteins in Arabidopsis. Plant J. 2010; 61:290–299. [PubMed: 19874541] 19. Grobler J, Bauer F, Subden RE, van Vuuren HJ. The MAE1 gene of Schizosaccharomyces pombe encodes a permease for malate and other C4 dicarboxylic acids. Yeast. 1995; 11:1485–1491. [PubMed: 8750236] 20. Park H, Bakalinsky AT. SSU1 mediates sulphite efflux in Saccharomyces cerevisiae. Yeast. 2000; 16:881–888. [PubMed: 10870099] 21. Léchenne B, et al. Sulphite efflux pumps in Aspergillus fumigatus and dermatophytes. Microbiology. 2007; 153:905–913. [PubMed: 17322211] 22. Walter EG, Weiner JH, Taylor DE. Nucleotide sequence and overexpression of the tellurite- resistance determinant from the IncHII plasmid pHH1508a. Gene. 1991; 101:1–7. [PubMed: 2060788] 23. Taylor DE, Hou Y, Turner RJ, Weiner JH. Location of a potassium tellurite resistance operon (tehA tehB) within the terminus of Escherichia coli K-12. J Bacteriol. 1994; 176:2740–2742. [PubMed: 8169225] 24. Daley DO, et al. Global topology analysis of the Escherichia coli inner membrane proteome. Science. 2005; 308:1321–1323. [PubMed: 15919996] 25. Unwin N. Refined structure of the nicotinic acetylcholine receptor at 4Å resolution. J. Mol. Biol. 2005; 346:967–989. [PubMed: 15701510] 26. Chen Y, Hu L, Siegelbaum SA, Hendrickson WA. Structure-based analysis of anion channel TehA and methyltransferase TehB implicated in bacterial tellurite resistance. Submitted. 27. Schmidt C, Schroeder JI. Anion Selectivity of Slow Anion Channels in the Plasma Membrane of Guard Cells (Large Nitrate Permeability). Plant Physiol. 1994; 106:383–391. [PubMed: 12232336] 28. Vahisalu T, et al. Ozone-triggered rapid stomatal response involves the production of reactive oxygen species, and is controlled by SLAC1 and OST1. Plant J. 2010; 62:442–53. [PubMed: 20128877] 29. Wright EM, Diamond JM. Anion selectivity in biological systems. Physiol. Rev. 1977; 57:109– 156. [PubMed: 834775] 30. Dutzler R, Campbell EB, MacKinnon R. Gating the selectivity filter in ClC chloride channels. Science. 2003; 300:108–112. [PubMed: 12649487] 31. Accardi A, Miller C. Secondary active transport mediated by a prokaryotic homologue of ClC Cl- channels. Nature. 2004; 427:803–807. [PubMed: 14985752] 32. Picollo A, Malvezzi M, Houtman JC, Accardi A. Basis of substrate binding and conservation of selectivity in the CLC family of channels and transporters. Nat. Struct. Mol. Biol. 2009; 16:294– 301. [PubMed: 19219045] 33. De Angeli A, et al. The nitrate/proton antiporter AtCLCa mediates nitrate accumulation in plant vacuoles. Nature. 2006; 442:939–942. [PubMed: 16878138] Chen et al. Page 13 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 34. Hiller S, Garces RG, Malia TJ, Orekhov VY, Colombini M, Wagner G. Solution structure of the integral human membrane protein VDAC-1 in detergent micelles. Science. 2008; 321:1206–1210. [PubMed: 18755977] 35. Bayrhuber M, et al. Structure of the human voltage-dependent anion channel. Proc. Natl. Acad. Sci. USA. 2008; 105:15370–15375. [PubMed: 18832158] 36. Ujwal R, et al. The crystal structure of mouse VDAC1 at 2.3 Å resolution reveals mechanistic insights into metabolite gating. Proc. Natl. Acad. Sci. USA. 2008; 105:17742–17747. [PubMed: 18988731] 37. Kouyama T, et al. Crystal structure of the light-driven chloride pump halorhodopsin from Natronomonas pharaonis. J. Mol. Biol. 2010; 396:564–579. [PubMed: 19961859] 38. Gadsby DC, Vergani P, Csanády L. The ABC protein turned chloride channel whose failure causes cystic fibrosis. Nature. 2006; 440:477–483. [PubMed: 16554808] 39. Miller PS, Smart TG. Binding, activation and modulation of Cys-loop receptors. Trends Pharmacol. Sci. 2010; 31:161–174. [PubMed: 20096941] 40. Yang YD, et al. TMEM16A confers receptor-activated calcium-dependent chloride conductance. Nature. 2008; 455:1210–5. [PubMed: 18724360] 41. Caputo A, et al. TMEM16A, a membrane protein associated with calcium-dependent chloride channel activity. Science. 2008; 322:590–594. [PubMed: 18772398] 42. Schroeder BC, Cheng T, Jan YN, Jan LY. Expression cloning of TMEM16A as a calcium- activated chloride channel subunit. Cell. 2008; 134:1019–1029. [PubMed: 18805094] 43. Schroeder JI, Keller BU. Two types of anion channel currents in guard cells with distinct voltage regulation. Proc. Natl. Acad. Sci. USA. 1992; 89:5025–5029. [PubMed: 1375754] 44. Meyer S, et al. AtALMT12 represents an R-type anion channel required for stomatal movement in Arabidopsis guard cells. Plant J. Jul 12.2010 Epub ahead of print. 45. Sasaki T, et al. Closing plant stomata requires a homolog of an aluminum-activated malate transporter. Plant Cell Physiol. 2010; 51:354–65. [PubMed: 20154005] 46. Punta M, et al. Structural genomics target selection for the New York Consortium on Membrane Protein Structure. J. Struct. Funct. Genomics. 2009; 4:255–268. [PubMed: 19859826] 47. Papadopoulos JS, Agarwala R. COBALT: constraint-based alignment tool for multiple protein sequences. Bioinformatics. 2007; 23:1073–1079. [PubMed: 17332019] 48. Landau M, et al. ConSurf 2005: the projection of evolutionary conservation scores of residues on protein structures. Nucleic Acids Res. 2005; 33:W299–W302. [PubMed: 15980475] 49. Rocchia W, et al. Rapid grid-based construction of the molecular surface for both molecules and geometric objects: applications to the finite difference Poisson-Boltzmann method. J. Comp. Chem. 2002; 23:128–137. [PubMed: 11913378] 50. Altschul SF, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997; 25:3389–3402. [PubMed: 9254694] 51. Kendrick BS, Kerwin BA, Chang BS, Philo JS. Online size-exclusion high-performance liquid chromatography light scattering and differential refractometry methods to determine degree of polymer conjugation to proteins and protein-protein or protein-ligand association states. Anal. Biochem. 2001; 299:136–146. [PubMed: 11730335] 52. Otwinowski Z, Minor W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 1997; 276:307–326. 53. Pape P, Schneider TR. HKL2MAP: a graphical user interface for phasing with SHELX programs. J. Appl. Cryst. 2004; 37:843–844. 54. Perrakis A, Morris R, Lamzin VS. Automated protein model building combined with iterative structure refinement. Nat. Struct. Biol. 1999; 6:458–463. [PubMed: 10331874] 55. Emsley P, Cowtan K. COOT: model-building tools for molecular graphics. Acta Crystallogr. D. 2004; 60:2126–2132. [PubMed: 15572765] 56. CCP4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D. 1994; 50:760– 763. [PubMed: 15299374] 57. Fatt P, Ginsborg BL. The ionic requirements for the production of action potentials in crustacean muscle fibers. J. Physiol. 1958; 142:516–543. [PubMed: 13576452] Chen et al. Page 14 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 58. DeLano, WL. The PyMOL molecular graphics system. DeLano Scientific; San Carlos, CA, USA: 2002. http://www.pymol.org Chen et al. Page 15 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 1. Sequence analysis for the SLAC1 superfamily a, Family tree. The presentation was computed by the program COBALT47 from representative subfamily sequences (Table S1), including Arabidopsis thaliana SLAC1 for SF1Ai, Haemophilus influenzae TehA for SF1Ci, Escherichia coli TehA for SF1Cii, Vibrio parahaemolyticus for SF2A, Staphylococcus aureus for SF2B, Aspergillus fumigatus Ssu1 for SF3A, and Schizosaccharomyces pombe Mae1 for SF3B. b, Structure-based sequence alignment of TehA from Haemophilus influenzae (HiTehA) and SLAC1 from Arabidopsis thaliana (AtSLAC1). The TehA structure has been used to restrict sequence gaps to inter- helical segments. Superior coils define extents of the HiTehA helical segments; red letters mark residue identities; red boxes are drawn for residues that are >95% identical within the plant subfamily SF1A for AtSLAC1 or within the TehA subfamily for HiTehA; red diamonds mark HiTehA residues that line the central pore; and the colored inferior bar encodes ConSurf sequence variability48 for the SF1 family of 204 non-redundant proteins. Chen et al. Page 16 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 2. Crystal structure of HiTehA and homology model of AtSLAC1 a, Electron density distribution from the HiTehA crystal structure at 1.2Å resolution. The map has (2Fo-Fc) coefficients based on the superimposed model. Contours are at 2.5σ. b, Ribbon diagram of the HiTehA trimer. Each protomer is colored spectrally from blue at its N-terminus to red at its C-terminus. c, DelPhi49 electrostatic potential at the extracellular surface. Electronegative and electropositive potential are colored in degrees of red and blue saturation, respectively. d, Electrostatic potential49 at the intracellular surface. e, Ribbon diagram of an HiTehA protomer viewed from outside the membrane. The ribbon is colored spectrally as in 2b. f, Ribbon diagram of an HiTehA protomer viewed from within the membrane from below the view of e. g, Surface of a homology model of AtSLAC1, viewed as in f, and colored by sequence variability48. h, Surface of AtSLAC1 as in g, but colored by electrostatic potential49. Chen et al. Page 17 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 3. Putative structure of the SLAC1 conductance pore a, Cross-section through the homology model of AtSLAC1. The model is viewed as in 2i, with the electrostatic potential49 shown on the external surface of the molecular envelope. The side chain of Phe450 is shown as a stick model (red) on the backbone ribbon colored yellow. b, Cross-section as in a, but colored by surface conservation48 as in 2h. c, Pore- lining residues in the SLAC1 homology model. A cylinder model (left), spectrally colored as in 2b, provides a key for viewing the rolled-open structure (left) with pore-lining residues of AtSLAC1 shown on the TModd helices. d, Ribbon diagrams of HiTehA TM9 (left) and TM7 (right) viewed from within the conductance pore. The side chains of Pro207 and Phe262 are shown as well as a kink-stabilizing water HOH25 that is coordinated by the NH of Gly263 and by C=O groups of Gly202 and Ala259. Density contours are shown for the water molecule. Chen et al. Page 18 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 4. Ionic conductance measurements a, Typical microelectrode voltage-clamp current traces from oocytes injected with various channel cRNAs. Left column, oocytes injected with cRNAs encoding wild-type HiTehA channels (WT), or F262A, G15D/F262A or G15D mutants. Right two columns, oocytes injected with cRNAs for WT AtSLAC1, or F450A, G194D/F450A or G194D mutants, with or without co-injection of AtOST1. Dotted lines represent zero current levels. Extracellular solution contained 30mM CsCl. Schematic keys at the left place phenyl and/or aspartyl gates. b, Effects of gating residue mutations. Mean chloride currents, measured at −90 mV, are shown comparing WT HiTehA and its mutant series F262A,G,T,V,L with WT AtSLAC1 and its corresponding series F450A,G,T,V,L, both alone and co-expressed with AtOST1. Full I-V relations are shown in Fig. S5. c, Effect of gating residues on relative AtSLAC1 anion permeabilities. Relative permeabilities (P[X]/P[Cl]) for chloride, nitrate, sulfite and malate of WT, F450A and F450T SLAC1 channels were measured from the change in current reversal potential with Cl− or anion X− as the sole permeant anion in the bath solution (Methods, Table S6). Chen et al. Page 19 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 5. Structural features at the SLAC1 homolog gate a,b, Cross-sections through the conductance pores of HiTehA F262A and HiTehA G15D. The view and presentations are as in 3a, except that helices are colored purple. c, Molecular basis for conformational strain in gating residue Phe262 of HiTehA. Helices TM1 (left), TM9 (center) and TM7 (right) are viewed from within the pore and presented as ribbon diagrams with selected side chains drawn in stick representation. The local low-energy conformation for the phenyl ring (χ2 = 90°) is shown in thin lines with short contacts indicated in dashed lines: Leu18 Cδ2 – Phe262 Cδ1, d = 2.4 Å; Val210 Cγ2 – Phe262 Cε2, d = 2.8 Å. d, Conformational shifts consequent to release of strain in gating residue Phe262. Cα backbone structures of F262A, F262V and F262L HiTehA were first superimposed onto WT HiTehA, and residues 258-266 were drawn in stereo as superimposed. All backbone atoms are shown for peptides 262±1, but only Cα atoms are shown otherwise. The WT backbone and phenyl group are green; other backbone are all magenta; side chains of Ala262, Val262 and Leu262 are cyan, blue and red, respectively; oxygen-directed bonds are red and nitrogen-directed bonds are blue. e, Typical microelectrode voltage-clamp current traces from oocytes injected with wild-type (WT) or F450L AtSLAC1 cRNA. Experimental conditions and displays are as in 4a. Chen et al. Page 20 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript
3M76
Crystal Structure of Plant SLAC1 homolog TehA
Homolog Structure of the SLAC1 Anion Channel for Closing Stomata in Leaves Yuhang Chen1,6, Lei Hu2, Marco Punta6,7, Renato Bruni6, Brandan Hillerich6, Brian Kloss6, Burkhard Rost1,6,1, James Love6, Steven A. Siegelbaum2,3,5, and Wayne A. Hendrickson1,4,5,6 1Department of Biochemistry and Molecular Biophysics, Columbia University, New York, NY 10032, USA 2Department of Neuroscience, Columbia University, New York, NY 10032, USA 3Department of Pharmacology, Columbia University, New York, NY 10032, USA 4Department of Physiology and Cellular Biophysics, Columbia University, New York, NY 10032, USA 5Howard Hughes Medical Institute, Columbia University, New York, NY 10032, USA 6NYCOMPS, New York Structural Biology Center 89 Convent Avenue, New York, NY 10027 USA 7Department of Computer Science and Institute for Advanced Study Technical University of Munich D-85748 Munich, Germany Summary The plant SLAC1 anion channel controls turgor pressure in the aperture-defining guard cells of plant stomata, thereby regulating exchange of water vapor and photosynthetic gases in response to environmental signals such as drought or high levels of carbon dioxide. We determined the crystal structure of a bacterial homolog of SLAC1 at 1.20Å resolution, and we have used structure- inspired mutagenesis to analyze the conductance properties of SLAC1 channels. SLAC1 is a symmetric trimer composed from quasi-symmetric subunits, each having ten transmembrane helices arranged from helical hairpin pairs to form a central five-helix transmembrane pore that is gated by an extremely conserved phenylalanine residue. Conformational features suggest a mechanism for control of gating by kinase activation, and electrostatic features of the pore coupled with electrophysiological characteristics suggest that selectivity among different anions is largely a function of the energetic cost of ion dehydration. Users may view, print, copy, download and text and data- mine the content in such documents, for the purposes of academic research, subject always to the full Conditions of use: http://www.nature.com/authors/editorial_policies/license.html#terms Correspondence and request for materials should be addressed to WAH (wayne@convex.hhmi.columbia.edu).. Supplementary Information is linked to the online version of the paper at www.nature.com/nature. Author Contributions YC, LH, MP, RB, BH, BK and JL performed experiments; YC, LH, BR, SAS, and WAH analyzed data; YC, LH, SAS, and WAH prepared the manuscript. Author Information Atomic coordinates and diffraction data are deposited in the Protein Data Bank with accession codes 3M71, 3M73, 3M74, 3M75, 3M76, and 3M7L. See Table S4 for identifications. Reprints and permissions information is available at www.Nature.com/reprints. HHS Public Access Author manuscript Nature. Author manuscript; available in PMC 2013 January 18. Published in final edited form as: Nature. 2010 October 28; 467(7319): 1074–1080. doi:10.1038/nature09487. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Stomatal pores in the leaves of plants permit the influx of atmospheric carbon dioxide in exchange for transpirational evaporation of water1,2. A pair of kidney-shaped guard cells define each pore aperture, and turgor pressure variation in these cells determines the degree of stomatal pore openness. Depending on diverse environmental factors, the stomata close to prevent H2O losses and open to admit CO2 for photosynthesis. Environmental stimuli that lead to stomatal closure include darkness, high CO2 levels, ozone, low air humidity and drought. The plant hormone abscisic acid (ABA) is critical for signal transduction from these stimuli. Mutational screens in Arabidopsis thaliana for CO2 and ozone sensitivity identified a protein with ten predicted transmembrane (TM) helices, now called slow anion channel 1 (SLAC1), as having a central role in the control of stomatal closure3-5. Recent studies proved that SLAC1 is indeed an anion channel6-7, with characteristics like those of slow anion channels found in guard cells8, and that it is activated by phosphorylation from the OST1 kinase9. OST1 activity is negatively regulated by the ABI1 phosphatase10,11, which is in turn inhibited by the stomatal ABA receptors PYR/RCAR12,13 when in the ternary hormone-receptor-phosphatase complex14-18. Thereby, ABA stimulates SLAC1 channel activity. Resulting Cl− efflux through SLAC1 causes membrane depolarization, which activates outward-rectifying K+ channels, leading to KCl and water efflux to further reduce turgor and cause stomatal closure. SLAC1 expression in Arabidopsis is confined to the guard cells of leaves, but other Arabidopsis tissues do have SLAC1 homologs3, named SLAH1-SLAH4. The identifying mutations slac1-14 and slac1-23 are, respectively, in predicted transmembrane segments 9 (S456F) and 1 (G194D) of a protein that includes substantial N- and C-terminal extensions outside a 10-helix transmembrane (TM) domain. SLAH1, which is absent from leaves and lacks the terminal extensions of SLAC1, fully complements the mutant phenotype in slac1-2 guard cell protoplasts3. SLAC1 and homologs are also present in other plant genomes, including nine in rice (Oryza sativa) and five in grape vines (Vitis vinifera). SLAC1 relatives, some quite remote, also occur in bacteria, archaea and fungi. Known prokaryotic homologs contain only the predicted transmembrane domain of SLAC1, but some fungal homologs do have N- and C-terminal extensions. One homolog, Mae1 from the yeast Schizosaccharomyces pombe, functions as a malate uptake transporter19; another, Ssu1 from Saccharomyces cerevisiae and other fungi including Aspergillus fumigatus, is characterized as a sulfite efflux pump20,21; and TehA from Escherichia coli is identified as a tellurite resistance protein by virtue of its association in the tehA/tehB operon22,23. Despite a lack of further biochemical characterization, many homologs are annotated as tellurite resistance/ dicarboxylate transporter (TDT) proteins. We have undertaken structural and functional characterizations of the SLAC1 anion channel. We first solved an atomic-resolution crystal structure of the TehA homolog from Haemophilus influenzae, and we then developed a homology model for Arabidopsis SLAC1. This model allowed us to conduct mutagenesis for functional testing of structure-inspired hypotheses on gating and selectivity. We expressed Haemophilus TehA and Arabidopsis SLAC1 in Xenopus oocytes to characterize channel properties of these proteins and mutant variants. We also determined crystal structures for several mutant variants, including the homolog of slac1-2. Chen et al. Page 2 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Structure of SLAC1 bacterial homolog TehA We performed a bioinformatic analysis of SLAC1-related proteins, first clustering nearly 900 non-redundant sequences into a superfamily at the PSI-BLAST level E≤10−3, then into three distinct families at an initial threshold of E≤10−30, and finally into subfamilies at a typical initial threshold of E≤10−55. Since previous annotation is not well founded in experiment and SLAC1 is now the best characterized member, we adopt a nomenclature defining a SLAC superfamily divided into families identified as SF1-SF3 and subfamilies SF1A, SF1B etc. Family SF1 comprises the plant SLAC proteins and close bacterial homologs, family SF2 comprises a distinct set of bacterial proteins often annotated as exfoliative toxins, and family SF3 comprises the fungal Mae1 and Ssu1 proteins and their archaeal or bacterial homologs, respectively. SLAC family SF1 has three large subfamilies: the plant SLAC and SLAH proteins are in subfamily SF1A, closest bacterial homologs are in SF1B, and the TehA homologs are in SF1C (Fig. 1a). The other families also divide into subfamilies as detailed in Table S1, and family SF1 is divided into sub-subfamilies (Table S2). Two pertinent SF1 sequences are aligned in Fig. 1b. We used a structural genomics approach to obtain structural information, testing expression and purification for 43 bacterial and archaeal likely homologs, assaying for detergent choice and stability on 8 of these, finding two with appropriate profiles by size exclusion chromatography, and obtaining suitable crystals for one. This protein, TehA from H. influenzae (HiTehA), was found to be trimeric both by size-exclusion multi-angle light scattering (SEC-MALS) measurements and by chemical cross-linking. When solubilized in β-octylglucoside, HiTehA crystallized in space group R3 with a=b=96.01Å and c=136.27Å. Each asymmetric unit contains one subunit and 65% solvent. The structure was solved by selenomethionyl (SeMet) SAD phasing, ultimately at 1.50Å resolution (Table S3, Fig. S1), and then refined at 1.20Å resolution (Fig. 2a) to R/Rfree values of 14.1%/15.7% for a model that includes ordered residues 6-313, 213 water molecules and four detergent molecules (Table S4). The crystal structure has TehA trimers aligned with three-fold axes of the lattice (Fig. 2b). Subunits are tightly associated, burying 8947 Å2 of total surface area within trimer interfaces. The electrostatic potential surface is largely negative on the extracellular surface (Fig. 2c) and largely positive on the cytoplasmic surface (Fig. 2d). The membrane orientation is specified experimentally from GFP tagging of E. coli TehA24. Each TehA protomer has ten TM helices, as predicted; however, the fold is novel. Tandemly repeated helical hairpins are arranged with quasi-five-fold symmetry (Figs. 2e,S2). Extracellular inter-helix loops are short (2-5 residues) whereas intracellular inter-helix connections are longer, including a nine-residue helix H2,3 between TM2 and TM3 (Fig. 2f). An inner pentad of outwardly directed, TModd, helices creates an apparent pore through each protomer perpendicular to the putative membrane plane. TMeven helices from the five hairpins surround the inner pore and make an outer layer. Chen et al. Page 3 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Homology model for plant SLAC1 Arabidopsis SLAC1 (AtSLAC1) is substantially similar to bacterial homologs, notably HiTehA (Fig. 1b). All HiTehA TM helices are fully aligned to predicted SLAC1 TM helices, but there are short inter-helical gaps (1-5 residues) in all five extracellular loops and in two of the intracellular loops. The TM domain of AtSLAC1 (residues 188-504) aligns to HiTehA with 19% sequence identity and with a PSI-BLAST E-value of 3×10−22. For comparison, and in keeping with the family tree (Fig. 1a), the TM domain of AtSLAC1 shares sequence identities of 76% with rice SLAC1, 41% with Arabidopsis SLAH1, 25% with an SL1B homolog from Halorhodospira halophila, 11% with S. cerevisiae Ssu1, and 9% with S. pombe Mae1. A conceptual model with the AtSLAC1 sequence transposed onto the HiTehA helices sufficed to guide most of our mutational tests, but a detailed AtSLAC1 homology model helped to refine our ideas. Surface variability and electrostatic potential are plotted onto the surface of this model (Fig. 2g,2h). The most remarkable feature of the TehA structure and corresponding SLAC1 model is the central pore through each protomer. As for acetylcholine receptors25, the SLAC1 pore is formed by five helices; but the SLAC1 helices come from one protein molecule rather than five. The SLAC1 pore has a relatively uniform diameter of approximately 5Å across nearly five helical turns (Fig. S3), except for a pronounced constriction in the middle of the membrane (Fig. 3a) where the pore is occluded by the side chain of Phe 450 (Phe 262 in HiTehA). This residue is the only absolutely conserved amino acid residue of the SLAC1 family. The pore is lined with highly conserved (86% identity among five SLAC1 orthologs; 32% identity between AtSLAC1 and HiTehA) and generally hydrophobic residues (Figs. 1b, 3b,3c,S4). Despite this hydrophobicity, the electrostatic potential on the pore surface is polarized (Fig. 3a), presumably due to an invaginated shape adjacent to charged residues outside the membrane. The generally electropositive character of the cytoplasmic surface likely contributes to anion efflux. Kinks in the pore helices contribute to formation of a relatively constant pore diameter across the membrane. Four of the five HiTehA inner helices have centrally located proline residues, which necessarily generate kinks, and TM9 is kinked at a backbone-coordinated water molecule (Fig. 3d). Proline replaces Gly263 in all SF1A and SF1B relatives, including Pro451 AtSLAC1. This water-displacing change is isostructural26. Centrally located prolines also prevail in TM3, TM5 and TM7 across the SF1 proteins. The outer helices are longer and straighter, but more inclined. The only outer-helix proline kink of HiTehA is in TM6 at the trimer three-fold axis. Two of the gene-identifying mutations in Arabadopsis SLAC1 are selected point mutations, others are disruptive transfer DNA (T-DNA) insertions4. In the AtSLAC1 homology model, the slac1-2 (G194D) mutation3 points into the pore from TM1 and can be accommodated structurally whereas the slac1-1 (S456F) mutation4 points away from the pore six residues after pore-blocking Phe450 on TM9 and would be expected to disruptive. Ser456 interacts with outer helix TM10 in the homology model, and the phenyl bulk from S456F would not fit. Position 456 has alanine in HiTehA, 66% of all 204 SF1 homologs and another 27% have threonine or serine (as in all SLAC1 channels); phenylalanine never occurs among all Chen et al. Page 4 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 814 SLAC superfamily members. Position 194 has glycine in 58% of SF1 homologs and alanine in another 22%. Asp194, which never occurs naturally, would block the pore and be expected to repel anions. Mutational tests of channel function Mutational studies corroborate the hypothesis that a TehA-based SLAC1 model is appropriate. First, as discussed above, the slac1-1 mutant (S456F) is expected to be structurally disruptive, and indeed it is inactive in guard cells4, and the slac1-2 mutation (G194D) is expected to block the pore, and we show below that this variant is also inactive. We have also shown that the introduction of SLAC1-conserved proline residues into HiTehA (A208P/G263P) is accommodated isomorphously26. Moreover, as shown below, channel conductance properties of several mutants are similar for AtSLAC1 and HiTehA. To examine characteristics of the SLAC1 channel in light of the structural model, we performed electrophysiological tests of membrane currents from voltage-clamped Xenopus oocytes following injection of wild-type or mutant AtSLAC1 or HiTehA cRNAs. We observed modest-sized chloride currents with wild-type AtSLAC1 cRNA, as found previously6,7, but did not detect any chloride current following injection of wild-type HiTehA cRNA. We found that SLAC1 Cl− conductance was enhanced when the OST1 kinase cRNA was co-injected with SLAC1 cRNA, but only to the levels found by Lee et al.6 and not to the much higher levels found by Geiger et al.7 with OST1 physically connected to SLAC1 by split YFP linkage. In keeping with structural evidence that Phe262 blocks the HiTehA pore, removal of the phenyl group in HiTehA F262A or in the homologous AtSLAC1 F450A resulted in very large Cl− currents relative to wild-type levels, and the SLAC1 currents were now less enhanced by presence of OST1 (Fig. 4a). The tempting interpretation of an opened gate will require validation with appropriately analyzed single- channel recordings27. In keeping with the slac1-2 phenotype3, neither the functionally impaired AtSLAC1 G194D nor its HiTehA G15D homolog showed any substantial conductance; moreover, further consistent with pore blockage by Asp194 in slac1-2, the large conductances of HiTehA F262A and AtSLAC1 F450A were stopped in the double mutants AtSLAC1 G194D/F450A and HiTehA G15D/F262A (Fig. 4a). Here again, the effects in SLAC1 were independent of OST1. We also tested conductance characteristics for a series of AtSLAC1 F450X substitution mutants – F450A,G,T,V,L– and for the corresponding HiTehA F262X series – F262A,G,T,V,L (Figs. 4b,S5, Table S5). Findings from the two series are roughly parallel; in particular, the alanine and glycine substitutions lead to large currents for both and in comparison to the others. There are distinctions, of course, including generally higher conductances for AtSLAC1 over HiTehA and less conductance of F262T TehA compared to F450T SLAC1. It is also noteworthy that OST1 activation is very muted for the F450A,G,L mutants, which is consistent with SLAC1 gating at Phe450. Crystal structures were also determined for several of the HiTehA mutant variants (Table S4). The structures of F262A (1.15Å), F262V (1.60Å), F262L (1.65Å) and G15D (1.50Å) are all essentially isomorphous with the wild-type TehA structure with changes localized Chen et al. Page 5 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript primarily at the sites of mutation; the same is true for the double mutations of F262A/G15D, F262G/G15D and A208P/G263P26. The F262A structure has a wide open pore (Fig. 5a) with a relatively uniform pore diameter of ~5Å through ~30Å across the membrane (Fig. S3) whereas G15D has a doubly occluded pore (Fig. 5b). The pores of other mutant variants are consistent with the sizes of constrictive residues and with the observed conductances. Gating and activation The crystal structures of TehA and its mutant variants when taken together with the functional studies in Xenopus oocytes point to a crucial role for Phe450 in gating of the SLAC1 anion channel. Conservation of this residue across the SLAC1 family implies functional importance. The occlusion of the pore by the presence of F262 in the structure of wild-type TehA and the openness of the pore upon its substitution by alanine in the structure of the F262A mutant provides physical evidence for a gating role of this residue. This interpretation is supported by the correlated conductance characteristics from variants of the AtSLAC1 and HiTehA channels (Figs. 4b,S5). While these observations may suffice for placing the gate within the channel pore, they do not by themselves suggest a mechanism for gating in response to physiological stimuli. Some insight does come from conformational details defined at high resolution. One important structural clue is that the side chain of Phe262 is in a high energy conformation in the HiTehA structure, with χ1/χ2 at −160°/-4°. Although χ1 is in a preferred trans conformation, the phenyl ring is restricted by contacts with Val210 and Leu18 to a χ2 value near 0° rather than near to the preferred 90° orientation (Fig. 5c). Further evidence that Phe262 is restrained from local equilibrium comes from shifts observed in crystal structures of the F262A, F262V, F262L and F262G/G15D variants, which all show consistent backbone movements that displace Cβ(262) by 0.47-0.78Å (Fig. 5d). By contrast, Leu262 in F262L is in a preferred trans/gauche+ conformation at χ1/χ2 = 177°/63° as is Val262 in F262V at χ1 = −176°. What might control activation of HiTehA is unclear, but for AtSLAC1 activation is by OST1 phosphorylation6,7. The molecular consequences of OST1 phosphorylation of SLAC1 remain unknown, but it is plausible that associated shifts in pore- helix orientations would unlatch Phe450 in SLAC1 from a TehA-like restrained orientation. By analogy with Leu262 in F262L, we expect a preferred rotameric state for Leu450 in AtSLAC1 F450L. Thus, the lack of appreciable OST1 activation of conductance in AtSLAC1 F450L (Fig. 5e) might be explained by the lack of a restraining latch, whereby the channel remains closed despite OST1 activation. Puzzles certainly remain since OST1 does substantially activate AtSLAC1 F450T or F450V, which like HiTehA F262V should also be unrestrained; presumably activating adjustments widen the pore enough for ion permeation past threonine and valine but not leucine. Phosphorylation sites have been discovered in the N- and C-terminal tails of AtSLAC16,7,28 (179 and 51 residues long, respectively), but these alone cannot explain OST1 activation of SLAC1. First SLAH1, which fully complements the slac1-1 mutation, does not have these cytoplasmic tails. Secondly, although OST1 phosphorylation of Ser120 in the N-terminal tail is necessary for SLAC1 activation, it is not sufficient7. Thus, we surmise that direct phosphorylation of the SLAC1 TM domain must be critical, and SLAC1 has four conserved Chen et al. Page 6 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Ser/Thr candidates in its cytoplasmic loops. Moreover, SLAC1 proteins have proline- mediated kinks at the putative Phe450 gate in helix TM9, and also in adjacent helix TM7; these features may play a role in phosphorylation-driven unlatching of the Phe450 gate in SLAC1. Ion selectivity and discrimination Our studies of SLAC1 channel relative ion permeabilities, based on measurements of current reversal potential, are consistent with earlier work demonstrating that AtSLAC1 conducts anions but not cations and is selective among anions, with greater permeability for nitrate than for chloride (as in Vicia faba guard cell protoplasts27) and much reduced permeability for malate, bicarbonate or sulfate6,7. We also find that SLAC1 has little permeability for sulfite. Additionally, we find that wild-type SLAC1, F450A and F450T all have similar relative permeabilities to chloride, sulfite and malate, despite having widely different conductance levels, but the gating mutants do show small but significant decreases in nitrate permeability (Fig. 4c, Table S6). The relative insensitivity of anion permeability to gating residue changes suggests that selectivity for these anions may occur away from the central constriction at the channel gate. To some extent, ionic discrimination must depend on pore geometry; thus, an organic anion such as malate may be simply too large to pass through the 5-Å wide pore. Although the SLAC1 pore is lined largely with hydrophobic sidechains (Fig. 3c), it is sprinkled with hydroxyl groups from serine and threonine residues (16%); their electropositive hydrogen atoms may facilitate conductance. Most strikingly, the electrostatic potential within the AtSLAC1 pore is electropositive throughout (Fig. 3a). This polarization, promoted by charges on extra-membranous loops, no doubt contributes significantly in discrimination against cations. The relative anion permeability sequence of SLAC1 determined by us and others, I− > NO3− > Br− > Cl− 6,7, corresponds to selectivity sequence 1 compiled by Wright and Diamond29 for a range of anion-selective proteins. This sequence correlates inversely with the hydration energies of monovalent anions – anions with a lower hydration energy have a greater channel permeability. It is thought to be generated in proteins with weak, low field-strength, anion binding sites, where selectivity is largely determined by the energetic cost of anion dehydration. These selectivity results are thus consistent with the SLAC1 structure, where the pore lacks any obvious anion binding site. Distinctiveness of the SLAC1 channel SLAC1 anion channels are entirely novel in structure and, apparently, in the mechanism for ion conductance. The best characterized of anion channels belong to the CLC family of Cl− channels and transporters30-32. CLC channels have an altogether different architecture from the SLAC1 channel, and the mechanism for selectivity is also very different. Bacterial CLC transporters bind halide ions at three sites in a highly constricted pore30. By contrast, the SLAC1 pore has a relatively uniform diameter across the membrane, except where closed by the gating phenyl group, and we do not find discrete ion binding sites. CLC selectivity is governed by specific residues surrounding these binding sites30,32. The anion selectivity Chen et al. Page 7 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript sequence for the CLC channels of Cl−>Br−>NO −3>I−, opposite of that in SLAC1, is consistent with the high field-strength anion binding sites in CLC channels29. Interestingly, as for AtSLAC1, an Arabidopsis CLCa channel also preferentially transports nitrate ions33, and an E. coli CLC channel is converted to preference of nitrate when a generally conserved serine at the central site is substituted with proline as in AtCLCa32. SLAC1 also differs radically from other structurally characterized anion channels and transporters. These include the VDAC1 voltage-gated anion channel from mitochondrial outer membranes, which has a porin-like beta-barrel structure34-36, and a light-driven halorhodopsin chloride pump, which has a transmembrane conductance pathway similar to that of the proton-pumping pore of bacteriorhodopsin37. Although its channel structure is still only known by homology to other ABC transporters, CFTR is another obviously distinct chloride channel38. Cys-loop receptors also include anion channels39, and these are similar to SLAC1 in having five-helix pores25, but here selectivity is governed by charged groups at the entrance to the pore, which distinguish the anion-selective GABAA and glycine receptors from the cation-selective acetylcholine and serotonin 5HT3 receptors39. Finally, the recently identified TMEM16A gene for calcium-activated chloride channels40-42 appears to encode an 8-TM protein that is again distinct from SLAC1. Stomatal guard cells show both rapidly activated (R-type) and slow (S-type) anion channel activity43. Although slac1 guard cells have very defective S-type activity, their R-type currents are normal4. Guard cell protoplasts from the slac1-2 mutant abnormally accumulate Cl−, K+, malate and fumarate3, whereas SLAC1 shows negligible malate conductance7. As for SLAC1-associated K+ movements, other channels or transporters must be responsible for SLAC1-associated malate movements. Recent studies indicate that AtALMT12, an aluminium-activated malate transporter (ALMT) family member, is a malate-dependent R- type anion channel44 needed for stomatal closure45. Conclusions We find that many functional properties of the plant SLAC1 anion channel are explained well by the structure of an uncharacterized bacterial TehA protein that has been associated with tellurite resistance. SLAC1 and TehA belong to distinct subfamilies within one branch of a larger SLAC1 superfamily, but AtSLAC1 and HiTehA are sufficiently similar (19% sequence identity) that the SLAC1 homology model is predictive for function, including a verified placement of the identifying slac1-2 mutation G194D and a phenylalanine gate. One remaining puzzle concerns the structural change that activating phosphorylation elicits in SLAC1, and another puzzle is the biochemical role of the TehA homologs in bacteria. In a companion paper26, we examine functional and structural properties of TehA in bacteria, showing that it is anion channel, although actually not conferring tellurite resistance, and identifying a mutant variant with properties suggestive of an activated state. Thus, SLAC1 and TehA likely represent a large family of selective anion channels controlled by environmental stimuli. Chen et al. Page 8 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript METHODS Selection of target sequences TehA from E. coli (EcTehA) was a centrally selected sequence for expansion as a NYCOMPS targeted family. Target selection criteria utilized at NYCOMPS are described in details elsewhere46. Briefly, the EcTehA sequence was run against a dataset of about 40,000 predicted alpha helical integral membrane protein sequences from prokaryotic genomes (NYCOMPS98 dataset46) using PSI-BLAST50. Sequences that matched EcTehA with an E- value lower than 10−3 in an alignment extending over at least 50% of both predicted TM regions and passing our post-seed-expansion filtering criteria46 were passed to the protein production pipeline. Protein expression screening Full-length homologs from the following 38 species, including 2 sequences each from 5 of these, were amplified from genomic DNA by PCR: Thermoplasma acidophilum, Lactococcus lactis subsp. lactis Il1403, Streptococcus pyogenes M1 GAS, Streptococcus pneumoniae TIGR4, Haemophilus influenzae Rd KW20, Methanocaldococcus jannaschii DSM 2661 (2), Escherichia coli K12, Salmonella typhimurium LT2, Clostridium perfringens ATCC 13124, Xanthomonas campestris pv. campestris str. ATCC 33913, Streptococcus agalactiae 2603V/R, Shewanella oneidensis MR-1, Streptococcus mutans UA159, Archaeoglobus fulgidus DSM 4304, Vibrio parahaemolyticus RIMD 2210633 (2), Pseudomonas syringae pv. tomato str. DC3000 (2), Enterococcus faecalis V583, Pyrococcus horikoshii OT3, Bordetella bronchiseptica RB50, Streptomyces coelicolor A3, Corynebacterium glutamicum ATCC 13032, Picrophilus torridus DSM 9790, Acinetobacter sp. ADP1, Vibrio fischeri ES114, Pseudomonas fluorescens Pf-5, Sulfolobus acidocaldarius DSM 639, Colwellia psychrerythraea 34H, Neisseria meningitidis MC58, Rhodobacter sphaeroides 2.4.1, Vibrio cholerae O1 biovar eltor str. N16961 (2), Marinobacter aquaeolei VT8, Acinetobacter baumannii ATCC 17978, Klebsiella pneumoniae subsp. pneumoniae MGH 78578, Streptomyces avermitili MA-4680, Bordetella parapertussis 12822 (2), Streptococcus thermophilus LMG 18311, Salmonella enterica subsp. enterica serovar Paratyphi A str. ATCC 9150, and Anaeromyxobacter dehalogenans 2CP-C. Selected cDNAs were cloned into a modified pET vector (Novagen, Inc) that fuses a FLAG and deca-histidine tag at the C-terminus, which are cleavable by TEV protease. Proteins were expressed in E. coli BL21(DE3) plysS by a high throughput format (0.6ml in a deep well block) and purified after lysis by sonication using metal affinity purification in a buffer containing N-Dodecyl-β-D-Maltopyranoside. Samples were passed over an analytical size exclusion column in 12 different detergent-containing mobile phases, which included N- dodecyl-β-D-maltopyranoside (DDM), N-decyl-β-D-altopyranoside (DM), N-nonyl-β-D- altopyranoside (NM), N-octyl-β-D-altopyranoside (OM), N-octyl-β-D-Glucopyranoside (OG), N-nonyl-β-D-Glucopyranoside (NG) and lauryl dimethyl amine oxide (LDAO). Multi-angle light scattering with refractive index detection was used to analyze the oligomeric state51. The E. coli and H. influenzae proteins were judged to be monodisperse and stable and were passed to scale up. Chen et al. Page 9 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Scaled-up production and purification For scale up, in brief, transformed BL21 plysS cells were grown at 37°C in 2× YT media to OD 0.6-0.8 after being inoculated with 1% of the overnight culture. The culture was induced with 0.4mM IPTG and continued to grow at 37 °C for another 4 hours. The cells were harvested by centrifugation and stored at −80°C before use. Selenomethionyl (SeMet) TehA was expressed in a similar way, but using containing SeMet in place of methionine in defined minimal media. Cells were resuspended in a buffer containing 50 mM Tris-HCl (pH 8.0) and 200 mM NaCl and lysed using a French Press with two passes at 15-20,000 psi. Cell debris were removed by centrifugation at 10,000g for 20min, and the membrane fraction was isolated from that supernatant by ultra-centrifugation at 150,000g for 1 hr. The membrane fraction was homogenized in a solubilization buffer containing 50 mM Tris (pH 8.0) and 200 mM NaCl, and incubated with final concentration of 1% (w/v) dodecyl-β- D-maltopyranoside (DDM, Anatrace, Inc) for 1 hr at 4°C. The non-dissolved matter was remove by ultracentrifugation at 150,000g for 30 min, and the supernatant was loaded to a 5ml Hitrap Ni2+-NTA affinity column (GE Healthcare), pre-equilibrated with the same solubilization buffer supplemented with 0.05% DDM. After 20 column volume buffer wash, the protein was eluted with 250mM imidazole in the solublization buffer. The Flag and 10- His tags were removed by adding super TEV at 1:100 mass ratio and incubating at 4°C overnight. Tag removal was confirmed by SDS-PAGE, and the resulting sample was concentrated to around 10mg/ml. Preparative size exclusion chromatography was carried out on a Superdex-200 column for further purification, removal of TEV protease and the cleaved tag, and for buffer and detergent exchange. The gel-filtration buffer contained 10 mM Tris (pH 8.0), 200 mM NaCl, 1mM EDTA, 0.5 mM Tris [2-carboxyethyl] phosphine (TCEP), and 2×CMC of detergent. In the case of HiTehA, the protein was well behaved and stable in nearly all tested detergents, and we have purified it from DDM, DM, NM, OM, OG and LDAO. Protein characterization We performed N-terminal amino-acid sequencing of purified HiTehA and EcTehA prior to TEV protease treatment. Results from these analyses proved that true initiating methionine residue is located 14 residues after one annotated as N-terminal. The intervening nucleotide sequence contains a Shine-Delgarno sequence. For cross-linking experiments, purified HiTehA was incubated with 10 mM disuccinimidyl glutarate at room temperate for 30 min, and 100mM Tris-HCl pH 7.5 was added to stop the reaction. The incubated sample when run on a 8-25% gradient SDS-PAGE gel showed a ladder consistent with a trimeric structure. Crystallization and data collection Purified protein was concentrated to ~10mg/ml for initial crystal trials in a Mosquito robot with commercial screens from Hampton research, Emerald Biosystems and Molecular Dimension. We obtained crystals of HiTehA from protein in detergent DDM, DM, NM, OM, OG and LDAO, but only those from LDAO and OG gave diffraction to beyond 4Å spacings. Crystals that proved useful were all grown at 4°C using the sitting-drop vapor Chen et al. Page 10 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript diffusion method. After extensive optimization we reached conditions supporting very high resolution. Wild-type HiTehA and most of its variants crystallized from 1mM ZnSO4, 50mM Hepes-Na pH 7.8 and 28% PEG600. HiTehA F262A crystallized from 200mM Li2SO4, 100mM glycine pH9.3 and 33% PEG400). Addition of 10mM spermidine as an additive helped to obtain slightly better diffracting crystals. Cryoprotection was achieved by adding 5% ethylene glycol or PEG400 to the crystallization solution. Structure determination and refinement Native and SeMet SAD data set were collected at NSLS beamline X4A and processed using the software HKL200052. Crystals of HiTehA grew in space group R3 with a=b=c = 85.0Å and α=β=γ = 93.5°. All subsequent manipulations were done in the hexagonal setting of this space group with a=b= 96.0Å and c = 136.7Å. The asymmetric unit contains one TehA protomer and 65% solvent volume. The structure was determined at 2.0Å resolution from single-wavelength anomalous diffraction (SAD) using selenomethionine-substituted protein crystals. Assessment of data quality for phasing, location of heavy atom sites and initial phases were calculated using the HKL2MAP interface to SHELX programs53. All the secondary structure elements were clearly visible in the experimental electron density map. Automatic model building was done in Arp/wArp54 and completed manually in the program COOT55. The model was refined against native data at 1.20Å resolution using the program Refmac5.5 in CCP456, with anisotropic B factor restrained refinement applied. Subsequent structural analyses of mutant variants were refined as isomorphous structures. Site-directed mutagenesis Site-directed mutants were constructed using the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA 92037) and expressed from pET vectors in E. coli BL21(DE3) plysS cells as for the wild-type protein. Electrophysiology All constructs were cloned into plasmid pGHME2, linearized, and transcribed into cRNA using T7 polymerase (mMessage mMachine, Ambion). Oocytes were injected with 50 nl of cRNA solution each, at a constant concentration of 0.5 mg/ml for HiTehA, or 0.5 mg/ml for AtSLAC1 constructs, with or without 0.5 mg/ml of AtOST1 constructs and recorded in voltage-clamp experiments 2 days after injection. For mixed expression experiments, 25nl of cRNA solution was injected for each AtSLAC1 component. Two-electrode voltage clamp recordings were performed to measure HiTehA or AtSLAC1 currents as described6,7. The pipette solutions contained 3 M KCl. For voltage-clamp current recordings, the bath solution contained 1 mM MgCl2, 1 mM CaCl2, 10 mM Mes/Tris (pH 5.6) and 30 mM CsCl; For anion selectivity measurements, the bath solutions contained 50 mM Cl−, NO −3, malate, or 30 mM SO 2-3 sodium salts, 45 mM Na-gluconate, 1 mM Ca-gluconate2, 1 mM Mg- gluconate2, 1 mM K-gluconate, as well as 10 mM Tris/Mes (pH 5.6). Osmolality was adjusted with D-mannitol to 220 mOsmol/kg. The bath electrode was a 3M KCl agar bridge. The testing potentials were +50 to −110 or −130 mV in 20 mV decrements with 7.5 s duration. The holding and return potentials were 0 mV with a 1.45 s or 2 s, respectively. I-V Chen et al. Page 11 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript relations of HiTehA or AtSLAC1 were generated from the current reached after 0.5 s in testing potentials. The Goldman-Hodgkin-Katz equation was applied to estimate permeability ratios for monovalent ions as described6. For divalent anions, the permeability ratios were derived according to Fatt and Ginsborg57. Bioinformatic analysis of SLAC-related proteins Sequences related to SLAC1 were analyzed comprehensively by PSI-BLAST50. Searches at E<10−3 starting from five disparate homologs each identified a common pool of over 900 proteins, which when pooled were used for sub-classification into families and subfamilies. Details of these analyses are reported in footnotes to Table S1. Molecular figures were produced in PyMOL58. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We thank Filippo Mancia, Larry Shapiro and Ming Zhou for helpful discussions; Patricia Rodriguez and Arianne Morrison for help with cloning and expression; Jonah Cheung, Mark Collins, Chris Min, Sheng Ye and Zhen Zhang for advice in protein chemistry and crystallography; John Schwanof and Randy Abramowitz for help with synchrotron experiments; John Riley for help with Xenopus oocyte injections; and Min Su for advice on the PSI- BLAST analysis of SLAC relatives. This project was supported in part by an award to the New York Consortium on Membrane Protein Structure (NYCOMPS) from the NIGMS Protein Structure Initiative. Beamline X4A at the National Synchrotron Light Source (NSLS), a DOE facility at Brookhaven National Laboratory, is supported by the New York Structural Biology Center. References 1. Hetherington AM, Woodward FI. The role of stomata in sensing and driving environmental change. Nature. 2003; 424:901–908. [PubMed: 12931178] 2. Sirichandra C, Wasilewska A, Vlad F, Valon C, Leung J. The guard cell as a single-cell model towards understanding drought tolerance and abscisic acid action. J. Exp. Bot. 2009; 60:1439–1463. [PubMed: 19181866] 3. Negi J, et al. CO2 regulator SLAC1 and its homologues are essential for anion homeostatsis in plant cells. Nature. 2008; 452:483–486. [PubMed: 18305482] 4. Vahisalu T, et al. SLAC1 is required for plant guard cell S-type anion channel function in stomatal signalling. Nature. 2008; 452:487–491. [PubMed: 18305484] 5. Saji S, et al. Disruption of a gene encoding C4-dicarboxylate transporter-like protein increases ozone sensitivity through deregulation of the stomatal response in Arabidopsis thaliana. Plant Cell Physiol. 2008; 49:2–10. [PubMed: 18084014] 6. Lee SC, Lan W, Buchanan BB, Luan S. A protein kinase-phophatase pair interacts with an ion channel to regulate ABA signaling in plant guard cells. Proc. Natl. Acad. Sci. USA. 2009; 106:21419–21424. [PubMed: 19955427] 7. Geiger D, et al. Activity of guard cell anion channel SLAC1 is controlled by drought-stress signaling kinase-phosphatase pair. Proc. Natl. Acad. Sci. USA. 2009; 106:21425–21430. [PubMed: 19955405] 8. Schroeder JI, Hagiwara S. Cytosolic calcium regulates ion channels in the plasma membrane of Vicia faba guard cells. Nature. 1989; 338:427–430. 9. Mustilli A, Merlot S, Vavasseur A, Fenzi F, Giraudat J. Arabidopsis OST1 protein kinase mediates the regulation of stomatal aperture by abscisic acid and acts upstream of reactive oxygen species production. Plant Cell. 2002; 14:3089–3099. [PubMed: 12468729] Chen et al. Page 12 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 10. Leung J, et al. Arabidopsis ABA response gene ABI1: features of a calcium-modulated protein phosphatase. Science. 1994; 264:1448–1452. [PubMed: 7910981] 11. Meyer K, Leube MP, Grill E. A protein phosphatase 2C involved in ABA signal transduction in Arabidopsis thaliana. Science. 1994; 264:1452–1455. [PubMed: 8197457] 12. Ma Y, et al. Regulators of PP2C phosphatase activity function as abscisic acid sensors. Science. 2009; 324:1064–1068. [PubMed: 19407143] 13. Park S, et al. Abscisic acid inhibits type 2C protein phosphatases via the PYR/PYL family of START proteins. Science. 2009; 324:1068–1071. [PubMed: 19407142] 14. Melcher K, et al. A gate-latch-lock mechanism for hormone signalling by abscisic acid receptors. Nature. 2009; 462:602–608. [PubMed: 19898420] 15. Miyazono K, et al. Structural basis of abscisic acid signalling. Nature. 2009; 462:609–614. [PubMed: 19855379] 16. Fujii H, et al. In vitro reconstitution of an abscisic acid signalling pathway. Nature. 2009; 462:660– 664. [PubMed: 19924127] 17. Yin P, et al. Structural insights into the mechanism of abscisic acid signaling by PYL proteins. Nat. Struct. Mol. Biol. 2009; 16:1230–1236. [PubMed: 19893533] 18. Nishimura N, et al. PYR/PYL/RCAR family members are major in-vivo ABI1 protein phosphatase 2C-interacting proteins in Arabidopsis. Plant J. 2010; 61:290–299. [PubMed: 19874541] 19. Grobler J, Bauer F, Subden RE, van Vuuren HJ. The MAE1 gene of Schizosaccharomyces pombe encodes a permease for malate and other C4 dicarboxylic acids. Yeast. 1995; 11:1485–1491. [PubMed: 8750236] 20. Park H, Bakalinsky AT. SSU1 mediates sulphite efflux in Saccharomyces cerevisiae. Yeast. 2000; 16:881–888. [PubMed: 10870099] 21. Léchenne B, et al. Sulphite efflux pumps in Aspergillus fumigatus and dermatophytes. Microbiology. 2007; 153:905–913. [PubMed: 17322211] 22. Walter EG, Weiner JH, Taylor DE. Nucleotide sequence and overexpression of the tellurite- resistance determinant from the IncHII plasmid pHH1508a. Gene. 1991; 101:1–7. [PubMed: 2060788] 23. Taylor DE, Hou Y, Turner RJ, Weiner JH. Location of a potassium tellurite resistance operon (tehA tehB) within the terminus of Escherichia coli K-12. J Bacteriol. 1994; 176:2740–2742. [PubMed: 8169225] 24. Daley DO, et al. Global topology analysis of the Escherichia coli inner membrane proteome. Science. 2005; 308:1321–1323. [PubMed: 15919996] 25. Unwin N. Refined structure of the nicotinic acetylcholine receptor at 4Å resolution. J. Mol. Biol. 2005; 346:967–989. [PubMed: 15701510] 26. Chen Y, Hu L, Siegelbaum SA, Hendrickson WA. Structure-based analysis of anion channel TehA and methyltransferase TehB implicated in bacterial tellurite resistance. Submitted. 27. Schmidt C, Schroeder JI. Anion Selectivity of Slow Anion Channels in the Plasma Membrane of Guard Cells (Large Nitrate Permeability). Plant Physiol. 1994; 106:383–391. [PubMed: 12232336] 28. Vahisalu T, et al. Ozone-triggered rapid stomatal response involves the production of reactive oxygen species, and is controlled by SLAC1 and OST1. Plant J. 2010; 62:442–53. [PubMed: 20128877] 29. Wright EM, Diamond JM. Anion selectivity in biological systems. Physiol. Rev. 1977; 57:109– 156. [PubMed: 834775] 30. Dutzler R, Campbell EB, MacKinnon R. Gating the selectivity filter in ClC chloride channels. Science. 2003; 300:108–112. [PubMed: 12649487] 31. Accardi A, Miller C. Secondary active transport mediated by a prokaryotic homologue of ClC Cl- channels. Nature. 2004; 427:803–807. [PubMed: 14985752] 32. Picollo A, Malvezzi M, Houtman JC, Accardi A. Basis of substrate binding and conservation of selectivity in the CLC family of channels and transporters. Nat. Struct. Mol. Biol. 2009; 16:294– 301. [PubMed: 19219045] 33. De Angeli A, et al. The nitrate/proton antiporter AtCLCa mediates nitrate accumulation in plant vacuoles. Nature. 2006; 442:939–942. [PubMed: 16878138] Chen et al. Page 13 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 34. Hiller S, Garces RG, Malia TJ, Orekhov VY, Colombini M, Wagner G. Solution structure of the integral human membrane protein VDAC-1 in detergent micelles. Science. 2008; 321:1206–1210. [PubMed: 18755977] 35. Bayrhuber M, et al. Structure of the human voltage-dependent anion channel. Proc. Natl. Acad. Sci. USA. 2008; 105:15370–15375. [PubMed: 18832158] 36. Ujwal R, et al. The crystal structure of mouse VDAC1 at 2.3 Å resolution reveals mechanistic insights into metabolite gating. Proc. Natl. Acad. Sci. USA. 2008; 105:17742–17747. [PubMed: 18988731] 37. Kouyama T, et al. Crystal structure of the light-driven chloride pump halorhodopsin from Natronomonas pharaonis. J. Mol. Biol. 2010; 396:564–579. [PubMed: 19961859] 38. Gadsby DC, Vergani P, Csanády L. The ABC protein turned chloride channel whose failure causes cystic fibrosis. Nature. 2006; 440:477–483. [PubMed: 16554808] 39. Miller PS, Smart TG. Binding, activation and modulation of Cys-loop receptors. Trends Pharmacol. Sci. 2010; 31:161–174. [PubMed: 20096941] 40. Yang YD, et al. TMEM16A confers receptor-activated calcium-dependent chloride conductance. Nature. 2008; 455:1210–5. [PubMed: 18724360] 41. Caputo A, et al. TMEM16A, a membrane protein associated with calcium-dependent chloride channel activity. Science. 2008; 322:590–594. [PubMed: 18772398] 42. Schroeder BC, Cheng T, Jan YN, Jan LY. Expression cloning of TMEM16A as a calcium- activated chloride channel subunit. Cell. 2008; 134:1019–1029. [PubMed: 18805094] 43. Schroeder JI, Keller BU. Two types of anion channel currents in guard cells with distinct voltage regulation. Proc. Natl. Acad. Sci. USA. 1992; 89:5025–5029. [PubMed: 1375754] 44. Meyer S, et al. AtALMT12 represents an R-type anion channel required for stomatal movement in Arabidopsis guard cells. Plant J. Jul 12.2010 Epub ahead of print. 45. Sasaki T, et al. Closing plant stomata requires a homolog of an aluminum-activated malate transporter. Plant Cell Physiol. 2010; 51:354–65. [PubMed: 20154005] 46. Punta M, et al. Structural genomics target selection for the New York Consortium on Membrane Protein Structure. J. Struct. Funct. Genomics. 2009; 4:255–268. [PubMed: 19859826] 47. Papadopoulos JS, Agarwala R. COBALT: constraint-based alignment tool for multiple protein sequences. Bioinformatics. 2007; 23:1073–1079. [PubMed: 17332019] 48. Landau M, et al. ConSurf 2005: the projection of evolutionary conservation scores of residues on protein structures. Nucleic Acids Res. 2005; 33:W299–W302. [PubMed: 15980475] 49. Rocchia W, et al. Rapid grid-based construction of the molecular surface for both molecules and geometric objects: applications to the finite difference Poisson-Boltzmann method. J. Comp. Chem. 2002; 23:128–137. [PubMed: 11913378] 50. Altschul SF, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997; 25:3389–3402. [PubMed: 9254694] 51. Kendrick BS, Kerwin BA, Chang BS, Philo JS. Online size-exclusion high-performance liquid chromatography light scattering and differential refractometry methods to determine degree of polymer conjugation to proteins and protein-protein or protein-ligand association states. Anal. Biochem. 2001; 299:136–146. [PubMed: 11730335] 52. Otwinowski Z, Minor W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 1997; 276:307–326. 53. Pape P, Schneider TR. HKL2MAP: a graphical user interface for phasing with SHELX programs. J. Appl. Cryst. 2004; 37:843–844. 54. Perrakis A, Morris R, Lamzin VS. Automated protein model building combined with iterative structure refinement. Nat. Struct. Biol. 1999; 6:458–463. [PubMed: 10331874] 55. Emsley P, Cowtan K. COOT: model-building tools for molecular graphics. Acta Crystallogr. D. 2004; 60:2126–2132. [PubMed: 15572765] 56. CCP4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D. 1994; 50:760– 763. [PubMed: 15299374] 57. Fatt P, Ginsborg BL. The ionic requirements for the production of action potentials in crustacean muscle fibers. J. Physiol. 1958; 142:516–543. [PubMed: 13576452] Chen et al. Page 14 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 58. DeLano, WL. The PyMOL molecular graphics system. DeLano Scientific; San Carlos, CA, USA: 2002. http://www.pymol.org Chen et al. Page 15 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 1. Sequence analysis for the SLAC1 superfamily a, Family tree. The presentation was computed by the program COBALT47 from representative subfamily sequences (Table S1), including Arabidopsis thaliana SLAC1 for SF1Ai, Haemophilus influenzae TehA for SF1Ci, Escherichia coli TehA for SF1Cii, Vibrio parahaemolyticus for SF2A, Staphylococcus aureus for SF2B, Aspergillus fumigatus Ssu1 for SF3A, and Schizosaccharomyces pombe Mae1 for SF3B. b, Structure-based sequence alignment of TehA from Haemophilus influenzae (HiTehA) and SLAC1 from Arabidopsis thaliana (AtSLAC1). The TehA structure has been used to restrict sequence gaps to inter- helical segments. Superior coils define extents of the HiTehA helical segments; red letters mark residue identities; red boxes are drawn for residues that are >95% identical within the plant subfamily SF1A for AtSLAC1 or within the TehA subfamily for HiTehA; red diamonds mark HiTehA residues that line the central pore; and the colored inferior bar encodes ConSurf sequence variability48 for the SF1 family of 204 non-redundant proteins. Chen et al. Page 16 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 2. Crystal structure of HiTehA and homology model of AtSLAC1 a, Electron density distribution from the HiTehA crystal structure at 1.2Å resolution. The map has (2Fo-Fc) coefficients based on the superimposed model. Contours are at 2.5σ. b, Ribbon diagram of the HiTehA trimer. Each protomer is colored spectrally from blue at its N-terminus to red at its C-terminus. c, DelPhi49 electrostatic potential at the extracellular surface. Electronegative and electropositive potential are colored in degrees of red and blue saturation, respectively. d, Electrostatic potential49 at the intracellular surface. e, Ribbon diagram of an HiTehA protomer viewed from outside the membrane. The ribbon is colored spectrally as in 2b. f, Ribbon diagram of an HiTehA protomer viewed from within the membrane from below the view of e. g, Surface of a homology model of AtSLAC1, viewed as in f, and colored by sequence variability48. h, Surface of AtSLAC1 as in g, but colored by electrostatic potential49. Chen et al. Page 17 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 3. Putative structure of the SLAC1 conductance pore a, Cross-section through the homology model of AtSLAC1. The model is viewed as in 2i, with the electrostatic potential49 shown on the external surface of the molecular envelope. The side chain of Phe450 is shown as a stick model (red) on the backbone ribbon colored yellow. b, Cross-section as in a, but colored by surface conservation48 as in 2h. c, Pore- lining residues in the SLAC1 homology model. A cylinder model (left), spectrally colored as in 2b, provides a key for viewing the rolled-open structure (left) with pore-lining residues of AtSLAC1 shown on the TModd helices. d, Ribbon diagrams of HiTehA TM9 (left) and TM7 (right) viewed from within the conductance pore. The side chains of Pro207 and Phe262 are shown as well as a kink-stabilizing water HOH25 that is coordinated by the NH of Gly263 and by C=O groups of Gly202 and Ala259. Density contours are shown for the water molecule. Chen et al. Page 18 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 4. Ionic conductance measurements a, Typical microelectrode voltage-clamp current traces from oocytes injected with various channel cRNAs. Left column, oocytes injected with cRNAs encoding wild-type HiTehA channels (WT), or F262A, G15D/F262A or G15D mutants. Right two columns, oocytes injected with cRNAs for WT AtSLAC1, or F450A, G194D/F450A or G194D mutants, with or without co-injection of AtOST1. Dotted lines represent zero current levels. Extracellular solution contained 30mM CsCl. Schematic keys at the left place phenyl and/or aspartyl gates. b, Effects of gating residue mutations. Mean chloride currents, measured at −90 mV, are shown comparing WT HiTehA and its mutant series F262A,G,T,V,L with WT AtSLAC1 and its corresponding series F450A,G,T,V,L, both alone and co-expressed with AtOST1. Full I-V relations are shown in Fig. S5. c, Effect of gating residues on relative AtSLAC1 anion permeabilities. Relative permeabilities (P[X]/P[Cl]) for chloride, nitrate, sulfite and malate of WT, F450A and F450T SLAC1 channels were measured from the change in current reversal potential with Cl− or anion X− as the sole permeant anion in the bath solution (Methods, Table S6). Chen et al. Page 19 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 5. Structural features at the SLAC1 homolog gate a,b, Cross-sections through the conductance pores of HiTehA F262A and HiTehA G15D. The view and presentations are as in 3a, except that helices are colored purple. c, Molecular basis for conformational strain in gating residue Phe262 of HiTehA. Helices TM1 (left), TM9 (center) and TM7 (right) are viewed from within the pore and presented as ribbon diagrams with selected side chains drawn in stick representation. The local low-energy conformation for the phenyl ring (χ2 = 90°) is shown in thin lines with short contacts indicated in dashed lines: Leu18 Cδ2 – Phe262 Cδ1, d = 2.4 Å; Val210 Cγ2 – Phe262 Cε2, d = 2.8 Å. d, Conformational shifts consequent to release of strain in gating residue Phe262. Cα backbone structures of F262A, F262V and F262L HiTehA were first superimposed onto WT HiTehA, and residues 258-266 were drawn in stereo as superimposed. All backbone atoms are shown for peptides 262±1, but only Cα atoms are shown otherwise. The WT backbone and phenyl group are green; other backbone are all magenta; side chains of Ala262, Val262 and Leu262 are cyan, blue and red, respectively; oxygen-directed bonds are red and nitrogen-directed bonds are blue. e, Typical microelectrode voltage-clamp current traces from oocytes injected with wild-type (WT) or F450L AtSLAC1 cRNA. Experimental conditions and displays are as in 4a. Chen et al. Page 20 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript
3M79
A tetrameric Zn-bound cytochrome cb562 complex with covalently and non-covalently stabilized interfaces crystallized in the presence of Cu(II) and Zn(II)
Evolution of Metal Selectivity in Templated Protein Interfaces Jeffrey D. Brodin†, Annette Medina-Morales†, Thomas Ni†, Eric N. Salgado†, Xavier I. Ambroggio‡, and F. Akif Tezcan†,* † Department of Chemistry and Biochemistry, University of California, San Diego, La Jolla, CA 92093-0356 ‡ Rosetta Design Group LLC, Fairfax, VA 22030 Abstract Selective binding by metalloproteins to their cognate metal ions is essential to cellular survival. How proteins originally acquired the ability to selectively bind metals and evolved a diverse array of metal-centered functions despite the availability of only a few metal-coordinating functionalities remains an open question. Using a rational design approach (Metal-Templated Interface Redesign), we describe the transformation of a monomeric electron transfer protein, cytochrome cb562, into a tetrameric assembly (C96RIDC-1) that stably and selectively binds Zn2+, and displays a metal-dependent conformational change reminiscent of a signaling protein. A thorough analysis of the metal binding properties of C96RIDC-14 reveals that it can also stably harbor other divalent metals with affinities that rival (Ni2+) or even exceed (Cu2+) those of Zn2+ on a per site basis. Nevertheless, this analysis suggests that our templating strategy also introduces an increased bias towards binding a higher number of Zn2+ ions (4 high affinity sites) versus Cu2+ or Ni2+ (2 high affinity sites), ultimately leading to the exclusive selectivity of C96RIDC-14 for Zn2 over those ions. More generally, our results indicate that an initial metal-driven nucleation event followed by the formation of a stable protein architecture around the metal provides a straightforward path for generating structural and functional diversity. Introduction The incorporation of metal ions into correct cellular targets is a formidable chemical task. Modern-day organisms use a number of strategies to ensure that a metal ion associates with the right target (frequently a protein), including the control of absolute and relative ambient metal concentrations,1 active delivery via chaperones,2 and compartmentalization.3 Nevertheless, these strategies still require the target protein to possess an intrinsic affinity and selectivity for the desired metal ion. This is achieved within the 3-D framework of proteins despite the availability of only a handful of metal-coordinating side chain functionalities. An outstanding question is how protein structures have evolved to stably and selectively bind metal ions and developed metal-dependent functions, such as signal transduction (which necessitates conformational flexibility) and electron transfer or catalysis (which generally require rigid architectures). Some contemporary metalloproteins likely were based on pre-existing protein folds that acquired the ability to bind metals through random genetic events and subsequently attained their current structures and functions in the course of tezcan@ucsd.edu. Supporting Information Available. Additional experimental details, figures and tables on protein preparation, crystallography, analytical ultracentrifugation and binding assays. This material is available free of charge via the Internet at http://pubs.acs.org. NIH Public Access Author Manuscript J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. Published in final edited form as: J Am Chem Soc. 2010 June 30; 132(25): 8610–8617. doi:10.1021/ja910844n. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript natural selection.4 In an alternative pathway, metal ions could first have templated the formation of a protein/peptide aggregate, followed by the evolution of the protein structure around the metal ion.5 Although this pathway would not have the benefit of re-using existing genetic material and a preformed scaffold, it may lead to greater flexibility towards generating bioinorganic diversity. We have developed a rational design approach (Metal- Templated Interface Redesign, MeTIR), which follows the time course of this latter hypothetical pathway (Scheme 1).6 Using MeTIR, we describe here the construction of an oligomeric protein architecture– C96RIDC-14–that stably and selectively binds Zn2+, and displays a large, metal-dependent conformational change akin to signaling/regulatory proteins. Results and Discussion Design Strategy and Initial Characterization of C96RIDC-14 We have previously shown that variants (termed MBPC’s) of a four-helix-bundle heme protein, cytochrome cb562, with appropriately installed metal-binding motifs on their surfaces (represented by Species 2 in Scheme 1) can adopt discrete supramolecular architectures upon metal coordination.7–9 Because the parent cyt cb562 (1) is a monomeric protein whose surface carries no bias towards oligomerization, its supramolecular assembly is largely dictated by metal coordination. Accordingly, all MBPC assemblies initially formed through metal binding (3) feature interfacial metal ions with saturated coordination spheres and ligand arrangements that obey the preferred stereochemistry of the metal ions: Zn2+ – tetrahedral, Cu2+ – tetragonal, Ni2+ – octahedral.9 If the protein assembly surrounding a particular metal ion can be stabilized without changing the overall supramolecular architecture, one would expect the binding affinity and specificity to increase for that metal ion. We decided to put this “template-and-stabilize” strategy to the test using a tetrameric Zn-mediated assembly, Zn4:MBPC-14 (Figure 1). MBPC-1 (2) is a cyt cb562 variant with two i, i+4 bis-His metal-binding motifs (H59/H63 and H73/H77) installed on its Helix3. Upon Zn coordination MBPC-1 forms a D2 (222) symmetrical tetramer, Zn4:MBPC-14, which is held together by four identical Zn ions coordinated to one bis-His motif (73/77) from one protomer, His63 from a second, and Asp74 from a third.7 Owing to its twofold dihedral symmetry, Zn4:MBPC-14 presents three pairs of C2-symmetric interfaces (i1, i2, i3) between its four protomeric constituents. Of these interfaces, only i1 presents an extensive surface (>1000 Å2) with close protein-protein contacts. We therefore undertook computationally-guided redesign of i1 to examine if a favorable set of interactions can be built into i1 to stabilize the entire Zn-driven assembly (Step b in Scheme 1). A construct, RIDC-1, which features six mostly hydrophobic mutations (R34A/L38A/ Q41W/K42S/D66W/V69I) in i1, indeed was found to form a considerably stabilized tetrameric assembly (Zn4:RIDC-14, 4) with an identical supramolecular geometry to the parent tetramer.6 Despite this stabilization, Zn4:RIDC-14 (like Zn4:MBPC-14) remains a dynamically exchanging assembly that does not stay intact, for instance, upon passage through a size-exclusion column. Consequently, it has not been possible to uncouple protein oligomerization from Zn binding and directly assess whether interface redesign has led to improvements in Zn affinity and selectivity. In order to obtain a stable tetrameric complex that would also form in the absence of Zn, we sought to further engineer RIDC-1. As the dihedral symmetry of the Zn-tetramer dictates, the concurrent stabilization of any two of the three interfaces could in principle lead to the formation of a persistent tetrameric assembly. While interface i2 is not as tightly packed as i1 and therefore is less amenable to redesign, it presents position 96 from two protomers within distance for disulfide (SS) Brodin et al. Page 2 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript crosslinking (Figure 1). Thus, we combined the six mutations in i1 with the T96C mutation in i2 to generate C96RIDC-1. Under oxidizing conditions C96RIDC-1 readily forms an SS- crosslinked dimer as shown by SDS-PAGE electrophoresis (Figure S1). Sedimentation velocity (SV) experiments indicate that the predominant oligomeric form of C96RIDC-1 in solution is a tetramer (C96RIDC-14, 6), even in the absence of metals (Figure 2a). The tetramer-dimer dissociation constant (Kd(4mer-2mer)) for C96RIDC-14 has been determined by sedimentation equilibrium (SE) measurements to be <100 nM (Figure 2b), which to our knowledge makes it one of the most stable engineered protein complexes. Zn-dependent changes in C96RIDC-14 conformation C96RIDC-14 remains tetrameric upon Zn binding. SV population distributions indicate that the resulting tetramer (5) forms at significantly lower protein and Zn concentrations and therefore is more stable than its progenitors Zn4:MBPC-14 and Zn4:RIDC-14 (Figure 2a). SV measurements also suggest that C96RIDC-14 undergoes a Zn-induced rearrangement, evidenced by a shift in its sedimentation coefficient from 4.25 S to 4.5 S. To elucidate this conformational change, we solved the crystal structures of C96RIDC-14 and its Zn adduct at 2.1 and 2.4-Å resolution, respectively (PDB ID’s 3IQ5 and 3IQ6). As illustrated in Figure 3, these structures reveal a remarkable double-clothespin motion of the four protomers upon Zn-coordination, measuring ~16 Å at the N-terminus of Helix3. One of the keys to the simultaneous stability and conformational plasticity of C96RIDC-14 lies with the redesigned interface i1. Each of two equivalent i1 interfaces in C96RIDC-14 is formed between two protomers oriented in an antiparallel fashion. These interfaces feature an extensive network of hydrophobic interactions (~1300 Å2 buried surface), the main contributors being the two pairs of engineered Trp (41 and 66) and His (59 and 77) residues (Figure S3). Not surprisingly, this hydrophobic interface was predicted and found to be rather fluid.6 Upon binding Zn, the fluidity of the engineered hydrophobic interactions allows the four protomers to pivot around i1 and undergo the double-clothespin motion. The resulting architecture of Zn4:C96RIDC-14 features a well-packed hydrophobic core in i1 (~1400 Å2 buried surface) built around the engineered residues originally proposed by computation. Zn4:C96RIDC-14 is superimposable onto both Zn4:MBPC-14 and Zn4:RIDC-14 structures with respective root-mean-square deviations of 0.59 and 0.63 Å over 424 Cα’s (Figure S4). Importantly, the four equivalent Zn coordination environments remain unchanged as intended by the template-and-stabilize strategy. The interfacial SS bonds incorporated into i2 are the second key component for the bistability of C96RIDC-14. The electron density maps for C96RIDC-14 and Zn4:C96RIDC-14 clearly outline the C96-C96′ linkages, which are found in distinct conformations to accommodate the two different supramolecular arrangements (Figure 3). Using the SS- bonds as “loose hinges”, the pairs of protomers that share i2 undergo a significant translational and rotational motion relative to one another, while maintaining the ideal bond distance (2.05 Å) and the stereochemical requirements for an SS bond. Operating together, the fluid non-covalent interactions in i1 and the adaptable SS bonds in i2 allow the formation of two stable and interconvertible tetrameric architectures in the absence or presence of Zn. Zn binding by C96RIDC-14 Having thus uncoupled protein oligomerization from metal binding, we examined if our interface templating strategy leads to increased Zn affinity. Because Zn is spectroscopically silent, its binding to C96RIDC-14 was assessed through two different indirect methods. In the first method modeled after a procedure used for determining uranyl binding to NikR,10 we used nitrilotriacetic acid (NTA) as a competing Zn ligand (Kd(Zn-NTA) = 18.2 nM at 22°C Brodin et al. Page 3 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript and μ = 0.15 M) and 4-(2-pyridylazo)resorcinol (PAR) as an indicator, which exhibits an increase in absorbance at 500 nm upon binding Zn.11 Briefly, increasing amounts of NTA were added to samples of nearly equimolar Zn and C96RIDC-14. The resulting mixtures were subjected to centrifugation in 10-kDa-cutoff protein concentrators, which allowed the separation of the Zn fraction bound to C96RIDC-14 from that bound to NTA at each NTA concentration. This, in turn, permitted the calculation of the extent of Zn binding to C96RIDC-14 versus free Zn concentration, which was inferred from the Zn-NTA binding equilibrium. The NTA titrations are reasonably well described by an n-equivalent binding sites model, and indicate that each C96RIDC-14 complex binds four Zn equivalents with an apparent dissociation constant of 3.3 nM (Figure 4a, see Figure S9 for fits to alternative binding models). In a second, “higher-resolution” method that obviates the separation step and the secondary indicator, we used the chromophore Fura-2 (Kd, Zn = 5.7 nM)12 as a competing Zn ligand/ indicator in a similar fashion to previously published procedures.13 In these experiments, increasing amounts of Zn were added to samples that contained C96RIDC-14 and Fura-2, and the Zn-dependent changes in the Fura-2 spectrum were directly monitored (Figure 4b). These titrations reveal that Zn binding to C96RIDC-14 is best described by four individual, consecutive binding equilibria or nearly equally well by two consecutive binding events by pairs of Zn2+ ions, with corresponding dissociation constants that range from 0.5 nM to 60 nM in the case of the 1+1+1+1 model and from 0.5 to 40 nM in the case of the 2+2 model (Table 1).14 Owing primarily to its flexibility, C96RIDC-14 apparently can accommodate Zn binding through several different modes; in other words, the intermediate Zn bound states may and most likely do utilize different ligand sets from one another. In any case, given that i, i+4 bis-His motifs on α-helices display Zn dissociation constants in the low μM range,15 the Zn binding titrations suggest that the pre-formation of a tetrameric, templated acceptor complex results in a ≥1000-fold increase in Zn-binding affinity relative to the monomeric parent species, MBPC-1. Zn binding selectivity of C96RIDC-14 over other divalent metal ions To elucidate if C96RIDC-14 also displays increased selectivity for Zn binding, its interactions with several other divalent metal ions (M2+) were examined, including the neighboring Co2+, Ni2+ and Cu2+, which typically are effective competitors for Zn binding sites. To this end, C96RIDC-14 was incubated with the metal ion of interest in a noncoordinating buffer solution (20 mM 3-(N-morpholino)propanesulfonic acid, MOPS), followed by the separation of the C96RIDC-14-metal complex via gel filtration and subsequent metal analysis by ICP-OES. These experiments show that C96RIDC-14 retains ~1 equivalent of Co2+, ~2 equivalents of Ni2+ and ~4 equivalents of Cu2+ (Figure S5a). In competition studies, the initial C96RIDC-14/metal mixture was additionally incubated with Zn at various metal/Zn ratios prior to gel filtration and metal analysis. Each competition experiment was also carried out in reverse order - incubation with Zn followed by addition of other metals - to ascertain the formation of thermodynamic products (Figure S5b). These experiments reveal that C96RIDC-14 displays significant Zn selectivity over all ions except Cu2+, especially relative to its parent structure, MBPC-1 (Figures 5, S5b and S7). Previous studies have shown the affinity of the i, i+4 bis-His motif for Zn2+ to be comparable to that for Ni2+ and 5–10 fold higher than that for Co2+,15,16 following the Irving-Williams(IW) series.17 In contrast, Zn2+ completely outcompetes Co2+ for C96RIDC-14 binding at all ratios measured (up to 100 Co:1 Zn), and has an effective affinity roughly 100-fold higher than Ni2+. Brodin et al. Page 4 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Cu2+ presents a special case in terms of Zn selectivity. Due to a combination of its d9 configuration and high Lewis acidity, Cu2+ is situated at the top of the IW series, leading to its higher affinity for most ligand platforms designed for specific Zn binding13,18 and even natural Zn enzymes.19 The results shown in Figure 5 initially indicated that neither Cu2+ nor Zn2+ outcompete each other for C96RIDC-14 binding; rather, each tetrameric C96RIDC-14 unit appeared to bind ~3 equivalents of each ion in the non- coordinating MOPS buffer solution. To ascertain whether this apparent oversaturation of C96RIDC-14 is due to the binding of Cu2+ or Zn2+ ions to the C96RIDC-14 surface (in addition to the core binding sites) or due to the formation of an actual Cu-Zn heterometallic core species, we obtained crystals of C96RIDC-14 grown in the presence of equimolar amounts of both ions. The resulting 2.1-Å resolution diffraction data reveal a structure (PDB ID: 3M79) identical to that of Zn4:C96RIDC-14. To identify the four observed interfacial metal ions in this complex, we collected full data sets at the Zn and Cu K edges (1.28 Å and 1.38 Å), respectively. The corresponding anomalous difference maps clearly show that the interfacial ions are Zn2+, and that there are no detectable Cu2+ ions associated with the core or the surface of the tetramer (Figures 6a and b), despite the fact the crystal clearly contained Cu as indicated by an X-ray fluorescence excitation scan (Figure 6c). These observations suggest that Zn2+ outcompetes Cu2+ completely for binding to the core sites. This is further supported by the finding that when the Zn-Cu competition experiments are carried out in a weakly coordinating buffer solution (20 mM Tris(hydroxymethyl)aminomethane, TRIS), the amount of Cu2+ associated with C96RIDC-14 is significantly diminished, whereas the amount of bound Zn2+ stays constant (see the last two rightmost bars in Figure 5). To describe the Zn selectivity of C96RIDC-14 in a more quantitative fashion, we examined its affinity for Co2+, Ni2+ and Cu2+, again using Fura-2 as a competing ligand (Kd, Fura-Co = 8.6 nM,20 Kd, Fura-Ni = 6.9 nM,12 Kd, Fura-Cu = 0.3 pM12). The titrations indicate that C96RIDC-14 has one weak Co2+ binding site that barely competes with Fura-2 (Figure S11 and Table 1), whereas it can accommodate two equivalents of Ni2+ and Cu2+ with comparable affinities to Fura-2 (Figure 7 and Table 1).21 Both Ni2+ and Cu2+ binding curves are well described by two consecutive binding equilibria. Although on a per site basis the derived affinities are either similar to those for Zn2+ (in the case of Ni2+) or considerably higher (in the case of Cu2+) in line with the IW series, the higher multiplicity for Zn2+ binding apparently results in a more favorable overall free energy (Table 1), and ultimately in the Zn selectivity of C96RIDC-14 over these ions.21 Thus, templated interface engineering leads to increased bias not only towards Zn coordination geometry but also towards Zn binding multiplicity, which, to our knowledge, is a rare, and perhaps unique, case in designed/synthetic systems. Studies are currently underway to elucidate the coordination modes/environments of Co2+, Ni2+ and Cu2+ to C96RIDC-14 (which we expect to be different from each other and from Zn2+) and associated changes in protein structure. Conclusions Metal-templated synthesis has been a powerful approach to construct ligands with enforced coordination geometries that provide stable and specific metal binding.22–24 We have shown here that such templating strategies used for smaller systems can also be applied to proteins, which, owing to their extensive surfaces rich in chemical functionality, allow the formation of an extensive network of covalent and non-covalent interactions around the template. Importantly, the templating of protein interfaces around the target metal ion, Zn2+, has led to the evolution of a flexible, multi-14 stable protein complex that not only presents an increased bias towards the tetrahedral Zn2+ coordination geometry, but also towards the Zn2+ binding multiplicity of four, ultimately resulting in significantly increased selectivity over other divalent ions including Cu2+. It is well established that Cu2+ can readily outcompete Zn2+ (or any other physiologically important metal ion) for binding even rigid, Brodin et al. Page 5 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript tetrahedral coordination environments that favor Zn2+.19 Such thermodynamic dominance by the Cu2+ ion indeed necessitates various cellular strategies – often operating under kinetic control – to be employed for the incorporation of other metal ions into their intended protein targets.25 Our findings suggest that control of metal binding multiplicity may be a viable thermodynamic strategy alongside the design and restraint of the inner-sphere coordination environment to enhance metal selectivity in natural or synthetic systems. Our study further demonstrates that cyt cb562 – a monomeric, putative electron transfer protein – can be transformed into a Zn-responsive complex through a minimal number of mutations that amount to less than 10% of its amino acid sequence. It is conceptually straightforward to envision how the Zn4:C96RIDC-14 architecture can be further rigidified and modified around the Zn centers to promote metal-based functions such as Lewis acid catalysis. While the role of metal templating in the evolution of metalloproteins can only be postulated, MeTIR clearly provides a practical route to generating structural and functional diversity. It remains to be seen if this approach can also be extended to non-metallic substrates, which could have served as nucleants in the early emergence of protein folds. 26,27 Experimental Section Site-Directed Mutagenesis and Protein Expression/Purification The T96C mutation was introduced into pET-ridc16 using QuikChange (Stratagene) site- directed mutagenesis and primers obtained from Integrated DNA technologies, yielding the expression vector pET-C96ridc1. pET-C96ridc1 was transformed into XL-1 blue E. coli cells, purified using the QIAprep Spin Miniprep Kit (QIAGEN) and sequenced by Retrogen. pET-C96ridc1 was transformed into BL21(DE3) E. coli cells along with the ccm heme maturation gene cassette plasmid, pEC86.28 Cells were plated on LB agar containing 100 μg/ml ampicillin and 34 μg/ml chloramphenicol and grown overnight. 3 mL starter cultures were inoculated from the resulting colonies, grown to an Abs600 of 0.6 and used to inoculate 1 L of LB medium. 1-L cultures were then incubated for 16 hours with rotary shaking at 250 rpm. No induction was necessary. Protein was obtained by sonicating cells in the presence of lysozyme, bringing the lysate to pH 5 with HCl and isolating the soluble fraction by centrifugation at 16,000 g, 4° C, for 1 hr. Initial purification was by cation-exchange chromatography on a CM-Sepharose matrix (Amersham Biosciences) using an NaCl step gradient in sodium acetate (pH 5). After dialysis into sodium phosphate (pH 8), the protein was further purified by anion exchange on an Uno-Q (BioRad) column using a DuoFlow chromatography workstation (BioRad) and a linear NaCl gradient. A protein sample was then exchanged into water, mixed 1:1 with sinapinic acid matrix (Agilent Technologies) and subjected to MALDI mass spectrometry to verify the T96C mutation (expected mass 12305 amu, observed 12299 amu). Dimeric C96RIDC-12 was separated from monomeric protein on a preparative scale size exclusion chromatography column (GE Healthcare) packed with Superdex 75 (GE Healthcare) resin equilibrated in 20 mM Tris HCl (pH 7) and 150 mM NaCl. Separation of dimer (which effectively is a tetramer at concentrations > 1 μM) from monomer was verified by non-reducing SDS-PAGE gel electrophoresis using a 15% acrylamide gel. Typical protein preparations yielded 8–10 mg of protein per liter of culture based on an extinction coefficient of 148,000 M−1cm−1 at 415 nm. After purification, dimeric C96RIDC-12 was concentrated to ~2 mM using an Amicon stirred cell (Millipore), flash frozen in liquid nitrogen and stored at −80° C. Brodin et al. Page 6 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Analytical Ultracentrifugation SV and SE samples were prepared in 20 mM Tris (pH 7) and 150 mM NaCl using appropriate volumes of 5 mM metal stock solutions or 50 mM EDTA. After incubating in the presence of M2+ or EDTA for 1 hour, SV measurements were carried out at 25° C on a Beckman XL-I Analytical Ultracentrifuge using an An-60 Ti rotor at 41,000 rpm for a total of 250 scans per sample. The following wavelengths were used to monitor C96RIDC-12 sedimentation at different protein concentrations: 415 nm (5 μM), 440 nm (30 μM). All data were processed using SEDFIT29 software with the following fixed parameters: buffer density (r) = 1.0049 g/ml; bufferviscosity = 0.010214 poise; Vbar = 0.73084. SE measurements were carried out at 25° C using speeds between 10,000 and 20,000 rpm. Scans were taken at 14 and 16 hours and visually inspected to verify that sedimentation equilibrium was achieved. The following wavelengths were used to monitor C96RIDC-12 sedimentation at different protein concentrations: 415 nm (1 μM), 420 nm (2.5 μM), and 500 nm (12.5 μM). 16-hr scans were fit to a monomer-dimer (where monomer = C96RIDC-12) or a dimer-only model using SEDPHAT.30 The molecular mass of C96RIDC-12 (24610 Da) and the menisci were fixed while floating the association constant. Standard deviation for the resulting log10(K) value was determined through Monte-Carlo analysis within SEDPHAT.30 Crystallography All crystals were grown by sitting drop vapor diffusion at room temperature (20–25° C) in drops consisting of 2 μL of protein and 1 μL of precipitant solution. For apo and Zn2+ crystals, a 2.1 mM protein stock solution in 20 mM Tris (pH 7) and 150 mM NaCl was used. The precipitant solution for apo-C96RIDC-14 crystals consisted of 100 mM Bis-Tris (pH 6.5) and 30% PEG 400. The precipitant solution for Zn4:C96RIDC-14 crystals was 100 mM Tris (pH 7.5), 20% PEG 2000 and 2.5 mM ZnCl2. Crystallization of C96RIDC-14 in the presence of Zn and Cu (Zn/CuC96RIDC-14) was performed by incubating 60 μM protein with one molar equivalent of Zn for one hour followed by addition of one equivalent of Cu and incubation overnight to ensure the formation of the thermodynamic product. The mixture was then concentrated to 2.6 mM using 4 mL Amicon Ultra (Millipore) centrifugal filters. Crystallization was performed as above with a precipitant solution consisting of 24% PEG 2000 and 100 mM Bis-Tris (pH 6.5). Crystals used for diffraction were exchanged stepwise into a solution containing 20% glycerol as a cryoprotectant and frozen in liquid nitrogen or directly in the cryostream. Diffraction data were collected at the Stanford Synchrotron Radiation Laboratory (SSRL) Beamline 9-2 for C96RIDC-14 and Zn4:C96RIDC-14 or Beamline 7-1 for Zn/CuC96RIDC-14 at 100 K. 1.0-Å radiation datasets were collected for all crystals and additional datasets were collected using 1.28-Å and 1.38-Å radiation, corresponding to Zn and Cu K edges, respectively, for Zn/CuC96RIDC-14. The data were processed using MOSFLM and SCALA. 31 The structures of C96RIDC-14, Zn4:C96RIDC-14 and Zn/CuC96RIDC-14 were determined at 2.05, 2.35 and 2.1 Å resolution, respectively, by molecular replacement with MOLREP,32 using the RIDC-1 monomeric structure (PDB ID: 3HNI) as the search model. Rigid-body, positional and thermal refinement with CNS33 or REFMAC,34 along with manual model rebuilding and water/ligand placement with XFIT35 or COOT36 produced the final models. For all structures, non-crystallographic symmetry (NCS) restraints (tight main-chain and medium side-chain restraints) were applied throughout the positional/thermal refinement process. The Ramachandran plots were calculated with PROCHECK.37 All figures were produced with PYMOL (www.pymol.org). Data collection and refinement statistics are summarized in Table S1. Brodin et al. Page 7 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Inductively-coupled plasma-optical emission (ICP-OES) spectroscopy ICP-OES samples were in 20 mM MOPS (pH 7) and 150 mM NaCl or 20 mM Tris (pH 7) and 150 mM NaCl. To determine the stoichiometry of metal binding to C96RIDC-14, 60 μM protein was incubated with one or two molar equivalents of the desired metal ion. For competition studies, 60 μM protein was incubated with one molar equivalent of Zn2+ for one hour, after which, one, ten or one hundred molar equivalents of competing metal (M2+) were added, bringing the final volume to 1 ml. The same experiments were also done in reverse order. All samples were allowed to equilibrate overnight at room temperature and free/ loosely bound metal was subsequently removed using a 10DG gel filtration column (Bio- Rad). Samples were prepared for ICP-OES by diluting 1.3 mL of protein solution collected off the 10DG column to a final volume of 2 ml and adding 90 μl of 69% reagent grade nitric acid (Fluka) to achieve a final concentration of 3%. Standards were prepared from 1000 ppm certified ICP-OES metal stock solutions (Ricca) by mixing equal volumes of all metal analytes and diluting to a final concentration of 200 ppm of each metal. A standard curve with eleven points between 0.05 and 10 ppm was then constructed by diluting appropriate volumes of the 200-ppm stock to 10 ml with 3% nitric acid in deionized water. Data were collected on a Perkin-Elmer Optima 3000 DV ICP-OES spectrometer located at the Analytical Facility of the Scripps Institute of Oceanography. Because each C96RIDC-1 monomer contains one Fe atom as part of the covalently linked heme molecule, the experimentally determined M(II):Fe ratios directly yielded the M2+:protein ratios. Wavelengths used for the detection of various metal ions were as follows: Mg (279.077, 280.271 and 285.213 nm), Ca (315.887 and 317.993), Fe (234.349, 238.204, 239.562, 259.939 and 273.055 nm), Co (228.616 and 238.892 nm), Ni (221.648 and 231.604 nm), Cu (224.7, 222.778, 221.459, 327.393 and 324.752 nm), and Zn (202.548, 206.2 and 213.857 nm). Values reported for each metal are averages of those for all wavelengths indicated. Competitive binding assays using nitrilotriacetic acid (NTA) The stability constant for Zn2+ binding to C96RIDC-14 was determined based on a previously published protocol for the determination of uranyl binding to NikR,10 with the exception that NTA was substituted as the competing ligand. Titrations were performed in 20 mM MOPS (pH 7) and 150 mM NaCl at 22° C. Trace metal was removed from all buffers by passage through a Chelex 100 (Bio-Rad) column. For each titration point, 22 μM protein was mixed with 28 μM of Zn and variable amounts of NTA. Samples were allowed to equilibrate overnight at room temperature and subsequently centrifuged in 0.5 mL Amicon Ultra (Millipore) protein concentrators (10 KDa MW cutoff) for ten minutes at 2000 rpm to allow free Zn, free NTA and NTA:Zn to pass through the membrane. Zn concentrations were determined for the flow-through and protein chamber at each titration point using 4-(2-pyridylazo)resorcinol (PAR), which exhibits an increase in aborbance at 500 nm upon binding Zn (Δε500 = 59100 M−1cm−1 in buffer and 54800 M−1cm−1 in 5 M guanidine hydrochloride), essentially as previously described.11 1800 μL of 1 mM PAR was added to a 1-cm pathlength cuvette, and the initial absorbance spectrum was recorded. Subsequently, 100 ml of flow-through was added, allowed to stir for a minimum of five minutes to ensure that equilibrium had been reached, and absorbance at 500 nm was again recorded. Metal content in the protein chamber was determined similarly to flow-through samples, except that 1800 μl of 5 M buffered guanidine hydrochloride was used to denature protein and allow the complete release of bound metal for measurement. Absorbance at 500 nm was corrected for dilution and protein absorbance, and Zn concentrations determined using the Δε500 values listed above. The fraction of protein bound to metal as a function of competing ligand concentration was determined by the following equation: Brodin et al. Page 8 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript where [M]P.C. is the concentration of metal in the protein chamber, [M]F.T. is the concentration of metal in the flow-through and [P]T is the total concentration of protein measured after centrifugation. For calculating dissociation constants, the concentration of free metal was determined based on the concentration and dissociation constant of the competing ligand (NTA) and metal in the flow-through using the program MaxChelator (http://maxchelator.stanford.edu). The data were fit to the following models38 using Igor Pro v. 6.02a (Wavemetrics, Inc.). (1) (2) (3) where B is the molar equivalents of metal bound per C96RIDC-14, n is the number of binding sites, [M] is the concentration of free metal and K1–4 are stoichiometric association constants. Equation (1) assumes four identical binding sites with invariant affinities. Equation (2) assumes two stepwise binding events each of which involves binding of two Zn2+. Equation (3) assumes four stepwise binding events. Binding affinity of Fura-2 for Ni2+, Cu2+ and Zn2+ 1 mg of lyophilized Fura-2 (Invitrogen) was suspended in 1 mL of deionized water and its concentration was determined based on a published extinction coefficient of 27,000 M−1cm−1 at 362 nm.39 EGTA competition binding assays were used to determine dissociation constants for Ni2+, Cu2+ and Zn2+ using the following EGTA:M2+ logK values obtained from the online program MaxChelator (http://maxchelator.stanford.edu): Cu2+:EGTA − 13.2, Ni2+ − 9.0, Zn2+ − 8.1. These values are corrected for pH, temperature and ionic strength. In a 3 mL quartz cuvette, 11 μM Fura-2 was mixed with 100 μM EGTA in 20 mM MOPS (pH 7) and 150 mM NaCl to a final volume of 2 mL at 22° C. M2+ was titrated into the solution with an equilibration time of 10 minutes at ambient conditions between each addition, and absorbance spectra were recorded from 190–800 nm using a Hewlett Packard 8452A diode array spectrophotometer (Figure S10). For Cu2+, a separate titration into a solution of 100 μM buffered EGTA was conducted and subtracted from the Fura-2 titration points to account for absorbance from aqueous Cu2+ and the Cu2+:EGTA complex. Absorbance and M2+ concentrations were corrected for dilution and data was fit to a 1:1 M2+:Fura-2 model using DynaFit40 scripts (Figure S10) modeling the competition between EGTA and Fura-2 for M2+. The Zn2+:Fura-2 value determined by this method (logK = 8.2) compares favorably with a value previously determined at pH 7.15 (logK = Brodin et al. Page 9 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 8.5).41 The logK values determined for the Fura-2 complexes of Cu2+ and Ni2+ complexes are 12.5 and 8.2, respectively. Competitive binding assays using Fura-2 Competition assays using the Fura-2:M2+ dissociation constants determined above were used to measure C96RIDC-14:M2+ affinities. In a 3-mL quartz cuvette, a stock solution of C96RIDC-14 was diluted to 7.5 μM in 20 mM MOPS (pH 7) and 150 mM NaCl at 22° C and absorbance was recorded. Fura-2 was added to a final concentration of 10 μM and absorbance again recorded. Stock metal solutions were then titrated into the Fura-2/C96RIDC-14 mixture in ~2 μM steps, allowed to equilibrate with stirring for 10 minutes at ambient conditions and absorbance recorded. Protein and metal concentrations were corrected for dilution, protein absorbance was subtracted from total absorbance and plots of Abs373 versus total metal added were fit to various models using custom DynaFit scripts (see Supporting Information). Co titrations were performed similarly to Zn, Ni and Cu, but with the following modifications due to its lower affinity for C96RIDC-14: C96RIDC-14 concentration was increased to 21 μM and the Fura-2 concentration was decreased to 8 μM. Also, titrations were performed in a 0.5 cm pathlength cuvette to allow for the use of a higher protein concentration without saturating the detector. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We thank Prof. Thomas O’Halloran for helpful discussions. Support for this work was provided by NIH (training grants to J.D.B. and A.M.M.; predoctoral fellowship to E.N.S.), NSF (CHE-0908115, structural work), DOE (DE- FG02-10ER46677, competitive metal binding titrations) and the Arnold and Mabel Beckman Foundation (F.A.T.). Portions of this research were carried out at SSRL, operated by Stanford University on behalf of DOE. References 1. Outten CE, O’Halloran TV. Science. 2001; 292:2488. [PubMed: 11397910] 2. Rosenzweig AC. Acc Chem Res. 2001; 34:119. [PubMed: 11263870] 3. Tottey S, Waldron KJ, Firbank SJ, Reale B, Bessant C, Sato K, Cheek TR, Gray J, Banfield MJ, Dennison C, Robinson NJ. Nature. 2008; 455:1138. [PubMed: 18948958] 4. Blundell, TL. The evolution of metal-binding sites in proteins. Symposium Press London; University of Sussex: 1977. 5. Liu CL, Xu HB. J Inorg Biochem. 2002; 88:77. [PubMed: 11750028] 6. Salgado EN, Ambroggio XI, Brodin JD, Lewis RA, Kuhlman B, Tezcan FA. Proc Natl Acad Sci USA. 2010; 107:1827. [PubMed: 20080561] 7. Salgado EN, Faraone-Mennella J, Tezcan FA. J Am Chem Soc. 2007; 129:13374. [PubMed: 17929927] 8. Salgado EN, Lewis RA, Faraone-Mennella J, Tezcan FA. J Am Chem Soc. 2008; 130:6082. [PubMed: 18422313] 9. Salgado EN, Lewis RA, Mossin S, Rheingold AL, Tezcan FA. Inorg Chem. 2009; 48:2726. [PubMed: 19267481] 10. Wegner SV, Boyaci H, Chen H, Jensen MP, He C. Angew Chem Intl Ed Eng. 2009; 48:2339. 11. McCall KA, Fierke CA. Anal Biochem. 2000; 284:307. [PubMed: 10964414] 12. The dissociation constants of Fura-2 complexes with Zn2+, Cu2+ and Ni2+ were determined separately using EGTA as a competing ligand. See Experimental Section for more details. 13. Walkup GK, Imperiali B. J Am Chem Soc. 1997; 119:3443. Brodin et al. Page 10 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 14. With the data in hand, we consider the 2+2 model to be more plausible than the 1+1+1+1 model, which would require a 3-Zn bound intermediate state that is not likely (though not impossible) to form within the D2 symmetry of the C96RIDC-14 scaffold. 15. Krantz BA, Sosnick TR. Nat Struct Biol. 2001; 8:1042. [PubMed: 11694889] 16. Ghadiri MR, Choi C. J Am Chem Soc. 1990; 112:1630. 17. Frausto da Silva, JJR.; Williams, RJP. The biological chemistry of the elements. Oxford University Press; Oxford: 2001. 18. Nolan EM, Ryu JW, Jaworski J, Feazell RP, Sheng M, Lippard SJ. J Am Chem Soc. 2006; 128:15517. [PubMed: 17132019] 19. Hunt JA, Ahmed M, Fierke CA. Biochemistry. 1999; 38:9054. [PubMed: 10413479] 20. Kwan CY, Putney JW. J Biol Chem. 1990; 265:678. [PubMed: 2404009] 21. As previously mentioned, ICP-OES results indicate two high-affinity binding sites for Ni and four for Cu on C96RIDC-14. The two Cu binding sites/modes not observed in the Fura-2 competition experiments may either be less tightly bound to the core or to the surface of C96RIDC-14. If they are bound to the core in a fashion that would compete with Zn binding, their average dissociation constants would have to be ~40 μM each in order to compensate for the free energy difference between the four-Zn-bound and two-Cu-bound species listed in Table 1. 22. Thompson MC, Busch DH. J Am Chem Soc. 1964; 86:3651. 23. Creaser II, Geue RJ, Harrowfield JM, Herlt AJ, Sargeson AM, Snow MR, Springborg J. J Am Chem Soc. 1982; 104:6016. 24. McMurry TJ, Raymond KN, Smith PH. Science. 1989; 244:938. [PubMed: 2658057] 25. Waldron KJ, Robinson NJ. Nature Rev Microbiol. 2009; 7:25. [PubMed: 19079350] 26. Pande VS, Grosberg AY, Tanaka T. Proc Natl Acad Sci USA. 1994; 91:12976. [PubMed: 7809158] 27. Saito S, Sasai M, Yomo T. Proc Natl Acad Sci USA. 1997; 94:11324. [PubMed: 9326608] 28. Braun M, Thony-Meyer L. Proc Natl Acad Sci USA. 2004; 101:12830. [PubMed: 15328415] 29. Schuck P. Biophys Chem. 2004; 108:187. [PubMed: 15043929] 30. Vistica J, Dam J, Balbo A, Yikilmaz E, Mariuzza RA, Rouault TA, Schuck P. Anal Biochem. 2004; 326:234. [PubMed: 15003564] 31. Collaborative Computational Project, Number 4. The CCP4 Suite: Programs for Protein Crystallography. Acta Cryst. 1994; D50:760–763. 32. Vagin A, Teplyakov A. J Appl Cryst. 1998; 30:1022. 33. Brünger AT, Adams PD, Clore GM, DeLano WL, Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges M, Pannu NS, Read RJ, Rice LM, Simonson T, Warren GL. Acta Crystallogr D. 1998; 54:905. [PubMed: 9757107] 34. Murshudov G, Vagin A, Dodson E. Acta Cryst. 1996; D53:240. 35. McRee DE. J Mol Graphics. 1992; 10:44. 36. Emsley P, Cowtan K. Acta Cryst. 2004; D60:2126. 37. Laskowski RA, Macarthur MW, Moss DS, Thornton JM. J Appl Crystallogr. 1993; 26:283. 38. Klotz, IM. Ligand-receptor energetics: a guide for the perplexed. John Wiley & Sons, Inc; New York: 1997. 39. Grynkiewicz G, Poenie M, Tsien RY. J Biol Chem. 1985; 260:3440. [PubMed: 3838314] 40. Kuzmic P. Anal Biochem. 1996; 237:260. [PubMed: 8660575] 41. Atar D, Backx PH, Appel MM, Gao WD, Marban E. J Biol Chem. 1995; 270:2473. [PubMed: 7852308] Brodin et al. Page 11 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Interprotomeric interfaces in the D2-symmetric Zn4:MBPC-14 complex. Residues subjected to redesign as well as those that coordinate Zn2+ ions are shown as sticks. Brodin et al. Page 12 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. (a) Sedimentation coefficient distributions for various cyt cb562 constructs at 5 μM monomeric concentration and equimolar Zn2+ where indicated. For MBPC-1 and RIDC-1, the Zn-induced tetrameric species are fully populated at >1 mM and >20 μM protein and Zn, respectively. At 5 μM protein and Zn, MBPC-1 is still predominantly monomeric (Smax= 1.8), while RIDC-1 is a mixture of dimeric (Smax= 2.8) and tetrameric (Smax= 4.5) forms. (b) Sedimentation equilibrium profile for 2.5 μM C96RIDC-12 obtained at 20000 rpm (see Figure S2 for other concentrations and speeds). The sedimentation data are equally well described by a dimer-tetramer equilibrium (Kd = 52 nM) or a tetramer-only model, which suggests the tetramer dissociation constant to be <100 nM. Brodin et al. Page 13 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Crystal structures of C96RIDC-14 (left) and Zn4:C96RIDC-14 (right). Redesigned residues in i1 and i2 are shown as sticks. Interfacial SS-bond configurations are shown below each structure along with corresponding Fo – Fc omit difference maps (cyan mesh − 4.5 σ (apo), 3 σ (Zn); purple mesh − 11 σ (apo), 5 σ (Zn)). See Figure S4 for detailed views of interfacial residues. Brodin et al. Page 14 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. (a) Zn-binding isotherm of C96RIDC-14 determined using NTA as a competing ligand. (b) Zn-binding isotherm for Fura-2–C96RIDC-14 competition experiments; corresponding changes in the Fura-2 absorbance spectrum are shown in the inset. The data are corrected for dilution and background absorbance by the protein. The sample contained 7.5 μM C96RIDC-14 and 11 μM Fura-2. The tick marks shown on the top x-axis correspond to theoretical endpoints for titration if C96RIDC-14 bound to one, two, three or four equivalents of Zn. The fits obtained using DynaFit are shown for the following different models: solid line, four consecutive Zn binding equilibria (1+1+1+1); dashed line, two consecutive binding equilibria (2+2); dotted line; single binding equilibrium (4×1). Equilibrium constants obtained with these different models are listed in Table 1. Brodin et al. Page 15 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Extent of divalent metal ion binding to C96RIDC-14 in competition experiments. Brodin et al. Page 16 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. Anomalous difference maps (4σ) for the C96RIDC-14 structures grown in the presence of equimolar Cu2+ and Zn2+ obtained at the Zn (a) or Cu (b) K edges. As expected, the heme Fe centers show anomalous signals at both wavelengths, whereas the core metal sites do the same only at the higher energy Zn edge, unambiguously identifying them as Zn ions. (c) X- ray fluorescence excitation scans of the same crystal – which was thoroughly washed with non-metal containing solutions – indicate the presence of both Zn and Cu. Brodin et al. Page 17 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 7. (a) Cu2+ and (b) Ni2+ binding isotherms for Fura-2–C96RIDC-14 competition experiments; corresponding changes in the Fura-2 absorbance spectrum are shown in the inset. The samples contained 7.5 μM C96RIDC-14 and 11 μM Fura-2. The tick marks shown on the top x-axis correspond to theoretical endpoints for titration if C96RIDC-14 bound to one, two, three or four equivalents of metal. The fits obtained using DynaFit are shown for the following different models: solid line, two consecutive binding equilibria (1+1); dashed line, single binding equilibrium (2). Equilibrium constants obtained with these different models are listed in Table 1. Brodin et al. Page 18 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Scheme 1. Brodin et al. Page 19 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Brodin et al. Page 20 Table 1 Association constants for various metal binding equilibrium models for C96RIDC-14 determined through competitive Fura-2 titrations (pH 7, 295 K). The total free energies for metal binding correspond to the free energy sums of individual equilibria (times their multiplicity) for every model. Corresponding titrations and fits are shown in Figure 4b (Zn2+), Figure 7 (Cu2+ and Ni2+) and Figure S11 (Co2+). Numbers in parentheses correspond to standard deviation in the last reported significant figure, were obtained through DynaFit, and do not include any experimental errors. Total metal equivalents Number of Consecutive Binding Equilibria Kd1 (M) Kd2 (M) Kd3 (M) Kd4 (M) Total -ΔG for metal binding (kJ mol−1) 4 Zn2+ 2 5.2(4) × 10−10 4.3(2) × 10−8 189 4 1.3(3) × 10−9 5.3(7) × 10−10 3.3(8) × 10−8 5.8(8) × 10−8 186 2 Cu2+ 1 1.0(1) × 10−12 136 2 2.5(3) × 10−13 1.4(1) × 10−12 138 2 Ni2+ 1 8.0(9) × 10−9 92 2 9.0(1) × 10−10 4.9(5) × 10−9 93 1 Co2+ 1 9(4) × 10−7 34 J Am Chem Soc. Author manuscript; available in PMC 2011 June 30.
3M7D
Crystal structure of an N-terminal 44 kDA fragment of topoisomerase V in the presence of dioxane
Structures of minimal catalytic fragments of topoisomerase V reveals conformational changes relevant for DNA binding Rakhi Rajan*, Bhupesh Taneja*,†, and Alfonso Mondragón*,‡ * Department of Biochemistry, Molecular Biology and Cell Biology, Northwestern University, 2205 Tech Dr, Evanston, IL 60208 Summary Topoisomerase V is an archaeal type I topoisomerase that is unique among topoisomerases due to presence of both topoisomerase and DNA repair activities in the same protein. It is organized as an N-terminal topoisomerase domain followed by 24 tandem helix hairpin helix (HhH) motifs. Structural studies have shown that the active site is buried by the (HhH) motifs. Here we show that the N-terminal domain can relax DNA in the absence of any HhH motifs and that the HhH motifs are required for stable protein-DNA complex formation. Crystal structures of various topoisomerase V fragments show changes in the relative orientation of the domains mediated by a long bent linker helix, and these movements are essential for the DNA to enter the active site. Phosphate ions bound to the protein near the active site helped model DNA in the topoisomerase domain and shows how topoisomerase V may interact with DNA. Introduction DNA topoisomerases are enzymes found in all forms of life (bacteria, eukarya, and archaea) and they regulate the topological state of DNA inside the cell. They form a transient break in a single or double stranded DNA and allow the passage of another single or double DNA strand through the break, before resealing the break (Champoux, 2001) (Schoeffler and Berger, 2008). As a result of this, topoisomerases can relax supercoiled DNA, help in the segregation of DNA strands following replication, and lead to the formation and resolution of knots and catenates (Gellert, 1981). Topoisomerases participate in many aspects of DNA metabolism, such as replication, recombination, and transcription (Champoux, 2001). In addition, they are targets of various anti-cancerous drugs and anti-bacterial agents (Maxwell, 1999; Pommier, 1998; Rothenberg, 1997; Wang et al., 1997). DNA topoisomerases are broadly classified into two types, type I and type II enzymes. Type I enzymes cleave a single strand of a DNA molecule and pass another single or double stranded DNA through the break before resealing the opening. Type II enzymes cleave both ‡Corresponding author: Phone: 847-491-7726, Fax: 847-467-6489, a-mondragon@northwestern.edu. †Present address: Institute of Genomics and Integrative Biology, CSIR, Delhi, India Protein data bank accession codes The final structure factors and coordinates of Topo-31, Topo-44 Form I, Form II, and Form III have been deposited in the Protein Data Bank with accession codes 3M7G, 3M7D, 3M6K, and 3M6Z respectively. Supplementary data Supplementary data are available at Structure Journal Online. Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. NIH Public Access Author Manuscript Structure. Author manuscript; available in PMC 2011 July 14. Published in final edited form as: Structure. 2010 July 14; 18(7): 829–838. doi:10.1016/j.str.2010.03.006. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript strands of a double stranded DNA in concert and pass another double stranded DNA through the break. Type I enzymes use the torsional energy stored in the supercoiled DNA to drive DNA relaxation and hence they do not require high energy cofactors, such as ATP, for their activity (Baker et al., 2009) Type II enzymes, on the other hand, require ATP and Mg2+ for their activity. Type I topoisomerases are further subdivided into three subtypes: IA, IB, and IC (Forterre et al., 2007). Type IA and IB enzymes have been studied extensively (Baker et al., 2009) and there is ample information available about their general mechanism of DNA relaxation and the mode of DNA binding. Type IC, on the other hand, is a relatively new subtype. Currently topoisomerase V is the only member of this family and it has been identified only in the Methanopyrus genus. Previously, topoisomerase V had been considered as a type IB enzyme based on its biochemical characteristics (Slesarev et al., 1993), but the crystal structure of an N-terminal 61 kDa of topoisomerase V (Topo-61) (Taneja et al., 2006) revealed a completely new fold without similarity to other topoisomerases or any other known protein. Furthermore, the orientation of the putative active site residues is also different from other type I topoisomerases, suggesting a different mechanism of cleavage and religation of DNA. These observations, together with the lack of sequence similarity, indicated that topoisomerase V defines a new subtype of type I enzymes (Forterre, 2006). Topoisomerase V was identified in Methanopyrus kandleri, an extremophile isolated from a deep-water ‘black smoker’ chimney in the Gulf of California (Huber et al., 1989). The enzyme is active at very high temperatures (122°C) and high salt concentrations (0.65 M NaCl and 3.1 M potassium glutamate). The unusual characteristic of topoisomerase V is that it has both topoisomerase and DNA repair activities in the same polypeptide (Belova et al., 2001). Based on the sequence analysis of topoisomerase V, it has been predicted that the protein contains 24 helix-hairpin-helix (HhH) DNA binding motifs arranged as 12 (HhH)2 domains around the N-terminal topoisomerase domain (Belova et al., 2002) (Figure 1A). Some of these (HhH)2 domains are involved in the apurinic/apyrimidinic (AP) site- processing activity, but the exact location of the repair active site is not known yet. Topoisomerase V can relax both positively and negatively supercoiled DNA without the need for metal cations or high energy cofactors. Single molecule experiments have shown that topoisomerase V relaxes DNA by a constrained swiveling mechanism, relaxing around 12 turns of DNA per relaxation cycle (Taneja et al., 2007). Type IB enzymes, which also use a constrained swiveling mechanism for DNA relaxation, relax around 19 turns of DNA per relaxation cycle (Koster et al., 2005). The structure of Topo-61 showed that the topoisomerase domain is mainly alpha helical and that the first four (HhH)2 domains curl around the topoisomerase domain (Taneja et al., 2006) (Figure 1B). The topoisomerase and (HhH)2 domains are joined by a long bent helix, termed the “linker helix”. Three of the five putative active site residues are present in a helix-turn-helix (HTH) domain and the other two are present in an intervening loop and a helix. The active site residues are buried by the first (HhH)2 domain and it has been suggested that large conformational changes will be needed for the DNA to access the active site of topoisomerase V (Taneja et al., 2007). Here we present data that shows that the N- terminal 31 kDa fragment of topoisomerase V (Topo-31) has topoisomerase activity, consistent with previous predictions based on the structure. In addition, we show that the Topo-44 fragment (N-terminal 44 kDa fragment of topoisomerase V) can form a stable protein-DNA complex, emphasizing the need of the (HhH)2 domains for binding DNA. We determined a crystal structure of (Topo-31) fragment, which has only the topoisomerase domain, and three different crystal structures of the Topo-44 fragment, which includes the topoisomerase domain and three tandem HhH motifs. In all structures, the topoisomerase domain is very similar. In contrast, the structures of Topo-44 show conformational changes in the linker helix resulting in variable orientations of the (HhH)2 domains when compared Rajan et al. Page 2 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript to the Topo-61 structure. Phosphate ions are present in the vicinity of the topoisomerase active site in two of the Topo-44 structures. Some of the catalytic residues interact with the phosphate ions and may mimic contacts with DNA. These observations suggest that the movement of the (HhH)2 domains is mediated by the linker helix and helps expose the topoisomerase active site to facilitate DNA binding. In addition, the location of the phosphate ions suggests a possible path for the DNA and the way the active site residues interact with it. Results The topoisomerase domain can relax DNA DNA relaxation assays using different topoisomerase V fragments showed that the topoisomerase domain alone is capable of relaxing DNA. Topoisomerase V fragments with different numbers of (HhH)2 domains, Topo-31, Topo-44, and Topo-78, were studied using relaxation assays. Topo-31 has no (HhH)2 domains, Topo-44 has one full and one partial (HhH)2 domain, while Topo-78 has eight full (HhH)2 domains, including a putative DNA repair domain. In addition to standard conditions, the effect of different pH conditions and presence of magnesium ions were also tested. The experiments show that Topo-31 is capable of relaxing DNA, despite the absence of the (HhH)2 domains (Figure 2B). A pH profile analysis for the DNA relaxation assays showed that Topo-78 relaxes DNA over a wider pH range (pH 5 to 9), while Topo-31 and Topo-44 relax DNA optimally at pH 5 (Figure 2A, 2B, 2C). In addition, magnesium is not required for the reaction, but stimulates it at all pH values (Figure 2B, 2C). Topo-78 can relax DNA to the same extent with lower amounts of protein (0.1 μg/reaction) compared to Topo-44 (~1.5 μg/reaction) and Topo-31 (~9 μg/reaction). This could be due to the enhanced DNA binding facilitated by the (HhH)2 domains. Together, these results suggest that, even though the (HhH)2 domains are dispensable for topoisomerase activity, they enhance DNA relaxation activity. In addition, the pH dependence of the DNA relaxation activity indicates that the reaction is likely to involve side chains with ionizable groups in the low pH range, such as glutamates. Finally, the magnesium independence of the reactions confirms that even the smallest fragments do not require metals for activity, although magnesium has a stimulatory effect. This may be due to favorable interactions of the cations with DNA. The (HhH)2 domains enhance DNA binding affinity EMSA experiments with different fragments of topoisomerase V and DNA showed that (HhH)2 domains could help in the formation of a stable protein-DNA complex. Various topoisomerase V fragments (Topo-31, Topo-44, and Topo-78) and single and double stranded DNA were analyzed by EMSA experiments. Topo-44 and Topo-78 formed stable complexes with a 39mer double stranded DNA (Figure 2D), while no DNA binding was observed for the Topo-31 fragment (data not shown). These observations indicate that (HhH)2 domains are necessary for a stable protein-DNA complex and that as few as one and half (HhH)2 domains are enough for formation of a stable protein-DNA complex. EMSA with single stranded DNA showed that Topo-31 and Topo-44 cannot bind to single stranded DNA, while Topo-78 can bind to single stranded DNA (data not shown). Overall Structures The topoisomerase domain of topoisomerase V is a helical-rich compact domain that has no structural similarity to any other known protein. The only recognizable structural element is a HTH that contains some of the active site residues. Not surprisingly, the topoisomerase domain of the four structures (Topo-31, Topo-44 (Forms I, II, and III)) superimpose very well on each other and also to that from the Topo-61 structure. In the Topo-31 structure, two surface loops, residues 39-49 and 120-124, adopt a different conformation compared to the Rajan et al. Page 3 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Topo-61 and Topo-44 structures. These two loops are not always visible in the Topo-44 structures, suggesting that they are mobile regions. The r.m.s.d. for the superposition of the topoisomerase core domain of all the new structures on to the Topo-61 structure range from 0.2 Å to 0.7 Å if the two mobile surface loops are not included (Figure 3A). In general, the topoisomerase domain remains unchanged and is identical in all structures. The (HhH)2 domains also remain largely unchanged, with r.m.s.d. for the superposition of only the (HhH)2 domains from the three Topo-44 crystal forms and equivalent domains in the Topo-61 structure ranging from 0.31 Å to 0.56 Å. The five crystallographically independent structures of Topo-44 (Form I, Form II A and B monomers, and Form III A and B monomers) were compared with each other and to the two crystallographically independent Topo-61 monomers to understand the conformational changes in the protein. The r.m.s.d. for the superposition of all the Topo-44 structures (residues 3-375) on to the Topo-61 fragment or on each other vary between 0.9 Å and 2.7 Å, with the majority above 1.5 Å, showing that in general the structures have slightly different conformations. As mentioned above, the different domains behave as rigid or almost rigid subunits and the only change in the structure is the relative orientation between the topoisomerase and the (HhH)2 domains. The change in orientation of the domains starts at the linker helix (residues 269-295), which acts as a hinge region, and follows into the (HhH)2 domains. At the start of the linker helix, the structures superimpose very well for all five Topo-44 and two Topo-61 structures. In the middle of the linker helix there is a kink after which the linker helix from all the structures shows different orientations (Figure 3B). The flexibility of the linker helix is also evident by the fact that the linker helix in the B subunit of Form III crystals appears in two alternate conformations. The change in the relative orientation of the (HhH)2 and topoisomerase domains (Figure 3C and 3D), suggests that these domains can adopt different orientations and these movements might be necessary for the DNA to access the active site. The topoisomerase domain has a positively charged groove adjacent to the active site The structure of the Topo-31 as well as the structures of the Topo-44 fragment reveals the presence of a positively charged groove in the protein that encompasses the active site region (shown later in Figure 6C). This charged groove had been observed before in the structure of the Topo-61 fragment, although several (HhH)2 motifs partially obstruct it (Taneja et al., 2006). The structure of the Topo-31 confirms the presence of the groove even in the absence of the (HhH)2 motifs. The groove is long and can be deep in some areas. It includes regions of the HTH motifs and extends all the way to the linker helix. All the residues forming the active site pentad point towards the groove. The active site tyrosine, Tyr226, is found near one of the ends of the groove, a region where it widens. The positively charged character of the groove and its presence by the active site strongly suggest that it may be involved in DNA binding. Phosphate ions bind in the groove near the topoisomerase active site An interesting observation stemming from the Form II and Form III Topo-44 structures is the presence of phosphate ions near the positively charged DNA binding groove. All three Topo-44 crystal forms were crystallized in the presence of phosphate-citrate buffer, but only Form II and Form III structures showed phosphate ions bound to the protein, which were assigned based on electron density consistent with a tetrahedral phosphate ion (Figure 4A). Form II and Form III crystals include 1–1.2 M guanidium hydrochloride in the crystallization solution. The high resolution Form III structure shows clear density for three guanidium ions bound to the protein, two very well ordered and one with weak density. The presence of guanidium hydrochloride in the crystals appears to trigger a conformational change allowing the binding of phosphate ions to the protein. It is interesting to note that Rajan et al. Page 4 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Form I crystals did not show any bound phosphate albeit its presence in the crystallization condition. This could be due to the absence of guanidium hydrochloride to trigger the binding of phosphate ions as observed in Form II and Form III structures. There are three phosphate ions in the B subunit and none in the A subunit of the Form II Topo-44 structure. Two of the phosphates are in the topoisomerase active site and one of them forms close contacts with the putative active site residues in the topoisomerase domain (Figure 4B). Form III crystal has seven phosphate ions, three in each subunit and one between both the subunits. In the Form III structure, the phosphate ion near the active site Tyr226 is absent, but it shows several new locations for phosphate ions, especially in the positively charged groove containing the topoisomerase active site (Figure 5A). An overlay of the A and B subunits of the Topo-44 Form III structure with the B subunit of Topo-44 Form II structure shows eight unique phosphate ions (Figure 5A). It clearly shows that there are more phosphate ions bound in the positively charged groove compared to other regions of the protein. Taking into account all structures, there are five unique phosphate ion binding sites in the putative DNA binding groove and an additional one near its end and close to the start of the linker helix. Several pairs of phosphates in the groove are separated by a distance of around 7 Å (Figure 5B), which would be consistent with the phosphate-phosphate distance in adjacent nucleotides in a DNA double helix. One of the phosphates (P1) is found near the active site tyrosine and is coordinated by Tyr226 and Arg131, two residues that have been implicated in cleavage and religation of the DNA (Taneja et al., 2006), and by Glu215, whose charge may be important for interactions with DNA (R.R. and A.M., unpublished observations). The side chains of the tyrosine and the glutamate residues are in contact with Arg144 and His200, the other putative active site residues, and these interactions may help to orient them for the catalytic reaction. Adjacent to P1, there is a second phosphate (P2) at a distance of 7.5 Å which is trapped between the topoisomerase domain and an HhH motif. P2 is coordinated by Arg131, an active site residue, in addition to Arg108 from the topoisomerase domain and Arg293 and serines 322 and 324 from the second HhH motif (Figure 6C). Three more phosphates are found in the groove (P3, P4, and P5) coordinated mainly by positively charged residues, such as Arg37, Lys47, Arg108, Lys134, and Arg135 from the topoisomerase domain and also residues from the linker helix such as Tyr289 and Arg293. In general, some of the side chains can contact more than one phosphate. The distance between P3 and P4 and P4 and P5 is 6.8 Å and 6.5 Å respectively. A final phosphate (P6) is located at the start of the linker helix and on the edge of the groove (Figure 5A). Discussion Topoisomerase V is active at very high temperatures (122°C) and high salt concentrations. DNA relaxation assays with various topoisomerase V fragments (Topo-44 and Topo-61) show that a temperature above 60° C is required for optimal activity, although longer fragments of topoisomerase V can relax DNA at lower temperatures (Taneja et al., 2007). Topo-44 was first identified by limited proteolytic digestion of the full length topoisomerase V protein (Belova et al., 2002) at 80°C. In contrast, Topo-61 is the shortest fragment showing topoisomerase activity when the proteolytic reaction is performed at 37°C (Belova et al., 2002). The N-terminal Topo-31 fragment, which contains neither HhH motifs nor the linker helix, was identified as the smallest region spanning the topoisomerase domain from the crystal structure of Topo-61 fragment (Taneja et al., 2006) and it was suggested that it could represent the minimal domain capable of relaxing DNA. Relaxation experiments with this minimal domain show that this is indeed the case, although the activity is not as robust as with longer fragments. As expected, Topo-31 does not require magnesium for activity, but magnesium enhances it, as is the case for type IB topoisomerases, which also uses a Rajan et al. Page 5 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript swiveling mechanism for DNA relaxation (Stewart et al., 1996). The optimal pH for activity for the Topo-31 and the Topo-44 fragments is around 5. This pH dependence is not observed for the Topo-78 fragment. The DNA relaxation by shorter fragments of topoisomerase V at pH 5 could point to the involvement of some ionizable side chains in the relaxation activity. It could also be simply due to the effects of various side chains on DNA binding. Further experiments with different active site mutations in both longer and shorter fragments of topoisomerase V will be required to probe the pH dependence of the relaxation reaction by shorter topoisomerase V fragments. Gel shift experiments show that Topo-44 and also longer fragments (Topo-78) can bind double stranded DNA. Surprisingly, Topo-31 does not show DNA binding activity in these assays even though it is still capable of relaxing DNA. It appears that the presence of the (HhH)2 domains stabilizes the DNA/protein complex. One possibility is that the (HhH)2 domains could play a similar role to the cap domain present in type IB enzymes, which helps to encircle the DNA during the swiveling reaction (Redinbo et al., 1998). In addition, both short fragments of topoisomerase V do not bind single stranded DNA, whereas Topo-78 can form a stable complex with single stranded DNA (data not shown). (HhH)2 domains binding to single stranded DNA has been observed before. For instance, the N-terminal 8 kDa of mammalian polymerase β, which contains a single HhH motif, binds to single stranded DNA through both helices (Kumar et al., 1990; Liu et al., 1994). The exact mode of single stranded DNA binding by Topo-78 or the possible role in relaxation or repair activities is not yet clear. The structure of Topo-61 showed that the topoisomerase active site of topoisomerase V is buried by one of the (HhH)2 domains suggesting that conformational changes are essential for the protein to bind DNA. The present structures of Topo-44 reinforce this observation and show that the (HhH)2 domains can change their position relative to the topoisomerase domain and that this change is mediated by the movement of the linker helix. The (HhH)2 domains act as rigid individual units, as evidenced by the fact that in different structures they show the same structure and relative orientation of the two HhH motifs. The topoisomerase domain also appears to be rigid showing the same structure even in the total absence of the rest of the protein. The linker helix (residues 269-295), which is a long bent helix, serves as a hinge for the movement of the (HhH)2 domains away from the rigid topoisomerase domain, possibly by responding to interactions with double stranded DNA. This movement has to be quite large. The Topo-44 structures in the absence of DNA capture the regions that move, but do not show the full extent of the movement or indicate the way the HhH motifs interact with DNA. As mentioned before, topoisomerase V binds double stranded DNA and has a groove wide enough to accommodate double strand DNA (Figure 6C). The presence of an HTH domain normally associated with DNA binding, the positively charged nature of it, and several phosphates bound along it suggest that this groove could be involved in DNA binding. In addition, the active site is found in this groove and some residues form part of the HTH domain. Previously, DNA was modeled bound to the topoisomerase domain (Taneja et al., 2006) based on the structures of HTH domains in complex with DNA but there was no evidence to support it. Using the phosphates present in the groove in the current structures, it is possible to refine this model. A superposition of the B subunit of Form II and the A and B subunits of Form III Topo-44 structures shows five different phosphate ions in the positively charged groove which are separated by a distance of around 7 Å, consistent with the distance of consecutive phosphates in B DNA of ~6.4 Å. A sixth phosphate ion is found outside the groove near the linker helix. A double stranded DNA molecule was modeled into the groove based on the positions of the phosphate ions (Figure 6). Only five out of the six phosphates could be placed on the DNA molecule, as one of them was inconsistent with a Rajan et al. Page 6 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript double stranded molecule. Phosphate ions P3, P4, and P5 would correspond to three adjacent phosphates in one DNA strand, while P1, located near the active site, would belong to the opposite strand. A final phosphate (P6) is away from the groove and near the linker helix (Figure 6A, 6B). The fit to the four inner phosphates is excellent and the DNA can be accommodated in the groove of the Topo-31 structure without the need for any major rearrangements of the protein backbone. The fifth phosphate (P6) does not fit as well and a better fit would require movement of either the protein or the DNA, but the change would be relatively modest. Several side chains would need to move, but these changes would also be minor. The major change needed to accommodate the DNA in the structures with the (HhH)2 domains present is the movement of the (HhH)2 domains away from the topoisomerase domain (Figure 6B). The movement of (HhH)2 domains should be feasible as is evident from the Topo-44 structures showing different orientations of the (HhH)2 domains. The location of the (HhH)2 domains after DNA binding is not evident, but one possibility is that they would help enclose the DNA to form a clamp around it, similar to the arrangement in type IB enzymes. In the model of the topoisomerase domain in complex with DNA, the active site residues are in close contact with the backbone of DNA. The catalytic Tyr226 is pointing towards the phosphate of the DNA backbone; Arg131 and Arg144 are positioned to stabilize the protein- DNA covalent complex. Surprisingly Glu215 also appears to interact directly with the DNA phosphate backbone. The other active site residues like His200 and Lys 218 are also near the DNA. The active site is located near the end of the groove, where it widens. At this end, the DNA fits loosely in the groove, which is spacious to accommodate the movement of the strands. The ‘constrained swiveling’ mechanism employed by type IB and IC enzymes necessitates rotation of one strand about the other after forming the covalent protein-DNA intermediate. The position of the active site at the wider end of the putative DNA binding groove would facilitate the rotation of the DNA strand at this end, while holding the rest of the DNA in place through extensive interactions along the groove. Even though type IB and IC enzymes have a similar overall mechanism of action, the structures of fragments of topoisomerase V suggest many differences. Type IB enzymes have two domains which come together to form a C-shaped clamp around the DNA (Perry et al., 2006; Redinbo et al., 1998; Stewart et al., 1998) The protein has an open stage where these domains are separate and this helps in the entry and release of the DNA from the protein active site. A wide DNA binding cavity is not observed in the topoisomerase V structures. Instead, the structures show a positively charged groove which is always present in the protein and does not require domain rearrangements to form. DNA can access this groove after a conformational change involving the movement of the (HhH)2 domains exposing the active site. The (HhH)2 domains could help enclose DNA during the swiveling of the DNA, forming a similar enclosure to the one observed for type IB enzymes. It is not known whether all HhH motifs contact DNA simultaneously, but this appears unlikely without a major rearrangement of the motifs. It is likely that only some of the HhH motifs contact DNA at any given time or that some of the motifs do not have the capacity to bind DNA. Finally, similar to type IB enzymes (Cheng and Shuman, 1998), the putative domain enclosing the DNA is dispensable for activity, although it enhances the relaxation activity markedly. Thus, it is likely that type IB and IC enzymes have several overall similarities in the way that they interact with DNA, but the atomic details are markedly different. There are still many details of the atomic mechanism of type IC topoisomerases that need to be understood. The present functional and structural studies provide new information about topoisomerase V including the observations that the Topo-31 is the minimal fragment capable of DNA relaxation, the (HhH)2 domains enhance binding of the protein to DNA, the changes in relative orientation of the domains is mediated by the linker helix, and several Rajan et al. Page 7 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript phosphate ions bind in a positively charged groove. Furthermore, the position of the phosphate ions in the groove helped in the placement of DNA in the topoisomerase domain and this provides an initial model of how topoisomerase V interacts with DNA. Thus the present study helps to establish the role of different domains more clearly, to illustrate a mechanism to drive the conformational changes needed for activity, and to suggest a possible manner of binding DNA. Additional work on structures of protein/DNA complexes and intermediates in the swiveling reaction are needed to understand the way this new type of topoisomerases interacts with DNA to perform a complex reaction. Experimental Procedures Protein purification The N-terminal 31 kDa (Topo-31: residues 1-269), and 44 kDa (Topo-44: residues 1 to 380) fragments of topoisomerase V protein were cloned into the pET15b plasmid and transformed into Escherichia coli BL21 Rosetta (DE3) cells. The N-terminal 78 kDa (Topo-78: residues 1 to 685) fragment of topoisomerase V protein was cloned into the pET14b plasmid (Belova et al., 2002) and transformed into Escherichia coli BL21(DE3) cells. For protein production, cells were grown at 37° C in LB medium containing 100 μg/ml ampicillin and 100 μg/ml chloramphenicol for Rosetta cells and LB medium with 100 μg/ml ampicillin for BL21(DE3) cells to an optical density (OD600) of 0.6. The cells were then cooled down on ice, followed by the addition of isopropyl β-D-1-thiogalactopyranoside (IPTG) to a final 0.5 mM concentration, and grown overnight at 16° C. Cells were harvested and resuspended in 50 mM Tris pH 8, 500 mM NaCl, 0.5 mM EDTA, 1 mM DTT, flash frozen in liquid nitrogen and stored at −80° C. After thawing the pellet, pepstatin (1μg/ml), benzamidine (1mM), PMSF (1mM), and Brij 58 (0.1%) were added to the cells and the protein was purified as described earlier (Taneja et al., 2006) The protein was further purified by anion exchange and gel filtration chromatography. Pure protein was concentrated and stored in 50 mM Tris pH 8, 250 mM NaCl, and 1 mM DTT. The seleno- methionine substituted Topo-44 was prepared from cells grown in a minimal medium supplemented with nutrients and salts (Doublie, 1997); protein purification followed the same procedure as for the native protein except that 5mM DTT was used in all the purification steps and for storage. Relaxation assays Relaxation assays with the different topoisomerase V fragments were carried out at pH values ranging from 4 to 10. The pH of the buffers was adjusted at 65 °C to account for the change in pH at higher temperature. The different buffers used were: sodium acetate for pH 4 and 5, MES for pH 6, HEPES for pH 7, TRIS for pH 8, CHES for pH 9, and CAPS for pH 10. Topoisomerase activity assays were performed by incubating varying amounts of protein (Topo-31, Topo-44 or Topo-78) with 0.2 μg negatively supercoiled pUC19 DNA in 50 mM of the required buffer, 30 mM NaCl, 0.2 mM or 5 mM EDTA or 1 mM MgCl2. The reactions were carried out at 65 °C for 15 min and terminated by cooling and addition of SDS to a final 1% concentration. The products were resolved on a 1% agarose gel and visualized by ethidium bromide staining. Electrophoretic Mobility Shift Assay For Electrophoretic Mobility Shift Assay (EMSA), 4 μM of a 39mer double stranded DNA oligonucleotide (5′ GCGACGCGAGGCTGGATGGCCTTCCCCATTATGATTCTT3′) was incubated with different concentrations of topoisomerase V fragments in 50 mM sodium acetate pH 5, 30 mM NaCl, 1 mM MgCl2 at 65 °C for 30 minutes. Glycerol was added to the reaction mixture to a final concentration of 8% and the products were separated on a 4 % acrylamide native gel. The gel was stained with ethidium bromide to detect the DNA. When Rajan et al. Page 8 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript a stable protein-DNA complex was formed, there was an upward shift in the band indicating a higher molecular weight complex. Crystallization Topo-31 crystals were grown using the sitting drop vapor diffusion method equilibrated against, 23% PEG 6000, 0.1 M Na citrate pH 5.5, at 22°C. For data collection, the Topo-31 crystals were cryo-protected by adding glycerol to the mother liquor to a final 20% concentration. Topo-44 was crystallized by the hanging drop vapor diffusion method under three different crystallization conditions (Forms I, II, and III). Crystal Form I grew under 0.1 M phosphate citrate pH 5, 0.2 M NaCl, 15% PEG 3350 and 8% dioxane. The crystals were cryo-protected by increasing the PEG concentration to 30%. Form II crystals grew under 0.1 M phosphate citrate pH 5, 0.2 M NaCl, 16% PEG 8000 and 1M guanidium hydrochloride. For cryo-protection, they were transferred to a solution with 1.5X reservoir solution and 20% 2,3 butanediol or 20% DMSO for 10 seconds and immediately flash frozen under liquid nitrogen. Form III crystals grew under 0.1 M phosphate citrate pH 5.5, 0.15 M sodium sulfate, 0.01 M MgCl2, 1 M guanidium hydrochloride, and 28 % PEG 3350. The crystals were grown at 30°C and were cryo-protected by increasing the PEG concentration to 40%. Further details of crystallization are presented in the Supplementary Information. Data collection and structure determination Diffraction data were collected at the Dupont Northwestern Dow and Life Science Collaborative Access Team stations (DND and LS CAT) at the Advanced Photon Source in Argonne National Laboratory. Data collection and refinement statistics are shown in Table I. All data were processed and integrated using XDS (Kabsch, 1993) and scaled with SCALA (Collaborative-Computational-Project-4, 1994). Data on the Topo-31 crystals were collected to 2.4 Å resolution. The structure was solved by Molecular Replacement (McCoy et al., 2007) using the topoisomerase domain from the Topo-61 structure (residues 1-266) (Taneja et al., 2006) as the search model. It was refined with refmac5 (Murshudov et al., 1997) and Phenix (Afonine et al., 2005) to a final Rwork of 20.0 % and Rfree of 24.8 %. Topo-44 Form I crystals diffract to 1.8 Å. The structure of Form I crystals was solved by Molecular Replacement (McCoy et al., 2007) using the topoisomerase domain from the Topo-61 structure as the search model. Model rebuilding was performed using coot (Emsley and Cowtan, 2004), and refinement using refmac5 (Murshudov et al., 1997). The final Rwork and Rfree are 17.5 % and 22.0 % respectively. For Topo-44 Form II and Form III crystals, seleno-methionine derivatized crystals were used for single-wavelength anomalous dispersion (SAD) phasing. AutoSharp (Vonrhein et al., 2007) was used for locating the selenium atoms; model building was done using coot (Emsley and Cowtan, 2004), and refinement was carried out using refmac5 (Murshudov et al., 1997) Three phosphate ions were noticed in the Form II structure; two of which present in the topoisomerase active site and are separated by a distance of ~7.5 Å. The structure was refined to a final Rwork of 24.1 % and Rfree of 28.9 %. Topo-44 Form III crystals diffracted to 1.4 Å. The final Rwork and Rfree are 16.5 % and 18.4%, respectively. An interesting observation is the presence of both phosphate and guanidium ions in the Form III Topo-44 structure. The linker helix and part of the first HhH motif of the B monomer show alternate conformations and were built as two separate chains with occupancy of 0.5 each. Further details on data collection and structure determination are given in the Supplementary Information. All figures were made with Pymol (DeLano, 2002) and the electrostatic surfaces were calculated with APBS (Baker et al., 2001). Rajan et al. Page 9 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We acknowledge staff and instrumentation support from the Keck Biophysics Facility and the Center for Structural Biology at Northwestern University, and DND and LS-CAT at the Advanced Photon Source (APS) at Argonne National Laboratory. Support from the R.H. Lurie Comprehensive Cancer Center of Northwestern University to the Structural Biology Facility is also acknowledged. DND-CAT is supported by Dupont, DOW and the NSF. LS-CAT was supported by the Michigan Economic Development Corporation and the Michigan Technology Tri-Corridor. Use of the APS is supported by the Department of Energy (DOE). Research was supported by NIH grant GM51350 (to AM). References Afonine PV, Grosse-Kunstleve RW, Adams PD. A robust bulk-solvent correction and anisotropic scaling procedure. Acta Crystallogr D Biol Crystallogr. 2005; 61:850–855. [PubMed: 15983406] Baker NA, Sept D, Joseph S, Holst MJ, McCammon JA. Electrostatics of nanosystems: application to microtubules and the ribosome. Proc Natl Acad Sci U S A. 2001; 98:10037–10041. [PubMed: 11517324] Baker NM, Rajan R, Mondragon A. Structural studies of type I topoisomerases. Nucleic Acids Res. 2009; 37:693–701. [PubMed: 19106140] Belova GI, Prasad R, Kozyavkin SA, Lake JA, Wilson SH, Slesarev AI. A type IB topoisomerase with DNA repair activities. Proc Natl Acad Sci U S A. 2001; 98:6015–6020. [PubMed: 11353838] Belova GI, Prasad R, Nazimov IV, Wilson SH, Slesarev AI. The domain organization and properties of individual domains of DNA topoisomerase V, a type 1B topoisomerase with DNA repair activities. J Biol Chem. 2002; 277:4959–4965. [PubMed: 11733530] Champoux JJ. DNA Topoisomerases: Structure, Function, and Mechanism. Annu Rev Biochem. 2001; 70:369–413. [PubMed: 11395412] Cheng C, Shuman S. A catalytic domain of eukaryotic DNA topoisomerase I. J Biol Chem. 1998; 273:11589–11595. [PubMed: 9565576] Collaborative-Computational-Project-4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D. 1994; 50:760–763. [PubMed: 15299374] Davis IW, Murray LW, Richardson JS, Richardson DC. MOLPROBITY: structure validation and all- atom contact analysis for nucleic acids and their complexes. Nucleic Acids Res. 2004; 32:W615– 619. [PubMed: 15215462] DeLano, WL. The PyMol Molecular Graphics System. San Carlos, CA: DeLano Scientific; 2002. Diederichs K, Karplus PA. Improved R-factors for diffraction data analysis in macromolecular crystallography. Nat Struct Biol. 1997; 4:269–275. [PubMed: 9095194] Doublie S. Preparation of selenomethionyl proteins for phase determination. Methods Enzymol. 1997; 276:523–530. [PubMed: 9048379] Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] Forterre P. DNA topoisomerase V: a new fold of mysterious origin. Trends Biotechnol. 2006; 24:245– 247. [PubMed: 16650908] Forterre P, Gribaldo S, Gadelle D, Serre MC. Origin and evolution of DNA topoisomerases. Biochimie. 2007; 89:427–446. [PubMed: 17293019] Gellert M. DNA Topoisomerases. Annu Rev Biochem. 1981; 50:879–910. [PubMed: 6267993] Huber R, Kurr M, Jannasch HW, Stetter KO. A novel group of abyssal methanogenic archaebacteria (Methanopyrus) growing at 110 °C. Nature. 1989; 342:833–834. Kabsch W. Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Crystallogr. 1993; 26:795–800. Rajan et al. Page 10 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Koster DA, Croquette V, Dekker C, Shuman S, Dekker NH. Friction and torque govern the relaxation of DNA supercoils by eukaryotic topoisomerase IB. Nature. 2005; 434:671–674. [PubMed: 15800630] Kumar A, Widen SG, Williams KR, Kedar P, Karpel RL, Wilson SH. Studies of the domain structure of mammalian DNA polymerase beta. Identification of a discrete template binding domain. J Biol Chem. 1990; 265:2124–2131. [PubMed: 2404980] Liu D, DeRose EF, Prasad R, Wilson SH, Mullen GP. Assignments of 1H, 15N, and 13C resonances for the backbone and side chains of the N-terminal domain of DNA polymerase beta. Determination of the secondary structure and tertiary contacts. Biochemistry. 1994; 33:9537– 9545. [PubMed: 8068628] Maxwell A. DNA gyrase as a drug target. Biochem Soc Trans. 1999; 27:48–53. [PubMed: 10093705] McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software. J Appl Crystallogr. 2007; 40:658–674. [PubMed: 19461840] Murshudov GN, Vagin AA, Dodson EJ. Refinement of macromolecular structures by the maximum- likelihood method. Acta Crystallogr D. 1997; 53:240–255. [PubMed: 15299926] Perry K, Hwang Y, Bushman FD, Van Duyne GD. Structural basis for specificity in the poxvirus topoisomerase. Mol Cell. 2006; 23:343–354. [PubMed: 16885024] Pommier Y. Diversity of DNA topoisomerases I and inhibitors. Biochimie. 1998; 80:255–270. [PubMed: 9615865] Redinbo MR, Stewart L, Kuhn P, Champoux JJ, Hol WG. Crystal structures of human topoisomerase I in covalent and noncovalent complexes with DNA. Science. 1998; 279:1504–1513. [PubMed: 9488644] Rothenberg ML. Topoisomerase I inhibitors: review and update. Ann Oncol. 1997; 8:837–855. [PubMed: 9358934] Schoeffler AJ, Berger JM. DNA topoisomerases: harnessing and constraining energy to govern chromosome topology. Q Rev Biophys. 2008; 41:41–101. [PubMed: 18755053] Slesarev AI, Stetter KO, Lake JA, Gellert M, Krah R, Kozyavkin SA. DNA topoisomerase V is a relative of eukaryotic topoisomerase I from a hyperthermophilic prokaryote. Nature. 1993; 364:735–737. [PubMed: 8395022] Stewart L, Ireton GC, Parker LH, Madden KR, Champoux JJ. Biochemical and biophysical analyses of recombinant forms of human topoisomerase I. J Biol Chem. 1996; 271:7593–7601. [PubMed: 8631793] Stewart L, Redinbo MR, Qiu X, Hol WG, Champoux JJ. A model for the mechanism of human topoisomerase I. Science. 1998; 279:1534–1541. [PubMed: 9488652] Taneja B, Patel A, Slesarev A, Mondragon A. Structure of the N-terminal fragment of topoisomerase V reveals a new family of topoisomerases. EMBO J. 2006; 25:398–408. [PubMed: 16395333] Taneja B, Schnurr B, Slesarev A, Marko JF, Mondragon A. Topoisomerase V relaxes supercoiled DNA by a constrained swiveling mechanism. Proc Natl Acad Sci U S A. 2007; 104:14670–14675. [PubMed: 17804808] Vonrhein C, Blanc E, Roversi P, Bricogne G. Automated structure solution with autoSHARP. Methods Mol Biol. 2007; 364:215–230. [PubMed: 17172768] Wang HK, Morris-Natschke SL, Lee KH. Recent advances in the discovery and development of topoisomerase inhibitors as antitumor agents. Med Res Rev. 1997; 17:367–425. [PubMed: 9211397] Rajan et al. Page 11 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Organization of topoisomerase V Topoisomerase V is a multi-domain protein consisting of 24 helix-hairpin-helix (HhH) DNA binding motifs arranged as 12 (HhH)2 domains following the N-terminal topoisomerase domain. A) Schematic diagram of various topoisomerase V fragments. The topoisomerase domain is shown in red, the (HhH)2 domains are shown in alternating colors of cyan and yellow. The (HhH)2 domains with repair activity are shown in green. All fragments shown have topoisomerase activity, but only the full length protein and the Topo78 fragment have repair activity. B) Crystal structure of Topo-61 fragment (Taneja et al., 2006). The coloring scheme is the same as in Figure 1A, except that the linker helix is shown in grey. Rajan et al. Page 12 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. DNA relaxation activity and EMSA for Topo-31, Topo-44 and Topo-78 fragments of topoisomerase V A) pH profile of the DNA relaxation activity of Topo-78 and Topo-44 fragments. 0.2 μg of pUC19 DNA were incubated with 0.1 μg of Topo-78 or 1.5 μg of Topo-44 proteins at 65°C for 15 minutes for the relaxation reaction. The reaction buffer contained 50 mM of the appropriate buffer, 30 mM NaCl and 0.2 mM EDTA. Topo-78 relaxes DNA at a wider pH range (5 to 9) than Topo-44, which relaxes DNA efficiently only at pH 5. DNA relaxation activity of Topo-31 (B) and Topo-44 (C) fragments in the absence and presence of MgCl2. 0.2 μg of pUC19 DNA were incubated with 9 μg of Topo-31 or 1.5 μg of Topo-44 proteins at 65°C for 15 minutes for the relaxation reaction. The reaction buffer contained 50 mM of the appropriate buffer, 30 mM NaCl and 5 mM EDTA or 1 mM MgCl2. Both Topo-31 and Topo-44 fragments can relax DNA in the absence of MgCl2, but MgCl2 enhances the DNA relaxation activity of the topoisomerase V fragments. The black triangle in panels A, B and C represents increasing pH from 4 to 10 by one pH unit. D) EMSA of Topo-44 and Topo-78 fragments with a 39mer double stranded DNA. Both Topo-44 and Topo-78 form stable complexes with DNA, although Topo-78 seems to saturate DNA binding while Topo-44 does not. In addition, Topo-44 shows some cleavage of the DNA (bottom free DNA band), while the cleavage is not apparent in Topo-78. The numbers at the bottom represent the molar ratio of protein to DNA. Rajan et al. Page 13 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Structure of Topo-44 fragments A) Overlay of Form I (green), Form II (B subunit: blue), and Form III (B subunit: magenta) structures onto the Topo-61 structure (B subunit: orange). The topoisomerase domains superimpose very well for all the structures, while the linker helix and (HhH)2 domains show differences in orientation. B) Overlay of the linker helices of Form I, II, and III structures with that of Topo-61. The color scheme is same for all the figures unless mentioned otherwise. Note that the linker helices have the same orientation at the start and they change as they move further down the helix. C) Superposition of Form I, II, and III Topo-44 structures with that of Topo-61. Only the (HhH)2 domains are colored while the remaining parts are shown in gray for clarity. The active site residues are shown as orange sticks. Note that the (HhH)2 domains adopt different orientations in all the structures. D) Orientation of the (HhH)2 domains of Form I, II and Topo-61 structures. In Form I and II structures, the (HhH)2 domains are moved away from the topoisomerase domain. For clarity, the (HhH)2 domains of Form III are not shown. In panels C and D, the topoisomerase domains were superposed to emphasize the different orientation of the other domains. Rajan et al. Page 14 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. Phosphate ions present near the active site of the Topo-44 structure A) Stereo view of a Form III difference electron density map calculated with a model not including the phosphates. The electron density is contoured at 3.7σ and shows the tetrahedral shape of the phosphate ions. The active site residues are shown in stick. B) Stereo view of the interaction of the phosphate ions with the putative active site residues. The B subunit of Form II structure was superimposed onto the B subunit of Form III structure and the phosphates ions from both structures are shown together with the Form II B subunit protein backbone. The interactions made by the phosphate ion with the active site residues and the corresponding distances in Å are represented as black dotted lines. Rajan et al. Page 15 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Representation of the unique phosphate ions from Form II and Form III Topo-44 structures A) An overlay of the A (magenta) and B (brown) subunits of the Form III structure and B (blue) subunit of Form II Topo-44 structures. The positions of eight unique phosphate ions (orange spheres) are shown. Note that most phosphate ions are found along the DNA binding groove of the topoisomerase domain. B) The phosphate ions in the DNA binding groove are separated by distances of around 7 Å. The protein backbone is that of the B subunit of Form III structure. The active site residues are represented as sticks and distances in Å between adjacent phosphate ions are shown as black dotted lines. Rajan et al. Page 16 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. Model showing DNA bound to the topoisomerase domain A) Model of a 17-mer double stranded DNA bound to the Topo-31 structure (teal). The DNA is represented as green sticks, where as phosphate ions are represented as orange sticks. DNA binds along the DNA binding groove and five of the eight phosphate ions noted in the Topo-44 structures coincide with the DNA backbone. B) Model of Topo-44 (Form II, B subunit: blue) binding to 17-mer double stranded DNA. Note that the linker helix and the (HhH)2 domains interfere with DNA binding to the topoisomerase domain and are likely to move away to allow binding. C) Electrostatic surface representation of the Topo-31 structure. The positively charged DNA binding groove is clearly visible and the phosphate ions are bound in this groove. The orientation corresponds to a 90° rotation of the one shown in Figure 6A in the direction of the arrow. Note that the DNA binding groove goes from one end of the molecule to the other and it is narrower at one end (start of the linker helix) and wider at the other end. The putative active site residues (green sticks) are located at the wider end of the groove. Other residues lining the groove and interacting with the phosphate ions are shown as cyan sticks. D) Electrostatic surface representation of Topo-31 with phosphate ions (orange) and DNA (green). Three phosphate ions (P3, P4, and P5) coincide with the phosphates of one of the DNA strands, where as P1 coincides with a phosphate of the opposite DNA strand. The model shows that the DNA binding groove of topoisomerase V is wide enough to bind DNA and that the movement of linker helix and (HhH)2 domains Rajan et al. Page 17 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript are required to accommodate the DNA. The electrostatic potential was calculated with a dielectric constant of 80 for solvent and 2 for protein. The surface is colored with a blue to red gradient from +10 to −10 KbT/ec. Rajan et al. Page 18 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Rajan et al. Page 19 Table 1 Data collection and refinement statistics Topo-31 Topo-44 Form I Topo-44 Form II Topo-44 Form III Data Collection Space group C2221 C121 P41212 P212121 Cell dimensions a=106.7 Å, b=119.4 Å, c=63.7 Å a=104.2 Å, b=47.7 Å, c=81.2 Å (β=112.48) a=b=70.1 Å, c=349.6 Å a=63.6 Å, b=80.1 Å, c=137.2 Å Resolution (Å)a 79.56 – 2.4 (2.53 – 2.4) 75.05 – 1.82 (1.91 – 1.82) 29.5- 2.6 (2.72-2.6) 28.9-1.4 (1.46-1.4) Number of observed reflections 78,729 (11,538 134,411 (13,220) 227,408 (19,917) 1,157,917 (126,319) Number of unique reflections 16,259 (2,346) 32,998 (4,301) 28,151 (3,331) 136,662 (15,986) Completeness (%) 99.8 (99.8) 98.3 (88.6) 99.9 (100.0) 98.8 (95.5) Multiplicity 4.8 (4.9) 4.1 (3.1) 8.1 (6.0) 8.5 (7.9) Rmerge (%)b 4.7 (71.1) 4.0 (16.3) 7.4 (52.2) 4.5 (37.9) Rmeas (%)c 5.3 (79.6) 4.6 (19.4) 7.9 (57.2) 4.8 (40.5) ≪I>/σ(<I>)>d 20.5 (2.5) 23.0 (6.8) 19 (3.2) 27.5 (5.3) Refinement Resolution (Å) 79.56 - 2.4 (2.46 - 2.4) 28.06 -1.82 (1.87 – 1.82) 29.14 – 2.6 (2.67 – 2.6) 28.9 - 1.4 (1.44 - 1.4) Number of reflections working/test 15,419/821 31,317/1,673 26,710/1,438 129,802/6,859 Rwork (%)e 20.0(24.3) 17.5 (17.9) 24.1(36.6) 16.5 (19.3) Rfree(%)f 24.8 (31.1) 22.0 (24.8) 28.9 (45.1) 18.4 (22.1) Protein residues/atomsg 269/2,203 376/3212 727/5,970 738/7,511 Atoms in alternate conformations 0 258 (20 protein residues) 8 (1 protein residue) 2846 (157 protein residues) Water molecules 29 238 30 573 Other atoms - - 3 PO4 7 PO4, 3 Gmh, 3 Mg++, 2 Cl− B-factor (Å2) Protein atoms (chain) 68.4 22.8 A:53.8; B:58.2 A:13.4; B:14.9 Water molecules 59.1 29.3 40.0 23.7 r.m.s. deviations bond lengths (Å) 0.015 0.006 0.01 0.009 bond angles (°) 1.42 0.920 1.2 1.2 Ramachandran ploti Favored regions (%) 94.3 98.9 96.2 98.5 Outliers (%) 0.0 0.0 0.3 0 aNumbers in parenthesis correspond to highest resolution shell. bRmerge= Σ|I − <I>|/ΣI, where I is the observed intensity and <I> the average intensity obtained from multiple measurements. Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Rajan et al. Page 20 cRmeas as described in Diederichs and Karplus (Diederichs and Karplus, 1997). d≪I>/σ(<I>)> = Mean Ih over the standard deviation of the mean Ih averaged over all reflections in a resolution shell. eRwork= Σ ||Fo| − |Fc||/Σ|Fo|, where |Fo| is the observed structure factor amplitude and |Fc| the calculated structure factor amplitude. fRfree: Rfactor based on 5% of the data excluded from refinement. gTotal number of protein atoms, including those in alternate conformations. hGm: guanidinum ion. iAs reported by Molprobity (Davis et al., 2004). Structure. Author manuscript; available in PMC 2011 July 14.
3M7F
Crystal structure of the Nedd4 C2/Grb10 SH2 complex
Structural Basis for the Interaction between the Growth Factor-binding Protein GRB10 and the E3 Ubiquitin Ligase NEDD4* Received for publication,May 10, 2010, and in revised form, October 11, 2010 Published, JBC Papers in Press,October 26, 2010, DOI 10.1074/jbc.M110.143412 Qingqiu Huang1 and Doletha M. E. Szebenyi From MacCHESS, Cornell University, Ithaca, New York 14853 In addition to inhibiting insulin receptor and IGF1R kinase activity by directly binding to the receptors, GRB10 can also negatively regulate insulin and IGF1 signaling by mediating insulin receptor and IGF1R degradation through ubiquitina- tion. It has been shown that GRB10 can interact with the C2 domain of the E3 ubiquitin ligase NEDD4 through its Src ho- mology 2 (SH2) domain. Therefore, GRB10 might act as a con- nector, bringing NEDD4 close to IGF1R to facilitate the ubiq- uitination of IGF1R by NEDD4. This is the first case in which it has been found that an SH2 domain could colocalize a ubiq- uitin ligase and its substrate. Here we report the crystal struc- ture of the NEDD4 C2-GRB10 SH2 complex at 2.0 A˚ . The structure shows that there are three interaction interfaces be- tween NEDD4 C2 and GRB10 SH2. The main interface centers on an antiparallel -sheet composed of the F -strand of GRB10 SH2 and the C -strand of NEDD4 C2. NEDD4 C2 binds at nonclassical sites on the SH2 domain surface, far from the classical phosphotyrosine-binding pocket. Hence, this in- teraction is phosphotyrosine-independent, and GRB10 SH2 can bind the C2 domain of NEDD4 and the kinase domain of IGF1R simultaneously. Based on these results, a model of how NEDD4 interacts with IGF1R through GRB10 has been pro- posed. This report provides further evidence that SH2 do- mains can participate in important signaling interactions be- yond the classical recognition of phosphotyrosine. The GRB7 (growth factor receptor-binding protein) family of adaptor proteins includes GRB7, GRB10, and GRB14. These proteins share a conserved molecular architecture: a proline-rich N-terminal region, a Ras-associating-like do- main, a pleckstrin homology domain, a family-specific BPS region, and a conserved C-terminal Src homology 2 (SH2)2 domain (1, 2). Their SH2 domains have the ability to recog- nize phosphotyrosine-containing peptides on a variety of acti- vated tyrosine kinase receptors. GRB7 has been shown to in- teract with EGF receptor, ErB2 receptor, EphB1, focal adhesion kinase, and platelet-derived growth factor receptor and to be involved in regulating cell migration (3, 4). GRB10 and GRB14 have been shown to interact with insulin receptor (IR), IGF1R (insulin-like growth factor 1 receptor), EGF re- ceptor, Raf1 kinase, and MEK1 kinase and to be involved in cell growth regulation (5–11). The Grb10 gene is maternally imprinted in mice. When the Grb10 gene was disrupted by a gene trap insertion, the mutant mice were 30% greater in size than normal, with disproportionately large livers (5, 12). As adults, these mutant mice had improved glucose tolerance, increased muscle mass, and reduced adiposity (13, 14). Fur- thermore, Grb10 transgenic mice overexpressing GRB10 showed growth retardation and insulin resistance (15). These results indicate that GRB10 plays a negative role in cell growth, as a consequence of hypernegative regulation of the IR and IGF1R. Mice lacking the Grb14 gene were of normal size and had improved glucose tolerance and increased insu- lin signaling in muscle and liver (16). Therefore, GRB10 and GRB14 are tissue-specific negative regulators of insulin and IGF1 signaling. Additional research results indicate that GRB10 and GRB14 might contribute to type 2 (non-insulin-dependent) diabetes in humans (5, 17, 18). From a genome-wide associa- tion scan in the Old Order Amish, the GRB10 gene has been identified as having the strongest association between type 2 diabetes and a single nucleotide polymorphism (SNP) (19). In subcutaneous adipose tissue, the GRB14 mRNA levels in type 2 diabetes patients were 43% higher than those in normal per- sons (18). The mechanisms of the negative regulation of IR and IGF1R by GRB10 and GRB14 are not yet clear. Biochemical studies have shown that GRB10 and GRB14 can bind IR and IGF1R through their BPS and SH2 domains (20–22). The GRB14 BPS region binds as a pseudosubstrate inhibitor in the tyrosine kinase domain of IR to suppress insulin signaling (20). Suppressing endogenous GRB10 expression led to in- creased IR protein levels, whereas overexpression of GRB10 led to reduced IR protein levels (7). Reduced IR levels were observed in cells with prolonged insulin treatment, and this reduction was inhibited in GRB10-deficient cells (7). The in- sulin-induced IR reduction was largely reversed by MG-132, a proteasomal inhibitor, but not by chloroquine, a lysosomal inhibitor. IR undergoes insulin-stimulated ubiquitination in cells, and this ubiquitination was inhibited in the GRB10-sup- * This work was supported, in whole or in part, by National Institutes of Health, NCRR, Grant RR-01646. This work is based upon research con- ducted at the Cornell High Energy Synchrotron Source (CHESS), which is supported by the National Science Foundation under Award DMR 0225180. The atomic coordinates and structure factors (code 3M7F) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformat- ics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). 1 To whom correspondence should be addressed: MacCHESS, Cornell Uni- versity, Ithaca, NY 14853. Tel.: 607-255-9386; Fax: 607-255-9001; E-mail: qh24@cornell.edu. 2 The abbreviations used are: SH2, Src homology 2; IR, insulin receptor; IPTG, isopropyl 1-thio--D-galactopyranoside; r.m.s., root mean square. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 53, pp. 42130–42139, December 31, 2010 © 2010 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. 42130 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 53•DECEMBER 31, 2010 pressed cell line (7). Therefore, in addition to inhibiting IR kinase activity by directly binding to IR through its BPS and SH2 domains, GRB10 also negatively regulates insulin signal- ing by affecting insulin-stimulated degradation of the recep- tor, possibly through regulation of ubiquitination of the receptor. GRB10 and GRB14 do not themselves have ubiquitin ligase activity. However, it has been shown that GRB10 binds the C2 domain of the E3 ubiquitin ligase NEDD4 through its SH2 domain, and formation of the complex promotes IGF1-stimu- lated multiubiquitination, internalization, and degradation of the IGF1R (23–25). GRB10 and NEDD4 remain associated with the IGF1R in early endosomes and caveosomes (23). Therefore, the SH2 domain of GRB10 might act as a connec- tor between the C2 domain of NEDD4 and the kinase domain of IGF1R; proximity of NEDD4 and IGF1R would facilitate the ubiquitination of IGF1R by NEDD4. However, SH2 is typ- ically highly specialized for the recognition of phosphoty- rosine and has only one binding site for phosphotyrosine- containing peptides (the Tyr(P)-binding pocket). It was not clear how the SH2 domain of GRB10 can bind the C2 domain of NEDD4 and the kinase domain of IGF1R simultaneously, although the possibility of a second binding site was suggested by the observation that the NEDD4 C2-GRB10 SH2 interac- tion is phosphorylation-independent (25). Here we report the crystal structure of the NEDD4 C2-GRB10 SH2 complex, which clearly shows that the SH2 domain of GRB10 binds the C2 domain of NEDD4 through a surface region distinct from the classical Tyr(P)-binding pocket. EXPERIMENTAL PROCEDURES Protein Expression and Purification—The cDNA encoding the mouse GRB10 SH2 domain (residues 429–536) was am- plified from the plasmid pcDNA-Grb10 (a kind gift from Dr. Junlin Guan) and inserted into the expression vector pQE80 to make a pQE80-SH2 expression construct. This construct was transferred into BL21(DE3) (Novagen) for protein expres- sion. Protein expression was induced at 37 °C for 4 h with 0.3 mM IPTG. The cells were harvested by centrifugation and then suspended in binding buffer (500 mM NaCl, 50 mM Tris- HCl, pH 8.5, 10 mM imidazole, 5 mM -mercaptoethanol, and 1 mM benzamidine chloride). Cell lysis was carried out by son- ication. After centrifugation, the supernatant was applied to a nickel affinity column. After protein binding, the column was washed thoroughly with 100 volumes of binding buffer fol- lowed by 10 volumes of washing buffer (500 mM NaCl, 50 mM Tris-HCl, pH 8.5, 40 mM imidazole, 5 mM -mercaptoethanol, and 1 mM benzamidine chloride). The protein was then eluted from the column with 5 volumes of elution buffer (200 mM NaCl, 300 mM imidazole-HCl, pH 7.5, 5 mM -mercaptoetha- nol, and 1 mM benzamidine chloride). The protein solution was concentrated and further purified by FPLC using a Superdex 200 column (GE Healthcare) with an elution buffer containing 0.15 M NaCl and 5 mM Tris-HCl (pH 7.5). The cDNA encoding the full-length mouse GRB10 (an isoform of GRB10) was inserted into the expression vector pMAL-2C to make a pMAL-2C-Grb10 expression construct. This construct was transferred into BL21(DE3) for protein expression. The harvested cells were lysed in the binding buffer (100 mM NaCl, 50 mM Tris-HCl, pH 8.5, 5 mM -mer- captoethanol, and 1 mM benzamidine chloride) by sonication, and the supernatant was applied to a maltose-agarose column. The column was washed thoroughly with 100 volumes of binding buffer, and the maltose-binding protein-GRB10 fu- sion protein was eluted with 3 volumes of elution buffer (100 mM NaCl, 50 mM maltose, 50 mM Tris-HCl, pH 8.0, 2 mM CaCl2). After the addition of Factor Xa, this elution solution was incubated at 4 °C for 48 h to cleave off the maltose-bind- ing protein tag. The protein solution was then concentrated and purified with a Superdex 200 column. The cDNA encoding the mouse NEDD4 C2 domain (resi- dues 108–287) was amplified from the plasmid pBS-Nedd4 (a kind gift from Dr. Andrea Morrione) and inserted into the expression vector pSUMO (Invitrogen) to make a pSUMO- Nedd4 C2 expression construct. This construct was trans- ferred into BL21(DE3) for protein expression. Protein expres- sion was induced at 22 °C overnight with 0.3 mM IPTG. The protein was purified with a nickel column using the method described above. After the column was washed thoroughly with 100 volumes of binding buffer followed by 10 volumes of washing buffer, the protease SUMOase was added to the col- umn and incubated at 4 °C for 24 h to cleave off the SUMOHis tag. The free protein C2 was eluted, concentrated, and further purified with a Superdex 200 column. Trunca- tions of NEDD4 C2 (residues 108–250, 200–300, and 115– 287, respectively) were cloned, expressed, and purified using the same procedure as for wild type NEDD4 C2. To express GRB10 SH2 without any tag, the cDNA encod- ing the mouse GRB10 SH2 domain (residues 429–536) was inserted into the expression vector pCDFDuet-1 to make pCDFDuet-SH2. To prepare the NEDD4 C2-GRB10 SH2 complex for crystallization, the plasmids pSUMO-C2 and pCDFDuet-SH2 were co-transferred into the host cell BL21(DE3) for co-expression. In this system, the expression level of NEDD4 C2-SUMOHis was higher than that of GRB10 SH2. Protein expression was induced at 15 °C for 24 h with 0.3 mM IPTG. The cell pellet was suspended in binding buffer (200 mM NaCl, 50 mM Tris-HCl, pH 8.5, 10 mM imidaz- ole, 5 mM -mercaptoethanol, and 1 mM benzamidine chlo- ride) and lysed by sonication. After centrifugation, the super- natant was applied to a nickel affinity column. After protein binding, the column was washed thoroughly with 100 vol- umes of binding buffer followed by 10 volumes of washing buffer (200 mM NaCl, 50 mM Tris-HCl, pH 8.5, 40 mM imidaz- ole, 5 mM -mercaptoethanol, and 1 mM benzamidine chlo- ride). Then protease SUMOase was added to the column and incubated at 4 °C for 24 h to cleave off the SUMOHis tag. The free proteins (NEDD4 C2-GRB10 SH2 complex and isolated NEDD4 C2) were eluted, concentrated, and purified with a Superdex 200 column. The fractions containing the NEDD4 C2-GRB10 SH2 complex were collected and further purified using a SourceQ column (GE Healthcare) for FPLC. Effect of Calcium on the Interaction—GRB10 SH2 (with an N-terminal His tag, GRB10 SH2His) was mixed with NEDD4 C2 (without tag) in a 1:1 molar ratio (as judged by A280) in the incubating buffer (150 mM NaCl and 50 mM Tris-HCl, pH 8.5) Structure of the NEDD4 C2-GRB10 SH2 Complex DECEMBER 31, 2010•VOLUME 285•NUMBER 53 JOURNAL OF BIOLOGICAL CHEMISTRY 42131 containing 20 mM EGTA or 20 mM CaCl2, on ice, for 2 h. Sim- ilarly, GRB10 (without tag) was mixed with NEDD4 C2SUMOHis in a 1:1 molar ratio in the incubating buffer containing 20 mM EGTA or 20 mM CaCl2, on ice, for 2 h. Then EGTA or CaCl2 in the mixtures was removed by a sizing column. The fraction containing proteins from the sizing col- umn was collected and applied onto a nickel affinity column. After protein binding, the column was washed with 100 vol- umes of incubating buffer and further washed with 10 vol- umes of washing buffer (150 mM NaCl, 50 mM Tris-HCl, and 40 mM imidazole, pH 8.5). 15 l of gel slurry was drawn for analysis by SDS-PAGE. Effect of Ionic Strength on the Interaction—GRB10 SH2His and NEDD4 C2 were mixed in a 1:1 molar ratio in the incu- bating buffer (50 mM Tris-HCl, pH 8.5). After being incubated on ice for 2 h, the NEDD4 C2-GRB10 SH2His complex was bound to a nickel column. Then the column was eluted with incubating buffer containing different NaCl concentrations (0, 100, 200, 300, 400, 500, and 1000 mM, sequentially). For each concentration, 2 volumes of incubating buffer were used. The eluted solutions were analyzed by SDS-PAGE. The Specificity of the Interaction—The cDNAs encoding mouse GRB14 SH2 domain and GRB7 SH2 domain were sep- arately inserted into the expression vector pGEX4T-1. The resulting constructs pGEX4T-GRB14 SH2 and pGEX4T- GRB7 SH2 were co-transferred, respectively, with pSUMO- NEDD4 C2 into the host cell BL21(DE3). Protein expression was induced with 0.3 mM IPTG at 15 °C for 24 h. The cells were collected and resuspended in PBS for lysis by sonication. The supernatant from the cell lysis was purified using a nickel affinity column and a glutathione-agarose column sequen- tially. After the column was washed thoroughly, 15 l of gel slurry was drawn and analyzed by SDS-PAGE. The cDNAs encoding mouse GRB14 and GRB7 were sepa- rately inserted into the expression vector pQE80 to express the target proteins with N-terminal His tags. Purified GRB14His (or GRB7His) was mixed with NEDD4 C2 (with- out tag) in a 1:1 molar ratio in the incubating buffer (150 mM NaCl and 50 mM Tris-HCl, pH 8.5). After being incubated on ice for 2 h, the mixture was applied to a nickel affinity col- umn. After protein binding, the column was washed by 100 volumes of incubating buffer followed by 10 volumes of wash- ing buffer (150 mM NaCl, 50 mM Tris-HCl, and 40 mM imid- azole, pH 8.5). Then 15 l of gel slurry was drawn for analysis by SDS-PAGE. Crystallization and Data Collection—The purified NEDD4 C2-GRB10 SH2 complex showed two protein bands on SDS- polyacrylamide gel, corresponding to NEDD4 C2 and GRB10 SH2, respectively (Fig. 1A). The NEDD4 C2-GRB10 SH2 com- plex solution was desalted and concentrated to 20 mg/ml. Crystals were grown by the hanging drop vapor diffusion method at 18 °C. Typically, 2 l of the protein stock was mixed with 2 l of the reservoir solution consisting of 35% MPD (v/v) and 0.1 M Hepes-HCl buffer, pH 7.5. Crystals were observed after 2 days and reached a typical size of 100  100  200 m3 1 month later. Diffraction data were collected at 100 K on beamline A1 at MacCHESS. The diffraction data were reduced using the HKL package (26), and the statistics of data collection and processing are summarized in Table 1. Structure Refinement—The structure of the NEDD4 C2- GRB10 SH2 complex was solved by the molecular replace- ment program Phaser (27) using the structure of human GRB10 SH2 (28) (Protein Data Bank code 1NRV) and the structure of human NEDD4 C2 (Protein Data Bank code 3B7Y) as the search models. The structure was refined using Refmac5 (29) and PHENIX (30). The refinement statistics are given in Table 1. The B factors are constant throughout the protein. For the GRB10 SH2 subunit (average B factor  33.7 Å2), the B factors of its N terminus (residues 429–438) and C terminus (residues 526–535) are 32.1 and 37.6 Å2, respec- tively. For the NEDD4 C2 subunit (average B factor  39.3 Å2), the B factors of its N terminus (residues 112–120) and C terminus (residues 281–287) are 41.1 and 40.5 Å2, respec- tively. The interface between GRB10 SH2 and NEDD4 C2 was analyzed using the program PISA (31) from the CCP4 suite (32). SDS-PAGE analysis of the NEDD4 C2-GRB10 SH2 crys- tals showed that the NEDD4 C2 and the GRB10 SH2 in the crystals had the same molecular weights as the original NEDD4 C2 and the original GRB10 SH2, respectively (Fig. 1B), suggesting that both NEDD4 C2 and GRB10 SH2 in the crystal are complete and have not been degraded. Mutation—Site-directed mutations of GRB10 SH2 (K505P, N519A, R533A, R431A and H434A, respectively) were carried out using the Phusion mutation kit (New England Biolabs). The interaction between the GRB10 SH2 mutants and NEDD4 C2 was analyzed using the co-expression system de- scribed above. RESULTS The NEDD4 C2-GRB10 SH2 Interaction Is Ca2-indepen- dent and Phosphorylation-independent—As shown by the yeast two-hybrid method, GRB10 can interact with the C2 domain of NEDD4 through its SH2 domain (25). The C2 domain is a conserved module of about 120 residues, which was first found in protein kinase C (33). In most proteins, the C2 do- main binds phospholipids in a Ca2-dependent manner (34, 35); in some proteins, it mediates protein-protein interac- tions (36). The C2 domain of NEDD4 has been shown to target NEDD4 to the plasma membrane in response to Ca2 (37). However, the interaction between NEDD4 and GRB10 has been demonstrated to be Ca2-independent by the co-immunoprecipitation method (25). We performed a series of in vitro interaction experiments using purified C2 domain and purified SH2 domain (or full-length GRB10). As shown in Fig. 1A, NEDD4 C2 could form a complex with GRB10 SH2 in the presence of either 20 mM EGTA or 20 mM CaCl2, suggesting that the interaction between NEDD4 C2 and GRB10 SH2 was Ca2-independent. NEDD4 C2 could also form a complex with full-length GRB10 (an isoform of GRB10), and this interaction was also Ca2-independent (Fig. 1A). The SH2 domain is a conserved module of about 100 resi- dues, which is highly specialized for the recognition of phos- photyrosine, with only a few exceptions to date (38). How- ever, GRB10 has been suggested to preferentially associate Structure of the NEDD4 C2-GRB10 SH2 Complex 42132 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 53•DECEMBER 31, 2010 with unphosphorylated NEDD4 (based on co-immunopre- cipitation experiments) (25). In our work, NEDD4 C2 was expressed in Escherichia coli, in which expressed proteins cannot be phosphorylated. The NEDD4 C2 domain produced in this system could form a complex with GRB10 SH2 or GRB10, indicating that binding of GRB10 SH2 to NEDD4 C2 is phosphorylation-independent. Both the NEDD4 C2-GRB10 SH2 complex and the NEDD4 C2-GRB10 complex were stable in low ionic strength solu- tion (200 mM NaCl). The complexes dissociated in a solu- tion containing more than 500 mM NaCl, in the presence or absence of 20 mM EGTA (or 20 mM CaCl2) (data not shown). These results suggested that the main interaction between NEDD4 C2 and GRB10 SH2 was not hydrophobic and was unaffected by Ca2. The molecular mass of GRB10 SH2 is about 12 kDa. The molecular mass of NEDD4 C2 is about 19 kDa. As determined by a Superdex200 sizing column, the molecular mass of the NEDD4 C2-GRB10 SH2 complex was about 30 kDa, suggest- ing that this complex is a heterodimer in solution (Fig. 1C). The Specificity of the Interaction—All members of the GRB7 adaptor protein family (GRB7, GRB10, and GRB14) contain a Ras-associating-like domain, a pleckstrin homology domain, a family-specific BPS region, and a C-terminal SH2 domain (1, 2). Because the SH2 domains are highly conserved among this family (Fig. 2), we investigated whether NEDD4 C2 could also bind the SH2 domains of GRB7 and GRB14, using the co- expression method. In the E. coli co-expression system, NEDD4 C2 was expressed with a SUMOHis tag, which could FIGURE 2. Alignment of the amino acid sequences of the SH2 domains of GRB7, GRB10, and GRB14. The -helices and -strands of GRB10 SH2 are highlighted with red lines and blue arrows, respectively. The underlined peptide of GRB10 is the F -strand, which forms an antiparallel -sheet with the C -strand of NEDD4 C2 in the NEDD4 C2-GRB10 SH2 complex. The resi- dues buried in the interface with NEDD4 C2 are shown with red letters. The two residues in interface I (Lys505 and Asn519) chosen for mutation are high- lighted as red boldface letters. TABLE 1 Data collection and structure refinement statistics Parameters Values Data collection Space group P212121 Cell dimensions a, b, c (Å) 52.31 70.30 86.47 Resolution (Å) 50-2.0 (2.03-2.00)a No. of unique observations 21,866 (1023) Redundancy 6.5 (4.4) Completeness (%) 99.4 (96.8) Average I/I 35.7 (3.4) Rmerge (%) 5.2 (36.2) Structure refinement No. of protein atoms/waters 2011/95 Resolution (Å) 50-2.0 Rwork (%) /Rfree (%) 19.2/23.2 r.m.s. deviations Bonds (Å)/Angles (degrees) 0.006/1.057 Average B factor (Å2) 36.78 Ramachandran plot Most favored (%) 95.34 Allowed (%) 4.66 a Values in parentheses are for the highest resolution shell. FIGURE 1. Interaction between NEDD4 C2 and GRB10 SH2. The two pro- teins were mixed and incubated in the incubating buffer for 2 h. After being desalted with a sizing column, the mixture was applied onto a nickel affinity column. After being washed with washing buffer, 15 l of gel slurry was drawn for analysis by SDS-PAGE (for details, see “Experimental Procedures”). A, GRB10 could form a stable complex with NEDD4 C2SUMOHis in the presence of 20 mM EGTA (lane 3) or 20 mM CaCl2 (lane 4) or without any ad- ditive (lane 5); GRB10 alone did not bind to the nickel beads (lane 2); NEDD4 C2SUMOHis alone is shown in lane 1. GRB10 SH2His could form a stable complex with NEDD4 C2 in the presence of 20 mM EGTA (lane 7) or 20 mM CaCl2 (lane 8) or without any additive (lane 9); NEDD4 C2 alone did not bind to the nickel beads (lane 10); GRB10 SH2His alone is shown in lane 11. Lane 6, standards. B, wild type (lane 1) and the mutants N519A (lane 3), R533A (lane 12), R431A (lane 13), and H434A (lane 14) of GRB10 SH2His could form a stable complex with NEDD4 C2, whereas the K505P mutant (lane 2) could not form a complex with NEDD4 C2. NEDD4 C2 alone could not bind to the nickel beads (lane 4); wild type GRB10 SH2His alone could bind to the nickel beads (lane 10). Neither the truncated NEDD4 C2 (resi- dues 108–250) (lane 5) nor the truncated NEDD4 C2 (residues 200–300) (lane 6) could bind to GRB10 SH2His; the truncated NEDD4 C2 (residues 115–287) could bind to GRB10 SH2His (lane 7). Crystals of the NEDD4 C2- GRB10 SH2 complex were picked out from the drop and analyzed by SDS- PAGE (lane 9). Lanes 8 and 11, standards. C, elution profile of the NEDD4 C2-GRB10 SH2His complex from a Superdex 200 sizing column. The inset shows the analysis of the peak fractions by SDS-PAGE. Structure of the NEDD4 C2-GRB10 SH2 Complex DECEMBER 31, 2010•VOLUME 285•NUMBER 53 JOURNAL OF BIOLOGICAL CHEMISTRY 42133 Structure of the NEDD4 C2-GRB10 SH2 Complex 42134 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 53•DECEMBER 31, 2010 bind to a nickel-agarose column, whereas GRB7 SH2 and GRB14 SH2 were expressed with a GST tag, which could bind to a glutathione-agarose column but not a nickel column. If the SH2 domain can form a complex with C2SUMOHis, it should be co-purified with C2SUMOHis by a nickel column. As shown by SDS-PAGE analysis, the protein purified by a nickel column contained only NEDD4 C2SUMOHis, whereas the protein purified by a glutathione-agarose column contained only GRB7 SH2GST or GRB14 SH2GST. Al- though both NEDD4 C2SUMOHis and GRB7 SH2GST (or GRB14 SH2GST) were co-expressed in the same cell at high levels, neither of the SH2 domains could be co-purified with NEDD4 C2SUMOHis by a nickel column, suggesting that neither GRB7 SH2 nor GRB14 SH2 could form a complex with NEDD4 C2. As a control, we showed that GRB10 SH2GST could form a complex with NEDD4 C2SUMOHis, suggesting that neither the GST tag nor the SUMO tag could inhibit the interaction between NEDD4 C2 and the SH2 domain. In vitro interactions between NEDD4 C2 and full-length GRB7 (or GRB14) were also investigated using purified pro- teins. As shown by SDS-PAGE analysis, NEDD4 C2 could not form a complex with GRB7 or GRB14. Crystal Structure of the NEDD4 C2-GRB10 SH2 Complex— To gain further insights into the interaction between NEDD4 C2 and GRB10 SH2, we co-expressed NEDD4 C2 and GRB10 SH2 in E. coli and determined the crystal structure of the NEDD4 C2-GRB10 SH2 complex at 2.0 Å resolution (Protein Data Bank code 3M7F). Data collection and refinement statis- tics are given in Table 1. The NEDD4 C2-GRB10 SH2 complex exists as a het- erodimer in the crystal. The most interesting features of the NEDD4 C2-GRB10 SH2 complex structure are the interaction interfaces between the two proteins (Fig. 3A). Interface I is the major interface, with a buried area of 530 Å2. Central to this interface is a small antiparallel -sheet formed from the F -strand (residues 502–508) of GRB10 SH2 and the C -strand (residues 153–160) of NEDD4 C2 (Fig. 3B). It has been shown that a Pro residue can disrupt a -strand in pro- teins (39). Therefore, the Lys505 residue of GRB10 SH2 was mutated to Pro to disrupt the F -strand. The resultant K505P mutant of GRB10 SH2 could not form a complex with NEDD4 C2 in the E. coli co-expression system as wild type GRB10 SH2 did (Fig. 1B). This result suggests that this antipa- rallel -sheet plays an important role in stabilizing the NEDD4 C2-GRB10 SH2 complex. Residues in -strand E and helix B of GRB10 SH2 and in -strands A, B, and D of NEDD4 C2 also contribute to interface I. There are a total of eight hydrogen bonds between SH2 and C2 in this interface, four main chain-main chain in the small -sheet and four side chain-main chain elsewhere. Two of the latter are from the side chain of Asn519 in helix B of SH2 to main chain nitrogen and oxygen atoms of Leu177 in the D -strand of C2 (Fig. 3C). The N519A mutant of GRB10 SH2 could form a complex with NEDD4 C2 (Fig. 1B), suggesting that this pair of hydro- gen bonds is not critical to the formation of the GRB10 SH2- NEDD4 C2 complex. There is a possible salt bridge between the acid residue Asp514 of GRB10 SH2 and the alkaline resi- due Arg179 of NEDD4 C2 (distance 4 Å). In this interface, there are also hydrophobic interactions between the two pro- teins, involving residues Phe496, Phe506, Leu512, Phe515, Tyr516, and Leu518 of GRB10 SH2 with Leu148, Ile155, Leu156, Val159, Ile176, Leu177, and Phe178 of NEDD4 C2 (Fig. 3C). Interface II is the smallest interface, with an area of 90 Å2, comprising the N terminus (residues 429–434) and part of the B helix (residue Ile510) of GRB10 SH2 and the N terminus (residues 112–114) of NEDD4 C2 (Fig. 3A). There is one hy- drogen bond between the side chain of residue Gln433 of GRB10 SH2 and the carbonyl group of residue Glu112 of NEDD4 C2 (Fig. 4B). A truncated NEDD4 C2 (residues 115– 287) lacking the N terminus could form a stable complex with GRB10 SH2 (Fig. 1B), suggesting that interface II does not play an important role in stabilizing the NEDD4 C2-GRB10 SH2 complex. Interface III is a medium-sized interface, with an area of 312 Å2. The proline-rich C terminus (residues 283–287) of NEDD4 C2 rests against a surface made up of an N-terminal part (residues 431–440) and a C-terminal part (residues 532– 535) of GRB10 SH2 (Fig. 3A). There is a hydrogen bond be- tween the side chain of residue Arg533 of GRB10 SH2 and the carbonyl group of residue Pro284 of NEDD4 C2 (Fig. 4C) and a salt bridge between the side chain of Arg431 in SH2 and the C-terminal carboxyl group of C2 (residue Pro287) (Fig. 4D). The particular importance of this interface is suggested by the observation that a 28-residue stretch preceding the C-termi- nal Pro-rich cluster of C2 is disordered (and therefore invisi- ble) in the crystal, but the cluster itself is well ordered and clearly defined in electron density maps. A truncated NEDD4 C2 (residues 108–250), lacking the C-terminal 37 residues, could not form a stable complex with GRB10 SH2 (Fig. 1B), suggesting that interface III plays an important role in stabi- lizing the NEDD4 C2-GRB10 SH2 complex. A single site mu- tation of SH2 at this interface (R533A, R431A, or H434A) could not disrupt the NEDD4 C2-GRB10 SH2 complex (Fig. 1B), indicating that interface III involves a number of inter- acting residues and is not easily disrupted. NEDD4 C2 trun- cated to residues 200–300, however, could not form a stable complex with GRB10 SH2 (Fig. 1B), suggesting that interface III by itself is not sufficient for the formation of the NEDD4 C2-GRB10 SH2 complex. FIGURE 3. Crystal structure of the NEDD4 C2-GRB10 SH2 complex. A, ribbon diagram of the entire complex. There are three interaction interfaces be- tween NEDD4 C2 and GRB10 SH2. Interface I includes the F -strand and the B helix of GRB10 SH2 and (primarily) the C and D -strands of NEDD4 C2. Inter- face II includes the N termini of both NEDD4 C2 and GRB10 SH2. Interface III includes the N- and C-terminal regions of GRB10 SH2 and the C terminus of NEDD4 C2. Both the Tyr(P)-binding pocket on GRB10 SH2 and the Ca2-binding site on NEDD4 C2 are far from these interfaces. The peptides without elec- tron density are shown as dashed lines. B, the antiparallel -sheet formed between the C -strand of NEDD4 C2 and the F -strand of GRB10 SH2 in interface I, containing four hydrogen bonds (shown as red lines with lengths in angstroms). Only the main chains are shown. C, the hydrophobic interactions be- tween NEDD4 C2 and GRB10 SH2 in interface I. The two hydrogen bonds between Asn519 of GRB10 SH2 and Leu177 of NEDD4 C2 are shown as red lines with lengths in angstroms. Only the side chains are shown (except for residue Leu177 of NEDD4 C2, whose main chain and side chain are shown). Structure of the NEDD4 C2-GRB10 SH2 Complex DECEMBER 31, 2010•VOLUME 285•NUMBER 53 JOURNAL OF BIOLOGICAL CHEMISTRY 42135 Structure of the NEDD4 C2-GRB10 SH2 Complex 42136 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 53•DECEMBER 31, 2010 Because the main interactions between NEDD4 C2 and GRB10 SH2 are hydrogen bonds and salt bridges, the complex is expected to be unstable in high salt solution. In fact, as shown above, the complex dissociates in a solution containing more than 500 mM NaCl. Phosphotyrosine-containing peptides/proteins bind to the Tyr(P)-binding pocket on the surface of an SH2 domain (the classical SH2 binding site) (Fig. 3A). On the surface of GRB10 SH2, this pocket is more than 15 Å away from the three NEDD4 C2-GRB10 SH2 interfaces and has no overlap with them (Fig. 3A); NEDD4 C2 binds to a nonclassical binding site on GRB10 SH2, which explains why the interaction between NEDD4 C2 and GRB10 SH2 is phosphorylation-independent. Usually, a C2 domain binds Ca2 through its AB and EF loops (Fig. 3A). This Ca2-binding site is more than 12 Å away from the three NEDD4 C2-GRB10 SH2 interaction in- terfaces, and no Ca2 was found in the interfaces. Hence, the NEDD4 C2-GRB10 SH2 interaction is expected to be Ca2- independent, as observed in binding studies (Fig. 1A). Structure Comparison—Superposition of the GRB10 SH2 subunit of the NEDD4 C2-GRB10 SH2 complex with the free GRB10 SH2 domain shows that binding of NEDD4 C2 has not resulted in significant conformational change of GRB10 SH2 (r.m.s. deviation  0.37 Å) (Fig. 4A), except that the side chain of residue Gln511 has moved from a position that would clash with the main chain of residue Leu177 of NEDD4 C2 to a location well clear of Leu177. The structure of GRB10 SH2 is similar to those of GRB14 SH2 (r.m.s. deviation 0.92 Å) and GRB7 SH2 (r.m.s. deviation 0.95 Å). Superposition of GRB14 SH2 and GRB7 SH2 onto the GRB10 SH2 subunit of the NEDD4 C2-GRB10 SH2 complex shows that the main differences among these three structures are located at the BC loop, the DE loop, and the N and the C termini (Fig. 4A). Almost all of the residues involved in interface I between SH2 and NEDD4 C2 are identical or very similar among the three species. One point of difference is the contact between residues 494–496 (QTF) of GRB10 SH2 and residues 153– 155 (SGI) of NEDD4 C2. The sequence of the corresponding SH2 residues is EMF in GRB14 and RLY in GRB7 (i.e. a charged rather than merely polar residue is introduced in the first position, and, in the case of GRB7, the large Tyr replaces Phe in the third position). Additionally, the position of the main chain is slightly shifted, relative to the central -sheet contact, in the other two proteins versus GRB10. In the other two interfaces, significant differences are ap- parent among the superposed structures. In interface II, the conformation of the N terminus of GRB7 SH2 is quite differ- ent from that of GRB10 and would clash with NEDD4 C2 (Fig. 4B). GRB14 SH2 is less discrepant but does show a posi- tional shift relative to GRB10 as well as a sequence difference in one position, which could affect binding. In interface III, both GRB7 and GRB14 exhibit significant differences in their SH2 C-terminal conformations from GRB10, such that the fit of the NEDD4 C2 Pro-rich cluster against SH2 becomes very poor (Fig. 4, C and D). The possi- bility exists that the GRB7 and/or GRB14 SH2 conformation could be modified in the presence of NEDD4 C2 to create more favorable interactions with C2, although in the case of GRB14, a sequence change from Gly to His at position 438 (GRB10 numbering) would create a clash with Pro286 in NEDD4 C2 even if the main chain conformation were the same as in GRB10 (Fig. 4D). In any case, the binding studies described above argue strongly against such a conformational change. NEDD4 C2 contains eight -strands and no -helices. Comparison of the NEDD4 C2 subunit of the NEDD4 C2- GRB10 SH2 complex with the free NEDD4 C2 domain (su- perposition r.m.s. deviation 0.53 Å) shows that the main dif- ferences between the two structures are located at the BC loop and the N-terminal part, regions that interact with the GRB10 SH2 subunit. DISCUSSION SH2 domains are well known as binding modules for phos- photyrosine-containing peptides. The classical SH2/Tyr(P)- containing peptide interactions play important roles in nu- merous cell signaling pathways (38, 40). A typical SH2 domain comprises a seven-stranded -sheet core flanked by two -helices (41). SH2 domains are highly specialized for the rec- ognition of Tyr(P) residues. On the classical substrate-binding site of an SH2 domain, the target Tyr(P)-containing peptide is usually bound in an extended conformation. The highly con- served Tyr(P)-binding site contains an invariant arginine and a second positively charged residue coordinating the phos- phate moiety (42). Residues in the target peptide that are lo- cated downstream of the Tyr(P) residue confer specificity on the interaction (43). It has been shown that the SH2 domains of GRB10 and GRB14 can bind the activation loop of the kinase domains of FIGURE 4. Superposition of free GRB10 SH2 (blue; Protein Data Bank code 1NRV), GRB7 SH2 (yellow; Protein Data Bank 2QMS), and GRB14 SH2 (red; Protein Data Bank 2AUG) with the GRB10 SH2 subunit (cyan) of the NEDD4 C2-GRB10 SH2 complex. The NEDD4 C2 subunit of the NEDD4 C2-GRB10 SH2 complex is shown in magenta. A, the structure of GRB10 SH2 is similar to those of GRB14 SH2 and GRB7 SH2, with significant conformational differences located at the BC loop, the DE loop, the N-terminal region, and the C-terminal region (marked with arrows). B, detailed structure of the N-terminal region around residue Gln433. The side chain of residue Gln433 of GRB10 SH2 can form a hydrogen bond with the carbonyl group of residue Glu112 of NEDD4 C2 (shown with a red line), whereas the side chain of residue Gln437 of GRB14 SH2 is too close to the carbonyl group of residue Glu112 of NEDD4 C2 (shown with a blue arrow, 1.49 Å). The N-terminal part of GRB7 SH2 (side chains of His426 and Arg427) clashes with the N-terminal part of NEDD4 C2 (residues Leu113 and His114). C, detailed structure of the C-terminal region around residue Arg533. The side chain of residue Arg533 of GRB10 SH2 can form a hydrogen bond with the carbonyl group of residue Pro284 of NEDD4 C2 (shown with a red line), whereas the side chain of residue Arg537 of GRB14 SH2 is too far away from the carbonyl group of residue Pro284 of NEDD4 C2 (shown with a blue arrow, 3.91 Å) to form a hydrogen bond. Moreover, the side chain of residue Arg537 of GRB14 SH2 is too close to residue Pro286 of NEDD4 C2 (1.5 Å). D, detailed structure of the N-terminal region around residue Arg431. There is a salt bridge interaction between the side chain of residue Arg431 of GRB10 SH2 and the carboxyl group of residue Pro287 of NEDD4 C2 (shown with a red line, 3.22 Å), whereas the side chain of residue Arg435 of GRB14 SH2 is too far away from the carboxyl group of residue Pro287 of NEDD4 C2 (shown with a blue arrow, 8.01 Å) to form a salt bridge interaction. Moreover, the side chain of residue Arg537 of GRB14 SH2 is too close to residue Pro286 of NEDD4 C2 (1.5Å). The side chain of residue His442 of GRB14 SH2 clashes with residue Pro286 of NEDD4 C2, whereas the corresponding residues in GRB10 SH2 and GRB7 SH2 are gly- cines that have no side chain. Structure of the NEDD4 C2-GRB10 SH2 Complex DECEMBER 31, 2010•VOLUME 285•NUMBER 53 JOURNAL OF BIOLOGICAL CHEMISTRY 42137 IR and IGF1R through classical Tyr(P)-SH2 interactions (44). On the other hand, there was evidence showing that GRB10 could form a complex with the E3 ubiquitin ligase NEDD4 through the GRB10 SH2-NEDD4 C2 interaction. This inter- action is phosphotyrosine-independent and Ca2-indepen- dent (25). Furthermore, there was also evidence suggesting the existence of a GRB10-NEDD4-IGF1R complex (23). The crys- tal structure of the GRB10 SH2-NEDD4 C2 complex, re- ported here, provides a structural basis for how the SH2 do- main of GRB10 can bind the C2 domain of NEDD4 and the kinase domain of IGF1R simultaneously. All of the three NEDD4 C2 recognition sites on GRB10 SH2 are far away (more than 15 Å) from, and do not overlap with, the classical Tyr(P)-containing peptide binding pocket (Fig. 3A); binding of the kinase domain of IGF1R at the Tyr(P)-binding pocket (the classical site) does not interfere with binding of the C2 domain of NEDD4. In the NEDD4-GRB10-IGF1R complex, GRB10 serves as a connector to form a bridge between NEDD4 and IGF1R. Although GRB10 can form a complex with the E3 ubiquitin ligase NEDD4, GRB10 is not ubiquitinated by NEDD4 inside the cell. However, IGF1R is ubiquitinated by NEDD4 inside the cell, and binding of GRB10 to NEDD4 is critical for this ubiquitination (25). This is explained by the predicted struc- ture of the NEDD4-GRB10-IGF1R complex, in which GRB10 acts as an adaptor to bring NEDD4 close enough to IGF1R to facilitate ubiquitination of IGF1R by NEDD4, through the C2-SH2-kinase domain interaction (Fig. 5A). There are a few other cases in which an SH2 domain binds proteins using binding sites other than the classical Tyr(P)- binding pocket (45, 46). For example, the Itk kinase domain docking site on the PLC1 SH2C domain surface, which in- cludes residues Glu709, Arg748, Met750, Lys751, and Arg753, is far from and does not overlap with the classical Tyr(P)-bind- ing pocket (47). Some instances have been reported in which SH2 domains use binding sites different from and not overlapping with the classical Tyr(P)-binding pocket to colocalize a kinase and sub- strate (45, 46). As shown in Fig. 5B, the SAP SH2 domain binds simultaneously to the SH3 domain of Fyn kinase and to the Tyr(P)-containing peptide of SLAM. This interaction re- sults in the colocalization of Fyn with its SLAM substrate to facilitate the phosphorylation of SLAM by Fyn (45). In the complex, SLAM binds to the classical Tyr(P)-binding pocket of SAP SH2, whereas Fyn SH3 binds SAP SH2 in a phosphoty- rosine-independent manner, at a binding site outside of the classical Tyr(P)-binding pocket, involving the DE loop and part of the B helix (45). Another example is the interaction between the fibroblast growth factor receptor kinase (FGFR1) and the N-terminal SH2 domain (SH2N) of PLC1 (Fig. 5C) (46). In this complex, there are two interaction sites between PLC1 SH2N and the kinase domain of FGFR1; the Tyr(P)- containing tail of the kinase domain binds the classical Tyr(P)-binding pocket on SH2N, whereas a second interac- tion site involves the BC and DE loops of SH2N. Unlike the above cases in which the SH2 domain colocal- izes a kinase and its substrate, in the GRB10-NEDD4-IGF1R complex, GRB10 SH2 colocalizes a ubiquitin ligase (NEDD4) and its substrate (IGF1R). This case provides further evidence that SH2 domains have a diverse set of other interaction sur- faces besides the classical Tyr(P)-binding pocket and that SH2 domains can colocalize an enzyme and its substrate to facili- tate the reaction between them. Our work also supports the FIGURE 5. Model for the interaction of NEDD4 with IGF1R through GRB10 and examples of SH2 domains that use binding sites different from and not overlapping with the classical Tyr(P)-binding pocket to colocalize a kinase and substrate. A, the interaction between GRB10 BPS (cyan) and the IGF1R kinase domain (green) is modeled based on the crystal structure of the GRB14 BPS-IR kinase complex (25). The interaction between GRB10 SH2 (cyan) and the IGF1R kinase domain is modeled based on the crystal structure of the APS SH2-IR kinase complex (48). The dashed red oval highlights the interaction between GRB10 SH2 and the activation loop of the kinase domain. The missing linker between the BPS and the SH2 do- main of GRB10 is shown as a dashed cyan line. B, structure of the SAP SH2- Fyn SH3-SLAM complex (45). The two dashed red ovals highlight the Tyr(P)- binding pocket and the interaction interface between SAP SH2 and Fyn SH3, respectively. C, structure of the PLC1 SH2N-SH2C-FGFR1 kinase do- main complex (46). The dashed blue oval highlights the Tyr(P)-binding pocket. The dashed purple oval highlights the secondary interaction inter- face between PLC1 SH2N and the FGFR1 kinase domain. Structure of the NEDD4 C2-GRB10 SH2 Complex 42138 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285•NUMBER 53•DECEMBER 31, 2010 conclusion that SH2 domains can participate in important signaling interactions beyond the recognition of phosphotyrosine. REFERENCES 1. Han, D. C., Shen, T. L., and Guan, J. L. (2001) Oncogene 20, 6315–6321 2. Holt, L. J., and Siddle, K. (2005) Biochem. J. 388, 393–406 3. Shen, T. L., and Guan, J. L. (2004) Front. Biosci. 9, 192–200 4. Han, D. C., Shen, T. L., Miao, H., Wang, B., and Guan, J. L. (2002) J. Biol. Chem. 277, 45655–45661 5. Holt, L. J., Lyons, R. J., Ryan, A. S., Beale, S. M., Ward, A., Cooney, G. J., and Daly, R. J. (2009) Mol. Endocrinol. 23, 1406–1414 6. Tezuka, N., Brown, A. M., and Yanagawa, S. (2007) Biochem. Biophys. Res. Commun. 356, 648–654 7. Ramos, F. J., Langlais, P. R., Hu, D., Dong, L. Q., and Liu, F. (2006) Am. J. Physiol. Endocrinol Metab. 290, E1262–E1266 8. Dufresne, A. M., and Smith, R. J. (2005) Endocrinology 146, 4399–4409 9. Murdaca, J., Treins, C., Monthoue¨l-Kartmann, M. N., Pontier-Bres, R., Kumar, S., Van Obberghen, E., and Giorgetti-Peraldi, S. (2004) J. Biol. Chem. 279, 26754–26761 10. Morrione, A. (2003) J. Cell. Physiol. 197, 307–311 11. Giorgetti-Peraldi, S., Murdaca, J., Mas, J. C., and Van Obberghen, E. (2001) Oncogene 20, 3959–3968 12. Charalambous, M., Smith, F. M., Bennett, W. R., Crew, T. E., Mackenzie, F., and Ward, A. (2003) Proc. Natl. Acad. Sci. U.S.A. 100, 8292–8297 13. Smith, F. M., Holt, L. J., Garfield, A. S., Charalambous, M., Koumanov, F., Perry, M., Bazzani, R., Sheardown, S. A., Hegarty, B. D., Lyons, R. J., Cooney, G. J., Daly, R. J., and Ward, A. (2007) Mol. Cell. Biol. 27, 5871–5886 14. Wang, L., Balas, B., Christ-Roberts, C. Y., Kim, R. Y., Ramos, F. J., Kikani, C. K., Li, C., Deng, C., Reyna, S., Musi, N., Dong, L. Q., DeFronzo, R. A., and Liu, F. (2007) Mol. Cell. Biol. 27, 6497–6505 15. Shiura, H., Miyoshi, N., Konishi, A., Wakisaka-Saito, N., Suzuki, R., Mu- guruma, K., Kohda, T., Wakana, S., Yokoyama, M., Ishino, F., and Kaneko-Ishino, T. (2005) Biochem. Biophys. Res. Commun. 329, 909–916 16. Cooney, G. J., Lyons, R. J., Crew, A. J., Jensen, T. E., Molero, J. C., Mitch- ell, C. J., Biden, T. J., Ormandy, C. J., James, D. E., and Daly, R. J. (2004) EMBO J. 23, 582–593 17. Di Paola, R., Wojcik, J., Succurro, E., Marucci, A., Chandalia, M., Pado- vano, L., Powers, C., Merla, G., Abate, N., Sesti, G., Doria, A., and Tris- chitta, V. (2010) J. Intern. Med. 267, 132–133 18. Cariou, B., Capitaine, N., Le Marcis, V., Vega, N., Be´re´ziat, V., Kergoat, M., Laville, M., Girard, J., Vidal, H., and Burnol, A. F. (2004) FASEB J. 18, 965–967 19. Rampersaud, E., Damcott, C. M., Fu, M., Shen, H., McArdle, P., Shi, X., Shelton, J., Yin, J., Chang, Y. P., Ott, S. H., Zhang, L., Zhao, Y., Mitchell, B. D., O’Connell, J., and Shuldiner, A. R. (2007) Diabetes 56, 3053–3062 20. Depetris, R. S., Hu, J., Gimpelevich, I., Holt, L. J., Daly, R. J., and Hub- bard, S. R. (2005) Mol. Cell. 20, 325–333 21. Stein, E. G., Gustafson, T. A., and Hubbard, S. R. (2001) FEBS Lett. 493, 106–111 22. He, W., Rose, D. W., Olefsky, J. M., and Gustafson, T. A. (1998) J. Biol. Chem. 273, 6860–6867 23. Monami, G., Emiliozzi, V., and Morrione, A. (2008) J. Cell. Physiol. 216, 426–437 24. Vecchione, A., Marchese, A., Henry, P., Rotin, D., and Morrione, A. (2003) Mol. Cell. Biol. 23, 3363–3372 25. Morrione, A., Plant, P., Valentinis, B., Staub, O., Kumar, S., Rotin, D., and Baserga, R. (1999) J. Biol. Chem. 274, 24094–24099 26. Otwinoski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307–326 27. McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Sto- roni, L. C., and Read, R. J. (2007) J. Appl. Crystallogr. 40, 658–674 28. Stein, E. G., Ghirlando, R., and Hubbard, S. R. (2003) J. Biol. Chem. 278, 13257–13264 29. Murshudov, G., Vagin, A., and Dodson, E. (1997) Acta Cryst. D53, 240–255 30. Adams, P. D., Afonine, P. V., Bunko´czi, G., Chen, V. B., Davis, I. W., Echols, N., Headd, J. J., Hung, L. W., Kapral, G. J., Grosse-Kunstleve, R. W., McCoy, A. J., Moriarty, N. W., Oeffner, R., Read, R. J., Richard- son, D. C., Richardson, J. S., Terwilliger, T. C., and Zwart, P. H. (2010) Acta Crystallogr. D 66, 213–221 31. Krissinel, E., and Henrick, K. (2007) J. Mol. Biol. 372, 774–797 32. Collaborative Computational Project 4 (1994) Acta Crystallogr. D 50, 760–763 33. Newton, A. C., and Johnson, J. E. (1998) Biochim. Biophys. Acta 1376, 155–172 34. Corbala´n-García, S., and Go´mez-Ferna´ndez, J. C. (2010) Biofactors 36, 1–7 35. Bai, J., and Chapman, E. R. (2004) Trends Biochem. Sci. 29, 143–151 36. Benes, C. H., Wu, N., Elia, A. E., Dharia, T., Cantley, L. C., and Soltoff, S. P. (2005) Cell 121, 271–280 37. Plant, P. J., Yeger, H., Staub, O., Howard, P., and Rotin, D. (1997) J. Biol. Chem. 272, 32329–32336 38. Pawson, T., Gish, G. D., and Nash, P. (2001) Trends Cell Biol. 11, 504–511 39. Li, S. C., Goto, N. K., Williams, K. A., and Deber, C. M. (1996) Proc. Natl. Acad. Sci. U.S.A. 93, 6676–6681 40. Filippakopoulos, P., Mu¨ller, S., and Knapp, S. (2009) Curr. Opin. Struct. Biol. 19, 643–649 41. Sawyer, T. K. (1998) Biopolymers 47, 243–261 42. Waksman, G., Kumaran, S., and Lubman, O. (2004) Expert Rev. Mol. Med. 6, 1–18 43. Sondermann, H., and Kuriyan, J. (2005) Cell 121, 158–160 44. Ceccarelli, D. F., and Sicheri, F. (2009) Nat. Struct. Mol. Biol. 16, 803–804 45. Chan, B., Lanyi, A., Song, H. K., Griesbach, J., Simarro-Grande, M., Poy, F., Howie, D., Sumegi, J., Terhorst, C., and Eck, M. J. (2003) Nat. Cell Biol. 5, 155–160 46. Bae, J. H., Lew, E. D., Yuzawa, S., Tome´, F., Lax, I., and Schlessinger, J. (2009) Cell 138, 514–524 47. Min, L., Joseph, R. E., Fulton, D. B., and Andreotti, A. H. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 21143–21148 48. Hu, J., Liu, J., Ghirlando, R., Saltiel, A. R., and Hubbard, S. R. (2003) Mol. Cell. 12, 1379–1389 Structure of the NEDD4 C2-GRB10 SH2 Complex DECEMBER 31, 2010•VOLUME 285•NUMBER 53 JOURNAL OF BIOLOGICAL CHEMISTRY 42139
3M7G
Structure of topoisomerase domain of topoisomerase V protein
Structures of minimal catalytic fragments of topoisomerase V reveals conformational changes relevant for DNA binding Rakhi Rajan*, Bhupesh Taneja*,†, and Alfonso Mondragón*,‡ * Department of Biochemistry, Molecular Biology and Cell Biology, Northwestern University, 2205 Tech Dr, Evanston, IL 60208 Summary Topoisomerase V is an archaeal type I topoisomerase that is unique among topoisomerases due to presence of both topoisomerase and DNA repair activities in the same protein. It is organized as an N-terminal topoisomerase domain followed by 24 tandem helix hairpin helix (HhH) motifs. Structural studies have shown that the active site is buried by the (HhH) motifs. Here we show that the N-terminal domain can relax DNA in the absence of any HhH motifs and that the HhH motifs are required for stable protein-DNA complex formation. Crystal structures of various topoisomerase V fragments show changes in the relative orientation of the domains mediated by a long bent linker helix, and these movements are essential for the DNA to enter the active site. Phosphate ions bound to the protein near the active site helped model DNA in the topoisomerase domain and shows how topoisomerase V may interact with DNA. Introduction DNA topoisomerases are enzymes found in all forms of life (bacteria, eukarya, and archaea) and they regulate the topological state of DNA inside the cell. They form a transient break in a single or double stranded DNA and allow the passage of another single or double DNA strand through the break, before resealing the break (Champoux, 2001) (Schoeffler and Berger, 2008). As a result of this, topoisomerases can relax supercoiled DNA, help in the segregation of DNA strands following replication, and lead to the formation and resolution of knots and catenates (Gellert, 1981). Topoisomerases participate in many aspects of DNA metabolism, such as replication, recombination, and transcription (Champoux, 2001). In addition, they are targets of various anti-cancerous drugs and anti-bacterial agents (Maxwell, 1999; Pommier, 1998; Rothenberg, 1997; Wang et al., 1997). DNA topoisomerases are broadly classified into two types, type I and type II enzymes. Type I enzymes cleave a single strand of a DNA molecule and pass another single or double stranded DNA through the break before resealing the opening. Type II enzymes cleave both ‡Corresponding author: Phone: 847-491-7726, Fax: 847-467-6489, a-mondragon@northwestern.edu. †Present address: Institute of Genomics and Integrative Biology, CSIR, Delhi, India Protein data bank accession codes The final structure factors and coordinates of Topo-31, Topo-44 Form I, Form II, and Form III have been deposited in the Protein Data Bank with accession codes 3M7G, 3M7D, 3M6K, and 3M6Z respectively. Supplementary data Supplementary data are available at Structure Journal Online. Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. NIH Public Access Author Manuscript Structure. Author manuscript; available in PMC 2011 July 14. Published in final edited form as: Structure. 2010 July 14; 18(7): 829–838. doi:10.1016/j.str.2010.03.006. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript strands of a double stranded DNA in concert and pass another double stranded DNA through the break. Type I enzymes use the torsional energy stored in the supercoiled DNA to drive DNA relaxation and hence they do not require high energy cofactors, such as ATP, for their activity (Baker et al., 2009) Type II enzymes, on the other hand, require ATP and Mg2+ for their activity. Type I topoisomerases are further subdivided into three subtypes: IA, IB, and IC (Forterre et al., 2007). Type IA and IB enzymes have been studied extensively (Baker et al., 2009) and there is ample information available about their general mechanism of DNA relaxation and the mode of DNA binding. Type IC, on the other hand, is a relatively new subtype. Currently topoisomerase V is the only member of this family and it has been identified only in the Methanopyrus genus. Previously, topoisomerase V had been considered as a type IB enzyme based on its biochemical characteristics (Slesarev et al., 1993), but the crystal structure of an N-terminal 61 kDa of topoisomerase V (Topo-61) (Taneja et al., 2006) revealed a completely new fold without similarity to other topoisomerases or any other known protein. Furthermore, the orientation of the putative active site residues is also different from other type I topoisomerases, suggesting a different mechanism of cleavage and religation of DNA. These observations, together with the lack of sequence similarity, indicated that topoisomerase V defines a new subtype of type I enzymes (Forterre, 2006). Topoisomerase V was identified in Methanopyrus kandleri, an extremophile isolated from a deep-water ‘black smoker’ chimney in the Gulf of California (Huber et al., 1989). The enzyme is active at very high temperatures (122°C) and high salt concentrations (0.65 M NaCl and 3.1 M potassium glutamate). The unusual characteristic of topoisomerase V is that it has both topoisomerase and DNA repair activities in the same polypeptide (Belova et al., 2001). Based on the sequence analysis of topoisomerase V, it has been predicted that the protein contains 24 helix-hairpin-helix (HhH) DNA binding motifs arranged as 12 (HhH)2 domains around the N-terminal topoisomerase domain (Belova et al., 2002) (Figure 1A). Some of these (HhH)2 domains are involved in the apurinic/apyrimidinic (AP) site- processing activity, but the exact location of the repair active site is not known yet. Topoisomerase V can relax both positively and negatively supercoiled DNA without the need for metal cations or high energy cofactors. Single molecule experiments have shown that topoisomerase V relaxes DNA by a constrained swiveling mechanism, relaxing around 12 turns of DNA per relaxation cycle (Taneja et al., 2007). Type IB enzymes, which also use a constrained swiveling mechanism for DNA relaxation, relax around 19 turns of DNA per relaxation cycle (Koster et al., 2005). The structure of Topo-61 showed that the topoisomerase domain is mainly alpha helical and that the first four (HhH)2 domains curl around the topoisomerase domain (Taneja et al., 2006) (Figure 1B). The topoisomerase and (HhH)2 domains are joined by a long bent helix, termed the “linker helix”. Three of the five putative active site residues are present in a helix-turn-helix (HTH) domain and the other two are present in an intervening loop and a helix. The active site residues are buried by the first (HhH)2 domain and it has been suggested that large conformational changes will be needed for the DNA to access the active site of topoisomerase V (Taneja et al., 2007). Here we present data that shows that the N- terminal 31 kDa fragment of topoisomerase V (Topo-31) has topoisomerase activity, consistent with previous predictions based on the structure. In addition, we show that the Topo-44 fragment (N-terminal 44 kDa fragment of topoisomerase V) can form a stable protein-DNA complex, emphasizing the need of the (HhH)2 domains for binding DNA. We determined a crystal structure of (Topo-31) fragment, which has only the topoisomerase domain, and three different crystal structures of the Topo-44 fragment, which includes the topoisomerase domain and three tandem HhH motifs. In all structures, the topoisomerase domain is very similar. In contrast, the structures of Topo-44 show conformational changes in the linker helix resulting in variable orientations of the (HhH)2 domains when compared Rajan et al. Page 2 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript to the Topo-61 structure. Phosphate ions are present in the vicinity of the topoisomerase active site in two of the Topo-44 structures. Some of the catalytic residues interact with the phosphate ions and may mimic contacts with DNA. These observations suggest that the movement of the (HhH)2 domains is mediated by the linker helix and helps expose the topoisomerase active site to facilitate DNA binding. In addition, the location of the phosphate ions suggests a possible path for the DNA and the way the active site residues interact with it. Results The topoisomerase domain can relax DNA DNA relaxation assays using different topoisomerase V fragments showed that the topoisomerase domain alone is capable of relaxing DNA. Topoisomerase V fragments with different numbers of (HhH)2 domains, Topo-31, Topo-44, and Topo-78, were studied using relaxation assays. Topo-31 has no (HhH)2 domains, Topo-44 has one full and one partial (HhH)2 domain, while Topo-78 has eight full (HhH)2 domains, including a putative DNA repair domain. In addition to standard conditions, the effect of different pH conditions and presence of magnesium ions were also tested. The experiments show that Topo-31 is capable of relaxing DNA, despite the absence of the (HhH)2 domains (Figure 2B). A pH profile analysis for the DNA relaxation assays showed that Topo-78 relaxes DNA over a wider pH range (pH 5 to 9), while Topo-31 and Topo-44 relax DNA optimally at pH 5 (Figure 2A, 2B, 2C). In addition, magnesium is not required for the reaction, but stimulates it at all pH values (Figure 2B, 2C). Topo-78 can relax DNA to the same extent with lower amounts of protein (0.1 μg/reaction) compared to Topo-44 (~1.5 μg/reaction) and Topo-31 (~9 μg/reaction). This could be due to the enhanced DNA binding facilitated by the (HhH)2 domains. Together, these results suggest that, even though the (HhH)2 domains are dispensable for topoisomerase activity, they enhance DNA relaxation activity. In addition, the pH dependence of the DNA relaxation activity indicates that the reaction is likely to involve side chains with ionizable groups in the low pH range, such as glutamates. Finally, the magnesium independence of the reactions confirms that even the smallest fragments do not require metals for activity, although magnesium has a stimulatory effect. This may be due to favorable interactions of the cations with DNA. The (HhH)2 domains enhance DNA binding affinity EMSA experiments with different fragments of topoisomerase V and DNA showed that (HhH)2 domains could help in the formation of a stable protein-DNA complex. Various topoisomerase V fragments (Topo-31, Topo-44, and Topo-78) and single and double stranded DNA were analyzed by EMSA experiments. Topo-44 and Topo-78 formed stable complexes with a 39mer double stranded DNA (Figure 2D), while no DNA binding was observed for the Topo-31 fragment (data not shown). These observations indicate that (HhH)2 domains are necessary for a stable protein-DNA complex and that as few as one and half (HhH)2 domains are enough for formation of a stable protein-DNA complex. EMSA with single stranded DNA showed that Topo-31 and Topo-44 cannot bind to single stranded DNA, while Topo-78 can bind to single stranded DNA (data not shown). Overall Structures The topoisomerase domain of topoisomerase V is a helical-rich compact domain that has no structural similarity to any other known protein. The only recognizable structural element is a HTH that contains some of the active site residues. Not surprisingly, the topoisomerase domain of the four structures (Topo-31, Topo-44 (Forms I, II, and III)) superimpose very well on each other and also to that from the Topo-61 structure. In the Topo-31 structure, two surface loops, residues 39-49 and 120-124, adopt a different conformation compared to the Rajan et al. Page 3 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Topo-61 and Topo-44 structures. These two loops are not always visible in the Topo-44 structures, suggesting that they are mobile regions. The r.m.s.d. for the superposition of the topoisomerase core domain of all the new structures on to the Topo-61 structure range from 0.2 Å to 0.7 Å if the two mobile surface loops are not included (Figure 3A). In general, the topoisomerase domain remains unchanged and is identical in all structures. The (HhH)2 domains also remain largely unchanged, with r.m.s.d. for the superposition of only the (HhH)2 domains from the three Topo-44 crystal forms and equivalent domains in the Topo-61 structure ranging from 0.31 Å to 0.56 Å. The five crystallographically independent structures of Topo-44 (Form I, Form II A and B monomers, and Form III A and B monomers) were compared with each other and to the two crystallographically independent Topo-61 monomers to understand the conformational changes in the protein. The r.m.s.d. for the superposition of all the Topo-44 structures (residues 3-375) on to the Topo-61 fragment or on each other vary between 0.9 Å and 2.7 Å, with the majority above 1.5 Å, showing that in general the structures have slightly different conformations. As mentioned above, the different domains behave as rigid or almost rigid subunits and the only change in the structure is the relative orientation between the topoisomerase and the (HhH)2 domains. The change in orientation of the domains starts at the linker helix (residues 269-295), which acts as a hinge region, and follows into the (HhH)2 domains. At the start of the linker helix, the structures superimpose very well for all five Topo-44 and two Topo-61 structures. In the middle of the linker helix there is a kink after which the linker helix from all the structures shows different orientations (Figure 3B). The flexibility of the linker helix is also evident by the fact that the linker helix in the B subunit of Form III crystals appears in two alternate conformations. The change in the relative orientation of the (HhH)2 and topoisomerase domains (Figure 3C and 3D), suggests that these domains can adopt different orientations and these movements might be necessary for the DNA to access the active site. The topoisomerase domain has a positively charged groove adjacent to the active site The structure of the Topo-31 as well as the structures of the Topo-44 fragment reveals the presence of a positively charged groove in the protein that encompasses the active site region (shown later in Figure 6C). This charged groove had been observed before in the structure of the Topo-61 fragment, although several (HhH)2 motifs partially obstruct it (Taneja et al., 2006). The structure of the Topo-31 confirms the presence of the groove even in the absence of the (HhH)2 motifs. The groove is long and can be deep in some areas. It includes regions of the HTH motifs and extends all the way to the linker helix. All the residues forming the active site pentad point towards the groove. The active site tyrosine, Tyr226, is found near one of the ends of the groove, a region where it widens. The positively charged character of the groove and its presence by the active site strongly suggest that it may be involved in DNA binding. Phosphate ions bind in the groove near the topoisomerase active site An interesting observation stemming from the Form II and Form III Topo-44 structures is the presence of phosphate ions near the positively charged DNA binding groove. All three Topo-44 crystal forms were crystallized in the presence of phosphate-citrate buffer, but only Form II and Form III structures showed phosphate ions bound to the protein, which were assigned based on electron density consistent with a tetrahedral phosphate ion (Figure 4A). Form II and Form III crystals include 1–1.2 M guanidium hydrochloride in the crystallization solution. The high resolution Form III structure shows clear density for three guanidium ions bound to the protein, two very well ordered and one with weak density. The presence of guanidium hydrochloride in the crystals appears to trigger a conformational change allowing the binding of phosphate ions to the protein. It is interesting to note that Rajan et al. Page 4 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Form I crystals did not show any bound phosphate albeit its presence in the crystallization condition. This could be due to the absence of guanidium hydrochloride to trigger the binding of phosphate ions as observed in Form II and Form III structures. There are three phosphate ions in the B subunit and none in the A subunit of the Form II Topo-44 structure. Two of the phosphates are in the topoisomerase active site and one of them forms close contacts with the putative active site residues in the topoisomerase domain (Figure 4B). Form III crystal has seven phosphate ions, three in each subunit and one between both the subunits. In the Form III structure, the phosphate ion near the active site Tyr226 is absent, but it shows several new locations for phosphate ions, especially in the positively charged groove containing the topoisomerase active site (Figure 5A). An overlay of the A and B subunits of the Topo-44 Form III structure with the B subunit of Topo-44 Form II structure shows eight unique phosphate ions (Figure 5A). It clearly shows that there are more phosphate ions bound in the positively charged groove compared to other regions of the protein. Taking into account all structures, there are five unique phosphate ion binding sites in the putative DNA binding groove and an additional one near its end and close to the start of the linker helix. Several pairs of phosphates in the groove are separated by a distance of around 7 Å (Figure 5B), which would be consistent with the phosphate-phosphate distance in adjacent nucleotides in a DNA double helix. One of the phosphates (P1) is found near the active site tyrosine and is coordinated by Tyr226 and Arg131, two residues that have been implicated in cleavage and religation of the DNA (Taneja et al., 2006), and by Glu215, whose charge may be important for interactions with DNA (R.R. and A.M., unpublished observations). The side chains of the tyrosine and the glutamate residues are in contact with Arg144 and His200, the other putative active site residues, and these interactions may help to orient them for the catalytic reaction. Adjacent to P1, there is a second phosphate (P2) at a distance of 7.5 Å which is trapped between the topoisomerase domain and an HhH motif. P2 is coordinated by Arg131, an active site residue, in addition to Arg108 from the topoisomerase domain and Arg293 and serines 322 and 324 from the second HhH motif (Figure 6C). Three more phosphates are found in the groove (P3, P4, and P5) coordinated mainly by positively charged residues, such as Arg37, Lys47, Arg108, Lys134, and Arg135 from the topoisomerase domain and also residues from the linker helix such as Tyr289 and Arg293. In general, some of the side chains can contact more than one phosphate. The distance between P3 and P4 and P4 and P5 is 6.8 Å and 6.5 Å respectively. A final phosphate (P6) is located at the start of the linker helix and on the edge of the groove (Figure 5A). Discussion Topoisomerase V is active at very high temperatures (122°C) and high salt concentrations. DNA relaxation assays with various topoisomerase V fragments (Topo-44 and Topo-61) show that a temperature above 60° C is required for optimal activity, although longer fragments of topoisomerase V can relax DNA at lower temperatures (Taneja et al., 2007). Topo-44 was first identified by limited proteolytic digestion of the full length topoisomerase V protein (Belova et al., 2002) at 80°C. In contrast, Topo-61 is the shortest fragment showing topoisomerase activity when the proteolytic reaction is performed at 37°C (Belova et al., 2002). The N-terminal Topo-31 fragment, which contains neither HhH motifs nor the linker helix, was identified as the smallest region spanning the topoisomerase domain from the crystal structure of Topo-61 fragment (Taneja et al., 2006) and it was suggested that it could represent the minimal domain capable of relaxing DNA. Relaxation experiments with this minimal domain show that this is indeed the case, although the activity is not as robust as with longer fragments. As expected, Topo-31 does not require magnesium for activity, but magnesium enhances it, as is the case for type IB topoisomerases, which also uses a Rajan et al. Page 5 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript swiveling mechanism for DNA relaxation (Stewart et al., 1996). The optimal pH for activity for the Topo-31 and the Topo-44 fragments is around 5. This pH dependence is not observed for the Topo-78 fragment. The DNA relaxation by shorter fragments of topoisomerase V at pH 5 could point to the involvement of some ionizable side chains in the relaxation activity. It could also be simply due to the effects of various side chains on DNA binding. Further experiments with different active site mutations in both longer and shorter fragments of topoisomerase V will be required to probe the pH dependence of the relaxation reaction by shorter topoisomerase V fragments. Gel shift experiments show that Topo-44 and also longer fragments (Topo-78) can bind double stranded DNA. Surprisingly, Topo-31 does not show DNA binding activity in these assays even though it is still capable of relaxing DNA. It appears that the presence of the (HhH)2 domains stabilizes the DNA/protein complex. One possibility is that the (HhH)2 domains could play a similar role to the cap domain present in type IB enzymes, which helps to encircle the DNA during the swiveling reaction (Redinbo et al., 1998). In addition, both short fragments of topoisomerase V do not bind single stranded DNA, whereas Topo-78 can form a stable complex with single stranded DNA (data not shown). (HhH)2 domains binding to single stranded DNA has been observed before. For instance, the N-terminal 8 kDa of mammalian polymerase β, which contains a single HhH motif, binds to single stranded DNA through both helices (Kumar et al., 1990; Liu et al., 1994). The exact mode of single stranded DNA binding by Topo-78 or the possible role in relaxation or repair activities is not yet clear. The structure of Topo-61 showed that the topoisomerase active site of topoisomerase V is buried by one of the (HhH)2 domains suggesting that conformational changes are essential for the protein to bind DNA. The present structures of Topo-44 reinforce this observation and show that the (HhH)2 domains can change their position relative to the topoisomerase domain and that this change is mediated by the movement of the linker helix. The (HhH)2 domains act as rigid individual units, as evidenced by the fact that in different structures they show the same structure and relative orientation of the two HhH motifs. The topoisomerase domain also appears to be rigid showing the same structure even in the total absence of the rest of the protein. The linker helix (residues 269-295), which is a long bent helix, serves as a hinge for the movement of the (HhH)2 domains away from the rigid topoisomerase domain, possibly by responding to interactions with double stranded DNA. This movement has to be quite large. The Topo-44 structures in the absence of DNA capture the regions that move, but do not show the full extent of the movement or indicate the way the HhH motifs interact with DNA. As mentioned before, topoisomerase V binds double stranded DNA and has a groove wide enough to accommodate double strand DNA (Figure 6C). The presence of an HTH domain normally associated with DNA binding, the positively charged nature of it, and several phosphates bound along it suggest that this groove could be involved in DNA binding. In addition, the active site is found in this groove and some residues form part of the HTH domain. Previously, DNA was modeled bound to the topoisomerase domain (Taneja et al., 2006) based on the structures of HTH domains in complex with DNA but there was no evidence to support it. Using the phosphates present in the groove in the current structures, it is possible to refine this model. A superposition of the B subunit of Form II and the A and B subunits of Form III Topo-44 structures shows five different phosphate ions in the positively charged groove which are separated by a distance of around 7 Å, consistent with the distance of consecutive phosphates in B DNA of ~6.4 Å. A sixth phosphate ion is found outside the groove near the linker helix. A double stranded DNA molecule was modeled into the groove based on the positions of the phosphate ions (Figure 6). Only five out of the six phosphates could be placed on the DNA molecule, as one of them was inconsistent with a Rajan et al. Page 6 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript double stranded molecule. Phosphate ions P3, P4, and P5 would correspond to three adjacent phosphates in one DNA strand, while P1, located near the active site, would belong to the opposite strand. A final phosphate (P6) is away from the groove and near the linker helix (Figure 6A, 6B). The fit to the four inner phosphates is excellent and the DNA can be accommodated in the groove of the Topo-31 structure without the need for any major rearrangements of the protein backbone. The fifth phosphate (P6) does not fit as well and a better fit would require movement of either the protein or the DNA, but the change would be relatively modest. Several side chains would need to move, but these changes would also be minor. The major change needed to accommodate the DNA in the structures with the (HhH)2 domains present is the movement of the (HhH)2 domains away from the topoisomerase domain (Figure 6B). The movement of (HhH)2 domains should be feasible as is evident from the Topo-44 structures showing different orientations of the (HhH)2 domains. The location of the (HhH)2 domains after DNA binding is not evident, but one possibility is that they would help enclose the DNA to form a clamp around it, similar to the arrangement in type IB enzymes. In the model of the topoisomerase domain in complex with DNA, the active site residues are in close contact with the backbone of DNA. The catalytic Tyr226 is pointing towards the phosphate of the DNA backbone; Arg131 and Arg144 are positioned to stabilize the protein- DNA covalent complex. Surprisingly Glu215 also appears to interact directly with the DNA phosphate backbone. The other active site residues like His200 and Lys 218 are also near the DNA. The active site is located near the end of the groove, where it widens. At this end, the DNA fits loosely in the groove, which is spacious to accommodate the movement of the strands. The ‘constrained swiveling’ mechanism employed by type IB and IC enzymes necessitates rotation of one strand about the other after forming the covalent protein-DNA intermediate. The position of the active site at the wider end of the putative DNA binding groove would facilitate the rotation of the DNA strand at this end, while holding the rest of the DNA in place through extensive interactions along the groove. Even though type IB and IC enzymes have a similar overall mechanism of action, the structures of fragments of topoisomerase V suggest many differences. Type IB enzymes have two domains which come together to form a C-shaped clamp around the DNA (Perry et al., 2006; Redinbo et al., 1998; Stewart et al., 1998) The protein has an open stage where these domains are separate and this helps in the entry and release of the DNA from the protein active site. A wide DNA binding cavity is not observed in the topoisomerase V structures. Instead, the structures show a positively charged groove which is always present in the protein and does not require domain rearrangements to form. DNA can access this groove after a conformational change involving the movement of the (HhH)2 domains exposing the active site. The (HhH)2 domains could help enclose DNA during the swiveling of the DNA, forming a similar enclosure to the one observed for type IB enzymes. It is not known whether all HhH motifs contact DNA simultaneously, but this appears unlikely without a major rearrangement of the motifs. It is likely that only some of the HhH motifs contact DNA at any given time or that some of the motifs do not have the capacity to bind DNA. Finally, similar to type IB enzymes (Cheng and Shuman, 1998), the putative domain enclosing the DNA is dispensable for activity, although it enhances the relaxation activity markedly. Thus, it is likely that type IB and IC enzymes have several overall similarities in the way that they interact with DNA, but the atomic details are markedly different. There are still many details of the atomic mechanism of type IC topoisomerases that need to be understood. The present functional and structural studies provide new information about topoisomerase V including the observations that the Topo-31 is the minimal fragment capable of DNA relaxation, the (HhH)2 domains enhance binding of the protein to DNA, the changes in relative orientation of the domains is mediated by the linker helix, and several Rajan et al. Page 7 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript phosphate ions bind in a positively charged groove. Furthermore, the position of the phosphate ions in the groove helped in the placement of DNA in the topoisomerase domain and this provides an initial model of how topoisomerase V interacts with DNA. Thus the present study helps to establish the role of different domains more clearly, to illustrate a mechanism to drive the conformational changes needed for activity, and to suggest a possible manner of binding DNA. Additional work on structures of protein/DNA complexes and intermediates in the swiveling reaction are needed to understand the way this new type of topoisomerases interacts with DNA to perform a complex reaction. Experimental Procedures Protein purification The N-terminal 31 kDa (Topo-31: residues 1-269), and 44 kDa (Topo-44: residues 1 to 380) fragments of topoisomerase V protein were cloned into the pET15b plasmid and transformed into Escherichia coli BL21 Rosetta (DE3) cells. The N-terminal 78 kDa (Topo-78: residues 1 to 685) fragment of topoisomerase V protein was cloned into the pET14b plasmid (Belova et al., 2002) and transformed into Escherichia coli BL21(DE3) cells. For protein production, cells were grown at 37° C in LB medium containing 100 μg/ml ampicillin and 100 μg/ml chloramphenicol for Rosetta cells and LB medium with 100 μg/ml ampicillin for BL21(DE3) cells to an optical density (OD600) of 0.6. The cells were then cooled down on ice, followed by the addition of isopropyl β-D-1-thiogalactopyranoside (IPTG) to a final 0.5 mM concentration, and grown overnight at 16° C. Cells were harvested and resuspended in 50 mM Tris pH 8, 500 mM NaCl, 0.5 mM EDTA, 1 mM DTT, flash frozen in liquid nitrogen and stored at −80° C. After thawing the pellet, pepstatin (1μg/ml), benzamidine (1mM), PMSF (1mM), and Brij 58 (0.1%) were added to the cells and the protein was purified as described earlier (Taneja et al., 2006) The protein was further purified by anion exchange and gel filtration chromatography. Pure protein was concentrated and stored in 50 mM Tris pH 8, 250 mM NaCl, and 1 mM DTT. The seleno- methionine substituted Topo-44 was prepared from cells grown in a minimal medium supplemented with nutrients and salts (Doublie, 1997); protein purification followed the same procedure as for the native protein except that 5mM DTT was used in all the purification steps and for storage. Relaxation assays Relaxation assays with the different topoisomerase V fragments were carried out at pH values ranging from 4 to 10. The pH of the buffers was adjusted at 65 °C to account for the change in pH at higher temperature. The different buffers used were: sodium acetate for pH 4 and 5, MES for pH 6, HEPES for pH 7, TRIS for pH 8, CHES for pH 9, and CAPS for pH 10. Topoisomerase activity assays were performed by incubating varying amounts of protein (Topo-31, Topo-44 or Topo-78) with 0.2 μg negatively supercoiled pUC19 DNA in 50 mM of the required buffer, 30 mM NaCl, 0.2 mM or 5 mM EDTA or 1 mM MgCl2. The reactions were carried out at 65 °C for 15 min and terminated by cooling and addition of SDS to a final 1% concentration. The products were resolved on a 1% agarose gel and visualized by ethidium bromide staining. Electrophoretic Mobility Shift Assay For Electrophoretic Mobility Shift Assay (EMSA), 4 μM of a 39mer double stranded DNA oligonucleotide (5′ GCGACGCGAGGCTGGATGGCCTTCCCCATTATGATTCTT3′) was incubated with different concentrations of topoisomerase V fragments in 50 mM sodium acetate pH 5, 30 mM NaCl, 1 mM MgCl2 at 65 °C for 30 minutes. Glycerol was added to the reaction mixture to a final concentration of 8% and the products were separated on a 4 % acrylamide native gel. The gel was stained with ethidium bromide to detect the DNA. When Rajan et al. Page 8 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript a stable protein-DNA complex was formed, there was an upward shift in the band indicating a higher molecular weight complex. Crystallization Topo-31 crystals were grown using the sitting drop vapor diffusion method equilibrated against, 23% PEG 6000, 0.1 M Na citrate pH 5.5, at 22°C. For data collection, the Topo-31 crystals were cryo-protected by adding glycerol to the mother liquor to a final 20% concentration. Topo-44 was crystallized by the hanging drop vapor diffusion method under three different crystallization conditions (Forms I, II, and III). Crystal Form I grew under 0.1 M phosphate citrate pH 5, 0.2 M NaCl, 15% PEG 3350 and 8% dioxane. The crystals were cryo-protected by increasing the PEG concentration to 30%. Form II crystals grew under 0.1 M phosphate citrate pH 5, 0.2 M NaCl, 16% PEG 8000 and 1M guanidium hydrochloride. For cryo-protection, they were transferred to a solution with 1.5X reservoir solution and 20% 2,3 butanediol or 20% DMSO for 10 seconds and immediately flash frozen under liquid nitrogen. Form III crystals grew under 0.1 M phosphate citrate pH 5.5, 0.15 M sodium sulfate, 0.01 M MgCl2, 1 M guanidium hydrochloride, and 28 % PEG 3350. The crystals were grown at 30°C and were cryo-protected by increasing the PEG concentration to 40%. Further details of crystallization are presented in the Supplementary Information. Data collection and structure determination Diffraction data were collected at the Dupont Northwestern Dow and Life Science Collaborative Access Team stations (DND and LS CAT) at the Advanced Photon Source in Argonne National Laboratory. Data collection and refinement statistics are shown in Table I. All data were processed and integrated using XDS (Kabsch, 1993) and scaled with SCALA (Collaborative-Computational-Project-4, 1994). Data on the Topo-31 crystals were collected to 2.4 Å resolution. The structure was solved by Molecular Replacement (McCoy et al., 2007) using the topoisomerase domain from the Topo-61 structure (residues 1-266) (Taneja et al., 2006) as the search model. It was refined with refmac5 (Murshudov et al., 1997) and Phenix (Afonine et al., 2005) to a final Rwork of 20.0 % and Rfree of 24.8 %. Topo-44 Form I crystals diffract to 1.8 Å. The structure of Form I crystals was solved by Molecular Replacement (McCoy et al., 2007) using the topoisomerase domain from the Topo-61 structure as the search model. Model rebuilding was performed using coot (Emsley and Cowtan, 2004), and refinement using refmac5 (Murshudov et al., 1997). The final Rwork and Rfree are 17.5 % and 22.0 % respectively. For Topo-44 Form II and Form III crystals, seleno-methionine derivatized crystals were used for single-wavelength anomalous dispersion (SAD) phasing. AutoSharp (Vonrhein et al., 2007) was used for locating the selenium atoms; model building was done using coot (Emsley and Cowtan, 2004), and refinement was carried out using refmac5 (Murshudov et al., 1997) Three phosphate ions were noticed in the Form II structure; two of which present in the topoisomerase active site and are separated by a distance of ~7.5 Å. The structure was refined to a final Rwork of 24.1 % and Rfree of 28.9 %. Topo-44 Form III crystals diffracted to 1.4 Å. The final Rwork and Rfree are 16.5 % and 18.4%, respectively. An interesting observation is the presence of both phosphate and guanidium ions in the Form III Topo-44 structure. The linker helix and part of the first HhH motif of the B monomer show alternate conformations and were built as two separate chains with occupancy of 0.5 each. Further details on data collection and structure determination are given in the Supplementary Information. All figures were made with Pymol (DeLano, 2002) and the electrostatic surfaces were calculated with APBS (Baker et al., 2001). Rajan et al. Page 9 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We acknowledge staff and instrumentation support from the Keck Biophysics Facility and the Center for Structural Biology at Northwestern University, and DND and LS-CAT at the Advanced Photon Source (APS) at Argonne National Laboratory. Support from the R.H. Lurie Comprehensive Cancer Center of Northwestern University to the Structural Biology Facility is also acknowledged. DND-CAT is supported by Dupont, DOW and the NSF. LS-CAT was supported by the Michigan Economic Development Corporation and the Michigan Technology Tri-Corridor. Use of the APS is supported by the Department of Energy (DOE). Research was supported by NIH grant GM51350 (to AM). References Afonine PV, Grosse-Kunstleve RW, Adams PD. A robust bulk-solvent correction and anisotropic scaling procedure. Acta Crystallogr D Biol Crystallogr. 2005; 61:850–855. [PubMed: 15983406] Baker NA, Sept D, Joseph S, Holst MJ, McCammon JA. Electrostatics of nanosystems: application to microtubules and the ribosome. Proc Natl Acad Sci U S A. 2001; 98:10037–10041. [PubMed: 11517324] Baker NM, Rajan R, Mondragon A. Structural studies of type I topoisomerases. Nucleic Acids Res. 2009; 37:693–701. [PubMed: 19106140] Belova GI, Prasad R, Kozyavkin SA, Lake JA, Wilson SH, Slesarev AI. A type IB topoisomerase with DNA repair activities. Proc Natl Acad Sci U S A. 2001; 98:6015–6020. [PubMed: 11353838] Belova GI, Prasad R, Nazimov IV, Wilson SH, Slesarev AI. The domain organization and properties of individual domains of DNA topoisomerase V, a type 1B topoisomerase with DNA repair activities. J Biol Chem. 2002; 277:4959–4965. [PubMed: 11733530] Champoux JJ. DNA Topoisomerases: Structure, Function, and Mechanism. Annu Rev Biochem. 2001; 70:369–413. [PubMed: 11395412] Cheng C, Shuman S. A catalytic domain of eukaryotic DNA topoisomerase I. J Biol Chem. 1998; 273:11589–11595. [PubMed: 9565576] Collaborative-Computational-Project-4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D. 1994; 50:760–763. [PubMed: 15299374] Davis IW, Murray LW, Richardson JS, Richardson DC. MOLPROBITY: structure validation and all- atom contact analysis for nucleic acids and their complexes. Nucleic Acids Res. 2004; 32:W615– 619. [PubMed: 15215462] DeLano, WL. The PyMol Molecular Graphics System. San Carlos, CA: DeLano Scientific; 2002. Diederichs K, Karplus PA. Improved R-factors for diffraction data analysis in macromolecular crystallography. Nat Struct Biol. 1997; 4:269–275. [PubMed: 9095194] Doublie S. Preparation of selenomethionyl proteins for phase determination. Methods Enzymol. 1997; 276:523–530. [PubMed: 9048379] Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] Forterre P. DNA topoisomerase V: a new fold of mysterious origin. Trends Biotechnol. 2006; 24:245– 247. [PubMed: 16650908] Forterre P, Gribaldo S, Gadelle D, Serre MC. Origin and evolution of DNA topoisomerases. Biochimie. 2007; 89:427–446. [PubMed: 17293019] Gellert M. DNA Topoisomerases. Annu Rev Biochem. 1981; 50:879–910. [PubMed: 6267993] Huber R, Kurr M, Jannasch HW, Stetter KO. A novel group of abyssal methanogenic archaebacteria (Methanopyrus) growing at 110 °C. Nature. 1989; 342:833–834. Kabsch W. Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Crystallogr. 1993; 26:795–800. Rajan et al. Page 10 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Koster DA, Croquette V, Dekker C, Shuman S, Dekker NH. Friction and torque govern the relaxation of DNA supercoils by eukaryotic topoisomerase IB. Nature. 2005; 434:671–674. [PubMed: 15800630] Kumar A, Widen SG, Williams KR, Kedar P, Karpel RL, Wilson SH. Studies of the domain structure of mammalian DNA polymerase beta. Identification of a discrete template binding domain. J Biol Chem. 1990; 265:2124–2131. [PubMed: 2404980] Liu D, DeRose EF, Prasad R, Wilson SH, Mullen GP. Assignments of 1H, 15N, and 13C resonances for the backbone and side chains of the N-terminal domain of DNA polymerase beta. Determination of the secondary structure and tertiary contacts. Biochemistry. 1994; 33:9537– 9545. [PubMed: 8068628] Maxwell A. DNA gyrase as a drug target. Biochem Soc Trans. 1999; 27:48–53. [PubMed: 10093705] McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software. J Appl Crystallogr. 2007; 40:658–674. [PubMed: 19461840] Murshudov GN, Vagin AA, Dodson EJ. Refinement of macromolecular structures by the maximum- likelihood method. Acta Crystallogr D. 1997; 53:240–255. [PubMed: 15299926] Perry K, Hwang Y, Bushman FD, Van Duyne GD. Structural basis for specificity in the poxvirus topoisomerase. Mol Cell. 2006; 23:343–354. [PubMed: 16885024] Pommier Y. Diversity of DNA topoisomerases I and inhibitors. Biochimie. 1998; 80:255–270. [PubMed: 9615865] Redinbo MR, Stewart L, Kuhn P, Champoux JJ, Hol WG. Crystal structures of human topoisomerase I in covalent and noncovalent complexes with DNA. Science. 1998; 279:1504–1513. [PubMed: 9488644] Rothenberg ML. Topoisomerase I inhibitors: review and update. Ann Oncol. 1997; 8:837–855. [PubMed: 9358934] Schoeffler AJ, Berger JM. DNA topoisomerases: harnessing and constraining energy to govern chromosome topology. Q Rev Biophys. 2008; 41:41–101. [PubMed: 18755053] Slesarev AI, Stetter KO, Lake JA, Gellert M, Krah R, Kozyavkin SA. DNA topoisomerase V is a relative of eukaryotic topoisomerase I from a hyperthermophilic prokaryote. Nature. 1993; 364:735–737. [PubMed: 8395022] Stewart L, Ireton GC, Parker LH, Madden KR, Champoux JJ. Biochemical and biophysical analyses of recombinant forms of human topoisomerase I. J Biol Chem. 1996; 271:7593–7601. [PubMed: 8631793] Stewart L, Redinbo MR, Qiu X, Hol WG, Champoux JJ. A model for the mechanism of human topoisomerase I. Science. 1998; 279:1534–1541. [PubMed: 9488652] Taneja B, Patel A, Slesarev A, Mondragon A. Structure of the N-terminal fragment of topoisomerase V reveals a new family of topoisomerases. EMBO J. 2006; 25:398–408. [PubMed: 16395333] Taneja B, Schnurr B, Slesarev A, Marko JF, Mondragon A. Topoisomerase V relaxes supercoiled DNA by a constrained swiveling mechanism. Proc Natl Acad Sci U S A. 2007; 104:14670–14675. [PubMed: 17804808] Vonrhein C, Blanc E, Roversi P, Bricogne G. Automated structure solution with autoSHARP. Methods Mol Biol. 2007; 364:215–230. [PubMed: 17172768] Wang HK, Morris-Natschke SL, Lee KH. Recent advances in the discovery and development of topoisomerase inhibitors as antitumor agents. Med Res Rev. 1997; 17:367–425. [PubMed: 9211397] Rajan et al. Page 11 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Organization of topoisomerase V Topoisomerase V is a multi-domain protein consisting of 24 helix-hairpin-helix (HhH) DNA binding motifs arranged as 12 (HhH)2 domains following the N-terminal topoisomerase domain. A) Schematic diagram of various topoisomerase V fragments. The topoisomerase domain is shown in red, the (HhH)2 domains are shown in alternating colors of cyan and yellow. The (HhH)2 domains with repair activity are shown in green. All fragments shown have topoisomerase activity, but only the full length protein and the Topo78 fragment have repair activity. B) Crystal structure of Topo-61 fragment (Taneja et al., 2006). The coloring scheme is the same as in Figure 1A, except that the linker helix is shown in grey. Rajan et al. Page 12 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. DNA relaxation activity and EMSA for Topo-31, Topo-44 and Topo-78 fragments of topoisomerase V A) pH profile of the DNA relaxation activity of Topo-78 and Topo-44 fragments. 0.2 μg of pUC19 DNA were incubated with 0.1 μg of Topo-78 or 1.5 μg of Topo-44 proteins at 65°C for 15 minutes for the relaxation reaction. The reaction buffer contained 50 mM of the appropriate buffer, 30 mM NaCl and 0.2 mM EDTA. Topo-78 relaxes DNA at a wider pH range (5 to 9) than Topo-44, which relaxes DNA efficiently only at pH 5. DNA relaxation activity of Topo-31 (B) and Topo-44 (C) fragments in the absence and presence of MgCl2. 0.2 μg of pUC19 DNA were incubated with 9 μg of Topo-31 or 1.5 μg of Topo-44 proteins at 65°C for 15 minutes for the relaxation reaction. The reaction buffer contained 50 mM of the appropriate buffer, 30 mM NaCl and 5 mM EDTA or 1 mM MgCl2. Both Topo-31 and Topo-44 fragments can relax DNA in the absence of MgCl2, but MgCl2 enhances the DNA relaxation activity of the topoisomerase V fragments. The black triangle in panels A, B and C represents increasing pH from 4 to 10 by one pH unit. D) EMSA of Topo-44 and Topo-78 fragments with a 39mer double stranded DNA. Both Topo-44 and Topo-78 form stable complexes with DNA, although Topo-78 seems to saturate DNA binding while Topo-44 does not. In addition, Topo-44 shows some cleavage of the DNA (bottom free DNA band), while the cleavage is not apparent in Topo-78. The numbers at the bottom represent the molar ratio of protein to DNA. Rajan et al. Page 13 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Structure of Topo-44 fragments A) Overlay of Form I (green), Form II (B subunit: blue), and Form III (B subunit: magenta) structures onto the Topo-61 structure (B subunit: orange). The topoisomerase domains superimpose very well for all the structures, while the linker helix and (HhH)2 domains show differences in orientation. B) Overlay of the linker helices of Form I, II, and III structures with that of Topo-61. The color scheme is same for all the figures unless mentioned otherwise. Note that the linker helices have the same orientation at the start and they change as they move further down the helix. C) Superposition of Form I, II, and III Topo-44 structures with that of Topo-61. Only the (HhH)2 domains are colored while the remaining parts are shown in gray for clarity. The active site residues are shown as orange sticks. Note that the (HhH)2 domains adopt different orientations in all the structures. D) Orientation of the (HhH)2 domains of Form I, II and Topo-61 structures. In Form I and II structures, the (HhH)2 domains are moved away from the topoisomerase domain. For clarity, the (HhH)2 domains of Form III are not shown. In panels C and D, the topoisomerase domains were superposed to emphasize the different orientation of the other domains. Rajan et al. Page 14 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. Phosphate ions present near the active site of the Topo-44 structure A) Stereo view of a Form III difference electron density map calculated with a model not including the phosphates. The electron density is contoured at 3.7σ and shows the tetrahedral shape of the phosphate ions. The active site residues are shown in stick. B) Stereo view of the interaction of the phosphate ions with the putative active site residues. The B subunit of Form II structure was superimposed onto the B subunit of Form III structure and the phosphates ions from both structures are shown together with the Form II B subunit protein backbone. The interactions made by the phosphate ion with the active site residues and the corresponding distances in Å are represented as black dotted lines. Rajan et al. Page 15 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Representation of the unique phosphate ions from Form II and Form III Topo-44 structures A) An overlay of the A (magenta) and B (brown) subunits of the Form III structure and B (blue) subunit of Form II Topo-44 structures. The positions of eight unique phosphate ions (orange spheres) are shown. Note that most phosphate ions are found along the DNA binding groove of the topoisomerase domain. B) The phosphate ions in the DNA binding groove are separated by distances of around 7 Å. The protein backbone is that of the B subunit of Form III structure. The active site residues are represented as sticks and distances in Å between adjacent phosphate ions are shown as black dotted lines. Rajan et al. Page 16 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. Model showing DNA bound to the topoisomerase domain A) Model of a 17-mer double stranded DNA bound to the Topo-31 structure (teal). The DNA is represented as green sticks, where as phosphate ions are represented as orange sticks. DNA binds along the DNA binding groove and five of the eight phosphate ions noted in the Topo-44 structures coincide with the DNA backbone. B) Model of Topo-44 (Form II, B subunit: blue) binding to 17-mer double stranded DNA. Note that the linker helix and the (HhH)2 domains interfere with DNA binding to the topoisomerase domain and are likely to move away to allow binding. C) Electrostatic surface representation of the Topo-31 structure. The positively charged DNA binding groove is clearly visible and the phosphate ions are bound in this groove. The orientation corresponds to a 90° rotation of the one shown in Figure 6A in the direction of the arrow. Note that the DNA binding groove goes from one end of the molecule to the other and it is narrower at one end (start of the linker helix) and wider at the other end. The putative active site residues (green sticks) are located at the wider end of the groove. Other residues lining the groove and interacting with the phosphate ions are shown as cyan sticks. D) Electrostatic surface representation of Topo-31 with phosphate ions (orange) and DNA (green). Three phosphate ions (P3, P4, and P5) coincide with the phosphates of one of the DNA strands, where as P1 coincides with a phosphate of the opposite DNA strand. The model shows that the DNA binding groove of topoisomerase V is wide enough to bind DNA and that the movement of linker helix and (HhH)2 domains Rajan et al. Page 17 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript are required to accommodate the DNA. The electrostatic potential was calculated with a dielectric constant of 80 for solvent and 2 for protein. The surface is colored with a blue to red gradient from +10 to −10 KbT/ec. Rajan et al. Page 18 Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Rajan et al. Page 19 Table 1 Data collection and refinement statistics Topo-31 Topo-44 Form I Topo-44 Form II Topo-44 Form III Data Collection Space group C2221 C121 P41212 P212121 Cell dimensions a=106.7 Å, b=119.4 Å, c=63.7 Å a=104.2 Å, b=47.7 Å, c=81.2 Å (β=112.48) a=b=70.1 Å, c=349.6 Å a=63.6 Å, b=80.1 Å, c=137.2 Å Resolution (Å)a 79.56 – 2.4 (2.53 – 2.4) 75.05 – 1.82 (1.91 – 1.82) 29.5- 2.6 (2.72-2.6) 28.9-1.4 (1.46-1.4) Number of observed reflections 78,729 (11,538 134,411 (13,220) 227,408 (19,917) 1,157,917 (126,319) Number of unique reflections 16,259 (2,346) 32,998 (4,301) 28,151 (3,331) 136,662 (15,986) Completeness (%) 99.8 (99.8) 98.3 (88.6) 99.9 (100.0) 98.8 (95.5) Multiplicity 4.8 (4.9) 4.1 (3.1) 8.1 (6.0) 8.5 (7.9) Rmerge (%)b 4.7 (71.1) 4.0 (16.3) 7.4 (52.2) 4.5 (37.9) Rmeas (%)c 5.3 (79.6) 4.6 (19.4) 7.9 (57.2) 4.8 (40.5) ≪I>/σ(<I>)>d 20.5 (2.5) 23.0 (6.8) 19 (3.2) 27.5 (5.3) Refinement Resolution (Å) 79.56 - 2.4 (2.46 - 2.4) 28.06 -1.82 (1.87 – 1.82) 29.14 – 2.6 (2.67 – 2.6) 28.9 - 1.4 (1.44 - 1.4) Number of reflections working/test 15,419/821 31,317/1,673 26,710/1,438 129,802/6,859 Rwork (%)e 20.0(24.3) 17.5 (17.9) 24.1(36.6) 16.5 (19.3) Rfree(%)f 24.8 (31.1) 22.0 (24.8) 28.9 (45.1) 18.4 (22.1) Protein residues/atomsg 269/2,203 376/3212 727/5,970 738/7,511 Atoms in alternate conformations 0 258 (20 protein residues) 8 (1 protein residue) 2846 (157 protein residues) Water molecules 29 238 30 573 Other atoms - - 3 PO4 7 PO4, 3 Gmh, 3 Mg++, 2 Cl− B-factor (Å2) Protein atoms (chain) 68.4 22.8 A:53.8; B:58.2 A:13.4; B:14.9 Water molecules 59.1 29.3 40.0 23.7 r.m.s. deviations bond lengths (Å) 0.015 0.006 0.01 0.009 bond angles (°) 1.42 0.920 1.2 1.2 Ramachandran ploti Favored regions (%) 94.3 98.9 96.2 98.5 Outliers (%) 0.0 0.0 0.3 0 aNumbers in parenthesis correspond to highest resolution shell. bRmerge= Σ|I − <I>|/ΣI, where I is the observed intensity and <I> the average intensity obtained from multiple measurements. Structure. Author manuscript; available in PMC 2011 July 14. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Rajan et al. Page 20 cRmeas as described in Diederichs and Karplus (Diederichs and Karplus, 1997). d≪I>/σ(<I>)> = Mean Ih over the standard deviation of the mean Ih averaged over all reflections in a resolution shell. eRwork= Σ ||Fo| − |Fc||/Σ|Fo|, where |Fo| is the observed structure factor amplitude and |Fc| the calculated structure factor amplitude. fRfree: Rfactor based on 5% of the data excluded from refinement. gTotal number of protein atoms, including those in alternate conformations. hGm: guanidinum ion. iAs reported by Molprobity (Davis et al., 2004). Structure. Author manuscript; available in PMC 2011 July 14.
3M7H
Crystal structure of the bacteriocin LLPA from Pseudomonas sp.
Structural Determinants for Activity and Specificity of the Bacterial Toxin LlpA Maarten G. K. Ghequire1., Abel Garcia-Pino2,3., Eline K. M. Lebbe1¤, Stijn Spaepen1, Remy Loris"2,3*, Rene´ De Mot"1* 1 Centre of Microbial and Plant Genetics, University of Leuven, Heverlee-Leuven, Belgium, 2 Molecular Recognition Unit, Department of Structural Biology, Vlaams Instituut voor Biotechnologie, Brussel, Belgium, 3 Structural Biology Brussels, Department of Biotechnology (DBIT), Vrije Universiteit Brussel, Brussel, Belgium Abstract Lectin-like bacteriotoxic proteins, identified in several plant-associated bacteria, are able to selectively kill closely related species, including several phytopathogens, such as Pseudomonas syringae and Xanthomonas species, but so far their mode of action remains unrevealed. The crystal structure of LlpABW, the prototype lectin-like bacteriocin from Pseudomonas putida, reveals an architecture of two monocot mannose-binding lectin (MMBL) domains and a C-terminal b-hairpin extension. The C-terminal MMBL domain (C-domain) adopts a fold very similar to MMBL domains from plant lectins and contains a binding site for mannose and oligomannosides. Mutational analysis indicates that an intact sugar-binding pocket in this domain is crucial for bactericidal activity. The N-terminal MMBL domain (N-domain) adopts the same fold but is structurally more divergent and lacks a functional mannose-binding site. Differential activity of engineered N/C-domain chimers derived from two LlpA homologues with different killing spectra, disclosed that the N-domain determines target specificity. Apparently this bacteriocin is assembled from two structurally similar domains that evolved separately towards dedicated functions in target recognition and bacteriotoxicity. Citation: Ghequire MGK, Garcia-Pino A, Lebbe EKM, Spaepen S, Loris R, et al. (2013) Structural Determinants for Activity and Specificity of the Bacterial Toxin LlpA. PLoS Pathog 9(2): e1003199. doi:10.1371/journal.ppat.1003199 Editor: Ambrose Cheung, Geisel School of Medicine at Dartmouth, United States of America Received August 22, 2012; Accepted January 3, 2013; Published February 28, 2013 Copyright:  2013 Ghequire et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was financially supported by FWO Vlaanderen (Research project G.0393.09), by The Onderzoeksraad of the VUB, by VIB and by the Hercules Foundation. The authors acknowledge support of the European Community - Research Infrastructure Action under the FP6 ‘‘Structuring the European Research Area Program’’, contract number: RII3-CT-2004-506008. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: remy.loris@vib-vub.be (RL); rene.demot@biw.kuleuven.be (RDM) ¤ Current address: Laboratory of Toxicology, University of Leuven, Leuven, Belgium. . These authors contributed equally to this work. " These authors also contributed equally to this work and should be considered joint senior authors. Introduction In most natural settings, complex interactions occur among microorganisms, ranging from nutritional co-operation to warfare among competitors. Examples of such interplay have been reported not only between unrelated microorganisms (e.g. fungi and bacteria [1,2]), but also between distant relatives (e.g. members of different bacterial genera [3]), and even between close relatives (e.g. at inter- and intra-species levels [4,5]). A major strategy in niche colonization is the production of growth inhibitors or toxins directed at microbial competitors [6]. While a huge variety of secondary metabolites is used to target phylogenetically-distant competitors, ribosome-synthesized pep- tides or proteins are typically active against close relatives. These protein toxins are collectively referred to as bacteriocins, and may either be released into the environment or transferred to the host via specialized contact-dependent delivery systems [7–9]. Bacteriocins are structurally and mechanistically very diverse. This is reflected in the bacteriocinogenic potential of the genus Pseudomonas [10]. Their R- and F-type pyocins are multi-subunit protein complexes evolutionarily related to contractile tails of bacteriophages [11–13]. R-pyocins attach to specific lipopolysac- charide moieties at the cell surface of susceptible cells and insert their core structure through the cell envelope, causing depolar- ization of the cytoplasmic membrane [14]. The S-type pyocins of Pseudomonas aeruginosa share structural and functional features with Escherichia coli colicins [15]. Following docking onto surface- exposed targets such as siderophore receptors [16,17], S-pyocins kill cells by nucleic acid degradation [10,17], cytoplasmic membrane damage [18], or inhibition of peptidoglycan synthesis [19,20]. Putidacin A (or LlpABW), first identified in Pseudomonas putida BW11M1 [21], represents a class of Pseudomonas-specific antibacterial proteins not related to any known bacteriocin. Additional llpA-like genes encoding functional bacteriocins were identified by genome mining in the biocontrol strain Pseudomonas fluorescens Pf-5 [22] and in the phytopathogen Pseudomonas syringae pv. syringae 642 [23]. Identification of this type of protein in two Xanthomonas pathovars extended its occurrence as a genus-specific killer protein [23]. The Xanthomonas LlpA precursor is proteolyt- ically processed by removal of a characteristic Type II secretion signal peptide, whereas such N-terminal sequence is lacking in Pseudomonas homologues, indicating that secretory routes may differ among LlpA producers. PLOS Pathogens | www.plospathogens.org 1 February 2013 | Volume 9 | Issue 2 | e1003199 The amino acid sequence of LlpA suggests the presence of two related domains belonging to the ‘monocot mannose- binding lectin’ (MMBL) family [24]. The MMBL domain consists of a b-prism fold containing three potential carbohy- drate-binding pockets, each generated by a QxDxNxVxY sequence (with x, any amino acid), but some sites may be inactive due to degeneracy of the signature motif [25]. This domain (Pfam domain: B_lectin - PF01453) was initially identified in lectins of monocot plants [26,27], but a more widespread occurrence of MMBL lectins has become evident and includes representatives in fungi [28,29], slime molds [30], sponges [31], and fishes [32–34]. The LlpA branch occupies a unique position among MMBL-domain proteins, harboring non-eukaryotic representatives and being equipped with the capacity to kill bacterial cells with bacteriocin-like specificity, a property not yet demonstrated for other family members [25]. Next to proteins with the LlpA-type tandem-MMBL organiza- tion, many other predicted MMBL proteins are encoded by bacterial genomes. Often the MMBL module is embedded in a larger protein. For one such protein, bacteriocin-like activity among Ruminococcus species, Gram-positive bacteria colonizing the rumen, was demonstrated [35]. Here we report on the crystal structure of LlpABW as the prototype of a novel family of antibacterial proteins and explore how domain architecture and specific structural elements contrib- ute to its activity and specificity. Results LlpA forms a rigid MMBL tandem The crystal structure of LlpABW from P. putida BW11M1 (LlpABW) shows it contains two b-prism MMBL domains, referred to as the N-domain and the C-domain following their position in the amino acid sequence (Figure 1A,B; Figure S1). The N-domain spans residues Arg4-Pro135 while the C-domain encompasses residues Ala136-Gln253. Each domain exhibits pseudo-threefold symmetry and the corresponding subdomains will be referred to as IN, IIN, IIIN, IC, IIC and IIIC, respectively (Figure 1A and Figure S1). Following these two domains, a b- hairpin extension is formed by residues Pro254-His275 (the numbering used in this paper corresponds to that of the wild-type protein without His-tag [21]). The two-domain architecture reflects the b-strand swapping that is typical in dimers of single-domain mannose-binding monocot lectins (Figure 1A,B) [36] and which apparently is retained after the ancestral fusion or duplication of the two domains, as is also the case in certain MMBL tandems or heterodimers from monocots [37,38]. Thus, residues Asp126- Pro135 from the first MMBL sequence complement the fold of the C-domain while residues Pro245-Gln253 from the second MMBL sequence complement the fold of the N-domain. However, in LlpABW, the relative orientation of both domains is different compared to what is observed in a canonical MMBL lectin dimer, such as snowdrop lectin [36], in the heterodimeric MMBL lectin ASA I from Allium sativum [38], or in the tandem MMBL SCAfet from Scilla campanulata [37] (Figure 1C and Figure S2). In contrast to these plant MMBL proteins, the resulting architecture of LlpABW does not obey pseudo-twofold symmetry (Figure 1C). LlpABW is a very rigid molecule. The two monomers present in the asymmetric unit are essentially identical with a root-mean- square deviation (RMSD) of 0.34 A˚ for 270 Ca atoms. This RMSD value does not change significantly when the individual domains are fitted separately (0.32 A˚ for 120 Ca’s of the N- domain and 0.22 A˚ for 115 Ca’s of the C-domain), indicating that the inter-domain orientation is fixed. This stems from three sets of interactions (Figure 2). Both domains are connected by a two- stranded anti-parallel b-sheet that is involved in the b-strand swapping mentioned above and that links both domains. The C- terminal b-hairpin extension makes extensive contacts, through hydrophobic and hydrogen bonds, with both domains. Finally, the stretch Val140-Asp145 of the C-domain makes extensive contacts with stretch Val115-Asp118 and with the side chains of Ser15 and Pro32 of the N-domain. Domains of LlpABW are shaped by differential evolutionary pressure A superposition of the Ca-trace of the N- and C-domain of LlpABW as well as the MMBL domain of snowdrop lectin is shown in Figure S3. Based on 79 Ca atoms that form the common b- sheet core of the MMBL domains, the RMSD between the N- and C-domains of LlpABW is 1.84 A˚ . While the secondary structure elements of the C-domain are restricted to the three four-stranded b-sheets of the b-prism fold, the N-domain contains three additional secondary structure elements (Figure 1A). A three-turn a-helix (a1) is inserted in the loop between strands b9 and b10, and sheet IIN contains two additional strands. Strand b69 is inserted in the loop between strands b6 and b7 and provides an anti-parallel extension to sheet II (hydrogen bonding to strand b9). Strand b19 is a short piece of b-strand that is part of the long N- terminus and forms a parallel extension on the opposite site of sheet IIN (hydrogen bonding to strand b2), making this b-sheet a mixed type six-stranded one rather than the canonical four- stranded anti-parallel sheet. Despite these additions to the b-prism fold, the common core of the N-domain more closely resembles that of the well-studied and highly conserved monocot lectins (e.g. RMSD of 1.35 A˚ with snowdrop lectin compared to 1.82 A˚ for the C-domain). This structural divergence is in contrast with the degree of conserva- Author Summary In their natural environments, microorganisms compete for space and nutrients, and a major strategy to assist in niche colonization is the deployment of antagonistic compounds directed at competitors, such as secondary metabolites (antibiotics) and antibacterial peptides or proteins (bacteriocins). The latter selectively kill closely related bacteria, which is also the case for members of the LlpA family. Here, we investigate the structure-function relationship for the prototype LlpABW from a saprophytic plant-associated Pseudomonas whose genus-specific tar- get spectrum includes several phytopathogenic pseudo- monads. By determining the 3D structure of this protein, we could assign LlpA to the so-called monocot mannose- binding lectin (MMBL) family, representing its first prokaryotic member, and also add a new type of protective function, as the eukaryotic MMBL members have been linked with antiviral, antifungal, nematicidal or insecticidal activities. For the protein containing two similarly folded domains, we constructed site-specific mutants affected in carbohydrate binding and domain chimers from LlpA homologues to show that mannose- specific sugar binding mediated by one domain is required for activity and that the other domain determines target strain specificity. The strategy that evolved for these bacteriocins is reminiscent of the one used by mammalian bactericidal proteins of the RegIII family that recruited a C- type lectin fold to kill bacteria. Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 2 February 2013 | Volume 9 | Issue 2 | e1003199 Figure 1. Overall structure of LlpABW. (A) Topology diagram of LlpABW. The N-domain is shown in red, the C-domain in blue and the C-terminal extension in green. The different strands and subdomains are labeled. Domain swapping involves b-strand segments b11b and b22b, which together with b-strand segments b11a and b22a link both MMBL domains. (B) Cartoon representation of LlpABW with the different domains colored as in panel A. The bound Me-Man residue is shown as an orange stick representation. (C) Domain orientations of LlpABW compared with the heterodimeric MMBL ASA I (Allium sativum agglutinin, PDB entry 1KJ1) and tandem MMBL SCAfet (Scilla campanulata fetuin-binding lectin, PDB entry 1DLP). In each case, the C-domain is shown in the same orientation, highlighting the different relative orientation of the N-domain in LlpABW. Domain-swapped dimers in homo-oligomeric plant MMBL lectins such as snowdrop lectin have their domain orientation similar to ASA I and SCAfet. doi:10.1371/journal.ppat.1003199.g001 Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 3 February 2013 | Volume 9 | Issue 2 | e1003199 tion of the carbohydrate-binding motif characteristic of the monocot lectins (QxDxNxVxY) in each of the three subdomains. In the N-domains of LlpA homologues, the surface-exposed motifs III and II are not well conserved and likely lost their function during evolution. In contrast they seem to be better conserved in the C-domains (Figure S4). Apparently, the two MMBL domains of LlpA experienced a differential evolutionary pressure resulting in different degrees of global and local (carbohydrate-binding motif) conservation, suggesting distinct functional roles for each domain. The C-domain of LlpABW further extends into a b-hairpin that helps to define the relative orientations of its two MMBL domains. This b-hairpin is highly bent due to a b-bulge inserted into its second b-strand (Figure 1B). It is absent in all plant representatives including tandem MMBL proteins such as SCAfet (Figure 1C). In bacteria it represents the most divergent part of LlpA homologues, both in primary sequence and in length (Figure S5). Most of these C-terminal extensions terminate with a phenylalanine residue. This is reminiscent of the conserved terminal phenylalanine of outer membrane proteins from Gram-negative bacteria such as PhoE, required for their translocation to the cell envelope [39]. An equivalent extension appears to be absent in the Xanthomonas and Arthrobacter sequences (Figure S5). LlpA is capable of binding mannose-containing carbohydrates Subdomains IIC and IIICof LlpABW contain the typical sugar- binding signature (QxDxNxVxY) of an active MMBL mannose- binding site (Figure S1 and S4). Soaking crystals of LlpABW with 200 mM methyl-a-D-mannopyranoside (Me-Man) led to clear electron density of a single Me-Man in site IIIC of each of the two LlpABW monomers in the asymmetric unit (Figure S6A). This site comprises the side chains from Gln171, Asp173, Asn175 and Tyr179, which contribute to hydrogen bond interactions and the side chains of residues Val177, Asn188, Gln192 and Ala185, which contribute to van der Waals contacts with the carbohydrate ligand (Figure 3A, Figure S7A,C). This architecture is very similar to what is observed for mannose bound to other MMBL-type lectins such as snowdrop and garlic lectin (Figure S7B). Soaks with oligomannoses revealed additional sugar-binding subsites. Binding site IIIC accommodates the disaccharide Mana(1–2)Man and the pentasaccharide GlcNAcb(1–2)Mana(1– 3)[GlcNAcb(1–2)Mana(1–6)]Man (Figure S6B,C). In the case of Figure 2. Domain interactions within LlpABW. (A) The C-terminal hairpin extension (green cartoon) covers the interface between the N-domain (red surface representation) and the C-domain (blue surface representation). (B) Stereo view of the interactions between loop segments Val140- Asp145 (cyan) of the C-domain and Val115-Asp118 (yellow) and Ser31-Gln34 (orange) of the N-domain. Other structural elements are colored according to panel A. (C) Stereo view of the two-stranded b-sheet formed by strands b11a,b and b22a,b that links the N- and the C-domains and gives rise to domain swapping. Colors according to panel A and B. doi:10.1371/journal.ppat.1003199.g002 Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 4 February 2013 | Volume 9 | Issue 2 | e1003199 the pentasaccharide, the central reducing mannose is located in the shallow Me-Man binding site and the two GlcNAcb(1–2)Man moieties stretch out over the surface making only a few additional hydrogen bonds or van der Waals contacts (Figure 3B). In the bound disaccharide, the non-reducing mannose is located in the Man-Me binding site while the reducing mannose faces the solvent and does not interact directly with the protein (Figure 3C). Site IIC of both LlpABW molecules in the asymmetric unit is involved in crystal packing interactions and the presence of Me- Man is therefore sterically excluded. All residues that form specific hydrogen bonds with Me-Man are retained but substitutions occur for three side chains that provide van der Waals contacts (Figure S4 and S8A). In contrast, site IC lost the conserved QxDxNxVxY motif (Figure S4) and is involved in inter-domain contacts and therefore inaccessible to ligands (Figure S8B). The putative carbohydrate-binding sites in the N-domain of LlpABW are less conserved. Similar to the C-domain, site IN is inaccessible and involved in inter-domain interactions (Figure S9A). In the IIN subdomain, the canonical mannose-binding motif QxDxNxVxY is essentially absent, with only the Gln residue of the motif being conserved as Gln82 (Figure S4). All other donors or acceptors required for hydrogen bonds with a mannose ligand are missing. In addition, the presence of Phe86 at the equivalent position of the expected Val sterically hinders the binding of mannose (Figure S9B). The potential carbohydrate-binding site on subdomain IIIN is only partially conserved (Figure S9C) and contains two relevant substitutions from the canonical signature: Figure 3. Carbohydrate binding in site IIIC of LlpABW. (A) Stereoview of methyl-a-D-mannopyranoside bound to subdomain IIIC. Methyl-a-D- mannopyranoside is shown in blue and indicated by M. Residues belonging to the QxDxNxVxY motif and hydrogen bonding to the sugar as well as Asn188 are labeled. Water molecules bridging protein and carbohydrate are shown in cyan (B) Similar view of the pentasaccharide GlcNAcb(1– 2)Mana(1–3)[GlcNAcb(1–2)Mana(1–6)]Man. The mannose residue occupying the primary binding site is shown in blue and labeled M. The additional two mannoses (labeled +1 and 21) and two N-acetyl glucosamine residues (labeled +2 and 22) are shown in green. Other colors are as in panel A. (C) Binding of the disaccharide Mana(1–2)Man. The non-reducing mannose residue occupying the primary binding site is shown in blue and labeled M. The second, reducing mannose is shown in green. Other colors are as in panel A. doi:10.1371/journal.ppat.1003199.g003 Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 5 February 2013 | Volume 9 | Issue 2 | e1003199 (1) the Tyr residue of the QxDxNxVxY motif is replaced by the shorter Gln49, thereby removing the canonical hydrogen bond between Man O4 and Tyr OH, and (2) a threonine at position 54 which may compensate the hydrogen bond lost due to the Tyr-to-Gln substitution in the canonical motif. The lack of electron density at this site in our Me-Man soak nevertheless indicates that this site does not recognize this ligand or that its affinity is so low that recognition would only be achieved in the context of a larger and as yet unidentified mannose-containing ligand. Alternatively, this putative site may possess specificity for a different monosaccharide. In order to evaluate this hypothesis, we soaked LlpABW crystals with D-galactose, N-acetyl-D-glucosamine and L-fucose. No electron density was observed for any of these sugars, suggesting that the N-domain has a function distinct from carbohydrate recognition (data not shown). Carbohydrate-binding capacity is required for LlpA toxicity The LlpABW motifs IIIN, IIIC and IIC create potential carbohydrate binding sites that may be involved in bacteriotoxicity of the protein. We therefore examined the role of carbohydrate binding in the bactericidal function of LlpABW. The presence of methyl-a-D-mannopyranoside up to 500 mM in the medium did not influence the activity of LlpABW on P. syringae GR12-2R3. Glycan array profiling did not highlight any specific oligosaccha- ride structure that could represent a natural ligand of LlpABW (Table S1). This could be due to the array design that is principally based on eukaryotic glycans and may therefore lack an appropri- ate carbohydrate for this prokaryotic toxin. Previously, it was observed that LlpABW from concentrated culture supernatant does not agglutinate rabbit red blood cells, nor binds to a mannose- agarose affinity matrix [21]. To assess whether the mannose-recognizing QxDxNxVxY motifs in LlpABW are nevertheless relevant for bactericidal activity, the conserved valine residue was mutated to tyrosine in subdomains IIIN, IIIC, and IIC. These mutations sterically preclude mannose or any other ligand to enter the binding sites (Figure S7C). Semi-quantitative activity assays with permeabilized E. coli cells expressing the LlpA variants in motifs IIIN, IIIC and IIC were used to assess the relationship between carbohydrate binding and bactericidal activity. Modification of the IIIN site, for which no mannose binding was observed, does not affect the antibacterial activity against P. syringae GR12-2R3 (Figure 4). In contrast, the altered IIIC pocket strongly diminishes activity, either alone or in pairwise combination with the other mutated sites (IIIN or IIC). A minor negative effect of the IIC mutation is only apparent in a double mutant, when combined with a modified IIIN motif. Purified proteins were prepared to further quantify these effects. Far UV CD spectra of these mutant forms are identical to that of native protein LlpABW, indicating that the mutations do not affect the overall structure of the protein. Isothermal titration calorim- etry (ITC) showed that LlpABW has an affinity of 2.1 mM for the pentasaccharide GlcNAcb(1–2)Mana(1–3)[GlcNAcb(1–2)Ma- na(1–6)]Man, the highest among all the tested oligo-mannosides (See Figure 5 and Table 1 for a summary of the experimentally validated LlpABW-carbohydrate interactions). This is in agreement with the crystal structures of the different complexes since this sugar is the one with the largest binding interface (Figure 3). Titrations of LlpABW, of the mutants LlpAV177Y (a site IIIC knockout), LlpAV208Y (a site IIC knockout) and of the double mutant LlpAV177Y-V208Y with a-methyl mannoside clearly pin- point site IIIC as the only responsible for the sugar binding activity. Point mutations in both sites or IIIC (V177Y) alone, completely abrogate sugar binding. However knocking out site IIC (V208Y) has little effect in binding and the affinities of LlpAV208Y for a- methyl mannoside and Mana(1–3)Man are very close to the ones measured for the wild-type protein (See Table 1 and Figure 5B). While the V208Y mutation in the IIC site has no observable effect on the MIC value for P. syringae GR12-2R3, the altered IIIC motif engenders a 5.2-fold increase in MIC (Figure 4). The mutant protein LlpAV177Y-V208Y suffers a further reduction in activity, yielding a 31.6-fold increased MIC compared to native LlpABW. The biological activities of LlpA and its mutants were further assessed by live/dead staining and subsequent flow cytometry analysis (Figure 6, Figure S10). Proportions of dead cells after 1 hour of exposure to LlpA or LlpAV208Y were comparable (10.1% and 9.7%, respectively). For LlpAV177Y this value was reduced to 6.1%, significantly lower than for LlpA. Killing activity was even further reduced for LlpAV177Y-V208Y (3.7%). These results are consistent with the MIC determination and ITC data, indicating that an active site IIIC is required to generate a fully active LlpA bacteriocin. The difference in bacteriotoxicity between LlpAV177Y and LlpAV177Y-V208Y suggests that site IIC has a supporting role in the LlpABW bacteriotoxicity. All domains are necessary for LlpABW functionality The site-directed mutagenesis approach revealed an important role for the C-domain’s carbohydrate-binding capacity in LlpABW toxicity. Considering the increased binding motif degeneration in the N-domain and the fact that a Ruminococcus bacteriocin composed of only a single MMBL domain fused to an unknown domain has been identified [35], the N-domain may fulfill a distinct function, different from that of the C-domain. In order to scrutinize the contribution of individual domains to overall activity, six domain deletion constructs of llpABW were engineered to potentially encode proteins lacking the first or second MMBL domain, a gene product devoid of the C-terminal hairpin, or a protein retaining only an individual domain (N-domain, C- domain, or hairpin) (Figure S11). To take the domain swapping into account, the constructs containing only a single MMBL domain were designed with a fusion of the swapped C-terminal b- strands to the corresponding domain via a short linker. None of these deletion constructs resulted in the production of an active protein, indicating that none of the domains are dispensable. Removal of the terminal phenylalanine residue still allows expression of a functional bacteriocin in E. coli (Figure 7), but a further C-terminal truncation (deletion of Trp-His-Phe tail) resulted in a negative bacteriocin assay (data not shown). From these data we conclude that both MMBL domains as well as the C-terminal hairpin extension are required for activity of LlpA. Whether the role of the C-terminal hairpin is any other than simply stabilization of the C-domain cannot be concluded. Target specificity of LlpA is hosted by the N-domain In order to investigate the role of the different domains in target specificity, we created hybrid LlpA proteins using the domains of LlpABW from P. putida BW11M1 and LlpA1 from P. fluorescens Pf-5. These two LlpA proteins share 45% sequence identity and differ in their target spectra. Strains P. syringae GR12-2R3 and P. fluorescens LMG1794 were identified as specific indicators for LlpABW [21] and LlpA1 [22], respectively. Six constructs carrying llpABW/ llpA1chimeric genes were made with domain exchanges involving the N-domain, C-domain, and hairpin region (Figure 7 and Figure S11). For four of these constructs activity against one of both indicators was detected. Only constructs retaining the original N- domain give rise to inhibition of the cognate indicator strain. The C-domain or the hairpin of LlpABW could be replaced with the corresponding LlpA1 domains without changing target specificity. Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 6 February 2013 | Volume 9 | Issue 2 | e1003199 Conversely, the original specificity of LlpA1 is retained upon replacement of its C-domain by the LlpABW equivalent. Discussion Structure elucidation of LlpABW from P. putida BW11M1 unequivocally assigns this bacteriocin to the MMBL lectin family, in which it constitutes the first prokaryotic member, representative for a group of bacterial proteins composed of two MMBL domains [21–23]. Systematic inactivation of the three potential carbohy- drate-binding sites present in the N-domain (IIIN) and in the C-domain (IIIC and IIC) of LlpABW, revealed that a non-occluded IIIC pocket is required to obtain a fully active LlpABW molecule. A negative co-operative effect on activity resulted when the IIC site was additionally modified. Although mannose-containing carbo- hydrates can bind to the IIIC pocket of LlpABW, it remains unclear Figure 4. Inhibitory activity of wild-type LlpABW and selected mutants with modified (potential) mannose-binding sites. The domain structure (N-domain in red, C-domain in blue and C-terminal extension in green) and the position of the MMBL motifs (potentially active binding sites in orange, inactive ones in grey) are shown. The positions of conserved valine residues converted to tyrosine residues by site-directed mutagenesis are indicated with a black bar. Inhibitory activity of E. coli strains expressing mutant LlpABW forms was assayed against P. syringae GR12-2R3 and semi- quantified according to the size (inner zone radius) of the growth inhibition halo relative to LlpABW (+++, native LlpABW; ++, halo size reduced; + halo size strongly reduced; 2, no halo; NT, not tested). For wild-type LlpABW and three purified His-tagged mutant forms (LlpAV177Y, LlpAV208Y and LlpAV177Y-V208Y) the MIC values were determined with indicator P. syringae GR12-2R3. Molar minimal inhibitory concentrations of recombinant proteins (with standard deviations): LlpA, 2.08 nM (60.58 nM); LlpAV177Y, 10.9 nM (60.66 nM); LlpAV208Y, 1.98 nM (60.066 nM); 65.72 nM (62.80 nM). doi:10.1371/journal.ppat.1003199.g004 Table 1. Binding affinities and thermodynamic parameters obtained from ITC titrations. Type of protein-carbohydrate interaction Kd (mM) DG6(kcal mol21) DH6 (kcal mol21) 2TDS6(kcal mol21) LlpABW Me-a-D-Man 45.9 21.8 25.4 3.6 LlpABW Mana(1–2)Man 42.4 21.9 23.6 1.7 LlpABW Mana(1–3)Man 18.2 22.4 25.9 3.5 LlpABW Mana(1–6)Man 17.2 22.4 25.5 3.1 LlpABW Mana(1–3)[Mana(1–6)]Man 10.1 22.6 26.4 3.8 LlpABW GlcNAcb(1–2)Mana(1–3)[GlcNAcb(1– 2)Mana(1–6)]Man 2.1 23.7 21.6 22.1 LlpAV208Y Me-a-D-Man 58.8 21.7 23.3 1.6 LlpAV208Y Mana(1–3)Man 23.0 22.2 25.1 2.9 The reported values for Kd, DGu, DHu and 2TDSu were determined from fitting a single site interaction model (n = 1) to the experimental ITC data. The interaction of the mutants LlpAV177Y and LlpAV177-V208Y with the different sugars is negligible and no heat effect was observed. Therefore they are not included in this table. doi:10.1371/journal.ppat.1003199.t001 Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 7 February 2013 | Volume 9 | Issue 2 | e1003199 if these are part of or mimic the natural ligand required for biological activity since bacteriocin activity is not impaired in the presence of excess mannose. A mutated IIIN site did not provoke a negative effect on antibacterial activity. However, the N-domain appears to play a major role in target selection. This was demonstrated by assessing the differential activity of domain chimers against two target strains, diagnostic for the LlpABW- and LlpA1- specific killing. The b-hairpin does not appear to be a specificity determi- nant, although it constitutes the most variable region among LlpA-like bacteriocins. Possibly, it is required for thermody- namic stability since it needs to be intact in LlpABW. An equivalent C-terminal stretch is absent from the Xanthomonas citri LlpA-like bacteriocin [23]. From our results relying on heterologous expression in E. coli and a bacteriocin assay with permeabilized cells, it cannot be excluded that this structural element may play a role in the way an LlpA protein is exported by its native producer cells. Figure 6. Killing activity of LlpABW and mutant proteins. Percentages of dead cells after live/dead staining as quantified by flow cytometry analysis (Figure S10). P. syringae GR12-2R3 was used as indicator strain and treated at a final concentration of 50 mg/ml for 1 h. Average values (with standard deviations; indicated by error bars): LlpA, 10.1 (61.04); LlpAV177Y, 6.1 (60.44); LlpAV208Y, 9.7 (61.39); LlpAV177Y-V208Y, 3.7 (60.90); buffer (control), 1.0 (60.11).Values are significantly different for (a) and (b), (b) and (c) (p,0.01). doi:10.1371/journal.ppat.1003199.g006 Figure 5. ITC analysis of carbohydrate binding to LlpABW and mutants. (A) Binding of LlpABW to the pentasaccharide GlcNAcb(1–2)Mana(1– 3)[GlcNAcb(1–2)Mana(1–6)]Man. (B) Binding of LlpABW (blue circles, wild type) and the mutants LlpAV177Y (green circles, site IIIC knockout), LlpAV208Y (red circles, site IIC knockout) and LlpAV177Y-V208Y (black circles, site IIC and IIIC knockout) to a-methyl mannoside. There is no heat exchanged in the titration of the double mutant or the site IIIC knockout LlpAV177Y, whereas the site IIC knockout LlpAV208Y, binds the monosaccharide in a ‘‘wildtype’’- like fashion, showing that only site IIIC is involved in sugar binding. doi:10.1371/journal.ppat.1003199.g005 Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 8 February 2013 | Volume 9 | Issue 2 | e1003199 In general, a defensive role has been proposed for the (oligo)mannose-binding MMBL lectins based on insecticidal, nematicidal, antifungal, or even antiviral activities demonstrated for several of these proteins that are abundantly found in monocot plants [40–46]. Some of these plant lectins trigger apoptosis in cancer cells [47]. Also their identification in fish mucus and epithelial cells is in line with a general protective (antimicrobial) function for MMBL domains [32]. LlpABW as a bactericidal protein fits within this picture of MMBL domains being involved in general defense mechanisms. Since no antibacterial activity has been assigned to the eukaryotic MMBL proteins, it is challenging to identify structural features that confer the intragenus-specific bacteriocin activity of LlpA, as shown for proteins from P. putida [21], P. fluorescens [22], P. syringae [23], and Xanthomonas citri [23]. Their target spectra are narrower than reported for the mammalian antibacterial C-type lectins of the RegIII family, such as mouse RegIIIc and its human homolog HIP/PAP that bind to the surface-exposed peptidoglycan layer of Gram-positive bacteria [48], and RegIIIb that also binds to the lipid A moiety of lipopolysaccharides on the cell envelope of Gram-negative bacteria [49]. The absence of any known secretory signal sequence in LlpABW and its homologues in other Pseudomonas species is intriguing in view of their extracellular location [21]. The translocation of the outer membrane-associated mannose/ fucose-specific lectin LecB of P. aeruginosa, that also lacks such signal sequence [50], is dependent on its glycosylation [51]. Contrary to LlpA that is exported to the culture supernatant to exert its antagonistic activity, LecB remains associated with the cell envelope through interaction with the major outer membrane protein OprF [52], in line with its role in biofilm formation. Materials and Methods Strains and culture conditions Bacterial strains and plasmids used in this study are listed in Table S2. Escherichia coli was routinely grown in shaken Luria- Bertani (LB, MP Biomedicals) broth at 37uC. Pseudomonas strains were grown in Tryptic Soy Broth (BD Biosciences) at 30uC with shaking. Media were solidified with 1.5% agar (Invitrogen) and supplemented with filter-sterilized antibiotics as required at following concentrations: ampicillin (Sigma-Aldrich), 100 mg/ml or kanamycin (Sigma-Aldrich), 50 mg/ml. Isopropyl b-D-thioga- lactoside (IPTG 40 mg/ml, ForMedium) and 5-bromo-4-chloro-3- indolyl-b-D-galactopyranoside (X-Gal 40 mg/ml, ForMedium) were added for blue/white screening of pUC18-derived plasmids in E. coli. Plasmids used for antibacterial testing and sequencing were propagated in E. coli TOP10F9 (Invitrogen). E. coli BL21(DE3) (Novagen) was used as a host for plasmids driving recombinant protein expression. Genomic DNA from P. putida BW11M1 was isolated using the Puregene Yeast/Bact. Kit B (Qiagen). Plasmid DNA was extracted using the QIAprep Spin Miniprep Kit (Qiagen). Stocks were stored at 280uC in the appropriate medium in 25% (v/v) glycerol. Figure 7. Differential inhibitory activity of wild-type LlpABW and LlpA/LlpA1 domain chimers. The domain structures of LlpABW (as in Figure 4) and of LlpA1 (inferred by pairwise alignment; N-domain in orange, C-domain in purple and C-terminal extension in grey) are depicted, along with those of chimeric forms (in dashed box). The LlpA variant lacking the terminal phenylalanine residue is marked with a yellow hexagon. Inhibitory activity of the respective E. coli recombinants was tested with diagnostic indicators for LlpABW (P. syringae GR12-2R3) and LlpA1 (P. fluorescens LMG 1794). Halo sizes are semi-quantified according to size of the growth inhibition halo (+++, native halo size of LlpABW and LlpA1; ++, halo size reduced; C, local clearing confined to producer colony spot; 2, no halo or clearing; NT, not tested). Additional chimeric and domain deletion constructs not conferring bacteriocin activity against one of the indicator strains are specified in Figure S11. doi:10.1371/journal.ppat.1003199.g007 Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 9 February 2013 | Volume 9 | Issue 2 | e1003199 Recombinant DNA methods Standard methods were used for preparation of competent E. coli cells, heat shock transformation of E. coli and DNA electrophoresis [53]. Restriction enzymes were used according to the supplier’s specifications (Roche Diagnostics and BIOKE´). DNA ligation was performed using T4 DNA ligase (Invitrogen). Plasmid sequencing was performed by GATC Biotech (Constance, Germany). Constructs that were generated are listed in Table S2 and primers are listed in Table S3. A 921-bp fragment containing llpABW was amplified by PCR with Platinum Pfx DNA polymerase (Invitrogen), using a C1000 Thermal Cycler (Bio-Rad). P. putida BW11M1 genomic DNA was taken as a template, and combined with primers PGPRB- 3155 and PGPRB-3156. The amplicon was purified using the QIAquick PCR Purification Kit (Qiagen), digested with KpnI and BamHI, ligated in pUC18, and transformed to E. coli TOP10F9. Transformants were verified for the presence of the insert by PCR using Taq Polymerase (BIOKE´ ) with primers PGPRB-2545 and PGPRB-2546. The cloned construct (pCMPG6129) was purified and its insert confirmed by sequencing. DpnI-mediated site-directed mutagenesis was performed to construct valine-to-tyrosine mutant forms of llpABW in pUC18 (pCMPG6129) and N-terminal His-tagged llpABW in pET28a (pCMPG6056 [54]). PCR conditions were: 2 min initial denatur- ation, followed by 16 cycles of denaturation (1 min), annealing (1 min, primer-dependent temperature) and elongation at 68uC (1 min./kb). Final elongation was for 8 min at 68uC. After PCR, samples were immediately treated with DpnI at 37uC for 2 h and transformed into E. coli TOP10F9 and selected on the appropriate medium. Plasmid inserts of selected transformants were verified by sequence analysis. Double mutants were constructed using plasmids with a single point mutation as a template. Domain deletants of llpABW were constructed using pCMPG6129 as a template (llpABW from P. putida BW11M1). Chimeric constructs were obtained using pCMPG6129 and pCMPG6053 (llpA1 from P. fluorescens Pf-5 [22]) as templates. Artificial ligation of gene fragments, generated with the PCR primers specified in Table S3, was performed by using splicing by overlap extension (SOE). The resulting recombinant amino acid sequences are listed in Table S4. Recombinant protein expression and purification Protein isolation and purification of N-terminal His6-tagged LlpABW, LlpAV177Y, LlpAV208Y, and LlpAV177Y-V208Y from E. coli BL21(DE3), carrying expression constructs pCMPG6056, pCMPG6149, pCMPG6150 and pCMPG6151 respectively, were performed as described by Parret and collaborators [54]. The presence of His-tagged protein was observed via immunodetection by Western blot, using monoclonal anti-His6 (IgG1 from mouse; Roche Diagnostics) as primary antibody. Fractions free of other proteins, as verified by SDS-PAGE and subsequent Coomassie Blue staining, were dialyzed against bis-TRIS propane buffer (20 mM, 200 mM NaCl, pH 7.0). Concentrations of purified proteins were determined by absorbance measurement at 280 nm using molar extinction coefficients of 62910 M21 cm21 for LlpABW, 64400 M21 cm21 for LlpAV177Y and LlpAV208Y, and 65890 M21 cm21 for LlpAV177Y-V208Y. Extinction coefficients were calculated according to Pace and collaborators [55]. Antibacterial assays Bacteriocin production by E. coli cells carrying pUC18-derived constructs was assayed as follows: 2-ml drops of an overnight stationary-phase culture were spotted onto LB agar plates and incubated for 8 h at 37uC. Next, plates were exposed to chloroform vapor (30 min), aerated and subsequently overlaid with 5 ml of soft agar (0.5%), seeded with 200 ml of a cell culture of an indicator strain (,108 CFU/ml), followed by overnight incubation at 30uC. Next day, plates were scored for the presence of halos in the confluently grown overlay. Antibacterial activity assays with purified recombinant His6- tagged proteins were performed as described by Ghequire and collaborators [23]. To assess the influence of added sugar, the same assay was carried out on agar medium supplemented with D- mannose (Sigma-Aldrich) or methyl-a-D-mannopyranoside (Sig- ma-Aldrich) to a final concentration of 0.01 M to 0.5 M. A Bioscreen C apparatus (Oy Growth Curves Ab Ltd, Finland) was used to determine the minimum inhibitory concentration (MIC). An overnight culture (16 h) of the indicator strain was diluted to 104–105 CFU/ml and incubated at 30uC, with a two- fold dilution series of recombinant His6-tagged LlpABW or mutant LlpABW. Bis-TRIS propane buffer was used as control. The MIC value was determined as the minimum concentration of protein at which no growth of the indicator strain (OD600,0.2) occurred after 24 h. At least three independent repeats, each with three replicates, were carried out. Glycan array His6-tagged LlpABW was lyophilized and verified for antibac- terial activity. After re-dissolving in MilliQ water, recombinant LlpABW was diluted to 200 mg/ml with binding buffer (20 mM TRIS-HCl pH 7.4, 150 mM NaCl, 2 mM CaCl2, 2 mM MgCl2, 0.05% Tween 20, 1% BSA), and used to probe the printed glycan arrays [56] following the standard procedures of Core H of the Consortium for Functional Glycomics (http://www. functionalglycomics.org/). Monoclonal anti-His6 antibodies (Roche Diagnostics) were used as primary antibodies, and fluorescently labeled anti-mouse IgG as secondary antibodies. Circular dichroism CD spectra were acquired on a Jasco J-715 spectropolarimeter. Curves were averaged over 8 scans, taken at 25uC using a 1 mm cuvette. Samples were dialyzed against bis-TRIS propane buffer (20 mM, NaCl 200 mM, pH 7.0), filtered and degassed prior to data acquisition. All proteins were assayed at 0.4 mg/ml. Isothermal titration calorimetry ITC titrations were carried out on an ITC200 apparatus (MicroCal). Prior to the measurement, LlpABW, LlpAV177Y, LlpAV208Y and LlpAV177Y-V208Y was dialyzed to bis-TRIS propane buffer. Sugars were directly dissolved into the same buffer. The samples were filtered and degassed for 10 min at 25uC before being examined in the calorimeter. The titrations were carried out at 25uC, injecting the sugars (methyl-a-D- mannoside, Mana(1–2)Man, Mana(1–3)Man, Mana(1–6)Man, Mana(1–3)[Mana(1–6)]Man and GlcNAcb(1–2)Mana(1– 3)[GlcNAcb(1–2)Mana(1–6)]Man) into a protein solution (pro- tein concentrations ranged from 2 mM to 4 mM depending on protein availability). All data were analyzed using the MicroCal Origin ITC 7.0 software. Binding affinities and thermodynamic parameters from all ITC titrations are reported in Table 1. X-ray data collection and structure determination Expression, purification and crystallization of recombinant His- tagged LlpABW have been described [54]. X-ray data for native and derivative crystals were collected on EMBL beamline BW7A of the DESY synchrotron (Hamburg, Germany). For each potential Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 10 February 2013 | Volume 9 | Issue 2 | e1003199 derivative, the wavelength was chosen to be at the high-energy side of the absorption edge in order to ensure a usable anomalous signal. All data were scaled and merged with the HKL package of programs. Data collection statistics are given in Table S5. The crystal structure of free LlpABW was solved combining single isomorphous replacement with anomalous scattering (SIRAS strategy) from a p-chloromercurybenzoate derivative. The heavy-atom substructure was determined with SHELXD [57] using a resolution cutoff of 4.0 A˚ . Heavy-atom refinement and phasing were performed with SHARP [58]. Phase improve- ment by solvent flattening was performed with SOLOMON [59]. Non-crystallographic symmetry averaging with density modifica- tion [60] further improved the electron density. A partial model (94% of the residues comprising the asymmetric unit) was automatically built with ARP/wARP [61] and the remainder was built manually over several cycles of model building with Coot [62], alternated with refinement using phenix.refine [63,64]. Phasing and refinement statistics are shown in Table S5. Carbohydrate soaks Crystals of LlpABW were transferred to artificial mother liquor (0.1 M imidazole pH 6.5, 1.3 M sodium acetate) enriched with either 200 mM methyl-a-D-mannopyranoside (Me-Man), Ma- na(1–2)Man, GlcNAcb(1–2)Mana(1–3)[GlcNAcb(1–2)Mana(1– 6)]Man (M592), D-galactose, L-fucose or N-acetyl-D-glucosamine and allowed to equilibrate overnight (all carbohydrates obtained from Dextra Laboratories, Reading, U.K.). Data were collected at room temperature on EMBL beamline X13, except for the N- acetyl-D-glucosamine soak collected at the PROXIMA-1 beam- line of the SOLEIL synchrotron (Gif-sur-Yvette, France) and the D-galactose soak collected at ESRF beam line ID14-1 (Grenoble, France). All data were scaled and merged using the HKL package. Refinement was started from the coordinates of the ligand-free structure using phenix.refine. Manual rebuilding, including the introduction of the carbohydrate ligand if present, was done with Coot [62]. Crystal structures of LlpABW from P. putida BW11M1 (PDB entry 3M7H) and LlpABW in complex with methyl-a-D- mannoside (PDB entry 3M7J), with Mana(1–2)Man (PDB entry 4GC1), and with GlcNAcb(1–2)Mana(1–3)[GlcNAcb(1–2)Ma- na(1–6)]Man (PDB entry 4GC2) have been deposited at the PDB. Flow cytometry Overnight cultures of P. syringae GR12-2R3 (16 h) were diluted to OD600 0.5 and washed twice with phosphate-buffered saline (PBS). Cells were treated with LlpA, mutant proteins or buffer (bis- TRIS propane buffer, negative control), at a final concentration of 50 mg/ml for 1 h, at 20uC. Next, PBS-washed bacteria were stained using the Live/Dead BacLight bacterial viability kit (Invitrogen), incubated for 15 minutes, and analyzed on a BD Influx (BD Biosciences). Excitation of the dyes was done at 488 nm, and fluorescence measured at 530 nm for SYTO 9 and at 610 nm for propidium iodide. Results were processed with FlowJo 10.0.4 software (Figure S10). Measurements were done indepen- dently and based on six biological repeats. Results are expressed as percentages of dead cells [dead/(live+dead) * 100)]. Supporting Information Figure S1 Amino acid sequence of LlpABW colored according to its domain structure. The N-domain is shown in red, the C-domain in blue and the C-terminal extension in green. Residues belonging to sequences equivalent to the mannose binding site signature motif QxDxNxVxY are in bold and underlined. (JPG) Figure S2 Quaternary structures and domain organi- zation of various MMBL family members. Individual domains or protomers are shown in different colours. The domain or protomer colored green (which in the tandem MMBLs of LlpA, ASA I and SCAfet corresponds to the N-terminal domain) is always shown in the same orientation. Bound carbohydrates are shown in black stick representation. For LlpA a single pentasaccharide is bound to site IIIC. In the case of Galanthus nivalis (snowdrop) lectin (PDB entry 1JPC), twelve trimannosides are bound to all QxDxNxVxY motifs (three on each monomer of the homotetrameric protein). The snowdrop lectin tetramer consists of the association of two domain-swapped dimers (green-blue and pink-yellow). In the case of Allium sativum (garlic lectin ASA I - PDB entry 1KJ1), again each QxDxNxVxY motif has a dimannose bound while an additional sugar (shown in red) is bound to a non-canonical site. The protein is synthesized as a single chain precursor and post-translationally cleaved into two MMBL domains that adopt the same domain-swapped dimer as found in snowdrop lectin. Gastrodianin is a monomeric MMBL family member from the orchid Gastrodia elata (PDB entry 1XD5). The location(s) of its carbohydrate-binding site(s) is (are) not known. The fetuin-binding tandem-MMBL SCAfet from Scilla campanulata (PDB entry 1DLP) consists of two covalently attached MMBL domains, whereas in LlpA the swap of the C- terminal b-strands is retained. The relative orientation in the two domains is as in ASA I. This lectin binds fetuin rather than oligomannosides, but the locations of the binding sites are not known. (JPG) Figure S3 Stereo view of the superpositions (Ca repre- sentations) of the N-domain of LlpABW (red), C-domain of LlpABW (blue) and Galanthus nivalis lectin (PDB entry 1MSA, black). The superposition is shown in two orientations rotated by 90u. (JPG) Figure S4 Sequence alignment of potential mannose- binding motifs in prokaryotic tandem MMBL proteins. The sequences corresponding to the consensus motif QxDxNxVxY, extracted from the N-domain and the C-domain of P. putida LlpABW and its homologues, are aligned per domain. Sequence conservation is visualized by differential shading. The sequence logo representation visualizes the degree of consensus for each residue. LlpA proteins with proven bacteriotoxic activity are labeled with an asterisk. Accession numbers: Arthrobacter sp. FB24 (YP_829274), Burkholderia ambifaria MEX-5 (ZP_02905572), Burk- holderia cenocepacia AU 1054 ([1], ABF75998; [2], ABF75999), Pseudomonas chlororaphis subsp. aureofaciens 30–84 (EJL08681), Pseudomonas putida BW11M1 (AAM95702), Pseudomonas fluorescens Pf-5 (LlpA1 [1], YP_258360; LlpA2 [2], YP_259234), Pseudomonas sp. GM80 ([1], ZP_10606046; [2], ZP_10606131), Pseudomonas syringae pv. aptata DSM 50252 (EGH77666), Pseudomonas syringae pv. syringae 642 (ZP_07263221), Xanthomonas axonopodis pv. citri str. 306 (AAM35756). (TIF) Figure S5 Sequence alignment of the carboxy-terminal sequences of LlpA-like proteins. The P. putida LlpABW sequence adopting a b-hairpin fold is delineated in Figure S1. The preceding conserved tryptophan residue is located C-terminally to IC (Figure S1). The sequence logo representation visualizes the degree of consensus for each residue. Accession numbers are listed in Figure S4. (TIF) Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 11 February 2013 | Volume 9 | Issue 2 | e1003199 Figure S6 Electron density for (A) Methyl-a-D-Man, (B) Mana(1–2)Man and (C) GlcNAcb(1–2)Mana(1– 3)[GlcNAcb(1–2)Mana(1–6)]Man. Difference electron-density maps are calculated by removing the sugar residues from the final coordinates and applying one round of slow-cool simulated annealing refinement to remove potential bias. The atomic model is superimposed in each case. (JPG) Figure S7 Mannose binding to LlpABW and garlic lectin. (A) Cartoon representation of subdomain IIIC of LlpABW (green) with residues implicated in carbohydrate binding showing in ball- and-stick representation and labeled (carbon green, oxygen red, nitrogen blue). The Me-Man residue is shown in red. Selected hydrogen bonds are shown as black dotted lines. (B) Stereoview of the superposition of subdomain IIIC of LlpABW (green) on the equivalent subdomain of garlic lectin (blue). The Me-Man residue bound to LlpABW is shown in red, the mannose bound to garlic lectin in blue. (C) Stereoview of the superposition of subdomain IIIC of LlpABW (green) identical as in panel A, but emphasizing the location of Val177 (shown as black sticks). The modeled Val177Tyr mutation is shown as orange sticks. Tyr177 makes a steric clash with the bound mannose (red) and is therefore expected to prevent binding, in agreement with our ITC experiments. (JPG) Figure S8 Sites II and I of the LlpABW C-domain. (A) Stereoview of site IIC of the C-domain (colored according to atom type) superimposed on site IIIC of the C-domain (dark gray). The Me-Man bound in site IIIC is shown in red. This site is very similar to site IIIC but in the crystal it is inaccessible due to crystal lattice interactions. Residue labels correspond to residues of site IIC. (B) Similar view showing site IC of the C-domain (colored according to atom type) superimposed on site IIIC of the C-domain (dark gray). The Me-Man bound in site IIIC is shown in red. The stretch of Ile137-Leu139 that provides a steric conflict preventing Me-Man binding in site IC, is highlighted with carbon atoms drawn in green. Residue labels correspond to residues of site IC. (JPG) Figure S9 Sites of the LlpABW N-terminal domain. (A) Stereoview of site IN of the N-domain (colored according to atom type) superimposed on site IIIC of the C-domain (dark gray). The Me-Man bound in site IIIC is shown in red. The stretch of Ile271- Trp274 that provides a steric conflict preventing Me-Man binding in site IN is highlighted with carbon atoms drawn in green. Residue numbering corresponds to residues of site IN. (B) Similar superposition for site IIN of the N-domain. Phe86 that prevents Me-Man binding to this site through a steric conflict is highlighted in green. Other residues belonging to site IIN are labeled in teal. Three residues of site IIIC for which site IIN has no structural equivalent are labeled in black. (C) Similar superposition for site IIIN of the N-domain. Residues belonging to site IIIN are labeled in teal. One residue of site IIIC for which site IIIN has no structural equivalent, is labeled in black. For this site there are no obvious steric conflicts that would prevent positioning of a Me-Man residue although none is observed experimentally. (JPG) Figure S10 Quantification of live and dead cells by flow cytometry. P. syringae GR12-2R3 cells were treated with LlpA (A), LlpAV177Y (B), LlpAV208Y (C), LlpAV177Y-V208Y (D), or buffer (E) at a final concentration of 50 mg/ml for 1 h at 20uC. After live/dead staining, cells were analysed by flow cytometry. Data processing allowed to distinguish populations of dead (left) and live (right) cells. Spot densities ranging from high to low are differentiated by a color gradient from red, yellow, green, teal to blue. Representative samples for LlpA, mutant proteins and buffer control are shown in panels A–E. (TIF) Figure S11 Overview of inactive LlpABW deletants and inactive LlpABW/LlpA1 chimers. The equivalent domains of LlpA1 are delineated based on pairwise sequence alignment with LlpABW: N-domain (orange), C-domain (purple), C-terminal extension (grey). No bacteriocin activity was conferred by these constructs upon recombinant E. coli cells tested against P. syringae GR12-2R3 (indicator strain for native LlpABW) and P. fluorescens LMG 1794 (indicator strain for native LlpA1). The small black rectangle represents an artificial linker sequence (DASRS). (TIF) Table S1 Glycan array profile of LlpABW as measured by fluorescence intensity. Results including a comprehensive list of oligosaccharides (array version PA_v5) are available from the Consortium of Functional Glycomics (CFG, www. functionalglycomics.org). (XLS) Table S2 Bacterial strains and plasmids used in this study. (DOC) Table S3 PCR primers used in this study. (DOCX) Table S4 Protein sequences of LlpABW deletants and LlpABW/LlpA1 chimers. (DOCX) Table S5 Structure determination and refinement. (DOCX) Acknowledgments The authors acknowledge the Consortium for Functional Glycomics for performing glycan array tests on LlpABW, and the use of the macromolecular crystallography beamlines at EMBL/DESY (Hamburg, Germany), ESRF (Grenoble, France) and SOLEIL (Gif-sur-Yvette, France) for X-ray data collection. Author Contributions Conceived and designed the experiments: MGKG AGP RL RDM. Performed the experiments: MGKG AGP EKML SS. Analyzed the data: MGKG AGP EKML SS RL RDM. Contributed reagents/materials/ analysis tools: MGKG AGP EKML SS RL RDM. Wrote the paper: MGKG AGP RL RDM. References 1. Frey-Klett P, Burlinson P, Deveau A, Barret M, Tarkka M, et al. (2011) Bacterial-fungal interactions: hyphens between agricultural, clinical, environ- mental, and food microbiologists. Microbiol Mol Biol Rev 75: 583–609. 2. Mela F, Fritsche K, de Boer W, van Veen JA, de Graaff LH, et al. (2011) Dual transcriptional profiling of a bacterial/fungal confrontation: Collimonas fungivorans versus Aspergillus niger. ISME J 5: 1494–1504. Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 12 February 2013 | Volume 9 | Issue 2 | e1003199 3. Garbeva P, Silby MW, Raaijmakers JM, Levy SB, de Boer W (2011) Transcriptional and antagonistic responses of Pseudomonas fluorescens Pf0-1 to phylogenetically different bacterial competitors. ISME J 5: 973–985. 4. Shank EA, Klepac-Ceraj V, Collado-Torres L, Powers GE, Losick R, et al. (2011) Interspecies interactions that result in Bacillus subtilis forming biofilms are mediated mainly by members of its own genus. Proc Natl Acad Sci U S A 108: E1236–E1243. 5. Rosenthal AZ, Matson EG, Eldar A, Leadbetter JR (2011) RNA-seq reveals cooperative metabolic interactions between two termite-gut spirochete species in co-culture. ISME J 5: 1133–1142. 6. Hibbing ME, Fuqua C, Parsek MR, Peterson SB (2010) Bacterial competition: surviving and thriving in the microbial jungle. Nat Rev Microbiol 8: 15–25. 7. Aoki SK, Diner EJ, de Roodenbeke CT, Burgess BR, Poole SJ, et al. (2010) A widespread family of polymorphic contact-dependent toxin delivery systems in bacteria. Nature 468: 439–442. 8. Poole SJ, Diner EJ, Aoki SK, Braaten BA, t’Kint de Roodenbeke C, et al. (2011) Identification of functional toxin/immunity genes linked to contact-dependent growth inhibition (CDI) and rearrangement hotspot (Rhs) systems. PLoS Genet 7: e1002217. 9. Russell AB, Hood RD, Bui NK, LeRoux M, Vollmer W, et al. (2011) Type VI secretion delivers bacteriolytic effectors to target cells. Nature 475: 343–347. 10. Michel-Briand Y, Baysse C (2002) The pyocins of Pseudomonas aeruginosa. Biochimie 84: 499–510. 11. Nakayama K, Takashima K, Ishihara H, Shinomiya T, Kageyama M, et al. (2000) The R-type pyocin of Pseudomonas aeruginosa is related to P2 phage, and the F-type is related to lambda phage. Mol Microbiol 38: 213–231. 12. Williams SR, Gebhart D, Martin DW, Scholl D (2008) Retargeting R-type pyocins to generate novel bactericidal protein complexes. Appl Environ Microbiol 74: 3868–3876. 13. Fischer S, Godino A, Quesada JM, Cordero P, Jofre´ E, et al. (2012) Characterization of a phage-like pyocin from the plant growth-promoting rhizobacterium Pseudomonas fluorescens SF4c. Microbiology 158: 1493–1503. 14. Ko¨hler T, Donner V, van Delden C (2010) Lipopolysaccharide as shield and receptor for R-pyocin-mediated killing in Pseudomonas aeruginosa. J Bacteriol 192: 1921–1928 15. Cascales E, Buchanan SK, Duche´ D, Kleanthous C, Lloube`s R, et al. (2007) Colicin biology. Microbiol Mol Biol Rev 71: 158–229. 16. Denayer S, Matthijs S, Cornelis P (2007) Pyocin S2 (Sa) kills Pseudomonas aeruginosa strains via the FpvA type I ferripyoverdine receptor. J Bacteriol 189: 7663–7668. 17. Elfarash A, Wei Q, Cornelis P (2012) The soluble pyocins S2 and S4 from Pseudomonas aeruginosa bind to the same FpvAI receptor. Microbiologyopen 1: 268–275. 18. Ling H, Saeidi N, Rasouliha BH, Chang MW (2010) A predicted S-type pyocin shows a bactericidal activity against clinical Pseudomonas aeruginosa isolates through membrane damage. FEBS Lett 584: 3354–3358. 19. Barreteau H, Tiouajni M, Graille M, Josseaume N, Bouhss A, et al. (2012) Functional and structural characterization of PaeM, a colicin M-like bacteriocin produced by Pseudomonas aeruginosa. J Biol Chem 287: 37395–37405. 20. Grinter R, Roszak AW, Cogdell RJ, Milner JJ, Walker D (2012) The crystal structure of the lipid II-degrading bacteriocin syringacin M suggests unexpected evolutionary relationships between colicin M-like bacteriocins. J Biol Chem 287: 38876–38888. 21. Parret AHA, Schoofs G, Proost P, De Mot R (2003) Plant lectin-like bacteriocin from a rhizosphere-colonizing Pseudomonas isolate. J Bacteriol 185: 897–908. 22. Parret AHA, Temmerman K, De Mot R (2005) Novel lectin-like bacteriocins of biocontrol strain Pseudomonas fluorescens Pf-5. Appl Environ Microbiol 71: 5197– 5207. 23. Ghequire MGK, Li W, Proost P, Loris R, De Mot R (2012) Plant lectin-like antibacterial proteins from phytopathogens Pseudomonas syringae and Xanthomonas citri. Environ Microbiol Rep 4: 373–380. 24. Van Damme EJM, Lannoo N, Peumans WJ (2008) Plant lectins. Adv Bot Res 48: 107–209. 25. Ghequire MGK, Loris R, De Mot R (2012) MMBL proteins: from lectin to bacteriocin. Biochem Soc Trans 40: 1553–1559. 26. Van Damme EJM, Kaku H, Perini F, Goldstein IJ, Peeters B, et al. (1991) Biosynthesis, primary structure and molecular cloning of snowdrop (Galanthus nivalis L.) lectin. Eur J Biochem 202: 23–30. 27. Hester G, Wright CS (1996) The mannose-specific bulb lectin from Galanthus nivalis (snowdrop) binds mono- and dimannosides at distinct sites. Structure analysis of refined complexes at 2.3 A˚ and 3.0 A˚ resolution. J Mol Biol 262: 516– 531. 28. Fouquaert E, Peumans WJ, Gheysen G, Van Damme EJM (2011) Identical homologs of the Galanthus nivalis agglutinin in Zea mays and Fusarium verticillioides. Plant Physiol Biochem 49: 46–54. 29. Shimokawa M, Fukudome A, Yamashita R, Minami Y, Yagi F, et al. (2012) Characterization and cloning of GNA-like lectin from the mushroom Marasmius oreades. Glycoconj J 29: 457–465. 30. Jung E, Fucini P, Stewart M, Noegel AA, Schleicher M (1996) Linking microfilaments to intracellular membranes: the actin-binding and vesicle- associated protein comitin exhibits a mannose-specific lectin activity. EMBO J 15: 1238–1246. 31. Wiens M, Belikov SI, Kaluzhnaya OV, Krasko A, Schro¨der HC, et al. (2006) Molecular control of serial module formation along the apical-basal axis in the sponge Lubomirskia baicalensis: silicateins, mannose-binding lectin and mago nashi. Dev Genes Evol 216: 229–242. 32. Tsutsui S, Tasumi S, Suetake H, Suzuki Y (2003) Lectins homologous to those of monocotyledonous plants in the skin mucus and intestine of pufferfish, Fugu rubripes. J Biol Chem 278: 20882–20889. 33. de Santana Evangelista K, Andrich F, Figueiredo de Rezende F, Niland S, Cordeiro MN, et al. (2009) Plumieribetin, a fish lectin homologous to mannose- binding B-type lectins, inhibits the collagen-binding a1b1 integrin. J Biol Chem 284: 34747–34759. 34. Rajan B, Fernandes JM, Caipang CM, Kiron V, Rombout JH, et al. (2011) Proteome reference map of the skin mucus of Atlantic cod (Gadus morhua) revealing immune competent molecules. Fish Shellfish Immunol 31: 224–231. 35. Chen J, Stevenson DM, Weimer PJ (2004) Albusin B, a bacteriocin from the ruminal bacterium Ruminococcus albus 7 that inhibits growth of Ruminococcus flavefaciens. Appl Environ Microbiol 70: 3167–3170. 36. Wright CS, Hester G (1996) The 2.0 A˚ structure of a cross-linked complex between snowdrop lectin and a branched mannopentaose: evidence for two unique binding modes. Structure 4: 1339–1352. 37. Wright LM, Reynolds CD, Rizkallah PJ, Allen AK, Van Damme EJM, et al. (2000) Structural characterisation of the native fetuin-binding protein Scilla campanulata agglutinin: a novel two-domain lectin. FEBS Lett 468: 19–22. 38. Ramachandraiah G, Chandra NR, Surolia A, Vijayan M (2002) Re-refinement using reprocessed data to improve the quality of the structure: a case study involving garlic lectin. Acta Crystallogr D Biol Crystallogr 58: 414–420. 39. Robert V, Volokhina EB, Senf F, Bos MP, Van Gelder P, et al. (2006) Assembly factor Omp85 recognizes its outer membrane protein substrates by a species- specific C-terminal motif. PLoS Biol 4: e377. 40. Wang X, Bauw G, Van Damme EJM, Peumans WJ, Chen ZL, et al. (2001) Gastrodianin-like mannose-binding proteins: a novel class of plant proteins with antifungal properties. Plant J 25: 651–661. 41. Balzarini J (2007) Targeting the glycans of glycoproteins: a novel paradigm for antiviral therapy. Nat Rev Microbiol 5: 583–597. 42. Tian Q, Wang W, Miao C, Peng H, Liu B, et al. (2008) Purification, characterization and molecular cloning of a novel mannose-binding lectin from rhizomes of Ophiopogon japonicus with antiviral and antifungal activities. Plant Sci 175: 877–884. 43. Bharathi Y, Vijaya Kumar S, Pasalu IC, Balachandran SM, Reddy VD, et al. (2011) Pyramided rice lines harbouring Allium sativum (asaI) and Galanthus nivalis (gna) lectin genes impart enhanced resistance against major sap-sucking pests. J Biotechnol 152: 63–71. 44. Hoorelbeke B, Van Damme EJM, Rouge´ P, Schols D, Van Laethem K, et al. (2011) Differences in the mannose oligomer specificities of the closely related lectins from Galanthus nivalis and Zea mays strongly determine their eventual anti- HIV activity. Retrovirology 8: 10. 45. Vandenborre G, Smagghe G, Van Damme EJM (2011) Plant lectins as defense proteins against phytophagous insects. Phytochemistry 72: 1538–1550. 46. Yang Y, Xu HL, Zhang ZT, Liu JJ, Li WW, et al. (2011) Characterization, molecular cloning, and in silico analysis of a novel mannose-binding lectin from Polygonatum odoratum (Mill.) with anti-HSV-II and apoptosis-inducing activities. Phytomedicine 18: 748–755. 47. Fu LL, Zhou CC, Yao S, Yu JY, Liu B, et al. (2011) Plant lectins: targeting programmed cell death pathways as antitumor agents. Int J Biochem Cell Biol 43: 1442–1449. 48. Lehotzky RE, Partch CL, Mukherjee S, Cash HL, Goldman WE, et al. (2010) Molecular basis for peptidoglycan recognition by a bactericidal lectin. Proc Natl Acad Sci U S A 107: 7722–7727. 49. Miki T, Holst O, Hardt WD (2012) The bactericidal activity of the C-type lectin RegIIIb against Gram-negative bacteria involves binding to Lipid A. J Biol Chem 287: 34844–34855. 50. Tielker D, Hacker S, Loris R, Strathmann M, Wingender J, et al. (2005) Pseudomonas aeruginosa lectin LecB is located in the outer membrane and is involved in biofilm formation. Microbiology 151: 1313–1323. 51. Bartels KM, Funken H, Knapp A, Brocker M, Wilhelm S, et al. (2011) Glycosylation is required for outer membrane localization of the lectin LecB in Pseudomonas aeruginosa. J Bacteriol 193: 1107–1113. 52. Funken H, Bartels KM, Wilhelm S, Brocker M, Bott M, et al. (2012) Specific association of lectin LecB with the surface of Pseudomonas aeruginosa: role of outer membrane protein OprF. PLoS One 7: e46857. 53. Green MR, Sambrook JR (2012) Molecular cloning: a laboratory manual. 4th edition. New York: Cold Spring Harbor Laboratory Press. 2028 p. 54. Parret AHA, Wyns L, De Mot R, Loris R (2004) Overexpression, purification and crystallization of bacteriocin LlpA from Pseudomonas sp. BW11M1. Acta Crystallogr D Biol Crystallogr 60: 1922–1924. 55. Pace CN, Vajdos F, Fee L, Grimsley G, Gray T (1995) How to measure and predict the molar absorption coefficient of a protein. Protein Sci 4: 2411–2423. 56. Blixt O, Head S, Mondala T, Scanlan C, Huflejt ME, et al. (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc Natl Acad Sci U S A 101: 17033–17038. 57. Schneider TR, Sheldrick GM (2002) Substructure solution with SHELXD. Acta Crystallogr D Biol Crystallogr 58: 1772–1779. 58. de la Fortelle E, Bricogne G (1997) Maximum likelihood heavy-atom parameter refinement for multiple isomorphous replacement and multiwavelength anomalous diffraction methods. In: Carter CW, Sweet RM, editors. Methods Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 13 February 2013 | Volume 9 | Issue 2 | e1003199 in Enzymology. Volume 276, Macromolecular Crystallography Part A. New York: Academic Press. 472–494. 59. Abrahams JP, Leslie AG (1996) Methods used in the structure determination of bovine mitochondrial F1 ATPase. Acta Crystallogr D Biol Crystallogr 52: 30– 42. 60. Cowtan K. (2010) Recent developments in classical density modification. Acta Crystallogr D Biol Crystallogr 66: 470–478. 61. Perrakis A, Morris R, Lamzin VS (1999) Automated protein model building combined with iterative structure refinement. Nat Struct Biol 6: 458–463. 62. Emsley P, Cowtan K (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132. 63. Adams PD, Gopal K, Grosse-Kunstleve RW, Hung LW, Ioerger TR, et al. (2004) Recent developments in the PHENIX software for automated crystallographic structure determination. J Synchrotron Radiat 11: 53–55. 64. Afonine PV, Grosse-Kunstleve RW, Adams PD (2005) A robust bulk-solvent correction and anisotropic scaling procedure. Acta Crystallogr D Biol Crystal- logr 61: 850–855. Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 14 February 2013 | Volume 9 | Issue 2 | e1003199
3M7J
Crystal structure of the bacteriocin LLPA from pseudomonas sp. in complex with Met-mannose
Structural Determinants for Activity and Specificity of the Bacterial Toxin LlpA Maarten G. K. Ghequire1., Abel Garcia-Pino2,3., Eline K. M. Lebbe1¤, Stijn Spaepen1, Remy Loris"2,3*, Rene´ De Mot"1* 1 Centre of Microbial and Plant Genetics, University of Leuven, Heverlee-Leuven, Belgium, 2 Molecular Recognition Unit, Department of Structural Biology, Vlaams Instituut voor Biotechnologie, Brussel, Belgium, 3 Structural Biology Brussels, Department of Biotechnology (DBIT), Vrije Universiteit Brussel, Brussel, Belgium Abstract Lectin-like bacteriotoxic proteins, identified in several plant-associated bacteria, are able to selectively kill closely related species, including several phytopathogens, such as Pseudomonas syringae and Xanthomonas species, but so far their mode of action remains unrevealed. The crystal structure of LlpABW, the prototype lectin-like bacteriocin from Pseudomonas putida, reveals an architecture of two monocot mannose-binding lectin (MMBL) domains and a C-terminal b-hairpin extension. The C-terminal MMBL domain (C-domain) adopts a fold very similar to MMBL domains from plant lectins and contains a binding site for mannose and oligomannosides. Mutational analysis indicates that an intact sugar-binding pocket in this domain is crucial for bactericidal activity. The N-terminal MMBL domain (N-domain) adopts the same fold but is structurally more divergent and lacks a functional mannose-binding site. Differential activity of engineered N/C-domain chimers derived from two LlpA homologues with different killing spectra, disclosed that the N-domain determines target specificity. Apparently this bacteriocin is assembled from two structurally similar domains that evolved separately towards dedicated functions in target recognition and bacteriotoxicity. Citation: Ghequire MGK, Garcia-Pino A, Lebbe EKM, Spaepen S, Loris R, et al. (2013) Structural Determinants for Activity and Specificity of the Bacterial Toxin LlpA. PLoS Pathog 9(2): e1003199. doi:10.1371/journal.ppat.1003199 Editor: Ambrose Cheung, Geisel School of Medicine at Dartmouth, United States of America Received August 22, 2012; Accepted January 3, 2013; Published February 28, 2013 Copyright:  2013 Ghequire et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was financially supported by FWO Vlaanderen (Research project G.0393.09), by The Onderzoeksraad of the VUB, by VIB and by the Hercules Foundation. The authors acknowledge support of the European Community - Research Infrastructure Action under the FP6 ‘‘Structuring the European Research Area Program’’, contract number: RII3-CT-2004-506008. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: remy.loris@vib-vub.be (RL); rene.demot@biw.kuleuven.be (RDM) ¤ Current address: Laboratory of Toxicology, University of Leuven, Leuven, Belgium. . These authors contributed equally to this work. " These authors also contributed equally to this work and should be considered joint senior authors. Introduction In most natural settings, complex interactions occur among microorganisms, ranging from nutritional co-operation to warfare among competitors. Examples of such interplay have been reported not only between unrelated microorganisms (e.g. fungi and bacteria [1,2]), but also between distant relatives (e.g. members of different bacterial genera [3]), and even between close relatives (e.g. at inter- and intra-species levels [4,5]). A major strategy in niche colonization is the production of growth inhibitors or toxins directed at microbial competitors [6]. While a huge variety of secondary metabolites is used to target phylogenetically-distant competitors, ribosome-synthesized pep- tides or proteins are typically active against close relatives. These protein toxins are collectively referred to as bacteriocins, and may either be released into the environment or transferred to the host via specialized contact-dependent delivery systems [7–9]. Bacteriocins are structurally and mechanistically very diverse. This is reflected in the bacteriocinogenic potential of the genus Pseudomonas [10]. Their R- and F-type pyocins are multi-subunit protein complexes evolutionarily related to contractile tails of bacteriophages [11–13]. R-pyocins attach to specific lipopolysac- charide moieties at the cell surface of susceptible cells and insert their core structure through the cell envelope, causing depolar- ization of the cytoplasmic membrane [14]. The S-type pyocins of Pseudomonas aeruginosa share structural and functional features with Escherichia coli colicins [15]. Following docking onto surface- exposed targets such as siderophore receptors [16,17], S-pyocins kill cells by nucleic acid degradation [10,17], cytoplasmic membrane damage [18], or inhibition of peptidoglycan synthesis [19,20]. Putidacin A (or LlpABW), first identified in Pseudomonas putida BW11M1 [21], represents a class of Pseudomonas-specific antibacterial proteins not related to any known bacteriocin. Additional llpA-like genes encoding functional bacteriocins were identified by genome mining in the biocontrol strain Pseudomonas fluorescens Pf-5 [22] and in the phytopathogen Pseudomonas syringae pv. syringae 642 [23]. Identification of this type of protein in two Xanthomonas pathovars extended its occurrence as a genus-specific killer protein [23]. The Xanthomonas LlpA precursor is proteolyt- ically processed by removal of a characteristic Type II secretion signal peptide, whereas such N-terminal sequence is lacking in Pseudomonas homologues, indicating that secretory routes may differ among LlpA producers. PLOS Pathogens | www.plospathogens.org 1 February 2013 | Volume 9 | Issue 2 | e1003199 The amino acid sequence of LlpA suggests the presence of two related domains belonging to the ‘monocot mannose- binding lectin’ (MMBL) family [24]. The MMBL domain consists of a b-prism fold containing three potential carbohy- drate-binding pockets, each generated by a QxDxNxVxY sequence (with x, any amino acid), but some sites may be inactive due to degeneracy of the signature motif [25]. This domain (Pfam domain: B_lectin - PF01453) was initially identified in lectins of monocot plants [26,27], but a more widespread occurrence of MMBL lectins has become evident and includes representatives in fungi [28,29], slime molds [30], sponges [31], and fishes [32–34]. The LlpA branch occupies a unique position among MMBL-domain proteins, harboring non-eukaryotic representatives and being equipped with the capacity to kill bacterial cells with bacteriocin-like specificity, a property not yet demonstrated for other family members [25]. Next to proteins with the LlpA-type tandem-MMBL organiza- tion, many other predicted MMBL proteins are encoded by bacterial genomes. Often the MMBL module is embedded in a larger protein. For one such protein, bacteriocin-like activity among Ruminococcus species, Gram-positive bacteria colonizing the rumen, was demonstrated [35]. Here we report on the crystal structure of LlpABW as the prototype of a novel family of antibacterial proteins and explore how domain architecture and specific structural elements contrib- ute to its activity and specificity. Results LlpA forms a rigid MMBL tandem The crystal structure of LlpABW from P. putida BW11M1 (LlpABW) shows it contains two b-prism MMBL domains, referred to as the N-domain and the C-domain following their position in the amino acid sequence (Figure 1A,B; Figure S1). The N-domain spans residues Arg4-Pro135 while the C-domain encompasses residues Ala136-Gln253. Each domain exhibits pseudo-threefold symmetry and the corresponding subdomains will be referred to as IN, IIN, IIIN, IC, IIC and IIIC, respectively (Figure 1A and Figure S1). Following these two domains, a b- hairpin extension is formed by residues Pro254-His275 (the numbering used in this paper corresponds to that of the wild-type protein without His-tag [21]). The two-domain architecture reflects the b-strand swapping that is typical in dimers of single-domain mannose-binding monocot lectins (Figure 1A,B) [36] and which apparently is retained after the ancestral fusion or duplication of the two domains, as is also the case in certain MMBL tandems or heterodimers from monocots [37,38]. Thus, residues Asp126- Pro135 from the first MMBL sequence complement the fold of the C-domain while residues Pro245-Gln253 from the second MMBL sequence complement the fold of the N-domain. However, in LlpABW, the relative orientation of both domains is different compared to what is observed in a canonical MMBL lectin dimer, such as snowdrop lectin [36], in the heterodimeric MMBL lectin ASA I from Allium sativum [38], or in the tandem MMBL SCAfet from Scilla campanulata [37] (Figure 1C and Figure S2). In contrast to these plant MMBL proteins, the resulting architecture of LlpABW does not obey pseudo-twofold symmetry (Figure 1C). LlpABW is a very rigid molecule. The two monomers present in the asymmetric unit are essentially identical with a root-mean- square deviation (RMSD) of 0.34 A˚ for 270 Ca atoms. This RMSD value does not change significantly when the individual domains are fitted separately (0.32 A˚ for 120 Ca’s of the N- domain and 0.22 A˚ for 115 Ca’s of the C-domain), indicating that the inter-domain orientation is fixed. This stems from three sets of interactions (Figure 2). Both domains are connected by a two- stranded anti-parallel b-sheet that is involved in the b-strand swapping mentioned above and that links both domains. The C- terminal b-hairpin extension makes extensive contacts, through hydrophobic and hydrogen bonds, with both domains. Finally, the stretch Val140-Asp145 of the C-domain makes extensive contacts with stretch Val115-Asp118 and with the side chains of Ser15 and Pro32 of the N-domain. Domains of LlpABW are shaped by differential evolutionary pressure A superposition of the Ca-trace of the N- and C-domain of LlpABW as well as the MMBL domain of snowdrop lectin is shown in Figure S3. Based on 79 Ca atoms that form the common b- sheet core of the MMBL domains, the RMSD between the N- and C-domains of LlpABW is 1.84 A˚ . While the secondary structure elements of the C-domain are restricted to the three four-stranded b-sheets of the b-prism fold, the N-domain contains three additional secondary structure elements (Figure 1A). A three-turn a-helix (a1) is inserted in the loop between strands b9 and b10, and sheet IIN contains two additional strands. Strand b69 is inserted in the loop between strands b6 and b7 and provides an anti-parallel extension to sheet II (hydrogen bonding to strand b9). Strand b19 is a short piece of b-strand that is part of the long N- terminus and forms a parallel extension on the opposite site of sheet IIN (hydrogen bonding to strand b2), making this b-sheet a mixed type six-stranded one rather than the canonical four- stranded anti-parallel sheet. Despite these additions to the b-prism fold, the common core of the N-domain more closely resembles that of the well-studied and highly conserved monocot lectins (e.g. RMSD of 1.35 A˚ with snowdrop lectin compared to 1.82 A˚ for the C-domain). This structural divergence is in contrast with the degree of conserva- Author Summary In their natural environments, microorganisms compete for space and nutrients, and a major strategy to assist in niche colonization is the deployment of antagonistic compounds directed at competitors, such as secondary metabolites (antibiotics) and antibacterial peptides or proteins (bacteriocins). The latter selectively kill closely related bacteria, which is also the case for members of the LlpA family. Here, we investigate the structure-function relationship for the prototype LlpABW from a saprophytic plant-associated Pseudomonas whose genus-specific tar- get spectrum includes several phytopathogenic pseudo- monads. By determining the 3D structure of this protein, we could assign LlpA to the so-called monocot mannose- binding lectin (MMBL) family, representing its first prokaryotic member, and also add a new type of protective function, as the eukaryotic MMBL members have been linked with antiviral, antifungal, nematicidal or insecticidal activities. For the protein containing two similarly folded domains, we constructed site-specific mutants affected in carbohydrate binding and domain chimers from LlpA homologues to show that mannose- specific sugar binding mediated by one domain is required for activity and that the other domain determines target strain specificity. The strategy that evolved for these bacteriocins is reminiscent of the one used by mammalian bactericidal proteins of the RegIII family that recruited a C- type lectin fold to kill bacteria. Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 2 February 2013 | Volume 9 | Issue 2 | e1003199 Figure 1. Overall structure of LlpABW. (A) Topology diagram of LlpABW. The N-domain is shown in red, the C-domain in blue and the C-terminal extension in green. The different strands and subdomains are labeled. Domain swapping involves b-strand segments b11b and b22b, which together with b-strand segments b11a and b22a link both MMBL domains. (B) Cartoon representation of LlpABW with the different domains colored as in panel A. The bound Me-Man residue is shown as an orange stick representation. (C) Domain orientations of LlpABW compared with the heterodimeric MMBL ASA I (Allium sativum agglutinin, PDB entry 1KJ1) and tandem MMBL SCAfet (Scilla campanulata fetuin-binding lectin, PDB entry 1DLP). In each case, the C-domain is shown in the same orientation, highlighting the different relative orientation of the N-domain in LlpABW. Domain-swapped dimers in homo-oligomeric plant MMBL lectins such as snowdrop lectin have their domain orientation similar to ASA I and SCAfet. doi:10.1371/journal.ppat.1003199.g001 Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 3 February 2013 | Volume 9 | Issue 2 | e1003199 tion of the carbohydrate-binding motif characteristic of the monocot lectins (QxDxNxVxY) in each of the three subdomains. In the N-domains of LlpA homologues, the surface-exposed motifs III and II are not well conserved and likely lost their function during evolution. In contrast they seem to be better conserved in the C-domains (Figure S4). Apparently, the two MMBL domains of LlpA experienced a differential evolutionary pressure resulting in different degrees of global and local (carbohydrate-binding motif) conservation, suggesting distinct functional roles for each domain. The C-domain of LlpABW further extends into a b-hairpin that helps to define the relative orientations of its two MMBL domains. This b-hairpin is highly bent due to a b-bulge inserted into its second b-strand (Figure 1B). It is absent in all plant representatives including tandem MMBL proteins such as SCAfet (Figure 1C). In bacteria it represents the most divergent part of LlpA homologues, both in primary sequence and in length (Figure S5). Most of these C-terminal extensions terminate with a phenylalanine residue. This is reminiscent of the conserved terminal phenylalanine of outer membrane proteins from Gram-negative bacteria such as PhoE, required for their translocation to the cell envelope [39]. An equivalent extension appears to be absent in the Xanthomonas and Arthrobacter sequences (Figure S5). LlpA is capable of binding mannose-containing carbohydrates Subdomains IIC and IIICof LlpABW contain the typical sugar- binding signature (QxDxNxVxY) of an active MMBL mannose- binding site (Figure S1 and S4). Soaking crystals of LlpABW with 200 mM methyl-a-D-mannopyranoside (Me-Man) led to clear electron density of a single Me-Man in site IIIC of each of the two LlpABW monomers in the asymmetric unit (Figure S6A). This site comprises the side chains from Gln171, Asp173, Asn175 and Tyr179, which contribute to hydrogen bond interactions and the side chains of residues Val177, Asn188, Gln192 and Ala185, which contribute to van der Waals contacts with the carbohydrate ligand (Figure 3A, Figure S7A,C). This architecture is very similar to what is observed for mannose bound to other MMBL-type lectins such as snowdrop and garlic lectin (Figure S7B). Soaks with oligomannoses revealed additional sugar-binding subsites. Binding site IIIC accommodates the disaccharide Mana(1–2)Man and the pentasaccharide GlcNAcb(1–2)Mana(1– 3)[GlcNAcb(1–2)Mana(1–6)]Man (Figure S6B,C). In the case of Figure 2. Domain interactions within LlpABW. (A) The C-terminal hairpin extension (green cartoon) covers the interface between the N-domain (red surface representation) and the C-domain (blue surface representation). (B) Stereo view of the interactions between loop segments Val140- Asp145 (cyan) of the C-domain and Val115-Asp118 (yellow) and Ser31-Gln34 (orange) of the N-domain. Other structural elements are colored according to panel A. (C) Stereo view of the two-stranded b-sheet formed by strands b11a,b and b22a,b that links the N- and the C-domains and gives rise to domain swapping. Colors according to panel A and B. doi:10.1371/journal.ppat.1003199.g002 Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 4 February 2013 | Volume 9 | Issue 2 | e1003199 the pentasaccharide, the central reducing mannose is located in the shallow Me-Man binding site and the two GlcNAcb(1–2)Man moieties stretch out over the surface making only a few additional hydrogen bonds or van der Waals contacts (Figure 3B). In the bound disaccharide, the non-reducing mannose is located in the Man-Me binding site while the reducing mannose faces the solvent and does not interact directly with the protein (Figure 3C). Site IIC of both LlpABW molecules in the asymmetric unit is involved in crystal packing interactions and the presence of Me- Man is therefore sterically excluded. All residues that form specific hydrogen bonds with Me-Man are retained but substitutions occur for three side chains that provide van der Waals contacts (Figure S4 and S8A). In contrast, site IC lost the conserved QxDxNxVxY motif (Figure S4) and is involved in inter-domain contacts and therefore inaccessible to ligands (Figure S8B). The putative carbohydrate-binding sites in the N-domain of LlpABW are less conserved. Similar to the C-domain, site IN is inaccessible and involved in inter-domain interactions (Figure S9A). In the IIN subdomain, the canonical mannose-binding motif QxDxNxVxY is essentially absent, with only the Gln residue of the motif being conserved as Gln82 (Figure S4). All other donors or acceptors required for hydrogen bonds with a mannose ligand are missing. In addition, the presence of Phe86 at the equivalent position of the expected Val sterically hinders the binding of mannose (Figure S9B). The potential carbohydrate-binding site on subdomain IIIN is only partially conserved (Figure S9C) and contains two relevant substitutions from the canonical signature: Figure 3. Carbohydrate binding in site IIIC of LlpABW. (A) Stereoview of methyl-a-D-mannopyranoside bound to subdomain IIIC. Methyl-a-D- mannopyranoside is shown in blue and indicated by M. Residues belonging to the QxDxNxVxY motif and hydrogen bonding to the sugar as well as Asn188 are labeled. Water molecules bridging protein and carbohydrate are shown in cyan (B) Similar view of the pentasaccharide GlcNAcb(1– 2)Mana(1–3)[GlcNAcb(1–2)Mana(1–6)]Man. The mannose residue occupying the primary binding site is shown in blue and labeled M. The additional two mannoses (labeled +1 and 21) and two N-acetyl glucosamine residues (labeled +2 and 22) are shown in green. Other colors are as in panel A. (C) Binding of the disaccharide Mana(1–2)Man. The non-reducing mannose residue occupying the primary binding site is shown in blue and labeled M. The second, reducing mannose is shown in green. Other colors are as in panel A. doi:10.1371/journal.ppat.1003199.g003 Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 5 February 2013 | Volume 9 | Issue 2 | e1003199 (1) the Tyr residue of the QxDxNxVxY motif is replaced by the shorter Gln49, thereby removing the canonical hydrogen bond between Man O4 and Tyr OH, and (2) a threonine at position 54 which may compensate the hydrogen bond lost due to the Tyr-to-Gln substitution in the canonical motif. The lack of electron density at this site in our Me-Man soak nevertheless indicates that this site does not recognize this ligand or that its affinity is so low that recognition would only be achieved in the context of a larger and as yet unidentified mannose-containing ligand. Alternatively, this putative site may possess specificity for a different monosaccharide. In order to evaluate this hypothesis, we soaked LlpABW crystals with D-galactose, N-acetyl-D-glucosamine and L-fucose. No electron density was observed for any of these sugars, suggesting that the N-domain has a function distinct from carbohydrate recognition (data not shown). Carbohydrate-binding capacity is required for LlpA toxicity The LlpABW motifs IIIN, IIIC and IIC create potential carbohydrate binding sites that may be involved in bacteriotoxicity of the protein. We therefore examined the role of carbohydrate binding in the bactericidal function of LlpABW. The presence of methyl-a-D-mannopyranoside up to 500 mM in the medium did not influence the activity of LlpABW on P. syringae GR12-2R3. Glycan array profiling did not highlight any specific oligosaccha- ride structure that could represent a natural ligand of LlpABW (Table S1). This could be due to the array design that is principally based on eukaryotic glycans and may therefore lack an appropri- ate carbohydrate for this prokaryotic toxin. Previously, it was observed that LlpABW from concentrated culture supernatant does not agglutinate rabbit red blood cells, nor binds to a mannose- agarose affinity matrix [21]. To assess whether the mannose-recognizing QxDxNxVxY motifs in LlpABW are nevertheless relevant for bactericidal activity, the conserved valine residue was mutated to tyrosine in subdomains IIIN, IIIC, and IIC. These mutations sterically preclude mannose or any other ligand to enter the binding sites (Figure S7C). Semi-quantitative activity assays with permeabilized E. coli cells expressing the LlpA variants in motifs IIIN, IIIC and IIC were used to assess the relationship between carbohydrate binding and bactericidal activity. Modification of the IIIN site, for which no mannose binding was observed, does not affect the antibacterial activity against P. syringae GR12-2R3 (Figure 4). In contrast, the altered IIIC pocket strongly diminishes activity, either alone or in pairwise combination with the other mutated sites (IIIN or IIC). A minor negative effect of the IIC mutation is only apparent in a double mutant, when combined with a modified IIIN motif. Purified proteins were prepared to further quantify these effects. Far UV CD spectra of these mutant forms are identical to that of native protein LlpABW, indicating that the mutations do not affect the overall structure of the protein. Isothermal titration calorim- etry (ITC) showed that LlpABW has an affinity of 2.1 mM for the pentasaccharide GlcNAcb(1–2)Mana(1–3)[GlcNAcb(1–2)Ma- na(1–6)]Man, the highest among all the tested oligo-mannosides (See Figure 5 and Table 1 for a summary of the experimentally validated LlpABW-carbohydrate interactions). This is in agreement with the crystal structures of the different complexes since this sugar is the one with the largest binding interface (Figure 3). Titrations of LlpABW, of the mutants LlpAV177Y (a site IIIC knockout), LlpAV208Y (a site IIC knockout) and of the double mutant LlpAV177Y-V208Y with a-methyl mannoside clearly pin- point site IIIC as the only responsible for the sugar binding activity. Point mutations in both sites or IIIC (V177Y) alone, completely abrogate sugar binding. However knocking out site IIC (V208Y) has little effect in binding and the affinities of LlpAV208Y for a- methyl mannoside and Mana(1–3)Man are very close to the ones measured for the wild-type protein (See Table 1 and Figure 5B). While the V208Y mutation in the IIC site has no observable effect on the MIC value for P. syringae GR12-2R3, the altered IIIC motif engenders a 5.2-fold increase in MIC (Figure 4). The mutant protein LlpAV177Y-V208Y suffers a further reduction in activity, yielding a 31.6-fold increased MIC compared to native LlpABW. The biological activities of LlpA and its mutants were further assessed by live/dead staining and subsequent flow cytometry analysis (Figure 6, Figure S10). Proportions of dead cells after 1 hour of exposure to LlpA or LlpAV208Y were comparable (10.1% and 9.7%, respectively). For LlpAV177Y this value was reduced to 6.1%, significantly lower than for LlpA. Killing activity was even further reduced for LlpAV177Y-V208Y (3.7%). These results are consistent with the MIC determination and ITC data, indicating that an active site IIIC is required to generate a fully active LlpA bacteriocin. The difference in bacteriotoxicity between LlpAV177Y and LlpAV177Y-V208Y suggests that site IIC has a supporting role in the LlpABW bacteriotoxicity. All domains are necessary for LlpABW functionality The site-directed mutagenesis approach revealed an important role for the C-domain’s carbohydrate-binding capacity in LlpABW toxicity. Considering the increased binding motif degeneration in the N-domain and the fact that a Ruminococcus bacteriocin composed of only a single MMBL domain fused to an unknown domain has been identified [35], the N-domain may fulfill a distinct function, different from that of the C-domain. In order to scrutinize the contribution of individual domains to overall activity, six domain deletion constructs of llpABW were engineered to potentially encode proteins lacking the first or second MMBL domain, a gene product devoid of the C-terminal hairpin, or a protein retaining only an individual domain (N-domain, C- domain, or hairpin) (Figure S11). To take the domain swapping into account, the constructs containing only a single MMBL domain were designed with a fusion of the swapped C-terminal b- strands to the corresponding domain via a short linker. None of these deletion constructs resulted in the production of an active protein, indicating that none of the domains are dispensable. Removal of the terminal phenylalanine residue still allows expression of a functional bacteriocin in E. coli (Figure 7), but a further C-terminal truncation (deletion of Trp-His-Phe tail) resulted in a negative bacteriocin assay (data not shown). From these data we conclude that both MMBL domains as well as the C-terminal hairpin extension are required for activity of LlpA. Whether the role of the C-terminal hairpin is any other than simply stabilization of the C-domain cannot be concluded. Target specificity of LlpA is hosted by the N-domain In order to investigate the role of the different domains in target specificity, we created hybrid LlpA proteins using the domains of LlpABW from P. putida BW11M1 and LlpA1 from P. fluorescens Pf-5. These two LlpA proteins share 45% sequence identity and differ in their target spectra. Strains P. syringae GR12-2R3 and P. fluorescens LMG1794 were identified as specific indicators for LlpABW [21] and LlpA1 [22], respectively. Six constructs carrying llpABW/ llpA1chimeric genes were made with domain exchanges involving the N-domain, C-domain, and hairpin region (Figure 7 and Figure S11). For four of these constructs activity against one of both indicators was detected. Only constructs retaining the original N- domain give rise to inhibition of the cognate indicator strain. The C-domain or the hairpin of LlpABW could be replaced with the corresponding LlpA1 domains without changing target specificity. Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 6 February 2013 | Volume 9 | Issue 2 | e1003199 Conversely, the original specificity of LlpA1 is retained upon replacement of its C-domain by the LlpABW equivalent. Discussion Structure elucidation of LlpABW from P. putida BW11M1 unequivocally assigns this bacteriocin to the MMBL lectin family, in which it constitutes the first prokaryotic member, representative for a group of bacterial proteins composed of two MMBL domains [21–23]. Systematic inactivation of the three potential carbohy- drate-binding sites present in the N-domain (IIIN) and in the C-domain (IIIC and IIC) of LlpABW, revealed that a non-occluded IIIC pocket is required to obtain a fully active LlpABW molecule. A negative co-operative effect on activity resulted when the IIC site was additionally modified. Although mannose-containing carbo- hydrates can bind to the IIIC pocket of LlpABW, it remains unclear Figure 4. Inhibitory activity of wild-type LlpABW and selected mutants with modified (potential) mannose-binding sites. The domain structure (N-domain in red, C-domain in blue and C-terminal extension in green) and the position of the MMBL motifs (potentially active binding sites in orange, inactive ones in grey) are shown. The positions of conserved valine residues converted to tyrosine residues by site-directed mutagenesis are indicated with a black bar. Inhibitory activity of E. coli strains expressing mutant LlpABW forms was assayed against P. syringae GR12-2R3 and semi- quantified according to the size (inner zone radius) of the growth inhibition halo relative to LlpABW (+++, native LlpABW; ++, halo size reduced; + halo size strongly reduced; 2, no halo; NT, not tested). For wild-type LlpABW and three purified His-tagged mutant forms (LlpAV177Y, LlpAV208Y and LlpAV177Y-V208Y) the MIC values were determined with indicator P. syringae GR12-2R3. Molar minimal inhibitory concentrations of recombinant proteins (with standard deviations): LlpA, 2.08 nM (60.58 nM); LlpAV177Y, 10.9 nM (60.66 nM); LlpAV208Y, 1.98 nM (60.066 nM); 65.72 nM (62.80 nM). doi:10.1371/journal.ppat.1003199.g004 Table 1. Binding affinities and thermodynamic parameters obtained from ITC titrations. Type of protein-carbohydrate interaction Kd (mM) DG6(kcal mol21) DH6 (kcal mol21) 2TDS6(kcal mol21) LlpABW Me-a-D-Man 45.9 21.8 25.4 3.6 LlpABW Mana(1–2)Man 42.4 21.9 23.6 1.7 LlpABW Mana(1–3)Man 18.2 22.4 25.9 3.5 LlpABW Mana(1–6)Man 17.2 22.4 25.5 3.1 LlpABW Mana(1–3)[Mana(1–6)]Man 10.1 22.6 26.4 3.8 LlpABW GlcNAcb(1–2)Mana(1–3)[GlcNAcb(1– 2)Mana(1–6)]Man 2.1 23.7 21.6 22.1 LlpAV208Y Me-a-D-Man 58.8 21.7 23.3 1.6 LlpAV208Y Mana(1–3)Man 23.0 22.2 25.1 2.9 The reported values for Kd, DGu, DHu and 2TDSu were determined from fitting a single site interaction model (n = 1) to the experimental ITC data. The interaction of the mutants LlpAV177Y and LlpAV177-V208Y with the different sugars is negligible and no heat effect was observed. Therefore they are not included in this table. doi:10.1371/journal.ppat.1003199.t001 Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 7 February 2013 | Volume 9 | Issue 2 | e1003199 if these are part of or mimic the natural ligand required for biological activity since bacteriocin activity is not impaired in the presence of excess mannose. A mutated IIIN site did not provoke a negative effect on antibacterial activity. However, the N-domain appears to play a major role in target selection. This was demonstrated by assessing the differential activity of domain chimers against two target strains, diagnostic for the LlpABW- and LlpA1- specific killing. The b-hairpin does not appear to be a specificity determi- nant, although it constitutes the most variable region among LlpA-like bacteriocins. Possibly, it is required for thermody- namic stability since it needs to be intact in LlpABW. An equivalent C-terminal stretch is absent from the Xanthomonas citri LlpA-like bacteriocin [23]. From our results relying on heterologous expression in E. coli and a bacteriocin assay with permeabilized cells, it cannot be excluded that this structural element may play a role in the way an LlpA protein is exported by its native producer cells. Figure 6. Killing activity of LlpABW and mutant proteins. Percentages of dead cells after live/dead staining as quantified by flow cytometry analysis (Figure S10). P. syringae GR12-2R3 was used as indicator strain and treated at a final concentration of 50 mg/ml for 1 h. Average values (with standard deviations; indicated by error bars): LlpA, 10.1 (61.04); LlpAV177Y, 6.1 (60.44); LlpAV208Y, 9.7 (61.39); LlpAV177Y-V208Y, 3.7 (60.90); buffer (control), 1.0 (60.11).Values are significantly different for (a) and (b), (b) and (c) (p,0.01). doi:10.1371/journal.ppat.1003199.g006 Figure 5. ITC analysis of carbohydrate binding to LlpABW and mutants. (A) Binding of LlpABW to the pentasaccharide GlcNAcb(1–2)Mana(1– 3)[GlcNAcb(1–2)Mana(1–6)]Man. (B) Binding of LlpABW (blue circles, wild type) and the mutants LlpAV177Y (green circles, site IIIC knockout), LlpAV208Y (red circles, site IIC knockout) and LlpAV177Y-V208Y (black circles, site IIC and IIIC knockout) to a-methyl mannoside. There is no heat exchanged in the titration of the double mutant or the site IIIC knockout LlpAV177Y, whereas the site IIC knockout LlpAV208Y, binds the monosaccharide in a ‘‘wildtype’’- like fashion, showing that only site IIIC is involved in sugar binding. doi:10.1371/journal.ppat.1003199.g005 Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 8 February 2013 | Volume 9 | Issue 2 | e1003199 In general, a defensive role has been proposed for the (oligo)mannose-binding MMBL lectins based on insecticidal, nematicidal, antifungal, or even antiviral activities demonstrated for several of these proteins that are abundantly found in monocot plants [40–46]. Some of these plant lectins trigger apoptosis in cancer cells [47]. Also their identification in fish mucus and epithelial cells is in line with a general protective (antimicrobial) function for MMBL domains [32]. LlpABW as a bactericidal protein fits within this picture of MMBL domains being involved in general defense mechanisms. Since no antibacterial activity has been assigned to the eukaryotic MMBL proteins, it is challenging to identify structural features that confer the intragenus-specific bacteriocin activity of LlpA, as shown for proteins from P. putida [21], P. fluorescens [22], P. syringae [23], and Xanthomonas citri [23]. Their target spectra are narrower than reported for the mammalian antibacterial C-type lectins of the RegIII family, such as mouse RegIIIc and its human homolog HIP/PAP that bind to the surface-exposed peptidoglycan layer of Gram-positive bacteria [48], and RegIIIb that also binds to the lipid A moiety of lipopolysaccharides on the cell envelope of Gram-negative bacteria [49]. The absence of any known secretory signal sequence in LlpABW and its homologues in other Pseudomonas species is intriguing in view of their extracellular location [21]. The translocation of the outer membrane-associated mannose/ fucose-specific lectin LecB of P. aeruginosa, that also lacks such signal sequence [50], is dependent on its glycosylation [51]. Contrary to LlpA that is exported to the culture supernatant to exert its antagonistic activity, LecB remains associated with the cell envelope through interaction with the major outer membrane protein OprF [52], in line with its role in biofilm formation. Materials and Methods Strains and culture conditions Bacterial strains and plasmids used in this study are listed in Table S2. Escherichia coli was routinely grown in shaken Luria- Bertani (LB, MP Biomedicals) broth at 37uC. Pseudomonas strains were grown in Tryptic Soy Broth (BD Biosciences) at 30uC with shaking. Media were solidified with 1.5% agar (Invitrogen) and supplemented with filter-sterilized antibiotics as required at following concentrations: ampicillin (Sigma-Aldrich), 100 mg/ml or kanamycin (Sigma-Aldrich), 50 mg/ml. Isopropyl b-D-thioga- lactoside (IPTG 40 mg/ml, ForMedium) and 5-bromo-4-chloro-3- indolyl-b-D-galactopyranoside (X-Gal 40 mg/ml, ForMedium) were added for blue/white screening of pUC18-derived plasmids in E. coli. Plasmids used for antibacterial testing and sequencing were propagated in E. coli TOP10F9 (Invitrogen). E. coli BL21(DE3) (Novagen) was used as a host for plasmids driving recombinant protein expression. Genomic DNA from P. putida BW11M1 was isolated using the Puregene Yeast/Bact. Kit B (Qiagen). Plasmid DNA was extracted using the QIAprep Spin Miniprep Kit (Qiagen). Stocks were stored at 280uC in the appropriate medium in 25% (v/v) glycerol. Figure 7. Differential inhibitory activity of wild-type LlpABW and LlpA/LlpA1 domain chimers. The domain structures of LlpABW (as in Figure 4) and of LlpA1 (inferred by pairwise alignment; N-domain in orange, C-domain in purple and C-terminal extension in grey) are depicted, along with those of chimeric forms (in dashed box). The LlpA variant lacking the terminal phenylalanine residue is marked with a yellow hexagon. Inhibitory activity of the respective E. coli recombinants was tested with diagnostic indicators for LlpABW (P. syringae GR12-2R3) and LlpA1 (P. fluorescens LMG 1794). Halo sizes are semi-quantified according to size of the growth inhibition halo (+++, native halo size of LlpABW and LlpA1; ++, halo size reduced; C, local clearing confined to producer colony spot; 2, no halo or clearing; NT, not tested). Additional chimeric and domain deletion constructs not conferring bacteriocin activity against one of the indicator strains are specified in Figure S11. doi:10.1371/journal.ppat.1003199.g007 Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 9 February 2013 | Volume 9 | Issue 2 | e1003199 Recombinant DNA methods Standard methods were used for preparation of competent E. coli cells, heat shock transformation of E. coli and DNA electrophoresis [53]. Restriction enzymes were used according to the supplier’s specifications (Roche Diagnostics and BIOKE´). DNA ligation was performed using T4 DNA ligase (Invitrogen). Plasmid sequencing was performed by GATC Biotech (Constance, Germany). Constructs that were generated are listed in Table S2 and primers are listed in Table S3. A 921-bp fragment containing llpABW was amplified by PCR with Platinum Pfx DNA polymerase (Invitrogen), using a C1000 Thermal Cycler (Bio-Rad). P. putida BW11M1 genomic DNA was taken as a template, and combined with primers PGPRB- 3155 and PGPRB-3156. The amplicon was purified using the QIAquick PCR Purification Kit (Qiagen), digested with KpnI and BamHI, ligated in pUC18, and transformed to E. coli TOP10F9. Transformants were verified for the presence of the insert by PCR using Taq Polymerase (BIOKE´ ) with primers PGPRB-2545 and PGPRB-2546. The cloned construct (pCMPG6129) was purified and its insert confirmed by sequencing. DpnI-mediated site-directed mutagenesis was performed to construct valine-to-tyrosine mutant forms of llpABW in pUC18 (pCMPG6129) and N-terminal His-tagged llpABW in pET28a (pCMPG6056 [54]). PCR conditions were: 2 min initial denatur- ation, followed by 16 cycles of denaturation (1 min), annealing (1 min, primer-dependent temperature) and elongation at 68uC (1 min./kb). Final elongation was for 8 min at 68uC. After PCR, samples were immediately treated with DpnI at 37uC for 2 h and transformed into E. coli TOP10F9 and selected on the appropriate medium. Plasmid inserts of selected transformants were verified by sequence analysis. Double mutants were constructed using plasmids with a single point mutation as a template. Domain deletants of llpABW were constructed using pCMPG6129 as a template (llpABW from P. putida BW11M1). Chimeric constructs were obtained using pCMPG6129 and pCMPG6053 (llpA1 from P. fluorescens Pf-5 [22]) as templates. Artificial ligation of gene fragments, generated with the PCR primers specified in Table S3, was performed by using splicing by overlap extension (SOE). The resulting recombinant amino acid sequences are listed in Table S4. Recombinant protein expression and purification Protein isolation and purification of N-terminal His6-tagged LlpABW, LlpAV177Y, LlpAV208Y, and LlpAV177Y-V208Y from E. coli BL21(DE3), carrying expression constructs pCMPG6056, pCMPG6149, pCMPG6150 and pCMPG6151 respectively, were performed as described by Parret and collaborators [54]. The presence of His-tagged protein was observed via immunodetection by Western blot, using monoclonal anti-His6 (IgG1 from mouse; Roche Diagnostics) as primary antibody. Fractions free of other proteins, as verified by SDS-PAGE and subsequent Coomassie Blue staining, were dialyzed against bis-TRIS propane buffer (20 mM, 200 mM NaCl, pH 7.0). Concentrations of purified proteins were determined by absorbance measurement at 280 nm using molar extinction coefficients of 62910 M21 cm21 for LlpABW, 64400 M21 cm21 for LlpAV177Y and LlpAV208Y, and 65890 M21 cm21 for LlpAV177Y-V208Y. Extinction coefficients were calculated according to Pace and collaborators [55]. Antibacterial assays Bacteriocin production by E. coli cells carrying pUC18-derived constructs was assayed as follows: 2-ml drops of an overnight stationary-phase culture were spotted onto LB agar plates and incubated for 8 h at 37uC. Next, plates were exposed to chloroform vapor (30 min), aerated and subsequently overlaid with 5 ml of soft agar (0.5%), seeded with 200 ml of a cell culture of an indicator strain (,108 CFU/ml), followed by overnight incubation at 30uC. Next day, plates were scored for the presence of halos in the confluently grown overlay. Antibacterial activity assays with purified recombinant His6- tagged proteins were performed as described by Ghequire and collaborators [23]. To assess the influence of added sugar, the same assay was carried out on agar medium supplemented with D- mannose (Sigma-Aldrich) or methyl-a-D-mannopyranoside (Sig- ma-Aldrich) to a final concentration of 0.01 M to 0.5 M. A Bioscreen C apparatus (Oy Growth Curves Ab Ltd, Finland) was used to determine the minimum inhibitory concentration (MIC). An overnight culture (16 h) of the indicator strain was diluted to 104–105 CFU/ml and incubated at 30uC, with a two- fold dilution series of recombinant His6-tagged LlpABW or mutant LlpABW. Bis-TRIS propane buffer was used as control. The MIC value was determined as the minimum concentration of protein at which no growth of the indicator strain (OD600,0.2) occurred after 24 h. At least three independent repeats, each with three replicates, were carried out. Glycan array His6-tagged LlpABW was lyophilized and verified for antibac- terial activity. After re-dissolving in MilliQ water, recombinant LlpABW was diluted to 200 mg/ml with binding buffer (20 mM TRIS-HCl pH 7.4, 150 mM NaCl, 2 mM CaCl2, 2 mM MgCl2, 0.05% Tween 20, 1% BSA), and used to probe the printed glycan arrays [56] following the standard procedures of Core H of the Consortium for Functional Glycomics (http://www. functionalglycomics.org/). Monoclonal anti-His6 antibodies (Roche Diagnostics) were used as primary antibodies, and fluorescently labeled anti-mouse IgG as secondary antibodies. Circular dichroism CD spectra were acquired on a Jasco J-715 spectropolarimeter. Curves were averaged over 8 scans, taken at 25uC using a 1 mm cuvette. Samples were dialyzed against bis-TRIS propane buffer (20 mM, NaCl 200 mM, pH 7.0), filtered and degassed prior to data acquisition. All proteins were assayed at 0.4 mg/ml. Isothermal titration calorimetry ITC titrations were carried out on an ITC200 apparatus (MicroCal). Prior to the measurement, LlpABW, LlpAV177Y, LlpAV208Y and LlpAV177Y-V208Y was dialyzed to bis-TRIS propane buffer. Sugars were directly dissolved into the same buffer. The samples were filtered and degassed for 10 min at 25uC before being examined in the calorimeter. The titrations were carried out at 25uC, injecting the sugars (methyl-a-D- mannoside, Mana(1–2)Man, Mana(1–3)Man, Mana(1–6)Man, Mana(1–3)[Mana(1–6)]Man and GlcNAcb(1–2)Mana(1– 3)[GlcNAcb(1–2)Mana(1–6)]Man) into a protein solution (pro- tein concentrations ranged from 2 mM to 4 mM depending on protein availability). All data were analyzed using the MicroCal Origin ITC 7.0 software. Binding affinities and thermodynamic parameters from all ITC titrations are reported in Table 1. X-ray data collection and structure determination Expression, purification and crystallization of recombinant His- tagged LlpABW have been described [54]. X-ray data for native and derivative crystals were collected on EMBL beamline BW7A of the DESY synchrotron (Hamburg, Germany). For each potential Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 10 February 2013 | Volume 9 | Issue 2 | e1003199 derivative, the wavelength was chosen to be at the high-energy side of the absorption edge in order to ensure a usable anomalous signal. All data were scaled and merged with the HKL package of programs. Data collection statistics are given in Table S5. The crystal structure of free LlpABW was solved combining single isomorphous replacement with anomalous scattering (SIRAS strategy) from a p-chloromercurybenzoate derivative. The heavy-atom substructure was determined with SHELXD [57] using a resolution cutoff of 4.0 A˚ . Heavy-atom refinement and phasing were performed with SHARP [58]. Phase improve- ment by solvent flattening was performed with SOLOMON [59]. Non-crystallographic symmetry averaging with density modifica- tion [60] further improved the electron density. A partial model (94% of the residues comprising the asymmetric unit) was automatically built with ARP/wARP [61] and the remainder was built manually over several cycles of model building with Coot [62], alternated with refinement using phenix.refine [63,64]. Phasing and refinement statistics are shown in Table S5. Carbohydrate soaks Crystals of LlpABW were transferred to artificial mother liquor (0.1 M imidazole pH 6.5, 1.3 M sodium acetate) enriched with either 200 mM methyl-a-D-mannopyranoside (Me-Man), Ma- na(1–2)Man, GlcNAcb(1–2)Mana(1–3)[GlcNAcb(1–2)Mana(1– 6)]Man (M592), D-galactose, L-fucose or N-acetyl-D-glucosamine and allowed to equilibrate overnight (all carbohydrates obtained from Dextra Laboratories, Reading, U.K.). Data were collected at room temperature on EMBL beamline X13, except for the N- acetyl-D-glucosamine soak collected at the PROXIMA-1 beam- line of the SOLEIL synchrotron (Gif-sur-Yvette, France) and the D-galactose soak collected at ESRF beam line ID14-1 (Grenoble, France). All data were scaled and merged using the HKL package. Refinement was started from the coordinates of the ligand-free structure using phenix.refine. Manual rebuilding, including the introduction of the carbohydrate ligand if present, was done with Coot [62]. Crystal structures of LlpABW from P. putida BW11M1 (PDB entry 3M7H) and LlpABW in complex with methyl-a-D- mannoside (PDB entry 3M7J), with Mana(1–2)Man (PDB entry 4GC1), and with GlcNAcb(1–2)Mana(1–3)[GlcNAcb(1–2)Ma- na(1–6)]Man (PDB entry 4GC2) have been deposited at the PDB. Flow cytometry Overnight cultures of P. syringae GR12-2R3 (16 h) were diluted to OD600 0.5 and washed twice with phosphate-buffered saline (PBS). Cells were treated with LlpA, mutant proteins or buffer (bis- TRIS propane buffer, negative control), at a final concentration of 50 mg/ml for 1 h, at 20uC. Next, PBS-washed bacteria were stained using the Live/Dead BacLight bacterial viability kit (Invitrogen), incubated for 15 minutes, and analyzed on a BD Influx (BD Biosciences). Excitation of the dyes was done at 488 nm, and fluorescence measured at 530 nm for SYTO 9 and at 610 nm for propidium iodide. Results were processed with FlowJo 10.0.4 software (Figure S10). Measurements were done indepen- dently and based on six biological repeats. Results are expressed as percentages of dead cells [dead/(live+dead) * 100)]. Supporting Information Figure S1 Amino acid sequence of LlpABW colored according to its domain structure. The N-domain is shown in red, the C-domain in blue and the C-terminal extension in green. Residues belonging to sequences equivalent to the mannose binding site signature motif QxDxNxVxY are in bold and underlined. (JPG) Figure S2 Quaternary structures and domain organi- zation of various MMBL family members. Individual domains or protomers are shown in different colours. The domain or protomer colored green (which in the tandem MMBLs of LlpA, ASA I and SCAfet corresponds to the N-terminal domain) is always shown in the same orientation. Bound carbohydrates are shown in black stick representation. For LlpA a single pentasaccharide is bound to site IIIC. In the case of Galanthus nivalis (snowdrop) lectin (PDB entry 1JPC), twelve trimannosides are bound to all QxDxNxVxY motifs (three on each monomer of the homotetrameric protein). The snowdrop lectin tetramer consists of the association of two domain-swapped dimers (green-blue and pink-yellow). In the case of Allium sativum (garlic lectin ASA I - PDB entry 1KJ1), again each QxDxNxVxY motif has a dimannose bound while an additional sugar (shown in red) is bound to a non-canonical site. The protein is synthesized as a single chain precursor and post-translationally cleaved into two MMBL domains that adopt the same domain-swapped dimer as found in snowdrop lectin. Gastrodianin is a monomeric MMBL family member from the orchid Gastrodia elata (PDB entry 1XD5). The location(s) of its carbohydrate-binding site(s) is (are) not known. The fetuin-binding tandem-MMBL SCAfet from Scilla campanulata (PDB entry 1DLP) consists of two covalently attached MMBL domains, whereas in LlpA the swap of the C- terminal b-strands is retained. The relative orientation in the two domains is as in ASA I. This lectin binds fetuin rather than oligomannosides, but the locations of the binding sites are not known. (JPG) Figure S3 Stereo view of the superpositions (Ca repre- sentations) of the N-domain of LlpABW (red), C-domain of LlpABW (blue) and Galanthus nivalis lectin (PDB entry 1MSA, black). The superposition is shown in two orientations rotated by 90u. (JPG) Figure S4 Sequence alignment of potential mannose- binding motifs in prokaryotic tandem MMBL proteins. The sequences corresponding to the consensus motif QxDxNxVxY, extracted from the N-domain and the C-domain of P. putida LlpABW and its homologues, are aligned per domain. Sequence conservation is visualized by differential shading. The sequence logo representation visualizes the degree of consensus for each residue. LlpA proteins with proven bacteriotoxic activity are labeled with an asterisk. Accession numbers: Arthrobacter sp. FB24 (YP_829274), Burkholderia ambifaria MEX-5 (ZP_02905572), Burk- holderia cenocepacia AU 1054 ([1], ABF75998; [2], ABF75999), Pseudomonas chlororaphis subsp. aureofaciens 30–84 (EJL08681), Pseudomonas putida BW11M1 (AAM95702), Pseudomonas fluorescens Pf-5 (LlpA1 [1], YP_258360; LlpA2 [2], YP_259234), Pseudomonas sp. GM80 ([1], ZP_10606046; [2], ZP_10606131), Pseudomonas syringae pv. aptata DSM 50252 (EGH77666), Pseudomonas syringae pv. syringae 642 (ZP_07263221), Xanthomonas axonopodis pv. citri str. 306 (AAM35756). (TIF) Figure S5 Sequence alignment of the carboxy-terminal sequences of LlpA-like proteins. The P. putida LlpABW sequence adopting a b-hairpin fold is delineated in Figure S1. The preceding conserved tryptophan residue is located C-terminally to IC (Figure S1). The sequence logo representation visualizes the degree of consensus for each residue. Accession numbers are listed in Figure S4. (TIF) Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 11 February 2013 | Volume 9 | Issue 2 | e1003199 Figure S6 Electron density for (A) Methyl-a-D-Man, (B) Mana(1–2)Man and (C) GlcNAcb(1–2)Mana(1– 3)[GlcNAcb(1–2)Mana(1–6)]Man. Difference electron-density maps are calculated by removing the sugar residues from the final coordinates and applying one round of slow-cool simulated annealing refinement to remove potential bias. The atomic model is superimposed in each case. (JPG) Figure S7 Mannose binding to LlpABW and garlic lectin. (A) Cartoon representation of subdomain IIIC of LlpABW (green) with residues implicated in carbohydrate binding showing in ball- and-stick representation and labeled (carbon green, oxygen red, nitrogen blue). The Me-Man residue is shown in red. Selected hydrogen bonds are shown as black dotted lines. (B) Stereoview of the superposition of subdomain IIIC of LlpABW (green) on the equivalent subdomain of garlic lectin (blue). The Me-Man residue bound to LlpABW is shown in red, the mannose bound to garlic lectin in blue. (C) Stereoview of the superposition of subdomain IIIC of LlpABW (green) identical as in panel A, but emphasizing the location of Val177 (shown as black sticks). The modeled Val177Tyr mutation is shown as orange sticks. Tyr177 makes a steric clash with the bound mannose (red) and is therefore expected to prevent binding, in agreement with our ITC experiments. (JPG) Figure S8 Sites II and I of the LlpABW C-domain. (A) Stereoview of site IIC of the C-domain (colored according to atom type) superimposed on site IIIC of the C-domain (dark gray). The Me-Man bound in site IIIC is shown in red. This site is very similar to site IIIC but in the crystal it is inaccessible due to crystal lattice interactions. Residue labels correspond to residues of site IIC. (B) Similar view showing site IC of the C-domain (colored according to atom type) superimposed on site IIIC of the C-domain (dark gray). The Me-Man bound in site IIIC is shown in red. The stretch of Ile137-Leu139 that provides a steric conflict preventing Me-Man binding in site IC, is highlighted with carbon atoms drawn in green. Residue labels correspond to residues of site IC. (JPG) Figure S9 Sites of the LlpABW N-terminal domain. (A) Stereoview of site IN of the N-domain (colored according to atom type) superimposed on site IIIC of the C-domain (dark gray). The Me-Man bound in site IIIC is shown in red. The stretch of Ile271- Trp274 that provides a steric conflict preventing Me-Man binding in site IN is highlighted with carbon atoms drawn in green. Residue numbering corresponds to residues of site IN. (B) Similar superposition for site IIN of the N-domain. Phe86 that prevents Me-Man binding to this site through a steric conflict is highlighted in green. Other residues belonging to site IIN are labeled in teal. Three residues of site IIIC for which site IIN has no structural equivalent are labeled in black. (C) Similar superposition for site IIIN of the N-domain. Residues belonging to site IIIN are labeled in teal. One residue of site IIIC for which site IIIN has no structural equivalent, is labeled in black. For this site there are no obvious steric conflicts that would prevent positioning of a Me-Man residue although none is observed experimentally. (JPG) Figure S10 Quantification of live and dead cells by flow cytometry. P. syringae GR12-2R3 cells were treated with LlpA (A), LlpAV177Y (B), LlpAV208Y (C), LlpAV177Y-V208Y (D), or buffer (E) at a final concentration of 50 mg/ml for 1 h at 20uC. After live/dead staining, cells were analysed by flow cytometry. Data processing allowed to distinguish populations of dead (left) and live (right) cells. Spot densities ranging from high to low are differentiated by a color gradient from red, yellow, green, teal to blue. Representative samples for LlpA, mutant proteins and buffer control are shown in panels A–E. (TIF) Figure S11 Overview of inactive LlpABW deletants and inactive LlpABW/LlpA1 chimers. The equivalent domains of LlpA1 are delineated based on pairwise sequence alignment with LlpABW: N-domain (orange), C-domain (purple), C-terminal extension (grey). No bacteriocin activity was conferred by these constructs upon recombinant E. coli cells tested against P. syringae GR12-2R3 (indicator strain for native LlpABW) and P. fluorescens LMG 1794 (indicator strain for native LlpA1). The small black rectangle represents an artificial linker sequence (DASRS). (TIF) Table S1 Glycan array profile of LlpABW as measured by fluorescence intensity. Results including a comprehensive list of oligosaccharides (array version PA_v5) are available from the Consortium of Functional Glycomics (CFG, www. functionalglycomics.org). (XLS) Table S2 Bacterial strains and plasmids used in this study. (DOC) Table S3 PCR primers used in this study. (DOCX) Table S4 Protein sequences of LlpABW deletants and LlpABW/LlpA1 chimers. (DOCX) Table S5 Structure determination and refinement. (DOCX) Acknowledgments The authors acknowledge the Consortium for Functional Glycomics for performing glycan array tests on LlpABW, and the use of the macromolecular crystallography beamlines at EMBL/DESY (Hamburg, Germany), ESRF (Grenoble, France) and SOLEIL (Gif-sur-Yvette, France) for X-ray data collection. Author Contributions Conceived and designed the experiments: MGKG AGP RL RDM. Performed the experiments: MGKG AGP EKML SS. Analyzed the data: MGKG AGP EKML SS RL RDM. Contributed reagents/materials/ analysis tools: MGKG AGP EKML SS RL RDM. Wrote the paper: MGKG AGP RL RDM. References 1. Frey-Klett P, Burlinson P, Deveau A, Barret M, Tarkka M, et al. (2011) Bacterial-fungal interactions: hyphens between agricultural, clinical, environ- mental, and food microbiologists. Microbiol Mol Biol Rev 75: 583–609. 2. Mela F, Fritsche K, de Boer W, van Veen JA, de Graaff LH, et al. (2011) Dual transcriptional profiling of a bacterial/fungal confrontation: Collimonas fungivorans versus Aspergillus niger. ISME J 5: 1494–1504. Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 12 February 2013 | Volume 9 | Issue 2 | e1003199 3. Garbeva P, Silby MW, Raaijmakers JM, Levy SB, de Boer W (2011) Transcriptional and antagonistic responses of Pseudomonas fluorescens Pf0-1 to phylogenetically different bacterial competitors. ISME J 5: 973–985. 4. Shank EA, Klepac-Ceraj V, Collado-Torres L, Powers GE, Losick R, et al. (2011) Interspecies interactions that result in Bacillus subtilis forming biofilms are mediated mainly by members of its own genus. Proc Natl Acad Sci U S A 108: E1236–E1243. 5. Rosenthal AZ, Matson EG, Eldar A, Leadbetter JR (2011) RNA-seq reveals cooperative metabolic interactions between two termite-gut spirochete species in co-culture. ISME J 5: 1133–1142. 6. Hibbing ME, Fuqua C, Parsek MR, Peterson SB (2010) Bacterial competition: surviving and thriving in the microbial jungle. Nat Rev Microbiol 8: 15–25. 7. Aoki SK, Diner EJ, de Roodenbeke CT, Burgess BR, Poole SJ, et al. (2010) A widespread family of polymorphic contact-dependent toxin delivery systems in bacteria. Nature 468: 439–442. 8. Poole SJ, Diner EJ, Aoki SK, Braaten BA, t’Kint de Roodenbeke C, et al. (2011) Identification of functional toxin/immunity genes linked to contact-dependent growth inhibition (CDI) and rearrangement hotspot (Rhs) systems. PLoS Genet 7: e1002217. 9. Russell AB, Hood RD, Bui NK, LeRoux M, Vollmer W, et al. (2011) Type VI secretion delivers bacteriolytic effectors to target cells. Nature 475: 343–347. 10. Michel-Briand Y, Baysse C (2002) The pyocins of Pseudomonas aeruginosa. Biochimie 84: 499–510. 11. Nakayama K, Takashima K, Ishihara H, Shinomiya T, Kageyama M, et al. (2000) The R-type pyocin of Pseudomonas aeruginosa is related to P2 phage, and the F-type is related to lambda phage. Mol Microbiol 38: 213–231. 12. Williams SR, Gebhart D, Martin DW, Scholl D (2008) Retargeting R-type pyocins to generate novel bactericidal protein complexes. Appl Environ Microbiol 74: 3868–3876. 13. Fischer S, Godino A, Quesada JM, Cordero P, Jofre´ E, et al. (2012) Characterization of a phage-like pyocin from the plant growth-promoting rhizobacterium Pseudomonas fluorescens SF4c. Microbiology 158: 1493–1503. 14. Ko¨hler T, Donner V, van Delden C (2010) Lipopolysaccharide as shield and receptor for R-pyocin-mediated killing in Pseudomonas aeruginosa. J Bacteriol 192: 1921–1928 15. Cascales E, Buchanan SK, Duche´ D, Kleanthous C, Lloube`s R, et al. (2007) Colicin biology. Microbiol Mol Biol Rev 71: 158–229. 16. Denayer S, Matthijs S, Cornelis P (2007) Pyocin S2 (Sa) kills Pseudomonas aeruginosa strains via the FpvA type I ferripyoverdine receptor. J Bacteriol 189: 7663–7668. 17. Elfarash A, Wei Q, Cornelis P (2012) The soluble pyocins S2 and S4 from Pseudomonas aeruginosa bind to the same FpvAI receptor. Microbiologyopen 1: 268–275. 18. Ling H, Saeidi N, Rasouliha BH, Chang MW (2010) A predicted S-type pyocin shows a bactericidal activity against clinical Pseudomonas aeruginosa isolates through membrane damage. FEBS Lett 584: 3354–3358. 19. Barreteau H, Tiouajni M, Graille M, Josseaume N, Bouhss A, et al. (2012) Functional and structural characterization of PaeM, a colicin M-like bacteriocin produced by Pseudomonas aeruginosa. J Biol Chem 287: 37395–37405. 20. Grinter R, Roszak AW, Cogdell RJ, Milner JJ, Walker D (2012) The crystal structure of the lipid II-degrading bacteriocin syringacin M suggests unexpected evolutionary relationships between colicin M-like bacteriocins. J Biol Chem 287: 38876–38888. 21. Parret AHA, Schoofs G, Proost P, De Mot R (2003) Plant lectin-like bacteriocin from a rhizosphere-colonizing Pseudomonas isolate. J Bacteriol 185: 897–908. 22. Parret AHA, Temmerman K, De Mot R (2005) Novel lectin-like bacteriocins of biocontrol strain Pseudomonas fluorescens Pf-5. Appl Environ Microbiol 71: 5197– 5207. 23. Ghequire MGK, Li W, Proost P, Loris R, De Mot R (2012) Plant lectin-like antibacterial proteins from phytopathogens Pseudomonas syringae and Xanthomonas citri. Environ Microbiol Rep 4: 373–380. 24. Van Damme EJM, Lannoo N, Peumans WJ (2008) Plant lectins. Adv Bot Res 48: 107–209. 25. Ghequire MGK, Loris R, De Mot R (2012) MMBL proteins: from lectin to bacteriocin. Biochem Soc Trans 40: 1553–1559. 26. Van Damme EJM, Kaku H, Perini F, Goldstein IJ, Peeters B, et al. (1991) Biosynthesis, primary structure and molecular cloning of snowdrop (Galanthus nivalis L.) lectin. Eur J Biochem 202: 23–30. 27. Hester G, Wright CS (1996) The mannose-specific bulb lectin from Galanthus nivalis (snowdrop) binds mono- and dimannosides at distinct sites. Structure analysis of refined complexes at 2.3 A˚ and 3.0 A˚ resolution. J Mol Biol 262: 516– 531. 28. Fouquaert E, Peumans WJ, Gheysen G, Van Damme EJM (2011) Identical homologs of the Galanthus nivalis agglutinin in Zea mays and Fusarium verticillioides. Plant Physiol Biochem 49: 46–54. 29. Shimokawa M, Fukudome A, Yamashita R, Minami Y, Yagi F, et al. (2012) Characterization and cloning of GNA-like lectin from the mushroom Marasmius oreades. Glycoconj J 29: 457–465. 30. Jung E, Fucini P, Stewart M, Noegel AA, Schleicher M (1996) Linking microfilaments to intracellular membranes: the actin-binding and vesicle- associated protein comitin exhibits a mannose-specific lectin activity. EMBO J 15: 1238–1246. 31. Wiens M, Belikov SI, Kaluzhnaya OV, Krasko A, Schro¨der HC, et al. (2006) Molecular control of serial module formation along the apical-basal axis in the sponge Lubomirskia baicalensis: silicateins, mannose-binding lectin and mago nashi. Dev Genes Evol 216: 229–242. 32. Tsutsui S, Tasumi S, Suetake H, Suzuki Y (2003) Lectins homologous to those of monocotyledonous plants in the skin mucus and intestine of pufferfish, Fugu rubripes. J Biol Chem 278: 20882–20889. 33. de Santana Evangelista K, Andrich F, Figueiredo de Rezende F, Niland S, Cordeiro MN, et al. (2009) Plumieribetin, a fish lectin homologous to mannose- binding B-type lectins, inhibits the collagen-binding a1b1 integrin. J Biol Chem 284: 34747–34759. 34. Rajan B, Fernandes JM, Caipang CM, Kiron V, Rombout JH, et al. (2011) Proteome reference map of the skin mucus of Atlantic cod (Gadus morhua) revealing immune competent molecules. Fish Shellfish Immunol 31: 224–231. 35. Chen J, Stevenson DM, Weimer PJ (2004) Albusin B, a bacteriocin from the ruminal bacterium Ruminococcus albus 7 that inhibits growth of Ruminococcus flavefaciens. Appl Environ Microbiol 70: 3167–3170. 36. Wright CS, Hester G (1996) The 2.0 A˚ structure of a cross-linked complex between snowdrop lectin and a branched mannopentaose: evidence for two unique binding modes. Structure 4: 1339–1352. 37. Wright LM, Reynolds CD, Rizkallah PJ, Allen AK, Van Damme EJM, et al. (2000) Structural characterisation of the native fetuin-binding protein Scilla campanulata agglutinin: a novel two-domain lectin. FEBS Lett 468: 19–22. 38. Ramachandraiah G, Chandra NR, Surolia A, Vijayan M (2002) Re-refinement using reprocessed data to improve the quality of the structure: a case study involving garlic lectin. Acta Crystallogr D Biol Crystallogr 58: 414–420. 39. Robert V, Volokhina EB, Senf F, Bos MP, Van Gelder P, et al. (2006) Assembly factor Omp85 recognizes its outer membrane protein substrates by a species- specific C-terminal motif. PLoS Biol 4: e377. 40. Wang X, Bauw G, Van Damme EJM, Peumans WJ, Chen ZL, et al. (2001) Gastrodianin-like mannose-binding proteins: a novel class of plant proteins with antifungal properties. Plant J 25: 651–661. 41. Balzarini J (2007) Targeting the glycans of glycoproteins: a novel paradigm for antiviral therapy. Nat Rev Microbiol 5: 583–597. 42. Tian Q, Wang W, Miao C, Peng H, Liu B, et al. (2008) Purification, characterization and molecular cloning of a novel mannose-binding lectin from rhizomes of Ophiopogon japonicus with antiviral and antifungal activities. Plant Sci 175: 877–884. 43. Bharathi Y, Vijaya Kumar S, Pasalu IC, Balachandran SM, Reddy VD, et al. (2011) Pyramided rice lines harbouring Allium sativum (asaI) and Galanthus nivalis (gna) lectin genes impart enhanced resistance against major sap-sucking pests. J Biotechnol 152: 63–71. 44. Hoorelbeke B, Van Damme EJM, Rouge´ P, Schols D, Van Laethem K, et al. (2011) Differences in the mannose oligomer specificities of the closely related lectins from Galanthus nivalis and Zea mays strongly determine their eventual anti- HIV activity. Retrovirology 8: 10. 45. Vandenborre G, Smagghe G, Van Damme EJM (2011) Plant lectins as defense proteins against phytophagous insects. Phytochemistry 72: 1538–1550. 46. Yang Y, Xu HL, Zhang ZT, Liu JJ, Li WW, et al. (2011) Characterization, molecular cloning, and in silico analysis of a novel mannose-binding lectin from Polygonatum odoratum (Mill.) with anti-HSV-II and apoptosis-inducing activities. Phytomedicine 18: 748–755. 47. Fu LL, Zhou CC, Yao S, Yu JY, Liu B, et al. (2011) Plant lectins: targeting programmed cell death pathways as antitumor agents. Int J Biochem Cell Biol 43: 1442–1449. 48. Lehotzky RE, Partch CL, Mukherjee S, Cash HL, Goldman WE, et al. (2010) Molecular basis for peptidoglycan recognition by a bactericidal lectin. Proc Natl Acad Sci U S A 107: 7722–7727. 49. Miki T, Holst O, Hardt WD (2012) The bactericidal activity of the C-type lectin RegIIIb against Gram-negative bacteria involves binding to Lipid A. J Biol Chem 287: 34844–34855. 50. Tielker D, Hacker S, Loris R, Strathmann M, Wingender J, et al. (2005) Pseudomonas aeruginosa lectin LecB is located in the outer membrane and is involved in biofilm formation. Microbiology 151: 1313–1323. 51. Bartels KM, Funken H, Knapp A, Brocker M, Wilhelm S, et al. (2011) Glycosylation is required for outer membrane localization of the lectin LecB in Pseudomonas aeruginosa. J Bacteriol 193: 1107–1113. 52. Funken H, Bartels KM, Wilhelm S, Brocker M, Bott M, et al. (2012) Specific association of lectin LecB with the surface of Pseudomonas aeruginosa: role of outer membrane protein OprF. PLoS One 7: e46857. 53. Green MR, Sambrook JR (2012) Molecular cloning: a laboratory manual. 4th edition. New York: Cold Spring Harbor Laboratory Press. 2028 p. 54. Parret AHA, Wyns L, De Mot R, Loris R (2004) Overexpression, purification and crystallization of bacteriocin LlpA from Pseudomonas sp. BW11M1. Acta Crystallogr D Biol Crystallogr 60: 1922–1924. 55. Pace CN, Vajdos F, Fee L, Grimsley G, Gray T (1995) How to measure and predict the molar absorption coefficient of a protein. Protein Sci 4: 2411–2423. 56. Blixt O, Head S, Mondala T, Scanlan C, Huflejt ME, et al. (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc Natl Acad Sci U S A 101: 17033–17038. 57. Schneider TR, Sheldrick GM (2002) Substructure solution with SHELXD. Acta Crystallogr D Biol Crystallogr 58: 1772–1779. 58. de la Fortelle E, Bricogne G (1997) Maximum likelihood heavy-atom parameter refinement for multiple isomorphous replacement and multiwavelength anomalous diffraction methods. In: Carter CW, Sweet RM, editors. Methods Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 13 February 2013 | Volume 9 | Issue 2 | e1003199 in Enzymology. Volume 276, Macromolecular Crystallography Part A. New York: Academic Press. 472–494. 59. Abrahams JP, Leslie AG (1996) Methods used in the structure determination of bovine mitochondrial F1 ATPase. Acta Crystallogr D Biol Crystallogr 52: 30– 42. 60. Cowtan K. (2010) Recent developments in classical density modification. Acta Crystallogr D Biol Crystallogr 66: 470–478. 61. Perrakis A, Morris R, Lamzin VS (1999) Automated protein model building combined with iterative structure refinement. Nat Struct Biol 6: 458–463. 62. Emsley P, Cowtan K (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132. 63. Adams PD, Gopal K, Grosse-Kunstleve RW, Hung LW, Ioerger TR, et al. (2004) Recent developments in the PHENIX software for automated crystallographic structure determination. J Synchrotron Radiat 11: 53–55. 64. Afonine PV, Grosse-Kunstleve RW, Adams PD (2005) A robust bulk-solvent correction and anisotropic scaling procedure. Acta Crystallogr D Biol Crystal- logr 61: 850–855. Bactericidal Lectin Structure PLOS Pathogens | www.plospathogens.org 14 February 2013 | Volume 9 | Issue 2 | e1003199
3M7K
Crystal structure of PacI-DNA Enzyme product complex
Unusual target site disruption by the rare-cutting HNH restriction endonuclease PacI Betty Shen1, Daniel F. Heiter2, Siu-Hong Chan2, Hua Wang2, Shuang-Yong Xu2, Richard D. Morgan2, Geoffrey G. Wilson2, and Barry L. Stoddard1,3 1Division of Basic Sciences, Fred Hutchinson Cancer Research Center, 1100 Fairview Ave. N. A3-025, Seattle, WA 98109, USA 2New England Biolabs, Inc., 240 County Road, Ipswich, MA 01938, USA Abstract The crystal structure of the rare-cutting HNH restriction endonuclease PacI in complex with its eight base pair target recognition sequence 5'-TTAATTAA-3' has been determined to 1.9 Å resolution. The enzyme forms an extended homodimer, with each subunit containing two zinc-bound motifs surrounding a ββα-metal catalytic site. The latter is unusual in that a tyrosine residue likely initiates strand-cleavage. PacI dramatically distorts its target sequence from Watson-Crick duplex DNA basepairing, with every base separated from its original partner. Two bases on each strand are unpaired, four are engaged in non-canonical A:A and T:T base pairs, and the remaining two bases are matched with new Watson-Crick partners. This represents a highly unusual DNA binding mechanism for a restriction endonuclease, and implies that initial recognition of the target site might involve significantly different contacts from those visualized in the DNA-bound cocrystal structures. Restriction endonucleases (REases) occur in all free-living bacteria and archaea and are believed to function to defend their hosts against invasion by foreign DNA, particularly from bacteriophage (Pingoud et al., 2005). REases vary in sequence, structure, oligomeric composition, substrate-specificity, and enzymatic behavior (Bujnicki, 2003). They range from compact monomers that act independently, to elaborate multifunctional protein assemblages, and typically recognize target sequences in duplex DNA ranging from four to eight specific base pairs in length. (Pingoud et al., 2005). These sequences can be symmetric or asymmetric, as well as continuous or discontinuous, depending upon the enzyme architecture. Several distinct catalytic site motifs and mechanisms have been identified among restriction endonucleases, suggesting this enzymatic and biological function has evolved independently several times. The most common catalytic motif, that of the 'PD…(D/E)xK' nuclease superfamily, is the the most wide-spread and best understood (Kosinski et al., 2005). Alternative catalytic motifs, associated with quite different core protein folds, have been identified in many additional restriction endonucleases, including the ‘HNH’ (Cymerman et al., 2006; Jakubauskas et al., 2007; Saravanan et al., 2004) and the ‘GIY-YIG’ (Ibryashkina et al., 2007) motifs (both of which are more commonly associated with mobile homing endonucleases from bacteriophage) (Stoddard, 2005). All three of these catalytic lineages are © 2010 Elsevier Inc. All rights reserved. 3Corresponding author, bstoddar@fhcrc.org 1-206-667-4031 (office) -4066 (lab) -3331 (fax). Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. NIH Public Access Author Manuscript Structure. Author manuscript; available in PMC 2011 June 9. Published in final edited form as: Structure. 2010 June 9; 18(6): 734–743. doi:10.1016/j.str.2010.03.009. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript also found in a much wider variety of enzymes involved in DNA metabolism and modification, including those responsible for DNA repair, recombination and fidelity (Cymerman et al., 2006; Dunin-Horkawicz et al., 2006; Kosinski et al., 2005). As well, isolated examples of two additional structural motifs (containing the phospholipase D and 'half-pipe' folds) have also observed for R.BfiI and R.PabI, respectively (Grazulis et al., 2005; Miyazono et al., 2007). The conserved structural core surrounding the HNH motif is termed the 'ββα-metal' fold. This protein topology consists of two anti-parallel β-strands connected by a loop of variable length, flanked by an α-helix (Mehta et al., 2004),(Kuhlmann et al., 1999). A binding site for a single catalytic metal ion—typically magnesium—is embedded within this catalytic fold. In some instances, significant insertions of additional structural elements are observed within this motif (Eastberg et al., 2007; Stoddard, 2005). The ββα-metal fold can exist as an independently folded catalytic domain (as observed in colicins) or it can be fused to additional protein domains that dictate DNA binding specificity and cleavage activity. The PD…(D/E)×K motif can be very well-suited for recognition of short DNA sequences with high fidelity, because the catalytic center is surrounded by a densely packed array of side chains that can contact neighboring base pairs in the major groove in a sequence-specific manner (Orlowski and Bujnicki, 2008). In contrast, the HNH motif and its associated ββα-metal fold appears less well-suited for this task. In order to target the scissile phosphate, the catalytic core motifs of these enzymes primarily interact with the DNA backbone where they contribute little to sequence-specificity and fidelity (Eastberg et al., 2007). Sequence-recognition by these enzymes is therefore usually carried out by additional protein domains that are tethered to the ββα-metal region, necessitating significant repackaging and augmentation of this catalytic motif. Recently, the structure of the ββα-metal restriction endonuclease Hpy99I was determined in complex with its DNA substrate, 5' - CGWCG - 3', at 1.5 Å resolution (Sokolowska et al., 2009) (W=A or T). Hpy99I binds as a homodimer and forms a ring-like structure that encircles the DNA. The protein contacts all four C:G base pairs within both the minor and major groove, and contacts the central base pair (A:T or T:A) in only the minor groove. The DNA is slightly bent in the complex. All nucleotides in the target site are found in canonical Watson-Crick basepair interactions. In contrast, PacI is a 'rare-cutting' homodimeric HNH restriction endonuclease found in the bacterium Pseudomonas alcaligenes. It recognizes the symmetric eight base pair duplex DNA sequence 5' – TTAAT/TAA - 3' and cleaves each strand between the internal thymine residues (as the position indicated by "/") to generate product fragments containing 2-base, 3’-overhangs (Roberts et al., 2010). PacI is one of the smallest REases known, comprising only 142 amino acids per subunit, eight of which are cysteines (Figure 1). Its gene resides within a super- integron, a chromosomal array that contains multiple gene cassettes each flanked by a large direct repeat sequence and mobilized by a common site-specific integrase (Vaisvila et al., 2001). Unlike the vast majority of REases, which are accompanied by DNA-methyltransferases that protect the cell’s own DNA from REase auto-digestion, PacI appears to be a solitary enzyme with no companion methyltransferase (see Supplementary Material). Host protection in this rare instance seems likely to depend not on the methylation of recognition sequences, but rather on the absence of such sequences in the P.alcaligenes genome. The length of the PacI recognition sequence (eight basepairs) places this enzyme in the company of NotI and SfiI, two other 'rare-cutting' endonucleases the co-crystal structures of which have been solved (Qiang and Schildkraut, 1987). NotI and SfiI belong to the PD…(D/ E)×K catalytic site superfamily, and in contrast to PacI, recognize sequences composed entirely of G:C base pairs. Shen et al. Page 2 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Bioinformatics analysis of PacI (Orlowski and Bujnicki, 2008), and independent analyses with online protein fold prediction servers such as PHYRE (Bennett-Lovsey et al., 2008) suggest the presence of an HNH-related catalytic site. The likely presence of this motif, combined with the opportunity to compare a ‘rare A:T-cutter’ to two ‘rare G:C-cutters’, led us to determine the structure of PacI bound to DNA. PacI displays little resemblance to either NotI or SfiI, and while it contains structural elements similar to those in Hpy99I, the arrangements of these elements and the overall folds of the two proteins are strikingly different. PacI binding induces an unusual distortion of its DNA target sequence that completely disrupts and reorganizes it normal Watson-Crick duplex structure. Results Overall protein structure and catalytic site The structure of PacI was determined in complex with its eight base pair cognate target site, within the context of an 18 base pair synthetic DNA duplex. The structure was determined both in the presence of calcium (yielding a co-crystal structure containing uncleaved DNA that extended to 2.0 Å resolution) and in the presence of magnesium (resulting in a bound product complex that was visualized at 1.9 Å resolution). The two structures are virtually identical, with the exception of the presence of free 5' phosphate and 3' hydroxyl DNA product ends in the endonuclease catalytic sites in the presence of magnesium. Data collection and refinement statistics are provided in Table 1, and a detailed description of materials and methods is provided in Supplementary Information. Examples of the experimental electron density, calculated using phases derived by a combination of the multiple isomorphous replacement (MIR) and single anomalous dispersion (SAD) methods, are shown in Supplementary Figure S1. The overall structure of the endonuclease homodimer bound to its DNA target is shown in Figure 1a; two separate views of a single enzyme subunit are shown in Figure 1b. The overall core topology of the PacI subunit corresponds to "β1–β2–α2–α3–β4–α4–α5", with the β3–β4– α4 secondary structure elements comprising the ββα-metal catalytic site motif. This core topology is further extended by very short β-hairpin motifs on the protein surface that are involved in DNA contacts. The PacI subunits display an extended structure containing a pair of bound zinc ions, each of which is coordinated by four cysteine residues. The first zinc ion is entirely sequestered within an N-terminal region (containing cysteines 4, 7, 24 and 27) that appears to be a unique feature of PacI: the only three homologues of PacI currently in Genbank (all from strains of the bacterium Campylobacter) display little sequence similarity to this region (Figure 1c). The second zinc ion is buried in the enzyme core, and is also coordinated by four cysteine residues (Cys 63, 66, 109 and 112). This zinc ion is located near the endonuclease catalytic site. Two of the cysteine residues involved in its coordination (Cys 109 and 112) extend from the α4 helix from the ββα-metal motif. The overall structural organization of the PacI enzyme resembles, at a superficial level, the organization of the homodimeric HNH restriction endonuclease Hpy99I (Sokolowska et al., 2009), and more distantly resembles the homodimeric HNH homing endonuclease I-PpoI (Flick et al., 1998). All three proteins contain a catalytic ββα-metal motif and contain two structural zinc ions embedded within each protein subunit, and all three position their active sites across the minor groove to produce 3' overhangs. However, the extended architecture and DNA binding modes of these enzymes are very different from one another (Figure 2 and supplementary Figure S2), indicating that they appear to have independently acquired and then optimized similar structural strategies for stabilization and catalysis, presumably after their divergence from a common ancestral endonuclease. Shen et al. Page 3 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The backbone conformation and metal coordination exhibited by the 'ββα-metal catalytic core of PacI is similar to those observed in other HNH endonucleases (Kuhlmann et al., 1999) (Figure 3a and Supplementary Figures S2 and S3). Structure-based alignments of this region with five separate ββα-metal endonucleases (E9 colicin, endonuclease VII, Hpy99I, I-HmuI and I-PpoI) gives RMSD values for all backbone atoms of 1.5 to 1.8 Å, with corresponding sequence identities ranging from as low at 6.7% (I-PpoI) to as high as 24% (I-HmuI and EndoVII). A single divalent cation is bound within the PacI catalytic motif, where it is coordinated by aspartate 92 and by asparagine 113, and also interacts with the 3' oxygen and a nonbridging oxygen of the scissile phosphate. The distance from the metal to each DNA atom is approximately 2.5 Å. In spite of the structural similarity of the ββα-metal motif described above, the PacI catalytic site displays a significant departure from the typical HNH motif. The position within the ββα- metal motif that is normally the site of a histidine general base that activates the water nucleophile (corresponding to His 149 in Hpy99I and His 98 in I-PpoI; Figure 3b and 3c) is instead occupied by an arginine residue (Arg 93), that interacts with the 3' leaving group . In place of the usual histidine, a neighboring tyrosine residue (Tyr 100) is instead positioned to either assist in activation of a nucleophilic water (which is not observed), or perhaps to act directly as a nucleophile itself. The distance from the tyrosine hydroxyl group to the phosphorus atom is 4.5 angstroms in the uncleaved calcium-bound complex, and is 3.3 angstroms in the cleaved magnesium-bound complex (the distance is reduced in the cleaved complex due to rotation of the 5' phosphate group after cleavage). The nearest histidine residue (His 42) is located approximately over 8 Å from the scissile phosphate, and like Arg 93, is located closer to the leaving group than to the site of the nucleophilic attack. Thus, the PacI endonuclease displays a dramatic alteration and rearrangement of the usual side chains found in a ββα-metal catalytic motif, and perhaps a change in the actual cleavage mechanism. To assess the relevance of Tyr 100, Arg 93 and His 42 in the PacI catalytic site, each was mutated by PCR and the mutant proteins were expressed in vitro and assayed for activity (Table 2). To the best of our ability to measure, the amounts of each protein construct generated in vitro, and then used in individual digest experiments, were comparable. Y100F was found to be inactive (less than 10−4 WT activity) indicating that the phenolic oxygen of this amino acid appears to be essential for catalysis . R93A and M were also inactive, but R93K displayed reduced activity, suggesting that a positively charged group in this position in the catalytic site is also essential. H42A displayed reduced activity (~10−2 WT activity). The putative metal- binding residues Asp 92 and Asn 113 were also mutated and assayed. D92L and N113L were inactive, whereas the D92A and N113A mutants displayed a low level of activity, indicating that the metal-binding residues are critical components of the catalytic site. The two bound zinc ions in PacI (each coordinated in a Cys4-Zn tetrahedral cluster) represent a widely distributed conserved structural motif, distinct from the conventional trinucleotide- specific Cys2-His2 ‘Zinc-finger’ domains of eukaryotic transcription factors (Supplementary Figure S3). The Cys4-Zn motif comprises a pair of CxxC sequences. The first two cysteines flank a loop, while the second two initiate an alpha helix (in some related sequences, the third Cys or the fourth Cys is replaced by His instead). The region between each CxxC pair varies in length and function, as does the helix. In many instances, this region includes catalytic residues that contribute to the HNH catalytic site. Approximately 200 HNH-like domains are aligned in pfam01844, and over one-third of these are embedded in Cys4-Zn motifs, indicating that this structural architecture is often associated with an HNH catalytic site. In contrast, this same region in the GATA family of transcription factors includes residues responsible for DNA sequence recognition (Bates et al., 2008) (Supplementary Figure S3, panel L). Shen et al. Page 4 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript DNA binding and recognition The mode of DNA binding displayed by PacI is very unusual. The ββα-metal catalytic sites from each protein subunit straddle the minor groove at the center of the DNA target, resulting in an overall bend angle of approximately 90 degrees (Figures 1a and 4a). This results in a dramatic widening of the minor groove (to approximately 18 Å) and a corresponding reduction in the width of the opposing major groove. This bend is accompanied by a radical alteration of the DNA duplex: every base throughout the target site is unpaired from its original Watson- Crick partner (Figure 4b). Within the eight base pair target site, two bases on each strand are completely unpaired, four are engaged in non-canonical A:A and T:T base pairs, and the remaining two bases are matched with new Watson-Crick partners. This disruption of the DNA duplex is entirely localized to the PacI target site; the base pairs immediately outside the 5' - TTAATTAA - 3' sequence still display canonical B-form interactions. It does not appear that crystal contacts play a role in these features of the protein-DNA complex: the solvent content of the crystals is not unusual (about 55%) and the regions of the protein and DNA involved in contacts and recognition are not located near symmetry mates in the crystal lattice. The bound conformation of the DNA target was analyzed using the online program 3DNA (Zheng et al., 2009) (Supplementary Figure S4). The perturbation of the DNA structure results from significant distortion of the individual ribose moieties and the corresponding glycosidic bonds between sugar C1' carbons and the corresponding nucleotide bases. Only three ribose sugars on each strand (corresponding to −4T, −2A and +3A) are found in their original C2'- endo pucker, while the remaining sugars are predominantly flipped into a C1'-exo conformation. The chi angles linking the ribose C1' carbons to the N1 nitrogen of the thymines, or to the N9 nitrogen of the adenines, deviate from their nominal B-form values by as much as +/− 40°, leading to a rotation of individual bases that allows non-canonical A:A or T:T base pairing. These base pairs still exhibit two intra-strand hydrogen bonds (Figure 4c), linking the thymine-thymine pairs via the O2-N3 and N3-O4 atoms of their pyrimidine rings, and the adenine-adenine pairs via the N6-N1 and N7-N6 atoms of their purine rings. These base pair interactions, while rarely observed in DNA duplexes, are often found in folded RNA structures (Olson et al., 2009). The deformation of the DNA in PacI is accompanied by a significant unwinding of the duplex at base step −2A in each DNA half-site (Supplementary Figure S4). That base, which is engaged in an A:A base pair with its −1A partner, exhibits a −40° tilt and unstacking from its neighboring (symmetry-related) A:A base pair. The local unwinding of each DNA half-site at these A:A base pairs is complemented by local over winding of the adjacent base steps and base pairs, allowing the rearranged DNA to maintain an overall duplex architecture. Although the DNA backbone and its base pairing interactions exhibits a dramatic rearrangement in the bound protein complex, all the individual base pairs (both Watson-Crick and non-Watson-Crick) exhibit near normal values of propeller twist and buckle angles. The PacI-DNA complex is further notable for the paucity of direct contacts between the protein and the nucleotides. The two unpaired bases in each half site (+1T and +4A) are in direct contact with amino acid side chains: +4A interacts in the major groove with Asn 32, and +1T interacts in the minor groove with Arg 114 (Figures 4b and 4c). The adenines in the reorganized A:T base pairs ( involving +3A in each DNA half-site) interact in the major groove with Asn 36. One adenine in each A:A base pair makes a nonspecific contact to Ser 117 in the minor groove, and the O4 groups of both thymines in each T:T base pair contact Lys 39 in the major groove. To assess the importance of the major groove contacts, Asn 32, Asn 36, and K 39 were changed by PCR to various other amino acids and the mutant proteins expressed in vitro and assayed (Supplementary Table S1). Mutation of Asn 32 or Asn36 abolished activity (<10−4 WT activity) indicating that these two amino acids are essential. Mutation of Lys 39 had little effect Shen et al. Page 5 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript indicating that this amino acid is unimportant in spite of the hydrogen bonds it forms with the T:T base pairs. Thus, across the eight nucleotides in each DNA half-site, PacI makes only eight direct hydrogen bond contacts: six in the major groove (N32 and N36 to adenine bases and K39 to thymine bases) and two more in the minor groove. This represents a radical departure from the usual strategy of restriction endonucleases which, under strong selective pressure for absolute cleavage fidelity, usually make more direct contacts than are strictly necessary for high fidelity sequence recognition. Discussion Diversity of site-specific HNH endonuclease scaffolds The overall organization of PacI is superficially similar to the HNH restriction endonuclease Hpy99I (Sokolowska et al., 2009), and more distantly related to the HNH (His-Cys box) homing endonuclease I-PpoI (Flick et al., 1998). All are homodimers containing one ββα-metal motif and two bound zinc ions per subunit. However, close examination of these three enzymes, which recognize target sites ranging from 5 base pairs to 14 base pairs in length, indicates that their folded structures, as well as their DNA binding modes and recognition mechanisms, differ significantly (Figure 2). Whereas the core of the Hpy99I protein forms a structure that encircles and binds almost orthogonally across and around its target site (with the helices from the catalytic site ββα-metal motif aligned almost perpendicular with the DNA duplex axis), PacI displays an elongated fold that associates with one face of the DNA target, with the two subunits and the ββα-metal motif aligned nearly parallel to the DNA duplex. The structure of the I-PpoI homing endonuclease is even more divergent: that protein relies upon extended β-sheet structures for the completion of the core protein fold and for formation of its DNA binding surface. Based on these observations, it seems likely that these site-specific HNH endonucleases are distantly related, but probably all descended from a common ββα-metal ancestor. That predecessor protein may have consisted of a nonspecific endonuclease folded around the common catalytic motif (perhaps resembling modern colicin nucleases). The details of the active site organization of PacI also indicate a significant divergence from the usual architecture and mechanism that is observed for an HNH active site (Figure 3). The presence of a tyrosine side chain at the position usually occupied by an imidazole base and nucleophilic water, combined with the requirement of the tyrosine phenolic oxygen for catalysis, indicates that this side chain might act as a direct nucleophile in DNA strand cleavage (although a covalently trapped phosphotyrosyl intermediate has not been observed in either of the structures determined in this study, and a role as a general base in more traditional mechanism involving water-mediated hydrolysis cannot be ruled out). While such a mechanism has not been observed previously for a restriction endonuclease, the BfiI enzyme is a member of the phospholipase D family of nucleases, that includes many enzymes that proceed via a phosphotyrosyl covalent intermediate, and is known to form a phospho-histidyl covalent intermediate during strand cleavage (Sasnauskas et al., 2010). DNA binding and perturbation The appropriate balance of specificity, fidelity and affinity for protein-DNA interactions is one of the most fundamental of biological requirements. Restriction endonucleases reside at one end of the spectrum of possible DNA recognition behaviors: they cleave relatively short DNA sequences that usually occur frequently within both the host genome and invasive DNA sequences , and display extremely high fidelity that spares the host from off-target cleavage (Pingoud et al., 2005). Shen et al. Page 6 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Protein-DNA recognition specificity is thought to depend upon a combination of direct readout of the nucleotide bases through contacts between the protein and the DNA that can be direct and/or water-mediated, and the additional 'indirect' exploitation of DNA conformational preferences by inducing a DNA structural perturbation or bend that is favored by a limited number of possible DNA sequences (Jones et al., 1999; Luscombe et al., 2001; vonHippel, 2007). Direct readout of DNA sequences is most effective within the major groove which is physically accessible and also provides chemically distinct combinations of hydrogen-bond partners from the four possible base pairs. However, many specific DNA binding proteins augment these contacts with additional interactions made within the minor groove (completely encircling the DNA), and in some cases can achieve specificity entirely via interactions within and across the minor groove (Bewley et al., 1998; Rohs et al., 2009). DNA-binding proteins that contact minor groove structural elements may rely on a combination of DNA bending and surface complementarity (for example, as displayed by the TATA binding protein) (Kim et al., 1993) and may also read out local sequence-dependent shape and charge characteristics of the minor groove (as observed for a variety of DNA-binding proteins, including the nucleosome core particle and the Drosophila Hox protein SCR) (Rohs et al., 2009). In these examples, the DNA is dramatically deformed, but the canonical Watson-Crick base pairing of the complementary strands is still preserved. Because of extreme pressure to maintain high fidelity of recognition, restriction endonucleases are notable for their propensity to fully exploit multiple avenues of DNA readout and specificity (a behavior that can be termed 'recognition overkill'). For example, the MunI restriction endonuclease (a PD‥(D/E)×K enzyme which recognizes the six base pair sequence 5' - GTTAAC - 3') establishes 16 direct hydrogen bonds to these bases in the major groove, 10 direct contacts to phosphates, and it induces a significant distortion between the central base pairs in the sequence (Deibert et al., 1999). That protein displays approximately four direct contacts per base pair--an accomplishment that is facilitated by its core fold, in which the catalytic sites are surrounded by a densely packed array of polar side chains that can fully read out the DNA target's sequence and its shape. In contrast, PacI is one of the smallest known restriction endonucleases, yet it recognizes a longer (eight basepair) target site while being folded around nonspecific catalytic motif that primarily interacts with the phosphate backbone. Evidently, PacI achieves a similarly high level of recognition-fidelity while forming far fewer hydrogen bonds to the nucleotides. Such a minimal protein-DNA interface—in which less than 50% of potential hydrogen bond partners within the target's major groove are engaged in direct contacts with the protein— would typically be expected to correspond to greatly reduced fidelity of recognition (Chevalier et al., 2003). The rearrangement of the DNA conformation and its interstrand base pair contacts may represent a mechanism that significantly increases specificity of recognition by PacI, without requiring an investment by the enzyme in a large number of base-specific contacts. It is known that the act of unstacking and/or unpairing consecutive base pairs can result in unfavorable increases in free energy of binding. Computational and direct biophysical analyses indicate that this energetic cost can differ by several kcal/mol per base step, depending on the sequence context of the bases involved (Delcourt and Blake, 1991; Hobza and Sponer, 2002). The sequestration of individual bases into unpaired conformations in the PacI complex and the partial unstacking of flanking base pairs, may therefore greatly favor the correct target sequence for binding and cleavage over closely related DNA sequences Shen et al. Page 7 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript While no sequence-specific DNA binding proteins or endonucleases have displayed the extreme basepair deformation and reorganization displayed by PacI, some restriction endonudleases have been observed to unpair and 'flip out' individual bases, while also greatly distorting the DNA backbone conformation. For example, the Ecl18kI enzyme flips the adenine and thymine bases from each strand of its cognate 5'-CCNGG-3' target site and sequesters both into protein binding pockets, as part of a mechanism that dramatically kinks the DNA and greatly reduces the value for the rise between the flanking inner C:G basepairs, thus decreasing the distance between scissile phosphates by several angstroms (Bochtler et al., 2006). A similar deformation is seen for the PspGI restriction endonuclease (Szczepanowski et al., 2008), and is presumably a common feature of many such REases that utilize base-flipping as part of their recognition mechanism. While the mechanism of nucleic acid recognition displayed by the PacI endonuclease appears very extreme as compared to most sequence-specific DNA-binding proteins , the distortion of the substrate and the contacts formed by the protein are in fact quite similar to the pattern of RNA recognition exhibited by archaeosine tRNA-guanine transglycosylase, which modifies a guanine base in the 'D arm' of its tRNA substrate. That enzyme disrupts all of the normal basepair and tertiary interactions in the tRNA D arm, leading to reorganization of the tRNA helical strucure and association of the G15 base with the enzyme active site (Ishitani et al., 2003). Initial cognate site recognition Finally, the observation of such a dramatic reorganization of the PacI target site, involving removal of each base from its complementary partner, begs the question of how the initial moment of cognate site recognition is related to the subsequent formation of the catalytic enzyme-substrate (ES) complex that is visualized in typical enzyme-DNA co crystal structures. A long history of biophysical studies of protein-DNA recognition (recently revisited and reviewed in (Halford, 2009)) indicates that DNA-binding proteins sample potential DNA binding sites by rapidly associating and dissociating from non-cognate DNA sequences (a process greatly accelerated by non-specific orientation and interaction between the oppositely charged molecules), while also sliding back and forth across regions covering approximately 50 base pairs around each initial 'landing site' in a limited 1-dimensional search of nearby DNA sequences. It is generally assumed that the contacts made within the initial encounter complex between a specific DNA-binding protein and its correct cognate target site are similar to those found in the enzyme-substrate complex, with additional conformational changes driven by the binding energy derived in the initial encounter with the cognate site. In this model, additional specificity of recognition, beyond that which is engendered by the contacts made between protein and DNA bases, can be derived by sequence-specific conformational preferences of the DNA. This model also allows for the possibility that the unbound sequence of a cognate DNA target site might be predisposed to physically sample a conformation that is similar to its final bound state, which would also enhance recognition and high affinity binding. However, the structure of the PacI endonuclease in complex with its cognate target site indicates that for this enzyme, and perhaps for other highly specific DNA binding proteins, the structure of the initial specific encounter complex might differ significantly from the subsequent biologically or catalytically active complex. Its seems unlikely that the 5' - TTAATTAA- 3' sequence recognized by PacI is predisposed to sample a conformation, in the absence of bound protein, in which several bases are completely unpaired from their Watson- Crick partners and the flanking base pairs are significantly unstacked. When examining the current collection of crystallographic structures of protein-DNA complexes, many examples can be found where the number of observable contacts between protein and DNA bases do not Shen et al. Page 8 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript obviously correspond to the actual specificity of the binding interaction. While such observations may be explained at least in part by the contribution of indirect readout to affinity and specificity, it may be that some specific DNA-protein binding events may be driven by the initial, transient formation of atomic contacts in the cognate complex that are difficult to visualize using traditional crystallographic methods, and are then significantly rearranged to produce the final catalytically or biologically active state. Experimental Procedures A detailed description of materials and methods is provided in Supplementary Information. Briefly, the gene encoding PacI was isolated from Pseudomonas alcaligenes chromosomal DNA and re-introduced into P.alcaligenes on a plasmid vector, resulting in a 48-fold increase in endonuclease expression. A 100-L culture was grown of this over expressing strain, from which 57 mg of homogeneous PacI was purified by FPLC column chromatography. The specific activity, monitored by conventional DNA-digestion and agarose gel electrophoresis, was approximately 6×105 units per mg. Crystals of the protein-DNA complex, using a synthetic 18 base pair DNA duplex corresponding to sequence 5' - GAGGCTTAATTAAGCCGC - 3' and a complementary bottom strand were grown by hanging drop geometry against a crystallization buffer containing 18 to 22% polyethylene glycol 3000 (PEG3K) 100 mM sodium citrate, pH 5.5 and either10 mM MgCl2 or 2 mM CaCl2. The structure of the complex in the presence of magnesium was determined using a combination of the multiple isomorphous replacement (MIR) and single anomalous dispersion (SAD) methods, using five independently generated heavy atom derivatives (two separate PtCl4 soaks, and one each of HgCN2, PIP and WO4). In-house wild-type and heavy atom MIR datasets, using a rotating anode generator, extended to approximately 2.6 Å resolution. In addition, a single SAD dataset (extending to 1.9 Å resolution) from a platinum-soaked was collected at the Advanced Light Source using beamline 5.0.2. The combination of in-house and synchrotron data was used to determine and refine the structure of the magnesium-bound, cleaved product complex. Subsequently, a second 2.0 Å resolution dataset of an unsoaked, wild-type crystal in the presence of calcium was also collected and refined, yielding a corresponding model of the uncleaved protein-DNA complex. Data and refinement statistics are provided in Table 1. Coordinates and Data Deposition The X-ray structure factor amplitudes and corresponding refined coordinates for the PacI/DNA complex, in the form of calcium-bound uncleaved DNA and magnesium-bound cleaved DNA structures, have been deposited in the RCSB database for immediate release (PDB ID code 3LDY and 3M7K). Requests for the PacI-overexpression clone should be direct to New England Biolabs (xus@neb.com). Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments X-ray data was collected at the Advanced Light Source (ALS) synchrotron facility at the Lawrence Berkeley National Laboratory (University of California) on beamline 5.0.2 with the assistance of ALS staff. We thank members of the laboratories of Roland Strong and Adrian Ferre-D'Amare for advice and assistance during structure determination. This work was supported by funding from the NIH to BLS (R01 GM49857) and by funding from the Fred Hutchinson Cancer Center to the Program in Structural Biology. Shen et al. Page 9 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript References Bates DL, Chen Y, Kim G, Guo L, Chen L. Crystal structures of multiple GATA zinc fingers bound to DNA reveal new insights into DNA recognition and self-association by GATA. J Mol Biol 2008;381:1292–1306. [PubMed: 18621058] Bennett-Lovsey RM, Herbert AD, Sternberg MJ, Kelley LA. Exploring the extremes of sequence/ structure space with ensemble fold recognition in the program Phyre. Proteins 2008;70:611–625. [PubMed: 17876813] Bewley CA, Gronenborn AM, Clore gM. Minor groove-binding architectural proteins: Structure, function and DNA recognition. Ann Rev Biophys Biomol Struct 1998;27:105–131. [PubMed: 9646864] Bochtler M, Szczepanowski RH, Tamulaitis G, Grazulis S, Czapinska H, Manakova E, Siksnys V. Nucleotide flips determine the specificity of the Ecl18kI restriction endonuclease. Embo J 2006;25:2219–2229. [PubMed: 16628220] Bujnicki JM. Crystallographic and bioinformatic studies on restriction endonucleases: inference of evolutionary relationships in the "midnight zone" of homology. Curr Protein Pept Sci 2003;4:327– 337. [PubMed: 14529527] Chevalier B, Turmel M, Lemieux C, Monnat RJ, Stoddard BL. Flexible DNA target site recognition by divergent homing endonuclease isoschizomers I-CreI and I-MsoI. J Mol Biol 2003;329:253–269. [PubMed: 12758074] Cymerman IA, Obarska A, Skowronek KJ, Lubys A, Bujnicki JM. Identification of a new subfamily of HNH nucleases and experimental characterization of a representative member, HphI restriction endonuclease. Proteins 2006;65:867–876. [PubMed: 17029241] Deibert M, Grazulis S, Janulaitis A, Siksnys V, Huber R. Crystal structure of MunI restriction endonuclease in complex with cognate DNA at 1.7 Å resolution. EMBO Journal 1999;18:5805–5816. [PubMed: 10545092] DeLano, W. The PYMOL molecular graphics system. San Carlos CA: DeLano Scientific; 2002. Delcourt SG, Blake RD. Stacking energies in DNA. J Biol Chem 1991;266:15160–15169. [PubMed: 1869547] Dunin-Horkawicz S, Feder M, Bujnicki JM. Phylogenomic analysis of the GIY-YIG nuclease superfamily. BMC Genomics 2006;7:98. [PubMed: 16646971] Eastberg JH, Eklund J, Monnat R Jr, Stoddard BL. Mutability of an HNH nuclease imidazole general base and exchange of a deprotonation mechanism. Biochemistry 2007;46:7215–7225. [PubMed: 17516660] Flick KE, Jurica MS, Monnat RJ Jr, Stoddard BL. DNA binding and cleavage by the nuclear intron- encoded homing endonuclease I-PpoI. Nature 1998;394:96–101. [PubMed: 9665136] Grazulis S, Manakova E, Roessle M, Bochtler M, Tamulaitiene G, Huber R, Siksnys V. Structure of the metal-independent restriction enzyme BfiI reveals fusion of a specific DNA-binding domain with a nonspecific nuclease. Proc Natl Acad Sci U S A 2005;102:15797–15802. [PubMed: 16247004] Halford SE. An end to 40 years of mistakes in DNA-protein association kinetics? Biochem Soc Trans 2009;37:343–348. [PubMed: 19290859] Hobza P, Sponer J. Toward true DNA base-stacking energies: MP2, CCSD(T), and complete basis set calculations. J Am Chem Soc 2002;124:11802–11808. [PubMed: 12296748] Ibryashkina EM, Zakharova MV, Baskunov vB, Bogdanova ES, Nagornykh MO, Denmukhamedov MM, Melnik BS, Kolinski A, Gront D, Feder M, et al. Type II restriction endonuclease R.Eco29kI is a member of the GIY-YIG nuclease superfamily. BMC Struct Biol 2007;7:48–56. [PubMed: 17626614] Ishitani R, Nureki O, Nameki N, Okada N, Nishimura S, Yokoyama S. Alternative tertiary structure of tRNA for recognition by a posttranscriptional modification enzyme. Cell 2003;113:383–394. [PubMed: 12732145] Jakubauskas A, Giedriene J, Bujnicki JM, Janulaitis A. Identification of a single HNH active site in type IIS restriction endonuclease Eco31I. J Mol Biol 2007;370:157–169. [PubMed: 17499273] Jones S, Heyningen Pv, Berman HM, Thornton JM. Protein-DNA interactions: a structural analysis. J Mol Biol 1999;287 Shen et al. Page 10 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Kim JL, Nikolov DB, Burley SK. Co-crystal struture of TBP recognizing the minor groove of a TATA element. Nature 1993;365:520–527. [PubMed: 8413605] Kosinski J, Feder M, Bujnicki JM. The PD-(D/E)XK superfamily revisited: identification of new members among proteins involved in DNA metabolism and functional predictions for domains of (hitherto) unknown function. BMC Bioinformatics 2005;6:172. [PubMed: 16011798] Kuhlmann UC, Moore GR, James R, Kleanthous C, Hemmings AM. Structural parsimony in endonuclease active sites: should the number of homing endonuclease families be redefined? FEBS Letters 1999;463:1–2. [PubMed: 10601625] Luscombe NM, Laskowski RA, Thornton JM. Amino acid-base interactions: a three-dimensional analysis of protein-DNA interactions at an atomic level. Nucleic Acids Res 2001;29:2860–2874. [PubMed: 11433033] Mehta P, Katta K, Krishnaswamy S. HNH family subclassification leads to identification of commonality in the His-Me endonuclease superfamily. Protein Science 2004;13:295–300. [PubMed: 14691243] Miyazono K, Watanabe M, Kosinski J, Ishikawa K, Kamo M, Sawasaki T, Nagata K, Bujnicki JM, Endo Y, Tanokura M, Kobayashi I. Novel protein fold discovered in the PabI family of restriction enzymes. Nucleic Acids Res 2007;35:1908–1918. [PubMed: 17332011] Olson WK, Esguerra M, Xin Y, Lu XJ. New information content in RNA base pairing deduced from quantitative analysis of high-resolution structures. Methods 2009;47:177–186. [PubMed: 19150407] Orlowski J, Bujnicki JM. Structural and evolutionary classification of Type II restriction enzymes based on theoretical and experimental analyses. Nucleic Acids Res 2008;36:1–13. [PubMed: 17962301] Pingoud A, Fuxreiter M, Pingoud V, Wende W. Type II restriction endonucleases: structure and mechanism. Cell Mol Life Sci 2005;62:685–707. [PubMed: 15770420] Qiang BQ, Schildkraut I. NotI and SfiI: restriction endonucleases with octanucleotide recognition sequences. Methods Enzymol 1987;155:15–21. [PubMed: 2828862] Roberts RJ, Vincze T, Posfai J, Macelis D. REBASE--a database for DNA restriction and modification: enzymes, genes and genomes. Nucleic Acids Res 2010;38:D234–D236. [PubMed: 19846593] Rohs R, West SM, Sosinsky A, Liu P, Mann RS, Honig B. The role of DNA shape in protein-DNA recognition. Nature 2009;461:1248–1253. [PubMed: 19865164] Saravanan M, Bujnicki JM, Cymerman IA, Rao DN, Nagaraja V. Type II restriction endonuclease R.KpnI is a member of the HNH nuclease superfamily. Nucleic Acids Res 2004;32:6129–6135. [PubMed: 15562004] Sasnauskas G, Zakrys L, Zaremba M, Cosstick R, Gaynor JW, Halford SE, Siksnys V. A novel mechanism for the scission of double-stranded DNA: BfiI cuts both 3'–5' and 5' –3' strands by rotating a single active site. Nucleic Acids Res. 2010 Advance Access published January 4, 2010. Sokolowska M, Czapinska H, Bochtler M. Crystal structure of the beta beta alpha-Me type II restriction endonuclease Hpy99I with target DNA. Nucleic Acids Res 2009;37:3799–3810. [PubMed: 19380375] Stoddard BL. Homing endonuclease structure and function. Quarterly Reviews of Biophysics 2005;38:49–95. [PubMed: 16336743] Szczepanowski RH, Carpenter MA, Czapinska H, Zaremba M, Tamulaitis G, Siksnys V, Bhagwat AS, Bochtler M. Central base pair flipping and discrimination by PspGI. Nucleic Acids Res 2008;36:6109–6117. [PubMed: 18829716] Vaisvila R, Morgan RD, Posfai J, Raleigh EA. Discovery and distribution of super-integrons among pseudomonads. Mol Microbiol 2001;42:587–601. [PubMed: 11722728] vonHippel PH. From "Simple" DNA-Protein interactions to the macromolecular machines of gene expression. Ann Rev Biophys Biomol Struct 2007;36:79–105. [PubMed: 17477836] Zheng G, Lu XJ, Olson WK. Web 3DNA--a web server for the analysis, reconstruction, and visualization of three-dimensional nucleic-acid structures. Nucleic Acids Res 2009;37:W240–W246. [PubMed: 19474339] Shen et al. Page 11 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. The structure of the PacI restriction endonuclease The protein is colored according to individual structural domains and motifs; the same color scheme is used in all panels. All figures were generated using the molecular graphics program PYMOL (DeLano, 2002). Panel a: The PacI homodimer is bound to ts palindromic DNA target sequence. The two protein subunits are colored green and cyan; the two bound zinc ions in each protein subunit are labeled and colored dark green, and the single bound divalent cation observed in each catalytic site is labeled 'Mg' and colored blue. The corresponding scissile phosphates, that yield 2 base, 3' cohesive overhangs when cleaved, are colored red. Panel b: A single subunit of PacI is shown, in two separate orientations. The left panel shows the same orientation as the cyan-colored subunit in panel a. The right panel shows the same subunit, Shen et al. Page 12 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript rotated by 90° around the horizontal axis. In both panels, the ββα-metal motif is colored dark green, and the unique N-terminal region (which binds zinc number 1) is colored blue. On the left, the eight cysteine ligands to the two zinc ions are labeled. On the right, the single bound divalent cation (magnesium in the product complex) and the two residues involved in its coordination (D92 and N113) are labeled. Panel c: Sequence homology between PacI, and its three recognizable homologues (hypothetical protein sequences from Campylobacter concisus, cowae and lari, respectively). The ββα-metal motif is indicated by the box. Residues that directly contact DNA are indicated with asterisks arrows; catalytic residues of the HNH endonuclease motif are indicated with blue font and asterisks; zinc-binding cysteine residues are indicated with red font and asterisks. The N-terminal regions, that harbors the first bound zinc ion, is indicated with light blue font, corresponding to the coloring of the same region in panel b. See also supplemental Figure S1. Shen et al. Page 13 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. ββα-metal (HNH) endonucleases that contain two bound zinc ions per subunit For each endonuclease structure, the left panels show comparable orientations of the bound DNA targets, while the right panels show comparable orientations of the protein subunits (with the dyad symmetry axis running vertically and the protein subunits oriented to the left and right side of that axis). For all three structures, the ββα-metal catalytic motif is colored blue, and the bound zinc ions are shown as teal spheres. Panel a: The PacI restriction endonuclease. While the protein's overall organization of secondary structure elements, relative to the catalytic sites and bound zinc ions, is similar to Hpy99I, the overall architecture of the individual ββα-metal repeats, as well as the mode of DNA binding, is significantly different. Panel b: The core of the Hpy99I restriction endonuclease, which cleaves a five base pair CGWCG target, generating Shen et al. Page 14 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript a 5-base, 3' overhang (Sokolowska et al., 2009). An N-terminal β-barrel domain (residues 1 to 53) that is not involved in DNA binding is removed for clarity. The remaining protein homodimer structure includes all catalytic and zinc-binding regions that correspond to those observed in PacI. Note that the orientation of the protein is roughly orthogonal to the major axis of the DNA duplex, which is relatively unbent. Panel c: The I-PpoI restriction endonuclease (a fourteen-base cutting homing endonuclease found in the eukaryotic amoeboid Physarum polycephalum) (Flick et al., 1998). The overall bend of the DNA target is similar to that displayed by PacI; however all base pairs are maintained in their original Watson-Crick base pairing arrangement. The protein fold, beyond the core ββα-metal motif, is completely different from that displayed by Hpy99I and PacI, indicating that divergence of these protein lineages may have occurred at an early stage, from a simple ββα-metal scaffold prior to subsequent elaboration and specialization. See also supplemental Figure S2. Shen et al. Page 15 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. ββα-metal motifs and catalytic sites for HNH restriction and homing endonucleases The core catalytic motif, consisting of a two-stranded antiparallel beta-sheet and single alpha- helix, is shown in approximately the same orientation for (Panel a) PacI; (Panel b) Hpy99I and (Panel c) I-PpoI . The scissile phosphate is shown in red, flanked by its 5' and 3' nucleosides in gray. The histidine general base is colored and labeled with red, and the corresponding water nucleophile is a small light blue sphere. In all three catalytic sites, a single bound divalent metal ion (dark blue larger sphere) is coordinated by two asparagine/aspartate residues. An additional polar residue (participating in cleavage as a Lewis acid, whereby it stabilizes the phosphoanion transition state) is present in each catalytic site that is positioned to help satisfy the charge on the phosphate during cleavage (R93 for PacI; H151 for Hpy99I, and R61 for I-PpoI). Note that Shen et al. Page 16 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript PacI displays a structural inversion in the positions of the histidine general base and the Lewis acid: His 42 is presented from a loop outside the ββα-metal motif, while R93 is found in the position that is usually occupied by an HNH histidine base. Finally, peripheral cysteine residues in all three ββα-metal motifs are involved in coordination of a structural zinc ion (C109 and C112 in PacI). However, the location of the bound zinc in the restriction endonucleases (panels a and b) differ from the homing endonuclease (panel c), indicating that these enzyme lineages may have diverged from a common ancestor prior to development of these metal binding sites. See also supplemental figure S3. Shen et al. Page 17 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. DNA recognition and binding by PacI Panel a: Conformation of the bound DNA target site. The individual bases of the 'TTAATTAA' target are colored both by their position in the original DNA half-site (left half-site = green; right = yellow) and by their identity (dark bases are adenine; light bases are thymine). The bound DNA is shown in a stick representation on the left, and in a cartoon representation, generated by the program 3DNA (Zheng et al., 2009), on the right. Those bases that lie outside the eight base pair target site are colored grey. Panel b: Cartoon representation of the base pairing interactions between the two target site strands and the contacts between the protein, DNA and solvent molecules. The individual bases are colored as shown in panel (a) above. The numbering of the bases corresponds to the position in the original unbound target sequence, Shen et al. Page 18 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript with bases −4T to −1A corresponding to the 5' half (left half) of the target site along one DNA strand, and bases +1T through +4A corresponding to the 3' (right) half of the target site along the same strand. In the bound complex, all bases in the target are removed from their original Watson-Crick partners, so that some bases are unpaired (+1T and +4A on each strand), some are found in noncanonical A:A and T:T base pairs (+2T, −1A, −2A and −3T on each strand) and some are found in new Watson-Crick base pairs (−4T with +3A from each strand). The scissile phosphates on each strand are red; well-ordered water molecules are blue. Protein residues (from only one of the two protein subunits, for clarity) that are involved in direct or water-mediated contacts to the DNA are indicated; those that form direct contacts to DNA bases are boxed. Panel c: Structural interactions between each unique DNA base or base pair in a single half-site to solvent and protein residues. The numbering of bases is consistent with panels a and b above. See also supplemental figure S4. Shen et al. Page 19 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Shen et al. Page 20 Table 1 Data collection and refinement statistics Data set Id Native-1 Native-2 PtCl4-1 PtCl4-2 HgCN2 PIP WO4 Wavelength (Å) 1.5418 0.97741 1.0719 1.5418 1.5418 1.5418 1.5418 Data collection Space group C2221 C2221 C2221 C2221 C2221 C2221 C2221 a (Å) 36.86 36.86 37.09 37.32 36.93 36.89 37.83 b (Å) 115.75 115.75 115.16 114.08 116.08 117.79 116.05 c (Å) 114.37 114.37 114.83 114.32 114.19 113.89 114.14 Resolution (Å) 50-2.64 38-1.97 50-1.92 25-3.2 50-3.0 50-2.6 50-3.07 Unique reflections 7568 17385 19094 4343 5193 7947 5014 Redundancy* 6.6(4.7) 10.7(4.2) 13.5(11.1) 6.9(7.1) 13.4(8.9) 7.1(7.0) 11.0(7.8) Completeness (%)* 99.8(98.7) 97.2(80.1) 98.6(92.4) 100(100) 99.8(99.6) 99.1(93.5) 99.4(94.4) I/σ* 20.4(5.0) 27.8(3.3) 44.2(3.4) 20.3(5.7) 28.7(7.1) 37.1(11.8) 32.9(13.9) Rmergea (%)* 9.2(31.2) 6.8(22.9) 6.2(22.6) 9.7(39.3) 8.6(31.9) 4.5(13.9) 6.0(12.5) B(iso)(Å2) 47.4 26.25 29.04 60.4 61.1 58.7 53.5 Refinement Protein atoms# 1108 1108 1108 DNA atoms# 366 366 367 Heavy atoms 2 Zn+2 2 Zn+2 2 Zn+2, Pt+2 Catalytic Metal ions Ca+2 Ca+2 Mg+2 Cations --- --- SO4−2 Solvent molecules 76 94 114 R-factorb (%)* 0.208(0.293) 0.184(0.227) 0.172(0.217) R-freeb (%)* 0.278(0.319) 0.217(0.361) 0.201(0.285) Rmsd Bond length (Å) 0.012 0.013 0.011 Angles (°) 1.667 1.541 1.305 Ramachandran (%) Core region 97.83 96.38 98.43 Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Shen et al. Page 21 Data set Id Native-1 Native-2 PtCl4-1 PtCl4-2 HgCN2 PIP WO4 Wavelength (Å) 1.5418 0.97741 1.0719 1.5418 1.5418 1.5418 1.5418 Allowed region 1.45 2.90 1.57 Outliers 0.72 0.72 0.00 *Highest resolution shell values in parenthesis. aRmerge = Σ|Ihi - <Ih> |/ΣIh, where Ihi is the ith measurement of reflection h, and <Ih> is the average measured intensity of reflection h. bR-factor/R-free = Σh|Fh(o) - Fh(c)|/Σh|Fh(o)|. Where R-free was calculated with 5% of the data excluded from refinement. Structure. Author manuscript; available in PMC 2011 June 9. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Shen et al. Page 22 Table 2 Activity of PacI site-directed mutants PacI variant Endonuclease activity Units per 25 µl in vitro reaction Wild-type PacI +++ ~200 Catalytic: H42A + ~1 R93K +/− ~0.2 R93A,M − <0.01 Y100F − <0.01 Mg2+-binding: D92A,L +/− <0.1 N113A,L +/− <0.1 Specificity: N32A,T,L,D − <0.01 N36A,T,L,D − <0.01 K39A,M ++ >10 Endonuclease activities of wild-type PacI and of mutant derivatives. Mutants were constructed by two-step PCR, expressed in vitro using the PURExpress™ transcription/translation system, and assayed by DNA-digestion and gel electrophoresis, as described in the Supplementary Material. The standard 25 µl PURExpress™ reaction produced approximately 200 units of endonuclease activity from the wild-type PacI gene template (1 unit completely digests 1 µg of substrate DNA to completion in 1 h at 37°C). The limit of endonuclease activity detectable in this assay corresponded to 10−4-fold less than wild-type, or approximately 0.01 units. In most cases, several different mutants were constructed for each amino acid targeted for alteration. Mutants yielding the same result are grouped together on a single line in the table; thus, ‘N36A,T,L,D’, for example, signifies that Asn 36 was individually changed to Ala, Thr, Leu, and Asp, and all four mutant enzymes behaved similarly—in this case displaying no detectable endonuclease activity. Plus and minus symbols in column 2 indicate the relative levels of endonuclease activity observed across several independent experiments. These levels are quantified approximately with respect to wild-type in column 3. Structure. Author manuscript; available in PMC 2011 June 9.
3M7L
Crystal Structure of Plant SLAC1 homolog TehA
Homolog Structure of the SLAC1 Anion Channel for Closing Stomata in Leaves Yuhang Chen1,6, Lei Hu2, Marco Punta6,7, Renato Bruni6, Brandan Hillerich6, Brian Kloss6, Burkhard Rost1,6,1, James Love6, Steven A. Siegelbaum2,3,5, and Wayne A. Hendrickson1,4,5,6 1Department of Biochemistry and Molecular Biophysics, Columbia University, New York, NY 10032, USA 2Department of Neuroscience, Columbia University, New York, NY 10032, USA 3Department of Pharmacology, Columbia University, New York, NY 10032, USA 4Department of Physiology and Cellular Biophysics, Columbia University, New York, NY 10032, USA 5Howard Hughes Medical Institute, Columbia University, New York, NY 10032, USA 6NYCOMPS, New York Structural Biology Center 89 Convent Avenue, New York, NY 10027 USA 7Department of Computer Science and Institute for Advanced Study Technical University of Munich D-85748 Munich, Germany Summary The plant SLAC1 anion channel controls turgor pressure in the aperture-defining guard cells of plant stomata, thereby regulating exchange of water vapor and photosynthetic gases in response to environmental signals such as drought or high levels of carbon dioxide. We determined the crystal structure of a bacterial homolog of SLAC1 at 1.20Å resolution, and we have used structure- inspired mutagenesis to analyze the conductance properties of SLAC1 channels. SLAC1 is a symmetric trimer composed from quasi-symmetric subunits, each having ten transmembrane helices arranged from helical hairpin pairs to form a central five-helix transmembrane pore that is gated by an extremely conserved phenylalanine residue. Conformational features suggest a mechanism for control of gating by kinase activation, and electrostatic features of the pore coupled with electrophysiological characteristics suggest that selectivity among different anions is largely a function of the energetic cost of ion dehydration. Users may view, print, copy, download and text and data- mine the content in such documents, for the purposes of academic research, subject always to the full Conditions of use: http://www.nature.com/authors/editorial_policies/license.html#terms Correspondence and request for materials should be addressed to WAH (wayne@convex.hhmi.columbia.edu).. Supplementary Information is linked to the online version of the paper at www.nature.com/nature. Author Contributions YC, LH, MP, RB, BH, BK and JL performed experiments; YC, LH, BR, SAS, and WAH analyzed data; YC, LH, SAS, and WAH prepared the manuscript. Author Information Atomic coordinates and diffraction data are deposited in the Protein Data Bank with accession codes 3M71, 3M73, 3M74, 3M75, 3M76, and 3M7L. See Table S4 for identifications. Reprints and permissions information is available at www.Nature.com/reprints. HHS Public Access Author manuscript Nature. Author manuscript; available in PMC 2013 January 18. Published in final edited form as: Nature. 2010 October 28; 467(7319): 1074–1080. doi:10.1038/nature09487. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Stomatal pores in the leaves of plants permit the influx of atmospheric carbon dioxide in exchange for transpirational evaporation of water1,2. A pair of kidney-shaped guard cells define each pore aperture, and turgor pressure variation in these cells determines the degree of stomatal pore openness. Depending on diverse environmental factors, the stomata close to prevent H2O losses and open to admit CO2 for photosynthesis. Environmental stimuli that lead to stomatal closure include darkness, high CO2 levels, ozone, low air humidity and drought. The plant hormone abscisic acid (ABA) is critical for signal transduction from these stimuli. Mutational screens in Arabidopsis thaliana for CO2 and ozone sensitivity identified a protein with ten predicted transmembrane (TM) helices, now called slow anion channel 1 (SLAC1), as having a central role in the control of stomatal closure3-5. Recent studies proved that SLAC1 is indeed an anion channel6-7, with characteristics like those of slow anion channels found in guard cells8, and that it is activated by phosphorylation from the OST1 kinase9. OST1 activity is negatively regulated by the ABI1 phosphatase10,11, which is in turn inhibited by the stomatal ABA receptors PYR/RCAR12,13 when in the ternary hormone-receptor-phosphatase complex14-18. Thereby, ABA stimulates SLAC1 channel activity. Resulting Cl− efflux through SLAC1 causes membrane depolarization, which activates outward-rectifying K+ channels, leading to KCl and water efflux to further reduce turgor and cause stomatal closure. SLAC1 expression in Arabidopsis is confined to the guard cells of leaves, but other Arabidopsis tissues do have SLAC1 homologs3, named SLAH1-SLAH4. The identifying mutations slac1-14 and slac1-23 are, respectively, in predicted transmembrane segments 9 (S456F) and 1 (G194D) of a protein that includes substantial N- and C-terminal extensions outside a 10-helix transmembrane (TM) domain. SLAH1, which is absent from leaves and lacks the terminal extensions of SLAC1, fully complements the mutant phenotype in slac1-2 guard cell protoplasts3. SLAC1 and homologs are also present in other plant genomes, including nine in rice (Oryza sativa) and five in grape vines (Vitis vinifera). SLAC1 relatives, some quite remote, also occur in bacteria, archaea and fungi. Known prokaryotic homologs contain only the predicted transmembrane domain of SLAC1, but some fungal homologs do have N- and C-terminal extensions. One homolog, Mae1 from the yeast Schizosaccharomyces pombe, functions as a malate uptake transporter19; another, Ssu1 from Saccharomyces cerevisiae and other fungi including Aspergillus fumigatus, is characterized as a sulfite efflux pump20,21; and TehA from Escherichia coli is identified as a tellurite resistance protein by virtue of its association in the tehA/tehB operon22,23. Despite a lack of further biochemical characterization, many homologs are annotated as tellurite resistance/ dicarboxylate transporter (TDT) proteins. We have undertaken structural and functional characterizations of the SLAC1 anion channel. We first solved an atomic-resolution crystal structure of the TehA homolog from Haemophilus influenzae, and we then developed a homology model for Arabidopsis SLAC1. This model allowed us to conduct mutagenesis for functional testing of structure-inspired hypotheses on gating and selectivity. We expressed Haemophilus TehA and Arabidopsis SLAC1 in Xenopus oocytes to characterize channel properties of these proteins and mutant variants. We also determined crystal structures for several mutant variants, including the homolog of slac1-2. Chen et al. Page 2 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Structure of SLAC1 bacterial homolog TehA We performed a bioinformatic analysis of SLAC1-related proteins, first clustering nearly 900 non-redundant sequences into a superfamily at the PSI-BLAST level E≤10−3, then into three distinct families at an initial threshold of E≤10−30, and finally into subfamilies at a typical initial threshold of E≤10−55. Since previous annotation is not well founded in experiment and SLAC1 is now the best characterized member, we adopt a nomenclature defining a SLAC superfamily divided into families identified as SF1-SF3 and subfamilies SF1A, SF1B etc. Family SF1 comprises the plant SLAC proteins and close bacterial homologs, family SF2 comprises a distinct set of bacterial proteins often annotated as exfoliative toxins, and family SF3 comprises the fungal Mae1 and Ssu1 proteins and their archaeal or bacterial homologs, respectively. SLAC family SF1 has three large subfamilies: the plant SLAC and SLAH proteins are in subfamily SF1A, closest bacterial homologs are in SF1B, and the TehA homologs are in SF1C (Fig. 1a). The other families also divide into subfamilies as detailed in Table S1, and family SF1 is divided into sub-subfamilies (Table S2). Two pertinent SF1 sequences are aligned in Fig. 1b. We used a structural genomics approach to obtain structural information, testing expression and purification for 43 bacterial and archaeal likely homologs, assaying for detergent choice and stability on 8 of these, finding two with appropriate profiles by size exclusion chromatography, and obtaining suitable crystals for one. This protein, TehA from H. influenzae (HiTehA), was found to be trimeric both by size-exclusion multi-angle light scattering (SEC-MALS) measurements and by chemical cross-linking. When solubilized in β-octylglucoside, HiTehA crystallized in space group R3 with a=b=96.01Å and c=136.27Å. Each asymmetric unit contains one subunit and 65% solvent. The structure was solved by selenomethionyl (SeMet) SAD phasing, ultimately at 1.50Å resolution (Table S3, Fig. S1), and then refined at 1.20Å resolution (Fig. 2a) to R/Rfree values of 14.1%/15.7% for a model that includes ordered residues 6-313, 213 water molecules and four detergent molecules (Table S4). The crystal structure has TehA trimers aligned with three-fold axes of the lattice (Fig. 2b). Subunits are tightly associated, burying 8947 Å2 of total surface area within trimer interfaces. The electrostatic potential surface is largely negative on the extracellular surface (Fig. 2c) and largely positive on the cytoplasmic surface (Fig. 2d). The membrane orientation is specified experimentally from GFP tagging of E. coli TehA24. Each TehA protomer has ten TM helices, as predicted; however, the fold is novel. Tandemly repeated helical hairpins are arranged with quasi-five-fold symmetry (Figs. 2e,S2). Extracellular inter-helix loops are short (2-5 residues) whereas intracellular inter-helix connections are longer, including a nine-residue helix H2,3 between TM2 and TM3 (Fig. 2f). An inner pentad of outwardly directed, TModd, helices creates an apparent pore through each protomer perpendicular to the putative membrane plane. TMeven helices from the five hairpins surround the inner pore and make an outer layer. Chen et al. Page 3 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Homology model for plant SLAC1 Arabidopsis SLAC1 (AtSLAC1) is substantially similar to bacterial homologs, notably HiTehA (Fig. 1b). All HiTehA TM helices are fully aligned to predicted SLAC1 TM helices, but there are short inter-helical gaps (1-5 residues) in all five extracellular loops and in two of the intracellular loops. The TM domain of AtSLAC1 (residues 188-504) aligns to HiTehA with 19% sequence identity and with a PSI-BLAST E-value of 3×10−22. For comparison, and in keeping with the family tree (Fig. 1a), the TM domain of AtSLAC1 shares sequence identities of 76% with rice SLAC1, 41% with Arabidopsis SLAH1, 25% with an SL1B homolog from Halorhodospira halophila, 11% with S. cerevisiae Ssu1, and 9% with S. pombe Mae1. A conceptual model with the AtSLAC1 sequence transposed onto the HiTehA helices sufficed to guide most of our mutational tests, but a detailed AtSLAC1 homology model helped to refine our ideas. Surface variability and electrostatic potential are plotted onto the surface of this model (Fig. 2g,2h). The most remarkable feature of the TehA structure and corresponding SLAC1 model is the central pore through each protomer. As for acetylcholine receptors25, the SLAC1 pore is formed by five helices; but the SLAC1 helices come from one protein molecule rather than five. The SLAC1 pore has a relatively uniform diameter of approximately 5Å across nearly five helical turns (Fig. S3), except for a pronounced constriction in the middle of the membrane (Fig. 3a) where the pore is occluded by the side chain of Phe 450 (Phe 262 in HiTehA). This residue is the only absolutely conserved amino acid residue of the SLAC1 family. The pore is lined with highly conserved (86% identity among five SLAC1 orthologs; 32% identity between AtSLAC1 and HiTehA) and generally hydrophobic residues (Figs. 1b, 3b,3c,S4). Despite this hydrophobicity, the electrostatic potential on the pore surface is polarized (Fig. 3a), presumably due to an invaginated shape adjacent to charged residues outside the membrane. The generally electropositive character of the cytoplasmic surface likely contributes to anion efflux. Kinks in the pore helices contribute to formation of a relatively constant pore diameter across the membrane. Four of the five HiTehA inner helices have centrally located proline residues, which necessarily generate kinks, and TM9 is kinked at a backbone-coordinated water molecule (Fig. 3d). Proline replaces Gly263 in all SF1A and SF1B relatives, including Pro451 AtSLAC1. This water-displacing change is isostructural26. Centrally located prolines also prevail in TM3, TM5 and TM7 across the SF1 proteins. The outer helices are longer and straighter, but more inclined. The only outer-helix proline kink of HiTehA is in TM6 at the trimer three-fold axis. Two of the gene-identifying mutations in Arabadopsis SLAC1 are selected point mutations, others are disruptive transfer DNA (T-DNA) insertions4. In the AtSLAC1 homology model, the slac1-2 (G194D) mutation3 points into the pore from TM1 and can be accommodated structurally whereas the slac1-1 (S456F) mutation4 points away from the pore six residues after pore-blocking Phe450 on TM9 and would be expected to disruptive. Ser456 interacts with outer helix TM10 in the homology model, and the phenyl bulk from S456F would not fit. Position 456 has alanine in HiTehA, 66% of all 204 SF1 homologs and another 27% have threonine or serine (as in all SLAC1 channels); phenylalanine never occurs among all Chen et al. Page 4 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 814 SLAC superfamily members. Position 194 has glycine in 58% of SF1 homologs and alanine in another 22%. Asp194, which never occurs naturally, would block the pore and be expected to repel anions. Mutational tests of channel function Mutational studies corroborate the hypothesis that a TehA-based SLAC1 model is appropriate. First, as discussed above, the slac1-1 mutant (S456F) is expected to be structurally disruptive, and indeed it is inactive in guard cells4, and the slac1-2 mutation (G194D) is expected to block the pore, and we show below that this variant is also inactive. We have also shown that the introduction of SLAC1-conserved proline residues into HiTehA (A208P/G263P) is accommodated isomorphously26. Moreover, as shown below, channel conductance properties of several mutants are similar for AtSLAC1 and HiTehA. To examine characteristics of the SLAC1 channel in light of the structural model, we performed electrophysiological tests of membrane currents from voltage-clamped Xenopus oocytes following injection of wild-type or mutant AtSLAC1 or HiTehA cRNAs. We observed modest-sized chloride currents with wild-type AtSLAC1 cRNA, as found previously6,7, but did not detect any chloride current following injection of wild-type HiTehA cRNA. We found that SLAC1 Cl− conductance was enhanced when the OST1 kinase cRNA was co-injected with SLAC1 cRNA, but only to the levels found by Lee et al.6 and not to the much higher levels found by Geiger et al.7 with OST1 physically connected to SLAC1 by split YFP linkage. In keeping with structural evidence that Phe262 blocks the HiTehA pore, removal of the phenyl group in HiTehA F262A or in the homologous AtSLAC1 F450A resulted in very large Cl− currents relative to wild-type levels, and the SLAC1 currents were now less enhanced by presence of OST1 (Fig. 4a). The tempting interpretation of an opened gate will require validation with appropriately analyzed single- channel recordings27. In keeping with the slac1-2 phenotype3, neither the functionally impaired AtSLAC1 G194D nor its HiTehA G15D homolog showed any substantial conductance; moreover, further consistent with pore blockage by Asp194 in slac1-2, the large conductances of HiTehA F262A and AtSLAC1 F450A were stopped in the double mutants AtSLAC1 G194D/F450A and HiTehA G15D/F262A (Fig. 4a). Here again, the effects in SLAC1 were independent of OST1. We also tested conductance characteristics for a series of AtSLAC1 F450X substitution mutants – F450A,G,T,V,L– and for the corresponding HiTehA F262X series – F262A,G,T,V,L (Figs. 4b,S5, Table S5). Findings from the two series are roughly parallel; in particular, the alanine and glycine substitutions lead to large currents for both and in comparison to the others. There are distinctions, of course, including generally higher conductances for AtSLAC1 over HiTehA and less conductance of F262T TehA compared to F450T SLAC1. It is also noteworthy that OST1 activation is very muted for the F450A,G,L mutants, which is consistent with SLAC1 gating at Phe450. Crystal structures were also determined for several of the HiTehA mutant variants (Table S4). The structures of F262A (1.15Å), F262V (1.60Å), F262L (1.65Å) and G15D (1.50Å) are all essentially isomorphous with the wild-type TehA structure with changes localized Chen et al. Page 5 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript primarily at the sites of mutation; the same is true for the double mutations of F262A/G15D, F262G/G15D and A208P/G263P26. The F262A structure has a wide open pore (Fig. 5a) with a relatively uniform pore diameter of ~5Å through ~30Å across the membrane (Fig. S3) whereas G15D has a doubly occluded pore (Fig. 5b). The pores of other mutant variants are consistent with the sizes of constrictive residues and with the observed conductances. Gating and activation The crystal structures of TehA and its mutant variants when taken together with the functional studies in Xenopus oocytes point to a crucial role for Phe450 in gating of the SLAC1 anion channel. Conservation of this residue across the SLAC1 family implies functional importance. The occlusion of the pore by the presence of F262 in the structure of wild-type TehA and the openness of the pore upon its substitution by alanine in the structure of the F262A mutant provides physical evidence for a gating role of this residue. This interpretation is supported by the correlated conductance characteristics from variants of the AtSLAC1 and HiTehA channels (Figs. 4b,S5). While these observations may suffice for placing the gate within the channel pore, they do not by themselves suggest a mechanism for gating in response to physiological stimuli. Some insight does come from conformational details defined at high resolution. One important structural clue is that the side chain of Phe262 is in a high energy conformation in the HiTehA structure, with χ1/χ2 at −160°/-4°. Although χ1 is in a preferred trans conformation, the phenyl ring is restricted by contacts with Val210 and Leu18 to a χ2 value near 0° rather than near to the preferred 90° orientation (Fig. 5c). Further evidence that Phe262 is restrained from local equilibrium comes from shifts observed in crystal structures of the F262A, F262V, F262L and F262G/G15D variants, which all show consistent backbone movements that displace Cβ(262) by 0.47-0.78Å (Fig. 5d). By contrast, Leu262 in F262L is in a preferred trans/gauche+ conformation at χ1/χ2 = 177°/63° as is Val262 in F262V at χ1 = −176°. What might control activation of HiTehA is unclear, but for AtSLAC1 activation is by OST1 phosphorylation6,7. The molecular consequences of OST1 phosphorylation of SLAC1 remain unknown, but it is plausible that associated shifts in pore- helix orientations would unlatch Phe450 in SLAC1 from a TehA-like restrained orientation. By analogy with Leu262 in F262L, we expect a preferred rotameric state for Leu450 in AtSLAC1 F450L. Thus, the lack of appreciable OST1 activation of conductance in AtSLAC1 F450L (Fig. 5e) might be explained by the lack of a restraining latch, whereby the channel remains closed despite OST1 activation. Puzzles certainly remain since OST1 does substantially activate AtSLAC1 F450T or F450V, which like HiTehA F262V should also be unrestrained; presumably activating adjustments widen the pore enough for ion permeation past threonine and valine but not leucine. Phosphorylation sites have been discovered in the N- and C-terminal tails of AtSLAC16,7,28 (179 and 51 residues long, respectively), but these alone cannot explain OST1 activation of SLAC1. First SLAH1, which fully complements the slac1-1 mutation, does not have these cytoplasmic tails. Secondly, although OST1 phosphorylation of Ser120 in the N-terminal tail is necessary for SLAC1 activation, it is not sufficient7. Thus, we surmise that direct phosphorylation of the SLAC1 TM domain must be critical, and SLAC1 has four conserved Chen et al. Page 6 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Ser/Thr candidates in its cytoplasmic loops. Moreover, SLAC1 proteins have proline- mediated kinks at the putative Phe450 gate in helix TM9, and also in adjacent helix TM7; these features may play a role in phosphorylation-driven unlatching of the Phe450 gate in SLAC1. Ion selectivity and discrimination Our studies of SLAC1 channel relative ion permeabilities, based on measurements of current reversal potential, are consistent with earlier work demonstrating that AtSLAC1 conducts anions but not cations and is selective among anions, with greater permeability for nitrate than for chloride (as in Vicia faba guard cell protoplasts27) and much reduced permeability for malate, bicarbonate or sulfate6,7. We also find that SLAC1 has little permeability for sulfite. Additionally, we find that wild-type SLAC1, F450A and F450T all have similar relative permeabilities to chloride, sulfite and malate, despite having widely different conductance levels, but the gating mutants do show small but significant decreases in nitrate permeability (Fig. 4c, Table S6). The relative insensitivity of anion permeability to gating residue changes suggests that selectivity for these anions may occur away from the central constriction at the channel gate. To some extent, ionic discrimination must depend on pore geometry; thus, an organic anion such as malate may be simply too large to pass through the 5-Å wide pore. Although the SLAC1 pore is lined largely with hydrophobic sidechains (Fig. 3c), it is sprinkled with hydroxyl groups from serine and threonine residues (16%); their electropositive hydrogen atoms may facilitate conductance. Most strikingly, the electrostatic potential within the AtSLAC1 pore is electropositive throughout (Fig. 3a). This polarization, promoted by charges on extra-membranous loops, no doubt contributes significantly in discrimination against cations. The relative anion permeability sequence of SLAC1 determined by us and others, I− > NO3− > Br− > Cl− 6,7, corresponds to selectivity sequence 1 compiled by Wright and Diamond29 for a range of anion-selective proteins. This sequence correlates inversely with the hydration energies of monovalent anions – anions with a lower hydration energy have a greater channel permeability. It is thought to be generated in proteins with weak, low field-strength, anion binding sites, where selectivity is largely determined by the energetic cost of anion dehydration. These selectivity results are thus consistent with the SLAC1 structure, where the pore lacks any obvious anion binding site. Distinctiveness of the SLAC1 channel SLAC1 anion channels are entirely novel in structure and, apparently, in the mechanism for ion conductance. The best characterized of anion channels belong to the CLC family of Cl− channels and transporters30-32. CLC channels have an altogether different architecture from the SLAC1 channel, and the mechanism for selectivity is also very different. Bacterial CLC transporters bind halide ions at three sites in a highly constricted pore30. By contrast, the SLAC1 pore has a relatively uniform diameter across the membrane, except where closed by the gating phenyl group, and we do not find discrete ion binding sites. CLC selectivity is governed by specific residues surrounding these binding sites30,32. The anion selectivity Chen et al. Page 7 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript sequence for the CLC channels of Cl−>Br−>NO −3>I−, opposite of that in SLAC1, is consistent with the high field-strength anion binding sites in CLC channels29. Interestingly, as for AtSLAC1, an Arabidopsis CLCa channel also preferentially transports nitrate ions33, and an E. coli CLC channel is converted to preference of nitrate when a generally conserved serine at the central site is substituted with proline as in AtCLCa32. SLAC1 also differs radically from other structurally characterized anion channels and transporters. These include the VDAC1 voltage-gated anion channel from mitochondrial outer membranes, which has a porin-like beta-barrel structure34-36, and a light-driven halorhodopsin chloride pump, which has a transmembrane conductance pathway similar to that of the proton-pumping pore of bacteriorhodopsin37. Although its channel structure is still only known by homology to other ABC transporters, CFTR is another obviously distinct chloride channel38. Cys-loop receptors also include anion channels39, and these are similar to SLAC1 in having five-helix pores25, but here selectivity is governed by charged groups at the entrance to the pore, which distinguish the anion-selective GABAA and glycine receptors from the cation-selective acetylcholine and serotonin 5HT3 receptors39. Finally, the recently identified TMEM16A gene for calcium-activated chloride channels40-42 appears to encode an 8-TM protein that is again distinct from SLAC1. Stomatal guard cells show both rapidly activated (R-type) and slow (S-type) anion channel activity43. Although slac1 guard cells have very defective S-type activity, their R-type currents are normal4. Guard cell protoplasts from the slac1-2 mutant abnormally accumulate Cl−, K+, malate and fumarate3, whereas SLAC1 shows negligible malate conductance7. As for SLAC1-associated K+ movements, other channels or transporters must be responsible for SLAC1-associated malate movements. Recent studies indicate that AtALMT12, an aluminium-activated malate transporter (ALMT) family member, is a malate-dependent R- type anion channel44 needed for stomatal closure45. Conclusions We find that many functional properties of the plant SLAC1 anion channel are explained well by the structure of an uncharacterized bacterial TehA protein that has been associated with tellurite resistance. SLAC1 and TehA belong to distinct subfamilies within one branch of a larger SLAC1 superfamily, but AtSLAC1 and HiTehA are sufficiently similar (19% sequence identity) that the SLAC1 homology model is predictive for function, including a verified placement of the identifying slac1-2 mutation G194D and a phenylalanine gate. One remaining puzzle concerns the structural change that activating phosphorylation elicits in SLAC1, and another puzzle is the biochemical role of the TehA homologs in bacteria. In a companion paper26, we examine functional and structural properties of TehA in bacteria, showing that it is anion channel, although actually not conferring tellurite resistance, and identifying a mutant variant with properties suggestive of an activated state. Thus, SLAC1 and TehA likely represent a large family of selective anion channels controlled by environmental stimuli. Chen et al. Page 8 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript METHODS Selection of target sequences TehA from E. coli (EcTehA) was a centrally selected sequence for expansion as a NYCOMPS targeted family. Target selection criteria utilized at NYCOMPS are described in details elsewhere46. Briefly, the EcTehA sequence was run against a dataset of about 40,000 predicted alpha helical integral membrane protein sequences from prokaryotic genomes (NYCOMPS98 dataset46) using PSI-BLAST50. Sequences that matched EcTehA with an E- value lower than 10−3 in an alignment extending over at least 50% of both predicted TM regions and passing our post-seed-expansion filtering criteria46 were passed to the protein production pipeline. Protein expression screening Full-length homologs from the following 38 species, including 2 sequences each from 5 of these, were amplified from genomic DNA by PCR: Thermoplasma acidophilum, Lactococcus lactis subsp. lactis Il1403, Streptococcus pyogenes M1 GAS, Streptococcus pneumoniae TIGR4, Haemophilus influenzae Rd KW20, Methanocaldococcus jannaschii DSM 2661 (2), Escherichia coli K12, Salmonella typhimurium LT2, Clostridium perfringens ATCC 13124, Xanthomonas campestris pv. campestris str. ATCC 33913, Streptococcus agalactiae 2603V/R, Shewanella oneidensis MR-1, Streptococcus mutans UA159, Archaeoglobus fulgidus DSM 4304, Vibrio parahaemolyticus RIMD 2210633 (2), Pseudomonas syringae pv. tomato str. DC3000 (2), Enterococcus faecalis V583, Pyrococcus horikoshii OT3, Bordetella bronchiseptica RB50, Streptomyces coelicolor A3, Corynebacterium glutamicum ATCC 13032, Picrophilus torridus DSM 9790, Acinetobacter sp. ADP1, Vibrio fischeri ES114, Pseudomonas fluorescens Pf-5, Sulfolobus acidocaldarius DSM 639, Colwellia psychrerythraea 34H, Neisseria meningitidis MC58, Rhodobacter sphaeroides 2.4.1, Vibrio cholerae O1 biovar eltor str. N16961 (2), Marinobacter aquaeolei VT8, Acinetobacter baumannii ATCC 17978, Klebsiella pneumoniae subsp. pneumoniae MGH 78578, Streptomyces avermitili MA-4680, Bordetella parapertussis 12822 (2), Streptococcus thermophilus LMG 18311, Salmonella enterica subsp. enterica serovar Paratyphi A str. ATCC 9150, and Anaeromyxobacter dehalogenans 2CP-C. Selected cDNAs were cloned into a modified pET vector (Novagen, Inc) that fuses a FLAG and deca-histidine tag at the C-terminus, which are cleavable by TEV protease. Proteins were expressed in E. coli BL21(DE3) plysS by a high throughput format (0.6ml in a deep well block) and purified after lysis by sonication using metal affinity purification in a buffer containing N-Dodecyl-β-D-Maltopyranoside. Samples were passed over an analytical size exclusion column in 12 different detergent-containing mobile phases, which included N- dodecyl-β-D-maltopyranoside (DDM), N-decyl-β-D-altopyranoside (DM), N-nonyl-β-D- altopyranoside (NM), N-octyl-β-D-altopyranoside (OM), N-octyl-β-D-Glucopyranoside (OG), N-nonyl-β-D-Glucopyranoside (NG) and lauryl dimethyl amine oxide (LDAO). Multi-angle light scattering with refractive index detection was used to analyze the oligomeric state51. The E. coli and H. influenzae proteins were judged to be monodisperse and stable and were passed to scale up. Chen et al. Page 9 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Scaled-up production and purification For scale up, in brief, transformed BL21 plysS cells were grown at 37°C in 2× YT media to OD 0.6-0.8 after being inoculated with 1% of the overnight culture. The culture was induced with 0.4mM IPTG and continued to grow at 37 °C for another 4 hours. The cells were harvested by centrifugation and stored at −80°C before use. Selenomethionyl (SeMet) TehA was expressed in a similar way, but using containing SeMet in place of methionine in defined minimal media. Cells were resuspended in a buffer containing 50 mM Tris-HCl (pH 8.0) and 200 mM NaCl and lysed using a French Press with two passes at 15-20,000 psi. Cell debris were removed by centrifugation at 10,000g for 20min, and the membrane fraction was isolated from that supernatant by ultra-centrifugation at 150,000g for 1 hr. The membrane fraction was homogenized in a solubilization buffer containing 50 mM Tris (pH 8.0) and 200 mM NaCl, and incubated with final concentration of 1% (w/v) dodecyl-β- D-maltopyranoside (DDM, Anatrace, Inc) for 1 hr at 4°C. The non-dissolved matter was remove by ultracentrifugation at 150,000g for 30 min, and the supernatant was loaded to a 5ml Hitrap Ni2+-NTA affinity column (GE Healthcare), pre-equilibrated with the same solubilization buffer supplemented with 0.05% DDM. After 20 column volume buffer wash, the protein was eluted with 250mM imidazole in the solublization buffer. The Flag and 10- His tags were removed by adding super TEV at 1:100 mass ratio and incubating at 4°C overnight. Tag removal was confirmed by SDS-PAGE, and the resulting sample was concentrated to around 10mg/ml. Preparative size exclusion chromatography was carried out on a Superdex-200 column for further purification, removal of TEV protease and the cleaved tag, and for buffer and detergent exchange. The gel-filtration buffer contained 10 mM Tris (pH 8.0), 200 mM NaCl, 1mM EDTA, 0.5 mM Tris [2-carboxyethyl] phosphine (TCEP), and 2×CMC of detergent. In the case of HiTehA, the protein was well behaved and stable in nearly all tested detergents, and we have purified it from DDM, DM, NM, OM, OG and LDAO. Protein characterization We performed N-terminal amino-acid sequencing of purified HiTehA and EcTehA prior to TEV protease treatment. Results from these analyses proved that true initiating methionine residue is located 14 residues after one annotated as N-terminal. The intervening nucleotide sequence contains a Shine-Delgarno sequence. For cross-linking experiments, purified HiTehA was incubated with 10 mM disuccinimidyl glutarate at room temperate for 30 min, and 100mM Tris-HCl pH 7.5 was added to stop the reaction. The incubated sample when run on a 8-25% gradient SDS-PAGE gel showed a ladder consistent with a trimeric structure. Crystallization and data collection Purified protein was concentrated to ~10mg/ml for initial crystal trials in a Mosquito robot with commercial screens from Hampton research, Emerald Biosystems and Molecular Dimension. We obtained crystals of HiTehA from protein in detergent DDM, DM, NM, OM, OG and LDAO, but only those from LDAO and OG gave diffraction to beyond 4Å spacings. Crystals that proved useful were all grown at 4°C using the sitting-drop vapor Chen et al. Page 10 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript diffusion method. After extensive optimization we reached conditions supporting very high resolution. Wild-type HiTehA and most of its variants crystallized from 1mM ZnSO4, 50mM Hepes-Na pH 7.8 and 28% PEG600. HiTehA F262A crystallized from 200mM Li2SO4, 100mM glycine pH9.3 and 33% PEG400). Addition of 10mM spermidine as an additive helped to obtain slightly better diffracting crystals. Cryoprotection was achieved by adding 5% ethylene glycol or PEG400 to the crystallization solution. Structure determination and refinement Native and SeMet SAD data set were collected at NSLS beamline X4A and processed using the software HKL200052. Crystals of HiTehA grew in space group R3 with a=b=c = 85.0Å and α=β=γ = 93.5°. All subsequent manipulations were done in the hexagonal setting of this space group with a=b= 96.0Å and c = 136.7Å. The asymmetric unit contains one TehA protomer and 65% solvent volume. The structure was determined at 2.0Å resolution from single-wavelength anomalous diffraction (SAD) using selenomethionine-substituted protein crystals. Assessment of data quality for phasing, location of heavy atom sites and initial phases were calculated using the HKL2MAP interface to SHELX programs53. All the secondary structure elements were clearly visible in the experimental electron density map. Automatic model building was done in Arp/wArp54 and completed manually in the program COOT55. The model was refined against native data at 1.20Å resolution using the program Refmac5.5 in CCP456, with anisotropic B factor restrained refinement applied. Subsequent structural analyses of mutant variants were refined as isomorphous structures. Site-directed mutagenesis Site-directed mutants were constructed using the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA 92037) and expressed from pET vectors in E. coli BL21(DE3) plysS cells as for the wild-type protein. Electrophysiology All constructs were cloned into plasmid pGHME2, linearized, and transcribed into cRNA using T7 polymerase (mMessage mMachine, Ambion). Oocytes were injected with 50 nl of cRNA solution each, at a constant concentration of 0.5 mg/ml for HiTehA, or 0.5 mg/ml for AtSLAC1 constructs, with or without 0.5 mg/ml of AtOST1 constructs and recorded in voltage-clamp experiments 2 days after injection. For mixed expression experiments, 25nl of cRNA solution was injected for each AtSLAC1 component. Two-electrode voltage clamp recordings were performed to measure HiTehA or AtSLAC1 currents as described6,7. The pipette solutions contained 3 M KCl. For voltage-clamp current recordings, the bath solution contained 1 mM MgCl2, 1 mM CaCl2, 10 mM Mes/Tris (pH 5.6) and 30 mM CsCl; For anion selectivity measurements, the bath solutions contained 50 mM Cl−, NO −3, malate, or 30 mM SO 2-3 sodium salts, 45 mM Na-gluconate, 1 mM Ca-gluconate2, 1 mM Mg- gluconate2, 1 mM K-gluconate, as well as 10 mM Tris/Mes (pH 5.6). Osmolality was adjusted with D-mannitol to 220 mOsmol/kg. The bath electrode was a 3M KCl agar bridge. The testing potentials were +50 to −110 or −130 mV in 20 mV decrements with 7.5 s duration. The holding and return potentials were 0 mV with a 1.45 s or 2 s, respectively. I-V Chen et al. Page 11 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript relations of HiTehA or AtSLAC1 were generated from the current reached after 0.5 s in testing potentials. The Goldman-Hodgkin-Katz equation was applied to estimate permeability ratios for monovalent ions as described6. For divalent anions, the permeability ratios were derived according to Fatt and Ginsborg57. Bioinformatic analysis of SLAC-related proteins Sequences related to SLAC1 were analyzed comprehensively by PSI-BLAST50. Searches at E<10−3 starting from five disparate homologs each identified a common pool of over 900 proteins, which when pooled were used for sub-classification into families and subfamilies. Details of these analyses are reported in footnotes to Table S1. Molecular figures were produced in PyMOL58. Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We thank Filippo Mancia, Larry Shapiro and Ming Zhou for helpful discussions; Patricia Rodriguez and Arianne Morrison for help with cloning and expression; Jonah Cheung, Mark Collins, Chris Min, Sheng Ye and Zhen Zhang for advice in protein chemistry and crystallography; John Schwanof and Randy Abramowitz for help with synchrotron experiments; John Riley for help with Xenopus oocyte injections; and Min Su for advice on the PSI- BLAST analysis of SLAC relatives. This project was supported in part by an award to the New York Consortium on Membrane Protein Structure (NYCOMPS) from the NIGMS Protein Structure Initiative. Beamline X4A at the National Synchrotron Light Source (NSLS), a DOE facility at Brookhaven National Laboratory, is supported by the New York Structural Biology Center. References 1. Hetherington AM, Woodward FI. The role of stomata in sensing and driving environmental change. Nature. 2003; 424:901–908. [PubMed: 12931178] 2. Sirichandra C, Wasilewska A, Vlad F, Valon C, Leung J. The guard cell as a single-cell model towards understanding drought tolerance and abscisic acid action. J. Exp. Bot. 2009; 60:1439–1463. [PubMed: 19181866] 3. Negi J, et al. CO2 regulator SLAC1 and its homologues are essential for anion homeostatsis in plant cells. Nature. 2008; 452:483–486. [PubMed: 18305482] 4. Vahisalu T, et al. SLAC1 is required for plant guard cell S-type anion channel function in stomatal signalling. Nature. 2008; 452:487–491. [PubMed: 18305484] 5. Saji S, et al. Disruption of a gene encoding C4-dicarboxylate transporter-like protein increases ozone sensitivity through deregulation of the stomatal response in Arabidopsis thaliana. Plant Cell Physiol. 2008; 49:2–10. [PubMed: 18084014] 6. Lee SC, Lan W, Buchanan BB, Luan S. A protein kinase-phophatase pair interacts with an ion channel to regulate ABA signaling in plant guard cells. Proc. Natl. Acad. Sci. USA. 2009; 106:21419–21424. [PubMed: 19955427] 7. Geiger D, et al. Activity of guard cell anion channel SLAC1 is controlled by drought-stress signaling kinase-phosphatase pair. Proc. Natl. Acad. Sci. USA. 2009; 106:21425–21430. [PubMed: 19955405] 8. Schroeder JI, Hagiwara S. Cytosolic calcium regulates ion channels in the plasma membrane of Vicia faba guard cells. Nature. 1989; 338:427–430. 9. Mustilli A, Merlot S, Vavasseur A, Fenzi F, Giraudat J. Arabidopsis OST1 protein kinase mediates the regulation of stomatal aperture by abscisic acid and acts upstream of reactive oxygen species production. Plant Cell. 2002; 14:3089–3099. [PubMed: 12468729] Chen et al. Page 12 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 10. Leung J, et al. Arabidopsis ABA response gene ABI1: features of a calcium-modulated protein phosphatase. Science. 1994; 264:1448–1452. [PubMed: 7910981] 11. Meyer K, Leube MP, Grill E. A protein phosphatase 2C involved in ABA signal transduction in Arabidopsis thaliana. Science. 1994; 264:1452–1455. [PubMed: 8197457] 12. Ma Y, et al. Regulators of PP2C phosphatase activity function as abscisic acid sensors. Science. 2009; 324:1064–1068. [PubMed: 19407143] 13. Park S, et al. Abscisic acid inhibits type 2C protein phosphatases via the PYR/PYL family of START proteins. Science. 2009; 324:1068–1071. [PubMed: 19407142] 14. Melcher K, et al. A gate-latch-lock mechanism for hormone signalling by abscisic acid receptors. Nature. 2009; 462:602–608. [PubMed: 19898420] 15. Miyazono K, et al. Structural basis of abscisic acid signalling. Nature. 2009; 462:609–614. [PubMed: 19855379] 16. Fujii H, et al. In vitro reconstitution of an abscisic acid signalling pathway. Nature. 2009; 462:660– 664. [PubMed: 19924127] 17. Yin P, et al. Structural insights into the mechanism of abscisic acid signaling by PYL proteins. Nat. Struct. Mol. Biol. 2009; 16:1230–1236. [PubMed: 19893533] 18. Nishimura N, et al. PYR/PYL/RCAR family members are major in-vivo ABI1 protein phosphatase 2C-interacting proteins in Arabidopsis. Plant J. 2010; 61:290–299. [PubMed: 19874541] 19. Grobler J, Bauer F, Subden RE, van Vuuren HJ. The MAE1 gene of Schizosaccharomyces pombe encodes a permease for malate and other C4 dicarboxylic acids. Yeast. 1995; 11:1485–1491. [PubMed: 8750236] 20. Park H, Bakalinsky AT. SSU1 mediates sulphite efflux in Saccharomyces cerevisiae. Yeast. 2000; 16:881–888. [PubMed: 10870099] 21. Léchenne B, et al. Sulphite efflux pumps in Aspergillus fumigatus and dermatophytes. Microbiology. 2007; 153:905–913. [PubMed: 17322211] 22. Walter EG, Weiner JH, Taylor DE. Nucleotide sequence and overexpression of the tellurite- resistance determinant from the IncHII plasmid pHH1508a. Gene. 1991; 101:1–7. [PubMed: 2060788] 23. Taylor DE, Hou Y, Turner RJ, Weiner JH. Location of a potassium tellurite resistance operon (tehA tehB) within the terminus of Escherichia coli K-12. J Bacteriol. 1994; 176:2740–2742. [PubMed: 8169225] 24. Daley DO, et al. Global topology analysis of the Escherichia coli inner membrane proteome. Science. 2005; 308:1321–1323. [PubMed: 15919996] 25. Unwin N. Refined structure of the nicotinic acetylcholine receptor at 4Å resolution. J. Mol. Biol. 2005; 346:967–989. [PubMed: 15701510] 26. Chen Y, Hu L, Siegelbaum SA, Hendrickson WA. Structure-based analysis of anion channel TehA and methyltransferase TehB implicated in bacterial tellurite resistance. Submitted. 27. Schmidt C, Schroeder JI. Anion Selectivity of Slow Anion Channels in the Plasma Membrane of Guard Cells (Large Nitrate Permeability). Plant Physiol. 1994; 106:383–391. [PubMed: 12232336] 28. Vahisalu T, et al. Ozone-triggered rapid stomatal response involves the production of reactive oxygen species, and is controlled by SLAC1 and OST1. Plant J. 2010; 62:442–53. [PubMed: 20128877] 29. Wright EM, Diamond JM. Anion selectivity in biological systems. Physiol. Rev. 1977; 57:109– 156. [PubMed: 834775] 30. Dutzler R, Campbell EB, MacKinnon R. Gating the selectivity filter in ClC chloride channels. Science. 2003; 300:108–112. [PubMed: 12649487] 31. Accardi A, Miller C. Secondary active transport mediated by a prokaryotic homologue of ClC Cl- channels. Nature. 2004; 427:803–807. [PubMed: 14985752] 32. Picollo A, Malvezzi M, Houtman JC, Accardi A. Basis of substrate binding and conservation of selectivity in the CLC family of channels and transporters. Nat. Struct. Mol. Biol. 2009; 16:294– 301. [PubMed: 19219045] 33. De Angeli A, et al. The nitrate/proton antiporter AtCLCa mediates nitrate accumulation in plant vacuoles. Nature. 2006; 442:939–942. [PubMed: 16878138] Chen et al. Page 13 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 34. Hiller S, Garces RG, Malia TJ, Orekhov VY, Colombini M, Wagner G. Solution structure of the integral human membrane protein VDAC-1 in detergent micelles. Science. 2008; 321:1206–1210. [PubMed: 18755977] 35. Bayrhuber M, et al. Structure of the human voltage-dependent anion channel. Proc. Natl. Acad. Sci. USA. 2008; 105:15370–15375. [PubMed: 18832158] 36. Ujwal R, et al. The crystal structure of mouse VDAC1 at 2.3 Å resolution reveals mechanistic insights into metabolite gating. Proc. Natl. Acad. Sci. USA. 2008; 105:17742–17747. [PubMed: 18988731] 37. Kouyama T, et al. Crystal structure of the light-driven chloride pump halorhodopsin from Natronomonas pharaonis. J. Mol. Biol. 2010; 396:564–579. [PubMed: 19961859] 38. Gadsby DC, Vergani P, Csanády L. The ABC protein turned chloride channel whose failure causes cystic fibrosis. Nature. 2006; 440:477–483. [PubMed: 16554808] 39. Miller PS, Smart TG. Binding, activation and modulation of Cys-loop receptors. Trends Pharmacol. Sci. 2010; 31:161–174. [PubMed: 20096941] 40. Yang YD, et al. TMEM16A confers receptor-activated calcium-dependent chloride conductance. Nature. 2008; 455:1210–5. [PubMed: 18724360] 41. Caputo A, et al. TMEM16A, a membrane protein associated with calcium-dependent chloride channel activity. Science. 2008; 322:590–594. [PubMed: 18772398] 42. Schroeder BC, Cheng T, Jan YN, Jan LY. Expression cloning of TMEM16A as a calcium- activated chloride channel subunit. Cell. 2008; 134:1019–1029. [PubMed: 18805094] 43. Schroeder JI, Keller BU. Two types of anion channel currents in guard cells with distinct voltage regulation. Proc. Natl. Acad. Sci. USA. 1992; 89:5025–5029. [PubMed: 1375754] 44. Meyer S, et al. AtALMT12 represents an R-type anion channel required for stomatal movement in Arabidopsis guard cells. Plant J. Jul 12.2010 Epub ahead of print. 45. Sasaki T, et al. Closing plant stomata requires a homolog of an aluminum-activated malate transporter. Plant Cell Physiol. 2010; 51:354–65. [PubMed: 20154005] 46. Punta M, et al. Structural genomics target selection for the New York Consortium on Membrane Protein Structure. J. Struct. Funct. Genomics. 2009; 4:255–268. [PubMed: 19859826] 47. Papadopoulos JS, Agarwala R. COBALT: constraint-based alignment tool for multiple protein sequences. Bioinformatics. 2007; 23:1073–1079. [PubMed: 17332019] 48. Landau M, et al. ConSurf 2005: the projection of evolutionary conservation scores of residues on protein structures. Nucleic Acids Res. 2005; 33:W299–W302. [PubMed: 15980475] 49. Rocchia W, et al. Rapid grid-based construction of the molecular surface for both molecules and geometric objects: applications to the finite difference Poisson-Boltzmann method. J. Comp. Chem. 2002; 23:128–137. [PubMed: 11913378] 50. Altschul SF, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997; 25:3389–3402. [PubMed: 9254694] 51. Kendrick BS, Kerwin BA, Chang BS, Philo JS. Online size-exclusion high-performance liquid chromatography light scattering and differential refractometry methods to determine degree of polymer conjugation to proteins and protein-protein or protein-ligand association states. Anal. Biochem. 2001; 299:136–146. [PubMed: 11730335] 52. Otwinowski Z, Minor W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 1997; 276:307–326. 53. Pape P, Schneider TR. HKL2MAP: a graphical user interface for phasing with SHELX programs. J. Appl. Cryst. 2004; 37:843–844. 54. Perrakis A, Morris R, Lamzin VS. Automated protein model building combined with iterative structure refinement. Nat. Struct. Biol. 1999; 6:458–463. [PubMed: 10331874] 55. Emsley P, Cowtan K. COOT: model-building tools for molecular graphics. Acta Crystallogr. D. 2004; 60:2126–2132. [PubMed: 15572765] 56. CCP4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D. 1994; 50:760– 763. [PubMed: 15299374] 57. Fatt P, Ginsborg BL. The ionic requirements for the production of action potentials in crustacean muscle fibers. J. Physiol. 1958; 142:516–543. [PubMed: 13576452] Chen et al. Page 14 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 58. DeLano, WL. The PyMOL molecular graphics system. DeLano Scientific; San Carlos, CA, USA: 2002. http://www.pymol.org Chen et al. Page 15 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 1. Sequence analysis for the SLAC1 superfamily a, Family tree. The presentation was computed by the program COBALT47 from representative subfamily sequences (Table S1), including Arabidopsis thaliana SLAC1 for SF1Ai, Haemophilus influenzae TehA for SF1Ci, Escherichia coli TehA for SF1Cii, Vibrio parahaemolyticus for SF2A, Staphylococcus aureus for SF2B, Aspergillus fumigatus Ssu1 for SF3A, and Schizosaccharomyces pombe Mae1 for SF3B. b, Structure-based sequence alignment of TehA from Haemophilus influenzae (HiTehA) and SLAC1 from Arabidopsis thaliana (AtSLAC1). The TehA structure has been used to restrict sequence gaps to inter- helical segments. Superior coils define extents of the HiTehA helical segments; red letters mark residue identities; red boxes are drawn for residues that are >95% identical within the plant subfamily SF1A for AtSLAC1 or within the TehA subfamily for HiTehA; red diamonds mark HiTehA residues that line the central pore; and the colored inferior bar encodes ConSurf sequence variability48 for the SF1 family of 204 non-redundant proteins. Chen et al. Page 16 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 2. Crystal structure of HiTehA and homology model of AtSLAC1 a, Electron density distribution from the HiTehA crystal structure at 1.2Å resolution. The map has (2Fo-Fc) coefficients based on the superimposed model. Contours are at 2.5σ. b, Ribbon diagram of the HiTehA trimer. Each protomer is colored spectrally from blue at its N-terminus to red at its C-terminus. c, DelPhi49 electrostatic potential at the extracellular surface. Electronegative and electropositive potential are colored in degrees of red and blue saturation, respectively. d, Electrostatic potential49 at the intracellular surface. e, Ribbon diagram of an HiTehA protomer viewed from outside the membrane. The ribbon is colored spectrally as in 2b. f, Ribbon diagram of an HiTehA protomer viewed from within the membrane from below the view of e. g, Surface of a homology model of AtSLAC1, viewed as in f, and colored by sequence variability48. h, Surface of AtSLAC1 as in g, but colored by electrostatic potential49. Chen et al. Page 17 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 3. Putative structure of the SLAC1 conductance pore a, Cross-section through the homology model of AtSLAC1. The model is viewed as in 2i, with the electrostatic potential49 shown on the external surface of the molecular envelope. The side chain of Phe450 is shown as a stick model (red) on the backbone ribbon colored yellow. b, Cross-section as in a, but colored by surface conservation48 as in 2h. c, Pore- lining residues in the SLAC1 homology model. A cylinder model (left), spectrally colored as in 2b, provides a key for viewing the rolled-open structure (left) with pore-lining residues of AtSLAC1 shown on the TModd helices. d, Ribbon diagrams of HiTehA TM9 (left) and TM7 (right) viewed from within the conductance pore. The side chains of Pro207 and Phe262 are shown as well as a kink-stabilizing water HOH25 that is coordinated by the NH of Gly263 and by C=O groups of Gly202 and Ala259. Density contours are shown for the water molecule. Chen et al. Page 18 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 4. Ionic conductance measurements a, Typical microelectrode voltage-clamp current traces from oocytes injected with various channel cRNAs. Left column, oocytes injected with cRNAs encoding wild-type HiTehA channels (WT), or F262A, G15D/F262A or G15D mutants. Right two columns, oocytes injected with cRNAs for WT AtSLAC1, or F450A, G194D/F450A or G194D mutants, with or without co-injection of AtOST1. Dotted lines represent zero current levels. Extracellular solution contained 30mM CsCl. Schematic keys at the left place phenyl and/or aspartyl gates. b, Effects of gating residue mutations. Mean chloride currents, measured at −90 mV, are shown comparing WT HiTehA and its mutant series F262A,G,T,V,L with WT AtSLAC1 and its corresponding series F450A,G,T,V,L, both alone and co-expressed with AtOST1. Full I-V relations are shown in Fig. S5. c, Effect of gating residues on relative AtSLAC1 anion permeabilities. Relative permeabilities (P[X]/P[Cl]) for chloride, nitrate, sulfite and malate of WT, F450A and F450T SLAC1 channels were measured from the change in current reversal potential with Cl− or anion X− as the sole permeant anion in the bath solution (Methods, Table S6). Chen et al. Page 19 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 5. Structural features at the SLAC1 homolog gate a,b, Cross-sections through the conductance pores of HiTehA F262A and HiTehA G15D. The view and presentations are as in 3a, except that helices are colored purple. c, Molecular basis for conformational strain in gating residue Phe262 of HiTehA. Helices TM1 (left), TM9 (center) and TM7 (right) are viewed from within the pore and presented as ribbon diagrams with selected side chains drawn in stick representation. The local low-energy conformation for the phenyl ring (χ2 = 90°) is shown in thin lines with short contacts indicated in dashed lines: Leu18 Cδ2 – Phe262 Cδ1, d = 2.4 Å; Val210 Cγ2 – Phe262 Cε2, d = 2.8 Å. d, Conformational shifts consequent to release of strain in gating residue Phe262. Cα backbone structures of F262A, F262V and F262L HiTehA were first superimposed onto WT HiTehA, and residues 258-266 were drawn in stereo as superimposed. All backbone atoms are shown for peptides 262±1, but only Cα atoms are shown otherwise. The WT backbone and phenyl group are green; other backbone are all magenta; side chains of Ala262, Val262 and Leu262 are cyan, blue and red, respectively; oxygen-directed bonds are red and nitrogen-directed bonds are blue. e, Typical microelectrode voltage-clamp current traces from oocytes injected with wild-type (WT) or F450L AtSLAC1 cRNA. Experimental conditions and displays are as in 4a. Chen et al. Page 20 Nature. Author manuscript; available in PMC 2013 January 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript
3M7N
archaeoglobus fulgidus exosome with RNA bound to the active site
Quantitative analysis of processive RNA degradation by the archaeal RNA exosome Sophia Hartung1, Theresa Niederberger1, Marianne Hartung2, Achim Tresch1 and Karl-Peter Hopfner1,* 1Center for Integrated Protein Sciences and Munich Center for Advanced Photonics at the Gene Center, Department of Biochemistry, Ludwig-Maximilians-University Munich, Feodor-Lynen-Strasse 25, 81377 Munich and 2General Electric - Global Research, Freisinger Landstrasse 50, 85748 Munich, Germany Received February 9, 2010; Revised March 18, 2010; Accepted March 21, 2010 ABSTRACT RNA exosomes are large multisubunit assemblies involved in controlled RNA processing. The archaeal exosome possesses a heterohexameric processing chamber with three RNase-PH-like active sites, capped by Rrp4- or Csl4-type subunits containing RNA-binding domains. RNA degradation by RNA exosomes has not been studied in a quan- titative manner because of the complex kinetics involved, and exosome features contributing to effi- cient RNA degradation remain unclear. Here we derive a quantitative kinetic model for degradation of a model substrate by the archaeal exosome. Markov Chain Monte Carlo methods for parameter estimation allow for the comparison of reaction kinetics between different exosome variants and substrates. We show that long substrates are degraded in a processive and short RNA in a more distributive manner and that the cap proteins influ- ence degradation speed. Our results, supported by small angle X-ray scattering, suggest that the Rrp4-type cap efficiently recruits RNA but prevents fast RNA degradation of longer RNAs by molecular friction, likely by RNA contacts to its unique KH-domain. We also show that formation of the RNase-PH like ring with entrapped RNA is not required for high catalytic efficiency, suggesting that the exosome chamber evolved for controlled processivity, rather than for catalytic chemistry in RNA decay. INTRODUCTION The eukaryotic and archaeal RNA exosomes and their distant relative, the bacterial degradosome, are large multiprotein assemblies that function as central cellular RNA processing and degradation machineries. The RNA exosome was originally found in yeast as an essen- tial protein complex with 30 ! 50 exonuclease activity. First, identified for the 30-processing of the yeast 5.8S ribo- somal RNA (1), the yeast RNA exosome subsequently turned out to be important for the trimming and degrad- ation of the 30-end of several nuclear RNA precursors (2). In addition, the exosome was shown to be also active in the cytoplasm by controlling mRNA turnover (3), and by its implication in various mRNA surveillance pathways like the non-sense-mediated and the non-stop decay pathways (4–7). Due to its involvement in all the different RNA processing and surveillance pathways the exosome is apparently one of the central exonucleases of a yeast cell [for reviews see for instance (8,9)]. Structural homologues of the yeast exosome were sub- sequently identified in humans, previously known as the PM-Scl (polymyositis–scleroderma overlap syndrome) complex, and in archaea (10–12). A variety of structural studies revealed a conserved architecture of exosome like complexes (13–18): exosomes consist of nine conserved core subunits, six RNase PH type subunits and three subunits with S1 and KH or zinc-ribbon domains. The six RNase-PH like domains form a ring, arranged as trimers of pseudo-dimers. In archaea, the ring is formed by three (archaeal)aRrp41:aRrp42 dimers, while human and yeast exosomes contain six different RNase PH type subunits. The archaeal exosome possesses a central chamber within the RNase PH ring which contains three phos- phorolytic active sites. The actual active site is located in the aRrp41 subunits, but the whole aRrp41:aRrp42 dimer is involved in positioning the RNA. These sites degrade single-stranded RNA (ssRNA) in a phosphate dependent manner in 30 ! 50 direction. They also catalyse the reverse reaction of adding nucleoside diphosphates to the 30-end of RNA (13), liberating inorganic phosphate. In archaea, this *To whom correspondence should be addressed. Tel: +49 89 2180 76953; Fax: +49 89 2180 76999; Email: hopfner@lmb.uni-muenchen.de 5166–5176 Nucleic Acids Research, 2010, Vol. 38, No. 15 Published online 14 April 2010 doi:10.1093/nar/gkq238  The Author(s) 2010. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/ by-nc/2.5), which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. activity has been attributed to formation of poly-A-rich tails on RNA (19). A proposed RNA entry pore at one side of the chamber restricts entry to mostly unstructured ssRNA, providing an explanation for controlled RNA deg- radation. Furthermore, three Csl4 or Rrp4 type putative RNA recognition subunits are located on top of the (Rrp41:Rrp42)3 ring and frame the proposed RNA entry pore. Current models suggest that these domains recognize RNA substrates and help to funnel them into the process- ing chamber. Although the human exosome is structurally related to the archaeal complex, including S1 and KH domain con- taining subunits (Csl4, Rrp4 and Rrp40), it has lost phosphorolytic activity (14). Instead, it gained additional ectopic subunits: the hydrolytic RNase Rrp44 (20,21) has both exonucleolytic and endonucleolytic activities, located in RNB and PIN domain subunits, respectively (14,20,22–24). A second hydrolytic RNase (Rrp6) was identified as transient part of the nuclear complex (10). Recent results indicate that despite the ectoptic placement of the nuclease active sites, RNA is still threaded through the nuclease deficient RNase PH type ring (25). A variety of groups have biochemically observed processive RNA degradation, in particular for the archaeal exosome. From structural studies, it was proposed that RNA is channelled through an entry pore between the S1 domains of aCsl4 or aRrp4 trimers into the processing chamber, where the 30-end of the RNA is pos- itioned in one of the three phosphorolytic active sites, subsequently degrading RNA base-per-base (16,26,27). The presence of the cap proteins Csl4 and Rrp4 in general increases the degradation efficiency of the exosome, but it is unclear how they do so. For instance, if the cap proteins recruit RNA, one would expect an increase in the general binding affinity. However, once RNA has entered the processing chamber, high affinity binding to the ectopic domains should slow down processive degradation. Another mechanism that is not yet understood is why processivity depends on the length of the RNA molecules (28). To address these questions and to develop means to quantitatively analyse processive degradation, we performed quantitative high-resolution RNase degradation activity assays with different variants of the Archaeoglobus fulgidus exosome. We evaluated dif- ferent kinetic models and developed a Markov Chain Monte Carlo (MCMC) analysis to fit the model to the data and derive appropriate rate constants of individual RNA degradation steps. Our data identify different struc- tural contributors to processivity, suggesting that ectopic RNA-binding domains, the entry pore and the active site are different contributors to processive degradation. The methods should be easily applicable also to other processive enzymes, including the hydrolytic nucleases of the eukaryotic exosomes. MATERIALS AND METHODS Protein expression and purification The Archaeaoglobus fulgidus RNA exosome with different Rrp4 and Csl4 caps, derivatives and mutants were expressed and purified as described (29). Site directed mu- tations were introduced using the QuickChange Site-directed Mutagenesis Kit (Stratagene) and verified by sequencing. Oligonucleotide sequences are provided in the Supplementary Table S2. Crystallization and structure determination An amount of 120 mM Csl4-exosome (Csl4:Rrp41:Rrp42)3 or its Y70ARrp42 mutant (= 27 g/l) were incubated with 400 mM RNA (3.3-fold excess, 6-mer CCCCUC) for 10 min on ice. Protein:RNA complexes were crystallized by sitting drop vapour diffusion technique by mixing 1 ml protein and 1 ml of reservoir solution (0.1 M NaAcetate, pH 4.6, 30% 3-Methyl-1,5-pentadiol (MPD), 100 mM NaCl) at 20C. Datasets were recorded at the ID-14-2 beamline (ESRF, Grenoble, France) to 2.4 A˚ (wild-type exosome) and at the PX I beamline (SLS, Villigen, Switzerland) to 3.0 A˚ (Y70ARrp42 mutant) and processed with X-ray Detector Software (XDS) (30). A model of the apo-Csl4-exosome complex (29) was used as a search model for molecular replacement using PHASER (31). Refinement to 2.4 A˚ and 3.0 A˚ , respectively was performed with CNS (32) and PHENIX (33). In the additional electron density RNA nucleotides were positioned using COOT (34). Refinement of the complete complexes was followed by iterative cycles of manual model completion with COOT and positional and B-factor refinement with CNS (Supplementary Table S1). Small angle X-ray scattering For small angle X-ray scattering (SAXS) studies, the (Rrp41:Rrp42:Rrp4)3 complex was purified as described above. To purify the exosome with endogenously bound Escherichia coli RNA the protocol was modified as follows: RNA was not washed offwith high salt, and in all buffers the salt concentration was 250 mM or lower. After the Ni-NTA affinity chromatography, the complex was loaded on an anion exchange column to remove unbound nucleic acids and the procedure was repeated to assure the total removal of free RNA. Not until only one distinct peak was eluted, the fractions were pooled, concentrated and flash frozen. The apo-complex was measured at 5, 10 and 15 mg/ml and the RNA complex was concentrated to an absorption at 280 nm of A280 = 55 and measured in a 1: 0, 1: 1 and 1: 2 dilution to evaluate the concentration dependency of scattering. Both complexes did not show concentration dependent aggregation and were not affected by long exposure to high-energy X-rays. SAXS data collection was performed in 20 mM Tris pH 7.4 and 200 mM NaCl buffer at the SIBYLS beamline (Advanced Light Source, Berkeley, CA, USA) (35). The radius of gyration was calculated using the Guinier plot in the linear region (constraint: s  Rg <1.3) and the calculation of the pair distribution function was done with GNOM within PRIMUS (36). Ab initio modelling of the solution structures was done with GASBORp (37) and more than 10 identically calculated models were aligned and averaged using DAMAVER and SUPCOMB (38). For analysis of the Nucleic Acids Research, 2010, Vol. 38, No. 15 5167 bound RNA, the protein was separated from the RNA by running the complex on a denaturing 6 M urea and 20% polyacrylamide gel and elution of the RNA from the gel. The pelleted RNA was sent to Vertis Biotechnologie AG, where the sample was poly(A)-tailed using poly(A) poly- merase followed by ligation of an RNA adapter to the 50-phosphate of the small RNAs. First-strand cDNA syn- thesis was then performed using an oligo(dT)-adapter primer and M-MLV H- reverse transcriptase. The result- ing cDNAs were PCR-amplified to 20–30 ng/ml in 19 cycles using standard Taq DNA polymerase. We cloned the cDNA products with EcoRV into pET21 vectors, transformed and amplified the plasmids and isolated and sequenced single clones. Cross-linking Site-specific crosslinking of the K37CRrp41:D143CRrp42 mutant was performed with a HBVS (1,6-Hexane-bis- vinylsulfon) crosslinker. The crosslinking reaction was performed with a 100-fold excess of crosslinker under oxygen-free conditions in a glove-box. We removed crosslinked protein from non-crosslinked protein complexes using a Superose 6 size-exclusion column, equilibrated with a running buffer containing 4 M guanidinium chloride. Protein from the peak correspond- ing to a crosslinked Rrp41/Rrp42 dimer was refolded in 50 mM Tris (pH 7.4), 200 mM NaCl, 500 mM arginine and 5% glycerol by dilution. The refolded protein was again applied onto a Superose 6 column in 50 mM Tris (pH 7.4) and 200 mM NaCl. Only the correctly refolded protein, verified by the formation of a hexamer in size exclusion chromatography, was used for further experiments. As a control sample, the same complex without crosslinker was partly denatured, purified and refolded in the same way as the crosslinked protein. RNase activity assays We carried out RNase activity assays using 32P-labelled poly(rA)-oligoribonucleotides with different lengths as substrate (26). RNA was incubated with [g-32P] ATP (Hartmann Analytics) and T4 polynucleotide kinase (NEB) for 45 min at 37C and purified by using MicroSpin G-25 columns (GE Healthcare). For each reaction, the protein (30 nM for the Csl4- and the Rrp4-capped wild-type exosome and the interface mutant; 60 nM for the cap-less exosome and the single site mutants R65ERrp41 and Y70ARrp42; 120 nM for the crosslinked cap-less exosome) was incubated with RNA in buffer containing 20 mM Tris (pH 7.8), 60 mM KCl, 10 mM MgCl2, 10% glycerol, 2 mM DTT, 0.1% PEG 8000, 10 mM NaH2PO4 (pH 7.8) and 0.8 U/ml RNasin (Promega) at 50C. Different time points were taken and the reaction was stopped by adding one volume of loading dye [0.75 g/l bromphenol blue, 0.75 g/l xylene cyanol, 25% (v/v) glycerol, 50% formamide]. The reaction products were resolved on a 20% polyacrylamide/6 M urea sequencing gel running at 50C and were analysed by phosphorimaging (GE Healthcare). The gel bands were quantified using the ImageQuant Software (GE Healthcare) and data analysis, simulation and fitting was done with MatLab (Mathworks). Models and kinetic data analysis Kinetic models are shown in Figure 3A. They are described by four parameters: association rate ka,i, dissoci- ation rate kd,i, cleavage rate kc,i and polymerization rate kp,i, one for each RNA of length i = 4, 5, . . ., 30. The cor- responding set of differential equations that quantitatively describe RNA degradation is shown in the supplement Data (Chapter 1). Since the reaction takes place in an excess of inorganic phosphate (10 mM phosphate compared to only 3.6 mM ADP at the time all RNA mol- ecules are totally degraded), we may assume no polymer- ization takes place, i.e. kp,i = 0 for all i. Consistently, we saw no synthesis of longer RNAs in our reactions. To obtain empirical estimates of the posterior parameter dis- tribution, we implemented a MCMC approach based on the Metropolis–Hastings algorithm. The key ingredients are the likelihood function, the prior, and the proposal distribution. The likelihood function penalizes the estima- tion error produced by a given model. More precisely, it penalizes the residuals, i.e. the deviation of the measured RNA amounts at each time point from the amounts that have been predicted from the current parameter set. We assume that the residuals are independent realizations of Gaussian distributions with zero mean. Since the vari- ances of these Gaussians are not known a priori, we assume a two-parameter error model with an additive and a multiplicative error component which has been proposed (39) in the context of spot quantification on arrays. We initialize the error model very conservatively (presuming large measurement errors). During the MCMC run, the error model is updated continuously by replacing it with an empirical estimate derived from the residuals that occurred in the Markov chain so far. The prior encodes prior knowledge/assumptions on the distribution of the parameters. It is sensible to require the kinetic parameters to vary smoothly with the RNA length i. This is made explicit by penalizing the difference of two successive kinetic parameters kx,i+1 and kx,i using a Gaussian prior on these differences. We em- phasize that this does not impose any restrictions on the absolute level of the parameter values. The comparison of the parameter levels obtained by different experiments is virtually unaffected by our prior choice and therefore practically unbiased. The proposal distribution generates a new parameter set as a candidate for the next MCMC step that is based on the current parameter set. We simply use a multivariate log-normal distribution with fixed diagonal covariance matrix, which is centred at the current parameter set. It turns out that the parameters of the model as stated above are not identifiable. We therefore fixed kd,i to one global constant kd, whereas the association param- eters ka,i are sampled individually. The parameters kc,i are set equal to one length-independent parameter kc. The details of this approach and its justification through extensive simulations are given in the supplementary Data (Chapters 2–4). 5168 Nucleic Acids Research, 2010, Vol. 38, No. 15 RESULTS RNA is not degraded with constant velocity Despite intense structural and biochemical research on RNA exosomes, a kinetic model, quantitative analysis of processive RNA degradation and a biochemical identifi- cation of elements that contribute to processive degrad- ation have not been studied, due to the complex kinetics involved. To address these issues, we performed RNase assays with 50-radioactively labelled 30-mer oligo(A) RNAs and the A. fulgidus Csl4- (Csl4:Rr41:Rrp42)3 and Rrp4-exosomes (Rrp4:Rr41:Rrp42)3. The reaction products and their time evolution were resolved on a denaturing sequencing gel and quantified by phosphorimaging (Figure 1), controls are shown in the supplemental material (Supplementary Figures S4 and S5). Several characteristic features of substrate degrad- ation by exosomes are revealed: First, RNA is not degraded at a constant speed, but the degradation of substrate has several phases and is distinct in different isoforms. In the Csl4 exosome (Figure 1A), after a slower first processing step, longer RNAs (>12–13 nt) are degraded very fast, seen by the low amount of intermediates in this range; shorter RNAs (<12–13 nt) are degraded slower and accumulate first before they are further degraded. On the contrary, the first processing step is faster in the Rrp4- than in the Csl4-exosome (Figure 1B). However, oligo-rA substrates >24 nt are degraded slower, intermediate substrates (24–13 nts) faster, and RNAs <13 nt slower again. This result is astonishing, considering homooligomeric se- quences are used and the effect is consequently not sequence dependent. In addition, the unexpected slow-fast-slow kinetics of the Rrp4 isoform reveals a quite complex length dependency of RNA degradation speeds. Second, the final degradation product is a 3-mer. Further degradation is extremely slow, comparable to spontaneous background hydrolytic cleavage under the present conditions. We hypothesized that features of the active site might interact specifically with the fourth base at the 30-end. Previous structural analysis with the Sulfolobus solfataricus exosome has shown that at least 4 nt are stably bound in the phosphoropytic active sites (26), but in the case of the Pyrococcus furiosus exosome some nucleotides were recognized (16). To get direct structural information for the A. fulgidus exosome:RNA interaction, used in this study, we crystallized our Csl4-exosome with a 6-mer RNA molecule (Figure 2; Supplementary Table S1). Four nu- cleotides from the 30-end are clearly visible in the unbiased Fo–Fc electron density, with weaker density for the two additional nucleotides. Interestingly, the side chain of Y70Rrp42 shows p-stacking with the fourth base (counting from the active site) and this seems to be a conserved feature among archaeal exosomes (16,26). This interaction specifically stabilizes the first 4 nt, while RNA positions+5 and+6 behind Y70Rrp42 appear not to be specifically recognized. To test the role of Y70Rrp42, we determined the co-crystal structure of the Csl4- exosome-Y70A mutant with a CCCCUC oligonucleotide. In fact, we only see clear electron density for 4 nt in the active site and the electron density at position+4 is weaker and less defined compared to the wild-type. Thus, the 3-mer as degradation end-product is likely the cause of inefficient recognition of RNA’s with <4 nt at the active site. Figure 1. Visualization of RNase activity of the archaeal exosome on denaturing polyacrylamide gels: the input (I) is a 30-mer polyA RNA radioactively labelled at the 50-end that is degraded from the 30-end to a final product (FP) of a 3-mer. Time points were taken in increasing intervals [in minutes: 0:10; 0:20; 0:30; 0:40; 0:50; 1:00; 1:10; 1:20; 1:40; 2:00; 2:20; 2:40; 3:00; 3:30; 4:00; 4:30; 5:00; 5:50; 6:00; 6:30; 7:00; 7:30; 8:00; 9:00; 10:00; 12:00; 14:00; 16:00; 18:00; 20:00; 25:00; 30:00; 35:00; 40:00; (B) ends at 8:00 min]. RNA degradation does clearly not occur with constant speed and the (Csl4:Rrp41:Rrp42)3 exosome (A) degrades RNA with a different time dependency than the (Rrp4:Rrp41:Rrp42)3 exosome (B). Nucleic Acids Research, 2010, Vol. 38, No. 15 5169 A kinetic model for processive RNA degradation by exosomes To obtain a comprehensive picture of the exosomal RNA decay, we need to analyse the reaction speeds in a quan- titative manner. The amount of RNA as function of time of intermediate i of an rA n-mer may be described by several rate constants (Figure 3A): an association rate constant ka,i of the 30-end of RNA to the active site; a corresponding dissociation rate constant kd,i; a rate of for- mation of intermediate i by cleavage of intermediate i+1, Figure 3. Three different models to describe the kinetics of RNA degradation by the exosome were tested: (A) scheme for the general kinetic model, which includes cleavage and polymerization rates kc and kp as well as association and dissociation rates ka and kd for all RNAs from 30–4 nt. (B–D) Quantified concentrations of RNA intermediates from Figure 1A, along with least square fits to different kinetic models. (B) Strict processivity considers only 27 different cleavage rates kc,30 –kc,4. (C) cleavage-and-polymerization considers 27 different cleavage rates kc,30 –kc,4, 27 different polymerization rates kp,30–kp,4 and one initial association rate ka,30 (=55 rates). With models (C) and (B), no reasonable fit could be obtained. (D) By including association, dissociation and cleavage and making rational simplifications (see text) we can convincingly fit the data with a model con- taining 28 different rate constants. Figure 2. Crystal structure of 6-mer RNA bound to the active site of the archaeal exosome. Rrp41 is shown in light and Rrp42 in dark green. The 2Fo–Fc electron density is contoured at 1.0s and only shown for the RNA and the side chain of Y70Rrp41. (A) In the wild-type exosome Y70 is stacking with the fourth base of the bound RNA, and only weak density can be seen for the fifth and sixth base. (B) Electron density for the fourth base of the RNA is much weaker in the Y70ARrp41 mutant compared to the wild-type and no density can be detected at this contour level for additional nucleotides. 5170 Nucleic Acids Research, 2010, Vol. 38, No. 15 kc,i+1; and by adenylation (polymerization) of intermedi- ate i1, kp,i-1; a rate of disappearance of intermediate i by cleavage of i, kc,i and by adenylation of i, kp,i. The system kinetics is then given by a set of differential equations (Supplementary Data). However, it is possible that this more general model can be further simplified. For instance, we likely can neglect adenylation (kp,i = 0) because our reaction conditions contain 10 mM phosphate compared to only 3.6 mM ADP at the time all RNA molecules are totally degraded, strongly shifting the reversible reaction towards degradation. In addition, the exosome might be strictly processive, i.e. association and dissociation rates of RNA intermediates are negligible compared to the cleavage rates (ka,i = kd,i = 0). Furthermore, all cleavage rates may be independent of the length of RNA, because they could be a local active site property (kc,i = kc,j). Hence we analysed three simplified models (Figure 3B–D). Once initial values for the rate constants, enzyme concentration and RNA substrate concentration (rA 30-mer) are provided, this corresponding set of differ- ential equations can be used to calculate the concentra- tions of all RNA intermediates over time. We then minimized the resulting least square deviations between the calculated and experimental concentrations of reaction intermediates by optimizing the rate constants using the ‘fminsearch’ parameter optimization procedure as implemented in Matlab. Using the ‘strict processivity’ model with 27 independ- ent variables (ka,i = kd,i = kp,i = 0) (Figure 3B), we obtained no reasonable fit of the experimental data. A second model including the adenylation reaction (55 inde- pendent variables, ka,i = kd,i = 0) could also not properly interpret the data (Figure 3C). Thus, simply adding more parameter does not automatically lead to reasonable fits and the RNA degradation activity cannot be convincingly explained by strict processivity. Consequently, we added association and dissociation of RNA intermediates to the equations and used the following alternative simplifica- tions: (i) adenylation is omitted (kp,i = 0; see above); (ii) the same cleavage rate is used for all RNA molecules (kc,i = kc,j), i.e. cleavage rate is a local active site property and not dependent on RNA length. We estimated starting values for kc and validated this simpli- fication from analysis of the initial exponential decay of RNAs substrates with different initial lengths (data not shown). (iii) Due to our experimental approach, we cannot experimentally distinguish between bound and free RNA since the gel bands represent the sum of free and exosome-bound RNA intermediates of length i. For that reason, we cannot reconstruct dissociation-, associ- ation- and cleavage rate constants independently of each other. Consequently, we do not treat the association and the dissociation rate constants independently, but analyse the ratio of ka,i/kd,i by setting kd,i’s to a constant low value, leaving ka,i free to vary. Variation of the value for kd did not result in significant changes in the analysis (Supplementary Data). These three reasonable simplifica- tions leave essentially one free parameter per intermediate plus one overall cleavage rate constant. Although this model has less degrees of freedom than the second model (55 versus 28), it can convincingly interpret the ex- perimental data for the both Csl4 and Rrp4 exosome variants and most mutants (Figure 3D). MCMC analysis of degradation To address the problem of multidimensional parameter fitting and to assess the variance in parameter esti- mation, we established MCMC simulations. Because of the difficulty in determination of separate values for the single rate constants, we defined an RNA length- dependent quantity vi vi ¼ kc,i Km,i ð1Þ with Km,i the Michaelis–Menten constant Km,i ¼ kc,i+kd,i ka,i ð2Þ vi is called ‘catalytic efficiency’ or ‘specificity constant’, as it is a measure of the velocity of RNA intermediate i deg- radation by the exosome. We are now in a position to test exosome features important for vi. We observe that for the Csl4 exosome, vi is highly dependent on the RNA length: the initial RNA processing step, likely determined by the initial association of RNA with the exosome, is generally slow. Once RNA is bound, vi is large and relatively constant for RNA lengths >13 nt. vi then progressively decreases for RNA lengths <13 nt until the final 3-mer appears (Figures 4B and 5A). This length dependency may be explained by the exosome structure: RNA molecules longer than 13 nt might still reach through the ‘neck’, and this topological interaction will induce a higher ‘local concentration’ of RNA at the active site with increased vi. Short RNAs, on the contrary, will lose this contact and due to their smaller size more easily diffuse out of the processing chamber, therefore decreasing vi. To test this idea, we analysed the Y70A mutant of the Csl4-exosome. The length profile of the catalytic efficiency has a similar shape than for the wild-type, although the catalytic efficiency is lower for all RNA intermediates (Figure 5B). For RNAs >13 nt, the difference in vi is 2- to 3-fold (about one log unit). However, the drop in vi for RNAs <13 nt is progressively more pronounced compared to the wild-type and towards short RNAs (<8 nt), the mutant is 20- to 150-fold (three to five log units) slower than the wild-type. This is consistent with the idea that for long RNAs the neck provides additional interaction and thus overcomes in part the destabilizing effect of Y70A. For shorter RNAs, the active site becomes the sole attachment, leading to a rapid drop of catalytic efficiency in the Y70A mutant. We also analysed the ‘neck’ mutant R65E, which has been shown to severely reduce exosome activity (16,26,27). This mutant exhibited a substantially delayed onset of degradation, presumably because RNA is unable to effi- ciently enter the active site (data not shown). A likely reason is the formation of non-productive RNA:protein complexes with RNA trapped on the outside of the exosome (29). At present, our model cannot deal with Nucleic Acids Research, 2010, Vol. 38, No. 15 5171 this scenario and we could not convincingly include—as only variant—the R65E mutant in the analysis. However, the data of the analysis of R65E are provided in the sup- plementary Figure S15 and following. Role of exosome ring formation and ring dynamics for RNA binding Although the initial binding of RNA appears slow, it seems unlikely that the 30-end directly finds its way through the small hole in the neck. It is perhaps more likely that the ring structure ‘breathes’—as observed e.g. in hexameric helicases—and allows some lateral entry at the neck. To explore this idea we analysed a crosslinked exosome, where the ring is rigidified by three site specific crosslinks, and a mutant that disrupts the ring structure into Rrp41:Rrp42 pairs. We compared these isoforms with the corresponding wild-type, the cap-less hexameric (Rrp41:Rrp42)3 ring (Figure 5C). From the structural analyses, it was observed that the Rrp41 and Rrp42 subunits possess two interfaces. One interface is larger, and characterized by contiguous b-sheets between Rrp41 and Rrp42 (40). The other interface is smaller, presumably more dynamic, and was chosen for the crosslinking and mutagenesis analysis (Supplementary Figure S2). (Rrp41:Rrp42)3 exhibit a biphasic length dependence of vi, similar to the Csl4-exosome. However, the catalytic efficiency of (Rrp41:Rrp42)3 is 5- to 10-fold higher Figure 5. Comparisons of the catalytic efficiency vi of different exosome variants versus RNA lengths: (A) differences in the cap proteins influence catalytic activity. This is shown by comparison of vi from the cap-less exosome (Rrp41:Rrp42)3 in magenta, the Csl4 capped exosome (Csl4:Rrp41:Rrp42)3 in red and the Rrp4 capped exosome (Rrp4:Rrp41:Rrp42)3 in blue. (B) Tyr70Rrp42 close to the active site is especially important to efficiently degrade small RNAs. The wild-typ Csl4 exosome is shown in red and the Y70ARrp42 mutant in green. (C) The role of the ring architecture and dynamics for cata- lytic activity is shown by comparing wild-type cap-less exosome (Rrp41:Rrp42)3 in magenta with the dimeric and open interface mutant (Rrp41:Rrp42)1 and a rigidified crosslinked variant that is less dynamic in yellow. A total of 1000 parameter sets have been randomly drawn from the stationary phase of the Markov chain. Thus for each RNA length and each timepoint, we obtained 1000 estimates whose distribution is displayed by boxplots. Figure 4. Catalytic efficiency vi for all RNA intermediates present during the degradation of a 30-mer RNA by the Csl4-Rrp41-Rrp42 exosome was determined with MCMC simulations. (A) shows the traceplot and (B) the final parameter set (burnin = 150 000). It can be seen that the MCMC chains vary in convergence speed as well as in variability. The boxplots in (B) illustrate the main advantage of the MCMC approach: it not only offers a set of parameters that best describe the measured data, but it also yields a posterior distribution for each catalytic efficiency parameter and thus provides a more com- prehensive summary of the data. 5172 Nucleic Acids Research, 2010, Vol. 38, No. 15 across the RNA spectrum than that of the Csl4-exosome, indicating that the Csl4 cap subunits do not substantially promote degradation of this model substrate. In addition, vi drops for RNAs <13 nt even for the (Rrp41:Rrp42)3 particle, indicating that the neck not e.g. the S1 domains of caps are responsible for the higher catalytic efficiency on longer RNAs. To explore the effect of the hexameric ring formation, we mutated Lys51 to Glu, located in the ‘smaller’ interface between alternating Rrp41 and Rrp42 pairs. This resulted in stable Rrp41:Rrp42 dimers that do not assemble into hexamers anymore (Supplementary Figure S3). The Rrp41:Rrp42 dimers exhibit catalytic efficiencies that are only slightly lower than the (Rrp41:Rrp42)3 particle for RNAs >13 nt, and almost identical to the corresponding hexamers for RNAs <13 nt. As a result, the drop around 13 nt from a faster to a slower degradation is much less pronounced in the Rrp41:Rrp42 dimer, further supporting the idea that encapsulation in the neck is responsible for higher degradation speeds. The relatively high activity of the dimers is possibly also a result of the effective ‘tripli- cation’ of active sites, i.e. only one RNA molecule can be degraded by a (Rrp41:Rrp42)3, while three RNA mol- ecules can be degraded by three Rrp41:Rrp42 dimers. In addition, while RNA probably dissociates faster from the dimers, this effect could be compensated by a faster asso- ciation of RNA to the readily accessible active sites in the open dimers. The opposite is observed, when the ring structure is crosslinked. We introduced cysteines on the outside of the RNase-PH ring and crosslinked the three Rrp41:Rrp42 dimers via a thiol specific bifunctional crosslinker. This procedure resulted in a hexameric RNase PH ring with wild-type-like size and shape accord- ing to gel filtration (Supplementary Figure S3). Comparison of ni between the crosslinked isoform with the (Rrp41:Rrp42)3 hexamer, revealed a dramatically reduced ni (500- to 2000-fold) indicating that rigidifying the exosome by the crosslink severely affects catalytic ef- ficiency. We cannot formally rule out that the crosslinking affects activity by other means, but considering that the hexamer disrupting mutation at the same interface does not severely reduce activity, a plausible scenario is that the rigidified exosome does not allow efficient association with RNA anymore. Thus, taken together, the self-compartmentalization of exosomes is probably not an evolution for high activity, but rather for controlled RNA degradation. Effect of the cap structures To learn about the role of the cap proteins in exosome activity, we compared the rate constants of the cap-less exosome with the Csl4 and the Rrp4 exosome. The initial rates for degradation of the 30-mer rA are similar for the Csl4-exosome and the cap-less version, but consid- erably faster for the Rrp4-exosome (Figure 5A). This in- dicates that cap proteins can influence recognition and recruitment of RNA substrates and that this step is more efficient for the Rrp4 exosome. However, while RNA degradation for medium and short RNAs is quite comparable between the Csl4 and Rrp4 exosomes, there is an interesting difference for long RNA molecules (>24 nt). The Rrp4 exosome is quite slower for RNAs >24 nt, faster for RNA between 24 and 13 nt, and then progressively slower for RNAs <13 nt. This remarkable length depen- dency is clearly evident in degradation profiles (Figure 1). The most likely explanation is that long RNAs might still have contacts with Rrp4, where a more specific binding site holds them partially back from rapid degradation. In principle, this could be viewed as molecular friction. When RNAs are shorter, they loose contact to Rrp4 and degrad- ation speed is increased. The Csl4 protein and the Rrp4 protein differ in their domain structures. While Csl4 contains a Zn-ribbon domain, Rrp4 possesses a KH-domain, which is a typical RNA-binding domain and could recognize the oligo-rA. Such a binding could be responsible for the faster first degradation step, because it more efficiently sequesters RNAs on the exosome surface, but may subsequently slow down degradation until RNAs are too short to maintain simultaneous contacts at the KH domain and active site. However, the Rrp4 isoform is more efficient for smaller RNA species than the Csl4 and capless isoforms. Since these shorter RNAs cannot form dual contacts with the active site and outside the caps, the Rrp4 could also influ- ence the dynamics or other properties of the RNase-PH ring, for instance to help in loading of RNA into the ring structure. SAXS structure of the Rrp4 exosome with endogenously purified bacterial RNA To explore the role of the Rrp4 cap in efficiently recruiting RNAs further, we performed SAXS studies with a nuclease deficient nine-subunit Rrp4 exosome bound to RNA: we had noticed that this nuclease deficient Rrp4-exosome (D180A in Rrp41) very efficiently co-purifies with E. coli RNA. To determine the kind of RNA that binds to the exosome we run it on a denaturing gel together with RNAs with known sizes and could estimate the size of the RNA to be between 55 and 65 nt (Supplementary Figure S1). Cloning and sequencing of bound RNA molecules revealed a set of much shorter in- homogeneous mixed sequences (Supplementary Table S3). It is possible that the bound RNAs are a mixture of various mRNAs from E. coli, although the isolated RNA is larger than the identified sequences and it is possible that highly structured RNAs such as tRNAs are underrepresented due to inefficient amplification and cloning. Comparison of the SAXS structure of apo-Rrp4-exosome with the RNA bound complex shows an increase in the radius of gyration from 39.6 A˚ to 46.8 A˚ when RNA is bound and the corresponding pair distribu- tion functions contains longer vectors (Figure 6A), likely because additional scattering elements from RNA protrude from the compact protein core. The resulting ab initio model of the complex overlaid with the crystal structure of the apo-complex clearly indicates additional mass from the bound RNA (Figure 6B and C). This clear additional mass is distributed in the centre of the cap structure on top of the neck region but also protrudes Nucleic Acids Research, 2010, Vol. 38, No. 15 5173 away from the complex. When looking at the overlay with the crystal structure it appears that the RNA is bound at the KH and the S1-domains. The SAXS analysis supports the model that RNA binds near the KH-domain on the outside of the caps and reveals a low-resolution image of trapped exosome–RNA complexes. DISCUSSION RNA exosomes are large, self-compartmentalized nucle- ases, implicated in processive, controlled degradation of a large variety of RNAs. While the archaeal exosome possesses three phosphorolytic active sites within the com- partment, the eukaryotic exosomes apparently have lost this activity but adopted hydrolytic RNase subunits that are bound at the outside of the evolutionary conserved core. Nevertheless, recent data suggest that RNA is still threaded through the eukaryotic core exosome before it is degraded in ectopic hydrolytic active sites, suggesting that the core particle retained critical ‘structural’ functions re- garding RNA degradation such as increased processivity or controlled RNA degradation (25). To be able to quantitatively address RNA exosome activities, we derived a kinetic model for the complex RNA degradation of the archaeal RNA exosome using Markov Chain Monte Carlo analysis. The kinetic model gives a realistic assessment of the velocity of the exosome and mutant variants during processive degradation of a rA 30-mer oligonucleotide. The considerable effort we had to put into the MCMC simulation pays offeventually. We are now able to derive a realistic joint posterior distribu- tion of kinetic parameters, enabling us to quantify the relation of different parameters in either the same or in distinct exosome mutants. This would have been impos- sible with a conventional least squares fit of the data, which produces very unstable parameter estimates (see Supplementary Data for a comparison), although the obtained fits are very good (Figure 3D). With this in hand, we find several interesting and unex- pected features of RNA degradation activities. First, kinetic evaluation of RNA degradation of exosomes needs to include association and dissociation rate con- stants. Thus the kinetics cannot be treated as strictly processive, at least for RNA species in the assessed Figure 6. SAXS structure of the Rrp4 exosome with endogenously purified bacterial RNA. (A) SAXS data of the Rrp4 exosome (green) and the Rrp4 exosome with RNA (orange) (curves show the scattering intensity I(q) as a function of the scattering angle 2y and X-ray wavelength , where q = (4sin/y)) and the pair-distribution function describing intramolecular distances; in the presence of RNA longer distances occur and the radius of gyration increases. (B) Average of 10 independent ab initio models for the apo exosome and the RNA-bound complex superimposed with the crystal structure. The additional density for the RNA is clearly visible. 5174 Nucleic Acids Research, 2010, Vol. 38, No. 15 length range. This does not necessarily imply that RNA dissociates and rebinds completely from the exosome. Longer RNAs may be retained within the neck as well as cap domains, while binding and dissociating from the active site in the processing chamber. The association and dissociation constants can thus be understood as ‘ratcheting’ constants that influence the rate of translation along the RNA to and from the active site. For short substrates that are unable to simultaneously bind neck and active site this connection is lost, the dissociation in- creases, degradation speeds drop and the exosome changes from being fast and processive to a slower distributive enzyme. Our results also show how neck region and active site features contribute to exosome activity. Although we could not quantitatively address the importance of Arg65 in the neck with the simplified model in hand, this residue appears to be important for loading RNA into the processing chamber, but not for efficient degrad- ation once RNA is bound. This conclusion is derived from the observation that while the initial degradation is sub- stantially delayed, the appearance of smaller RNA species is qualitatively similar to the wild-type Csl4 exosome. Taken together with the observation that crosslinking severely reduces processing and the RNase PH ring needs to breath or display some conformational dynamics, it is unlikely that RNA is simply threaded into the processing chamber like a yarn through the eye of a needle. Rather, we propose that initial RNA binding includes some lateral entry near the neck. We are also in the position now to address the influence of the cap proteins Rrp4 and Csl4. These proteins possess a variety of domains with unclear function in exosome activity. While eukaryotic exosomes have defined heterotrimeric caps, the stoichiometry of cap proteins in archaeal exosomes is not defined in vitro and perhaps variable in vivo. For the archaeal exosome, the Csl4 capped isoform displays similar degradation kinetics to that of the capless variant, and the function of this type of cap remains to be shown. However, the Rrp4 isoform substantially differs from the other two variants and our analysis suggests that Rrp4 more efficiently recruits RNA to the exosome. In fact, RNA from the heterologous expression in E. coli is very tightly bound to the Rrp4 exosome. It must be noted that the gene coding for Rrp4 is in the same operon as genes for Rrp41 and Rrp42, indicating that this cap is perhaps a ‘default’ isoform of the exosome, while the Csl4 cap, located else- where in the genome, might be differentially regulated. The cap structures, however, also influence the degrad- ation of short RNAs. This is to some extent surprising, since short RNAs (<13 nt) cannot bind to the caps and the active site at the same time. However, the Rrp4 subunit more intimately interacts with the RNase PH ring than the Csl4 protein and might influence also the dynamics of the RNase PH type ring. Likewise, binding of RNA to the KH domains, consistent with the lateral density of RNA in the SAXS models, may position it better for loading into the processing chamber. In sum, we present here a robust method to analyse the complex degradation kinetics of a partially processive degradation enzyme in a quantitative manner, with esti- mates of the posterior distribution of the model param- eters. We applied this analysis to degradation of RNA by several isoforms and variants of the archaeal exosome and could reveal a variety of features that are important for catalytic efficiency. The objective of this manuscript is to derive a general method that can now be used to unravel the biochemistry of exosomes in a more quantitative manner. The method can now form a basis for compre- hensive analysis of different substrates, other RNA se- quences, as well as mutants of this system or related systems such as the eukaryotic exosome. ACCESSION NUMBERS 3M7N, 3M85. SUPPLEMENTARY DATA Supplementary Data are available at NAR Online. ACKNOWLEDGEMENTS The authors thank Christian Luginsland for help in protein purification, Katharina Bu¨ ttner for exosome con- structs and Katja Lammens and Gregor Witte for helpful discussions. The authors thank the staffof the European Synchrotron Radiation Facility (beamline 14–2) and the Swiss Light Source (beamline PX I) for help with diffrac- tion data collection and Michal Hammel from the Advanced Light Source (SIBYLS beamline) for help with scattering data collection. FUNDING Deutsche Forschungsgemeinschaft (HO2489/3 and SFB646); Center for Integrated Protein Science Munich. Funding for open access charge: Deutsche Forschungsgemeinschaft. Conflict of interest statement. None declared. REFERENCES 1. Mitchell,P., Petfalski,E., Shevchenko,A., Mann,M. and Tollervey,D. (1997) The exosome: a conserved eukaryotic RNA processing complex containing multiple 30–>50 exoribonucleases. Cell, 91, 457–466. 2. Allmang,C., Kufel,J., Chanfreau,G., Mitchell,P., Petfalski,E. and Tollervey,D. (1999) Functions of the exosome in rRNA, snoRNA and snRNA synthesis. EMBO J., 18, 5399–5410. 3. Anderson,J.S. and Parker,R.P. (1998) The 30 to 50 degradation of yeast mRNAs is a general mechanism for mRNA turnover that requires the SKI2 DEVH box protein and 30 to 50 exonucleases of the exosome complex. EMBO J., 17, 1497–1506. 4. Isken,O. and Maquat,L.E. (2007) Quality control of eukaryotic mRNA: safeguarding cells from abnormal mRNA function. Genes Dev., 21, 1833–1856. 5. Thompson,D.M. and Parker,R. (2007) Cytoplasmic decay of intergenic transcripts in Saccharomyces cerevisiae. Mol. Cell. Biol., 27, 92–101. 6. Mitchell,P. and Tollervey,D. (2003) An NMD pathway in yeast involving accelerated deadenylation and exosome-mediated 30–>50 degradation. Mol. Cell., 11, 1405–1413. Nucleic Acids Research, 2010, Vol. 38, No. 15 5175 7. Vasudevan,S., Peltz,S.W. and Wilusz,C.J. (2002) Non-stop decay– a new mRNA surveillance pathway. Bioessays, 24, 785–788. 8. Houseley,J., LaCava,J. and Tollervey,D. (2006) RNA-quality control by the exosome. Nat. Rev. Mol. Cell. Biol., 7, 529–539. 9. Schmid,M. and Jensen,T.H. (2008) The exosome: a multipurpose RNA-decay machine. Trends Biochem. Sci., 33, 501–510. 10. Allmang,C., Petfalski,E., Podtelejnikov,A., Mann,M., Tollervey,D. and Mitchell,P. (1999) The yeast exosome and human PM-Scl are related complexes of 30 –> 50 exonucleases. Genes Dev., 13, 2148–2158. 11. Koonin,E.V., Wolf,Y.I. and Aravind,L. (2001) Prediction of the archaeal exosome and its connections with the proteasome and the translation and transcription machineries by a comparative-genomic approach. Genome Res., 11, 240–252. 12. Evguenieva-Hackenberg,E., Walter,P., Hochleitner,E., Lottspeich,F. and Klug,G. (2003) An exosome-like complex in Sulfolobus solfataricus. EMBO Rep., 4, 889–893. 13. Bu¨ ttner,K., Wenig,K. and Hopfner,K.P. (2006) The exosome: a macromolecular cage for controlled RNA degradation. Mol. Microbiol., 61, 1372–1379. 14. Liu,Q., Greimann,J.C. and Lima,C.D. (2006) Reconstitution, activities, and structure of the eukaryotic RNA exosome. Cell, 127, 1223–1237. 15. Lorentzen,E., Basquin,J., Tomecki,R., Dziembowski,A. and Conti,E. (2008) Structure of the active subunit of the yeast exosome core, Rrp44: diverse modes of substrate recruitment in the RNase II nuclease family. Mol. Cell., 29, 717–728. 16. Navarro,M.V., Oliveira,C.C., Zanchin,N.I. and Guimaraes,B.G. (2008) Insights into the mechanism of progressive RNA degradation by the archaeal exosome. J. Biol. Chem., 283, 14120–14131. 17. Ramos,C.R., Oliveira,C.L., Torriani,I.L. and Oliveira,C.C. (2006) The Pyrococcus exosome complex: structural and functional characterization. J. Biol. Chem., 281, 6751–6759. 18. Wang,H.W., Wang,J., Ding,F., Callahan,K., Bratkowski,M.A., Butler,J.S., Nogales,E. and Ke,A. (2007) Architecture of the yeast Rrp44 exosome complex suggests routes of RNA recruitment for 30 end processing. Proc. Natl Acad. Sci. USA, 104, 16844–16849. 19. Portnoy,V., Evguenieva-Hackenberg,E., Klein,F., Walter,P., Lorentzen,E., Klug,G. and Schuster,G. (2005) RNA polyadenylation in Archaea: not observed in Haloferax while the exosome polynucleotidylates RNA in Sulfolobus. EMBO Rep., 6, 1188–1193. 20. Dziembowski,A., Lorentzen,E., Conti,E. and Seraphin,B. (2007) A single subunit, Dis3, is essentially responsible for yeast exosome core activity. Nat. Struct. Mol. Biol., 14, 15–22. 21. Schneider,C., Anderson,J.T. and Tollervey,D. (2007) The exosome subunit Rrp44 plays a direct role in RNA substrate recognition. Mol. Cell., 27, 324–331. 22. Liu,Q., Greimann,J.C. and Lima,C.D. (2007) Reconstruction, activities, and structure of the eukaryotic RNA exosome. Erratum. Cell, 131, 188–189. 23. Lebreton,A., Tomecki,R., Dziembowski,A. and Seraphin,B. (2008) Endonucleolytic RNA cleavage by a eukaryotic exosome. Nature, 456, 993–996. 24. Schaeffer,D., Tsanova,B., Barbas,A., Reis,F.P., Dastidar,E.G., Sanchez-Rotunno,M., Arraiano,C.M. and van Hoof,A. (2009) The exosome contains domains with specific endoribonuclease, exoribonuclease and cytoplasmic mRNA decay activities. Nat. Struct. Mol. Biol., 16, 56–62. 25. Bonneau,F., Basquin,J., Ebert,J., Lorentzen,E. and Conti,E. (2009) The yeast exosome functions as a macromolecular cage to channel RNA substrates for degradation. Cell, 139, 547–559. 26. Lorentzen,E. and Conti,E. (2005) Structural basis of 30 end RNA recognition and exoribonucleolytic cleavage by an exosome RNase PH core. Mol. Cell., 20, 473–481. 27. Lorentzen,E., Dziembowski,A., Lindner,D., Seraphin,B. and Conti,E. (2007) RNA channelling by the archaeal exosome. EMBO Rep., 8, 470–476. 28. Walter,P., Klein,F., Lorentzen,E., Ilchmann,A., Klug,G. and Evguenieva-Hackenberg,E. (2006) Characterization of native and reconstituted exosome complexes from the hyperthermophilic archaeon Sulfolobus solfataricus. Mol. Microbiol., 62, 1076–1089. 29. Bu¨ ttner,K., Wenig,K. and Hopfner,K.P. (2005) Structural framework for the mechanism of archaeal exosomes in RNA processing. Mol. Cell., 20, 461–471. 30. Kabsch,W. (1993) Automatic data processing of rotation diffraction data from crystals of initially unkonwn symmetry and cell constants. J. Appl. Cryst., 26, 795–800. 31. CCP4. (1994) The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D Biol. Crystallogr., 50, 760–763. 32. Brunger,A.T., Adams,P.D., Clore,G.M., DeLano,W.L., Gros,P., Grosse-Kunstleve,R.W., Jiang,J.S., Kuszewski,J., Nilges,M., Pannu,N.S. et al. (1998) Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr. D Biol. Crystallogr., 54, 905–921. 33. Afonine,P.V., Grosse-Kunstleve,R.W. and Adams,P.D. (2005) A robust bulk-solvent correction and anisotropic scaling procedure. Acta Crystallogr. D Biol. Crystallogr., 61, 850–855. 34. Emsley,P. and Cowtan,K. (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr., 60, 2126–2132. 35. Hura,G.L., Menon,A.L., Hammel,M., Rambo,R.P., Poole,F.L., 2nd, Tsutakawa,S.E., Jenney,F.E. Jr, Classen,S., Frankel,K.A., Hopkins,R.C. et al. (2009) Robust, high-throughput solution structural analyses by small angle X-ray scattering (SAXS). Nat. Methods, 6, 606–612. 36. Konarev,P.V., Volkov,V.V., Sokolova,A.V., Koch,M.H.J. and Svergun,D.I. (2003) PRIMUS: a Windows PC-based system for small-angle scattering data analysis. J. Appl. Cryst., 36, 1277–1282. 37. Svergun,D.I., Petoukhov,M.V. and Koch,M.H. (2001) Determination of domain structure of proteins from X-ray solution scattering. Biophys. J., 80, 2946–2953. 38. Volkov,V.V. and Svergun,D.I. (2003) Uniqueness of ab initio shape determination in small-angle scattering. J. Appl. Cryst., 36, 860–864. 39. Rocke,D.M. and Durbin,B. (2001) A model for measurement error for gene expression arrays. J. Comput. Biol., 8, 557–569. 40. Lorentzen,E., Walter,P., Fribourg,S., Evguenieva-Hackenberg,E., Klug,G. and Conti,E. (2005) The archaeal exosome core is a hexameric ring structure with three catalytic subunits. Nat. Struct. Mol. Biol., 12, 575–581. 5176 Nucleic Acids Research, 2010, Vol. 38, No. 15
3M7W
Crystal Structure of Type I 3-Dehydroquinate Dehydratase (aroD) from Salmonella typhimurium LT2 in Covalent Complex with Dehydroquinate
Insights into the Mechanism of Type I Dehydroquinate Dehydratases from Structures of Reaction Intermediates* Received for publication,October 8, 2010, and in revised form, October 28, 2010 Published, JBC Papers in Press,November 18, 2010, DOI 10.1074/jbc.M110.192831 Samuel H. Light‡§, George Minasov‡§, Ludmilla Shuvalova‡§, Mark-Eugene Duban‡, Michael Caffrey¶, Wayne F. Anderson‡§, and Arnon Lavie¶1 From the ‡Center for Structural Genomics of Infectious Diseases and §Department of Molecular Pharmacology and Biological Chemistry, Feinberg School of Medicine, Northwestern University, Chicago, Illinois 60611 and the ¶Department of Biochemistry and Molecular Genetics, University of Illinois, Chicago, Illinois 60607 The biosynthetic shikimate pathway consists of seven en- zymes that catalyze sequential reactions to generate choris- mate, a critical branch point in the synthesis of the aromatic amino acids. The third enzyme in the pathway, dehydro- quinate dehydratase (DHQD), catalyzes the dehydration of 3-dehydroquinate to 3-dehydroshikimate. We present three crystal structures of the type I DHQD from the intestinal pathogens Clostridium difficile and Salmonella enterica. Structures of the enzyme with substrate and covalent pre- and post-dehydration reaction intermediates provide snapshots of successive steps along the type I DHQD-catalyzed reaction coordinate. These structures reveal that the position of the substrate within the active site does not appreciably change upon Schiff base formation. The intermediate state structures reveal a reaction state-dependent behavior of His-143 in which the residue adopts a conformation proximal to the site of cata- lytic dehydration only when the leaving group is present. We speculate that His-143 is likely to assume differing catalytic roles in each of its observed conformations. One conformation of His-143 positions the residue for the formation/hydrolysis of the covalent Schiff base intermediates, whereas the other conformation positions the residue for a role in the catalytic dehydration event. The fact that the shikimate pathway is ab- sent from humans makes the enzymes of the pathway potential targets for the development of non-toxic antimicrobials. The structures and mechanistic insight presented here may inform the design of type I DHQD enzyme inhibitors. Present in bacteria, fungi, and plants but absent in higher eukaryotes, the seven enzymes of the shikimate pathway cata- lyze sequential reactions to generate chorismate. Chorismate serves as a precursor of many biologically important aromatic compounds including the ubiquinones, folates, and aromatic amino acids (1). The essential nature of these proteins in a number of species, in combination with an absence of human homologs, makes the shikimate pathway an attractive target for the development of non-toxic antimicrobials, anti-fungals, and herbicides (2–4). Given the complexities inherent to in- hibitor design, a comprehensive understanding of the struc- tural and mechanistic framework underlying the function of the shikimate pathway enzymes should aid in the develop- ment of novel shikimate-targeting inhibitors. The third step in the shikimate pathway, consisting of the dehydration of dehydroquinate to dehydroshikimate (Fig. 1), can be catalyzed by two unrelated enzymes, termed type I and type II dehydroquinate dehydratases (DHQDs).2 These two enzyme families lack sequence or structural homology and employ distinct reaction mechanisms (5–10). The type I DHQDs utilize a covalent Schiff base (imine) intermediate that results in a cis-elimination, whereas the type II reaction lacks a covalent intermediate and undergoes a trans-elimina- tion (5–10). The phylogenetic distribution of the two enzyme types is somewhat disorderly, with closely related species of- ten possessing different types. In general, the type I enzyme is found in plants and fungi as a domain within a multifunc- tional protein and in some bacteria as an 29-kDa mono- functional protein that assembles into a homo-dimer. In con- trast to the type I DHQD, the type II enzyme is found within a mostly non-overlapping subset of bacteria and exists as an 17-kDa protein that assembles into a homo-dodecamer (8, 10–12). Previously reported structures of the Salmonella typhi, Staphylococcus aureus, and Archaeoglobus fulgidus type I DHQD have characterized the enzyme in an apo and co- valently bound post-dehydration intermediate state (8, 13– 15). Here we present crystal structures of the type I DHQD from the two intestinal pathogens, the Gram-positive Clos- tridium difficile (cdDHQD) and Gram-negative Salmonella enterica (seDHQD), and characterize previously unobserved substrate and pre-dehydration reaction intermediate bound states of the enzyme. The similar mode of substrate and reac- tion intermediate binding and the reaction state-dependent behavior of a conserved active site histidine are identified, and their functional implications are discussed. EXPERIMENTAL PROCEDURES Cloning, Protein Overexpression, and Purification—Follow- ing published protocols (16), seDHQD and cdDHQD genes * This work was supported, in whole or in part, by National Institutes of Health Contract HHSN272200700058C (to W. F. A.) through the NIAID. The atomic coordinates and structure factors (codes 3M7W, 3NNT, and 3JS3) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/). 1 To whom correspondence should be addressed: 900 South Ashland Ave., MBRB Room 1108, Chicago, IL 60607. Fax: 312-355-4535; E-mail Lavie@uic.edu. 2 The abbreviations used are: DHQD, dehydroquinate dehydratase; cd- DHQD, C. difficile DHQD; seDHQD, S. enterica DHQD. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 286, NO. 5, pp. 3531–3539, February 4, 2011 © 2011 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A. FEBRUARY 4, 2011•VOLUME 286•NUMBER 5 JOURNAL OF BIOLOGICAL CHEMISTRY 3531 from C. difficile strain 630 and S. enterica subspecies enterica serovar typhimurium strain LT2 genomic DNA were PCR- amplified and cloned into pMCSG7 and pMCSG19 vectors, respectively, each of which contains N-terminal hexahistidine tag followed by a tobacco etch virus protease cleavage se- quence. Insert-containing plasmids were transformed into Escherichia coli BL21 (DE3) strain from Agilent (Santa Clara, CA). Protein expression and purification were performed us- ing standard Center for Structural Genomics of Infectious Diseases protocols (17, 18). Crystallization screens were set up immediately following purification, and the remaining en- zyme was aliquoted, frozen in liquid nitrogen, and stored at 80 °C for future use. DHQD Assays—Immediately before performing assays, protein was thawed and diluted to the appropriate concentra- tion. Assays were performed at 37 °C in potassium phosphate buffer (100 mM, pH 7.5). Reaction was initiated by the addi- tion of protein to a mixture of buffer and 3-dehydroquinic acid (Sigma-Aldrich). Formation of the conjugated enone carboxylate in dehydroshikimate was followed by mea- suring the increase in absorbance at   234 nm relative to substrate (  12 mM1cm1) (4, 19). To determine the kinetic parameters, triplicate measurements of the reaction rate were determined at varying concentrations of 3-dehy- droquinic acid. Data were fitted to the Michaelis-Menten equation using the enzyme kinetics module in SigmaPlot version 8.02. Protein Crystallization and Data Collection—Protein con- centrated to 7.5 mg/ml in a buffer containing 0.5 M NaCl and 10 mM Tris-HCl (pH 8.3) was used to set up sitting drops at a ratio of 1:1 protein to reservoir. Further detail regarding crys- tallization conditions is presented in Table 1. The substrate- bound structure was obtained by co-crystallization of the Lys- 170 3 Met (K170M) mutant protein with 2 mM dehydroquinic acid. The pre-dehydration reaction intermedi- ate bound structure was obtained by soaking a crystal in mother liquor containing 5 mM dehydroquinic acid for 15 min prior to freezing in liquid nitrogen. The post-dehydration re- action intermediate bound structure was obtained by co-crys- tallization with 1 mM dehydroshikimic acid. All crystals where immersed in mother liquor before being frozen in liquid ni- trogen. Diffraction data were collected at 100° K at the Life Sciences Collaborative Access Team at the Advanced Photon Source, Argonne, IL. Structure Determination and Refinement—Data were pro- cessed using HKL-3000 for indexing, integration, and scaling (20). Structures were solved with Phaser (21) using the S. typhi apo DHQD structure (Protein Data Bank (PDB) code 1GQN) as a starting model for the cdDHQD structure and the apo seDHQD structure (PDB code 3L2I) for determination of seDHQD structures. Structures were refined with Refmac (22) and manually corrected based on electron density maps displayed in Coot (23). All figures were prepared in the PyMOL Molecular Graphics System, Version 1.3 (Schro¨- dinger, LLC). Atomic coordinates and structure factors were deposited in the PDB under codes 3M7W (pre-dehydration complex), 3NNT (K170M substrate complex), and 3JS3 (post- dehydration complex). RESULTS AND DISCUSSION C. difficile and S. enterica DHQDs Display a Similar Overall Structure—The cdDHQD and seDHQD share 56% sequence identity (Fig. 2A). Structures of the reaction intermediate bound seDHQD and cdDHQD protein were solved by molec- ular replacement and refined to a resolution of 1.95 and 2.20 Å, respectively. The DHQD monomer from each bacterium is characterized by an eight-stranded -barrel motif and is ob- served to dimerize with helices 6, 7, 8, and 9, making anti- parallel intermolecular contacts, in a manner similar to previ- ously reported DHQDs (8, 13–15). The overlay of cdDHQD and seDHQD shows a high degree of structural conservation (Fig. 2B). Differing Side Chain Conformation of His-143 in Pre- and Post-dehydration States—To gain insight into the reaction mechanism of DHQD, we obtained structures of DHQD com- plexed with ligand. A soak of seDHQD crystals with the sub- strate, 3-dehydroquinic acid, produced a structure in which the substrate covalently linked at its 3-position to the Schiff base-forming Lys-170 was observed (Fig. 3, A and B). The bound ligand displays clear electron density corresponding to the 1-hydroxyl leaving group (Fig. 3A, arrow), and therefore this structure represents a pre-dehydration intermediate state of the reaction. Co-crystallization of the cdDHQD with the product, 3-dehydroshikimic acid, also produced a crystal structure in which a Lys-1703-bound covalent species is ob- served (Fig. 3, C and D). In this case, the bound ligand lacks electron density corresponding to the leaving group. Interest- ingly, an ordered water molecule is located directly above where the leaving group was observed in the pre-dehydration ligand (Fig. 3C, arrow) and is thus positioned to be the mole- cule dehydrated from the substrate. This structure, which is similar to previously reported type I DHQD intermediate complexes (8, 14), represents a post-dehydration intermediate state of the reaction. Together, these structures provide snap- shots of successive steps in the reaction and thus allow for the identification of structural changes that occur over the course of the reaction. A particularly striking reaction state-dependent structural behavior is revealed by an overlay of the pre- and post-dehy- dration reaction intermediate bound active sites. In the pre- 3 cdDHQD and seDHQD have different chain lengths, but for the sake of consistency, all residue numbering within the text refers to the position of the residue in the seDHQD protein. FIGURE 1. The substrate and product of the dehydration reaction cata- lyzed by DHQDs. Indicated is the numbering convention used throughout the text for the substrate, intermediate, and product and the pro-R and pro-S hydrogen relevant to the mechanism of elimination of the 1-hydroxyl group. Multiple Roles for Active Site Histidine in Type I DHQDs 3532 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 286•NUMBER 5•FEBRUARY 4, 2011 dehydration complex, His-143 is proximal to both elements of the catalytic dehydration. The His-143 N2 atom is 2.7 Å from the modeled pro-R C-2 hydrogen of the substrate and forms a 2.8 Å hydrogen bond with the 1-hydroxyl leaving group (Fig. 4A). In the post-dehydration covalent complex, the histidine side chain is displaced 1.5 Å away from the site of catalytic dehydration to the position it adopts in all previously reported apo and post-dehydration reaction intermediate DHQD structures (8, 13–15). In this position, the His-143 N1 atom forms a 2.9 Å hydrogen bond with the conserved residue Glu-86 (Fig. 4B). To address the possibility that the displacement of His-143 observed in the pre-dehydration reaction intermediate bound structure is a non-biologically relevant artifact of the low pH of 4.6 of the crystallization condition, we used a similar crystal soaking protocol to test a crystal grown at neutral pH. A com- parable behavior of His-143 was observed in this neutral pH crystal structure (data not shown), suggesting that the confor- mation of His-143 seen in the pre-dehydration structure is representative of the behavior of the residue over a wide range of physiological pH values. Having ruled out pH in explaining the displacement of His-143, the only difference between the pre- and post- dehydration reaction intermediate bound structures is the presence of the leaving group in the pre-dehydration li- gand. As such, the differential behavior of His-143 between the two structures can be attributed to the presence of the leaving group in the pre-dehydration reaction state. In par- ticular, induction of the pre-dehydration conformation of His-143 is likely due to the potential for formation of a fa- vorable hydrogen-bonding interaction between the N2 atom of the histidine and the leaving hydroxyl group (Fig. 4). In the apo and post-dehydration reaction states, with- out the potential for His-143 to hydrogen-bond with the leaving group, adoption of the His-143 post-dehydration conformation is likely due to the potential for the N1 atom of the histidine to hydrogen-bond with Glu-86 in this posi- tion. These structures define a never before characterized leaving group-dependent conformation of His-143 in which the residue is proximal to the site of dehydration only when the leaving group is present. Intermediate State Crystal Structures Suggest Role for His- 143 in Catalysis—Several lines of experimental evidence suggest that this conserved active site histidine may act to shuttle the proton from the C-2 position in the ring to the 1-hydroxyl leaving group. Hanson and Rose (9) used ste- reo-specific labeling experiments to demonstrate that type I dehydroquinate dehydration results in cis-elimination. That is, the abstracted proton comes from the same side of the ring as the leaving group. This finding is consistent with a single residue both abstracting the proton from the ring and protonating the leaving group (9). Based on di- ethyl polycarbonate treatment resulting in an inactivation of the protein, Deka et al. (24) hypothesized that the con- served histidine might be the proton-transferring residue. Mutagenesis studies might be expected to conclusively ad- dress the role of this residue in proton shuttling. Although Leech et al. (25) showed that the E. coli His-143 3 Ala (H143A) mutant had a profound loss of activity, potentially supportive of this histidine acting as the proton-shuttling entity, it was unclear whether or not the effect of the muta- tion was due solely to the role of this residue in Schiff base formation and hydrolysis. Based on the body of kinetic evi- dence, the proximity of the pre-dehydration conformation TABLE 1 Crystallization conditions, data collection, and refinement statistics Values for highest resolution shell are in parentheses. Species S. enterica S. enterica C. difficile Variant Wild type K170M Wild type PDB code 3M7W 3NNT 3JS3 Color in figures Cyan Gray Pink Active site ligand Pre-dehydration covalent intermediate 3-Dehydroquinate Post-dehydration covalent intermediate Crystallization conditions 170 mM NH4OAc, pH 4.6, 25.5% (w/v) PEG 4000, 15% (v/v) glycerol 50 mM K2PO4, pH 6, 20% (w/v) PEG 8000 100 mM Tris, pH 8.5, 30% (w/v) PEG 500 Space group C2 P1 P21 Unit cell dimensions a, b, c (Å) a  184.33, b  66.58, c  128.23 a  36.73, b  43.55, c  79.94 a  60.47, b  139.62, c  66.77 , ,  (°) a  90.00,   119.08,   90.00   91.18,   101.27,   109.05   90.00,   90.63,   90.00 Resolution (Å) 29.69–1.95 (2.00–1.95) 29.65–1.60 (1.64–1.60) 29.55–2.20 (2.26–2.20) Number of reflections 99109 (7189) 55405 (3978) 55907 (4028) Completeness (%) 99.9% (99.1) 96.8 (95.2) 99.8% (98.0) Redundancy 3.7 (3.6) 2.0 (2.3) 3.8 (3.8) Rsym I 0.086 (0.466) 0.030 (0.357) 0.066 (0.614) I/(I) 14.9 (2.8) 23.2 (2.3) 17.3 (2.2) Molecules per aua 6 2 4 No. of atoms Protein 11646 3992 8116 Water 1128 525 322 Ligand 156 26 44 Rwork/Rfree 0.154/0.206 0.158/0.187 0.189/0.241 r.m.s.b deviations Bond lengths (Å) 0.012 0.011 0.012 Bond angles (°) 1.45 1.41 1.54 Average B-factors Protein 25.4 25.5 26.9 Waters 35.9 37.3 28.2 Ligand 21.3 25.6 37.4 a asu, asymmetric unit. b r.m.s., root mean square. Multiple Roles for Active Site Histidine in Type I DHQDs FEBRUARY 4, 2011•VOLUME 286•NUMBER 5 JOURNAL OF BIOLOGICAL CHEMISTRY 3533 of His-143 to both the pro-R C-2 proton and the 1-hy- droxyl leaving group, and the absence of another feasible proton-transferring residue, we propose a reaction mecha- nism consistent with previously hypothesized models (25– 28) but in which His-143 catalyzes the dehydration event by its N2 atom abstracting the pro-R C-2 proton and do- nating it to the 1-hydroxyl leaving group (Fig. 5A). Reaction State-dependent His-143 Behavior Suggests Con- formation-dependent Catalytic Roles for the Residue—In addi- tion to its role in the catalytic dehydration event, His-143 has a well established role in the formation and hydrolysis of the Schiff base intermediates. Leech et al. (25) found that the H143A mutant had a profoundly deficient reaction rate. They observed that the mutant enzyme slowly accumulated co- FIGURE 2. Sequence and structure similarity between the cdDHQD and seDHQD type I DHQDs. A, sequence alignment of cdDHQD and seDHQD. Align- ment was done in ClustalW2 version 2 using the default settings. Secondary features were defined using the cdDHQD structure in ESPript version 2.2. B, superposition of the biological dimer of the seDHQD pre-dehydration complex (cyan) and the cdDHQD post-dehydration complex (pink) structures (back- bone root mean square deviation  1.16 Å). Lys-170 and the reaction intermediates to which it is covalently bound are shown as sticks. Helices of the dimer interface are labeled. Multiple Roles for Active Site Histidine in Type I DHQDs 3534 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 286•NUMBER 5•FEBRUARY 4, 2011 valently bound reaction intermediate at the active site and interpreted this result to mean that the mutant enzyme has a diminished capacity both to form and later to hydrolyze the covalent Schiff base. Based on these findings, Leech et al. (25) proposed a model in which His-143 acts to facilitate the pro- tonation or deprotonation of the carbinolamine reaction in- termediate depending upon whether the Schiff base is being formed or hydrolyzed (Fig. 5B). Considering that adoption of the His-143 pre-dehydration conformation appears to be leaving group-dependent, the hydrolysis of the Schiff base, which occurs post-dehydration, must initiate with His-143, adopting its post-dehydration conformation. As such, His-143 would appear to assume dif- fering functionalities in each of its conformations. In its pre- dehydration conformation, the residue catalyzes the catalytic dehydration event (Fig. 5A), whereas in its post-dehydration conformation, the residue catalyzes Schiff base hydrolysis (Fig. 5B). Although the conformation His-143 adopts when it catalyzes Schiff base formation cannot be inferred from these structures, clearly, His-143 plays a complex and highly tuned role in which it shifts between two conformations to catalyze multiple aspects of the type I DHQD reaction coordinate. Similar Positioning of Substrate and Reaction Intermediates within the Active Site Supports a Catalytic Role for the Schiff Base—Stereoelectronic principles dictate that the Lys-170 amine nucleophile will form the covalent Schiff base after ap- proaching the carbonyl carbon of the substrate at the Bu¨rgi- Dunitz (N–C–O) angle of 107° (29, 30). According to this model of bond formation, the position of the Lys-170 amine relative to the 3-carbon of the substrate should differ mark- edly during its approach to the carbonyl when compared with after the Schiff base has formed. To meet these positional re- quirements, either Lys-170 or the substrate must undergo substantial movement as the substrate transitions from a non- covalent to Schiff base bound state. The simplest mechanism by which the enzyme might accommodate both the Bu¨rgi- Dunitz approach and the corresponding active site rearrange- ments would be for the substrate to initially dock in an orien- tation that positioned it for the nucleophilic approach. Subsequently, in a process correlated with bond formation, the substrate could move to its covalent intermediate bound position. In fact, this expectation is supported by studies of several enzymes that generate a similar Schiff base intermediate by the nucleophilic attack of a lysine on a substrate ketone. Crys- tallographic studies of dihydrodipicolinate synthase (31), fruc- tose 1,6-bisphosphate (32), and 2-keto-3-deoxy-6-phospho- gluconate aldolase (PDB ID 3LAB) all show that the orientation of the covalent Schiff base intermediate is rotated 180° within the active site when compared with the non- covalently bound substrate. These observations are consistent with the substrate initially docking in a conformation that FIGURE 3. Crystal structures of DHQD in pre- and post-dehydration covalent intermediate states. A, structure of seDHQD active site with covalent pre- dehydration reaction intermediate. Carbon atoms are depicted in cyan, oxygens are in red, nitrogens are in blue, and sulfurs are in yellow. Difference maps were calculated with the reaction intermediate omitted from the model. The Fo  Fc map is contoured at the 3.0 level (red), and the 2Fo  Fc map is con- toured at the 1 level (blue). An arrow indicates the 1-hydroxyl leaving group. B, schematic rendering of the pre-dehydration reaction intermediate shown in A. Distances between atoms are shown in angstroms. C, structure of the cdDHQD active site with covalent post-dehydration reaction intermediate. Model and maps are the same as A, except that carbons are shown in pink. The arrow points toward an ordered water molecule at a position consistent with that dehydrated from the substrate. D, schematic rendering of the post-dehydration reaction intermediate shown in C. Distances between atoms are shown in angstroms. Multiple Roles for Active Site Histidine in Type I DHQDs FEBRUARY 4, 2011•VOLUME 286•NUMBER 5 JOURNAL OF BIOLOGICAL CHEMISTRY 3535 allows for a Bu¨rgi-Dunitz approach before moving to the in- termediate bound position in a process concurrent with bond formation. Whether this mode of substrate nucleophile ap- proach is specific to these enzymes or is general to the Schiff base-forming enzymes is presently unknown. Based on the severe effect of the Lys-170 3 Ala mutation on kcat but minimal effect on Km, it has been argued that the Schiff base is likely to play a direct catalytic role in the reac- tion mechanism of DHQD (25). Suggestions on the mecha- nism by which the Schiff base might promote catalysis have focused on how it could distort the ring of the substrate in a manner that would stereoelectronically promote elimination or function to stabilize the carbanion reaction intermediate (25, 33, 34). However, in light of the stereoelectronic argu- mentation and experimental data describing substrate bind- ing in other Schiff base-forming enzymes, in DHQD, it is also likely that prior to Schiff base formation, the substrate adopts an orientation that is different from the covalent reaction in- termediate within the active site. In that case, it is conceivable that Schiff base formation is required to position the substrate in the proper orientation in which for catalysis to occur, pre- senting a scenario wherein the Schiff base plays only an indi- rect role in catalysis. To gain insight into the initial substrate binding event within Schiff base-forming enzymes generally and to deter- mine whether the Schiff base is directly involved in DHQD catalysis, we set out to capture a complex of DHQD with sub- strate. We reasoned that if the non-covalently bound sub- strate was similarly positioned relative to the covalently bound reaction intermediate, then formation of the Schiff base must not result in significant change in the position of the substrate within the active site and therefore cannot be promoting catalysis by positioning the substrate into the cata- lytically required orientation. In that case, having ruled out the proposed non-catalytic (orientation-determining) role of the Schiff base, it could be concluded that the Schiff base is likely to have only a direct involvement in catalysis. To obtain a crystal structure of seDHQD in a non-covalent complex with substrate, site-directed mutagenesis was uti- lized to convert the reactive Schiff base-forming Lys-170 3 Met (K170M). Methionine was chosen because it is the resi- due closest in shape to a lysine but cannot form a Schiff base. As previously reported (25), mutation of Lys-170 resulted in a dramatic loss in reactivity (Table 2). Co-crystallization of the K170M mutant enzyme with dehydroquinic acid produced a structure in which the substrate was observed non-covalently bound at the active site (Fig. 6A). A comparison of the pre- dehydration intermediate bound structure and K170M sub- strate-bound structure shows the ring core of the substrate to be similarly positioned but slightly shifted (0.3 Å) within the active site, likely to avoid a clash with the Met-170 terminal methyl group (Fig. 6B). The similar orientation of substrate and reaction intermediate demonstrates a differential mode of substrate binding when compared with other Schiff base- forming enzymes. The dispensability of Lys-170 in substrate binding, as reasoned above, provides support for the Schiff base playing a direct catalytic role in the reaction mechanism of the protein. K170M Substrate-bound Structure Provides Insight but Also Generates New Questions Regarding Schiff Base Formation— Although a previous study has shown a role for His-143 in Schiff base formation (25), the mechanism of His-143 involve- ment remains ill-defined. Based on mutagenesis studies, Leech et al. (25) proposed that His-143 catalyzes the conver- sion of the carbinolamine intermediate to the Schiff base in- termediate (Fig. 5B). However, it is possible that His-143 af- fects other steps in Schiff base formation. The K170M substrate-bound structure reveals that His-143 adopts partial occupancies of both of the conformations observed in the wild type structures (Fig. 6A). Both positions of His-143 are proximal to the 3-carbonyl oxygen of the substrate (Fig. 6C). This structural observation positions the residue to protonate the carbonyl oxygen and/or to accentuate the dipole of the carbonyl by drawing charge away from the carbonyl carbon, which might promote the Lys-170 nucleophilic attack. The proximity of His-143 to the carbonyl oxygen prior to the for- mation of the covalent intermediate suggests that the residue may play a functional role in an earlier step of Schiff base for- mation than has previously been recognized. FIGURE 4. The conformation of His-143 is dependent on the presence of the 1-hydroxyl group of the reaction intermediate. A, in the pre-dehy- dration reaction intermediate structure (cyan), the N2 atom of His-143 is within interaction distance to the 1-hydroxyl (2.8 Å) and the pro-R C-2 hy- drogen (2.7 Å). Thus, it can facilitate dehydration by transferring the hydro- gen atom to the leaving hydroxyl group. B, superposition of the pre- (cyan) and post- (pink) dehydration states of DHQD. In the pre-dehydration state, His-143 interacts with the 1-hydroxyl group. Upon leaving of this group, as revealed by the post-dehydration intermediate structure, His-143 rotates to a position where it interacts with Glu-86 (2.9 Å). Multiple Roles for Active Site Histidine in Type I DHQDs 3536 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 286•NUMBER 5•FEBRUARY 4, 2011 Although providing new insight into the significance of the Schiff base in catalysis, the K170M substrate-bound structure presents new questions regarding how formation of the Schiff base occurs. The observed position of the substrate is incon- sistent with the Lys-170 N atom approaching the carbonyl carbon of the substrate at an ideal stereoelectronic angle. Modeling Lys-170 into the K170M substrate-bound structure reveals that, without significant perturbations to the main chain, the maximal approach angle that the Lys-170 amine can achieve falls well short of the 107° Bu¨rgi-Dunitz angle (Fig. 6D). As such, the bond must form either by a non-Bu¨rgi- Dunitz approach of the substrate in its observed position or by a Bu¨rgi-Dunitz approach of the substrate in an unobserved position within the active site. Possible exceptions to the Bu¨rgi-Dunitz nucleophilic ap- proach in the context of enzyme catalysis are not without precedent (35). In conjunction with previous observations, the mode of substrate binding observed here potentially sup- ports a more widespread exception to the established mode of nucleophilic approach in enzyme catalysis. Alternatively, the approach of Lys-170 and formation of the covalent bond may occur when the substrate transiently adopts a conformation that places its carbonyl carbon at the Bu¨rgi-Dunitz angle; this requirement may be met during the process of substrate docking into the active site. In that case, in the wild type en- zyme, the covalent adduct may form before the substrate reaches its observed position in the K170M substrate-bound structure. The mechanism of Schiff base formation is the topic of ongoing study. Conclusions—We report three crystal structures that ad- dress several issues about how type I DHQDs function. The substrate and pre-dehydration covalent intermediate bound structures provide the first view of these reaction states. The different conformation of His-143 in pre- and post-dehydra- tion intermediate states defines a leaving group-dependent behavior of the residue. The proximity of His-143 to the C-2 proton and leaving group of the pre-dehydration reaction in- termediate supports a role for this residue in the transport of the proton to the leaving group in the catalytic elimination and provides a rare example of the requirement of the leaving group for a residue to adopt the conformation consistent with its presumed catalytic role. Previous kinetic studies and the structural data presented here suggest a reaction mechanism in which His-143 moves between two conformations while undergoing a series of protonation/deprotonation events to FIGURE 5. Proposed role of His-143 in type I DHQD-catalyzed reaction. A, putative role of His-143 in the catalytic dehydration. Following forma- tion of the covalent Schiff base linker between Lys-170 and the substrate 3-dehydroquinate 1, His-143 assumes its pre-dehydration position, where its N2 atom forms a key hydrogen-bonding interaction with the 1-hydroxyl group of the reaction intermediate. In this position, the His-143 N2 atom abstracts the C-2 pro-R proton of the substrate to generate the carbanion intermediate 2, which gives rise to the enamine intermediate 3. The proto- nated His-143 then delivers its N2 proton to facilitate departure of the 1-hydroxyl leaving group, which generates the ene-iminium intermediate 4. Because the H143 N2 atom can no longer form the critical interaction with the 1-hydroxyl leaving group, a shift to the post-dehydration position of the residue ensues. Finally, following Schiff base hydrolysis, the formally dehydrated product 3-dehydroshikimate is released. Boxed intermediates 1 and 4 represent likely states of the reaction captured by pre- and post-dehydration crystal structures. B, role of His-143 in Schiff base formation and hydrolysis based on Leech et al. (25). In the formation of the Schiff base, attack by the Lys-170 N atom on the 3-carbonyl carbon leads to formation of the carbinolamine intermediate. His-143 then delivers a proton to facilitate departure of the hydroxyl leaving group to generate the Schiff base intermediate. Following the catalytic hydrolysis described in A, His-143 adopts its post-dehydration conformation and catalyzes the reverse reaction to hydrolyze the Schiff base and regenerate the active site. TABLE 2 Kinetic characterization of seDHQD and cdDHQD Errors were calculated by fitting of the kinetic data to the Michaelis-Menten equation and expressed as  S.D. kcat Km kcat/Km s1 M s1M1 seDHQD wild type 210  5 21  3 10  1 seDHQD K170M 0.015  0.007 33  4 4.5  104  2.3  104 cdDHQD wild type 125  4 36  9 3.5  0.9 Multiple Roles for Active Site Histidine in Type I DHQDs FEBRUARY 4, 2011•VOLUME 286•NUMBER 5 JOURNAL OF BIOLOGICAL CHEMISTRY 3537 catalyze multiple steps in the formation and hydrolysis of the Schiff base as well as the catalytic dehydration. The K170M substrate-bound structure provides insight into the role of the Schiff base in catalysis. The similar mode of binding of substrate in the K170M variant and reaction intermediate in the wild type enzyme eliminates the possibil- ity that the functional role of the Schiff base is to orient the substrate within the active site. Based on these results, we conclude that the Schiff base must have a direct role in ca- talysis. It is anticipated that the more detailed knowledge of the DHQD kinetic pathway provided by this work will aid in the design of novel anti-bacterials, which are partic- ularly relevant for the emerging pathogens C. difficile and S. enterica. Acknowledgments—We thank Dr. Elisabetta Sabini for facilitating the communications with the groups involved in this work and Dr. Scott Peterson and Dr. Keehwan Kwon for providing DHQD expres- sion clones. Use of the Advanced Photon Source at Argonne Na- tional Laboratory was supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract DE-AC02-06CH11357. Use of the Life Sciences Collaborative Access Team (LS-CAT) Sector 21 was supported in part by the Michigan Economic Development Corporation and the Michigan Technology Tri-Corridor for the support of this research program (Grant 085P1000817). REFERENCES 1. Bentley, R. (1990) Crit. Rev. Biochem. Mol. Biol. 25, 307–384 2. Kishore, G. M., and Shah, D. M. (1988) Annu. Rev. Biochem. 57, 627–663 3. Marques, M. R., Pereira, J. H., Oliveira, J. S., Basso, L. A., de Azevedo, W. F., Jr., Santos, D. S., and Palma, M. S. (2007) Curr. Drug Targets 8, 445–457 4. Noble, M., Sinha, Y., Kolupaev, A., Demin, O., Earnshaw, D., Tobin, F., West, J., Martin, J. D., Qiu, C., Liu, W. S., DeWolf, W. E., Jr., Tew, D., and Goryanin, I. I. (2006) Biotechnol. Bioeng. 95, 560–573 5. Butler, J. R., Alworth, W. L., and Nugent, M. J. (1974) J. Am. Chem. Soc. 96, 1617–1618 6. Kleanthous, C., Campbell, D. G., and Coggins, J. R. (1990) J. Biol. Chem. 265, 10929–10934 7. Kleanthous, C., Deka, R., Davis, K., Kelly, S. M., Cooper, A., Harding, S. E., Price, N. C., Hawkins, A. R., and Coggins, J. R. (1992) Biochem. J. 282, 687–695 8. Gourley, D. G., Shrive, A. K., Polikarpov, I., Krell, T., Coggins, J. R., Hawkins, A. R., Isaacs, N. W., and Sawyer, L. (1999) Nat. Struct. Biol. 6, 521–525 9. Hanson, K. R., and Rose, I. A. (1963) Proc. Natl. Acad. Sci. U.S.A. 50, 981–988 10. White, P. J., Young, J., Hunter, I. S., Nimmo, H. G., and Coggins, J. R. (1990) Biochem. J. 265, 735–738 11. Lumsden, J., and Coggins, J. R. (1977) Biochem. J. 161, 599–607 12. Polley, L. D. (1978) Biochim. Biophys. Acta 526, 259–266 13. Lee, W. H., Perles, L. A., Nagem, R. A., Shrive, A. K., Hawkins, A., Saw- yer, L., and Polikarpov, I. (2002) Acta Crystallogr. D Biol. Crystallogr. 58, 798–804 14. Nichols, C. E., Lockyer, M., Hawkins, A. R., and Stammers, D. K. (2004) FIGURE 6. Structure of the seDHQD K170M mutant with 3-dehydroquinate bound. A, active site of K170M mutant structure showing difference maps calculated and colored as described in the legend for Fig. 2. Each His-143 conformation is modeled at 50% occupancy. B, superposition of K170M substrate-bound (gray) and pre-dehydration reaction intermediate bound (cyan) structures. C, the two H143 side chain conformations are proximal to the 3-carbonyl oxygen of the 3-dehydroquinate. D, Lys-170 side chain modeled in to the active site of the K170M substrate-bound struc- ture with the goal of maximizing the angle of approach of Lys-170 N atom to the carbonyl carbon of the substrate (the angle defined by N–C–O, i.e. the Bu¨rgi-Dunitz angle). Multiple Roles for Active Site Histidine in Type I DHQDs 3538 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 286•NUMBER 5•FEBRUARY 4, 2011 Proteins 56, 625–628 15. Smith, N. N., and Gallagher, D. T. (2008) Acta Crystallogr. Sect. F Struct. Biol. Cryst. Commun. 64, 886–892 16. Stols, L., Gu, M., Dieckman, L., Raffen, R., Collart, F. R., and Donnelly, M. I. (2002) Protein Expr. Purif. 25, 8–15 17. Godsey, M. H., Minasov, G., Shuvalova, L., Brunzelle, J. S., Vorontsov, I. I., Collart, F. R., and Anderson, W. F. (2007) Protein Sci. 16, 1285–1293 18. Kim, Y., Bigelow, L., Borovilos, M., Dementieva, I., Duggan, E., Eschen- feldt, W., Hatzos, C., Joachimiak, G., Li, H., Maltseva, N., Mulligan, R., Quartey, P., Sather, A., Stols, L., Volkart, L., Wu, R., Zhou, M., and Joachimiak, A. (2008) Adv. Protein Chem. Struct. Biol 75, 85–105 19. Chaudhuri, S., Lambert, J. M., McColl, L. A., and Coggins, J. R. (1986) Biochem. J. 239, 699–704 20. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307–326 21. McCoy, A. J., Grosse-Kunstleve, R. W., Storoni, L. C., and Read, R. J. (2005) Acta Crystallogr. D Biol. Crystallogr. 61, 458–464 22. Murshudov, G. N., Vagin, A. A., Lebedev, A., Wilson, K. S., and Dodson, E. J. (1999) Acta Crystallogr. D Biol. Crystallogr. 55, 247–255 23. Emsley, P., and Cowtan, K. (2004) Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 24. Deka, R. K., Kleanthous, C., and Coggins, J. R. (1992) J. Biol. Chem. 267, 22237–22242 25. Leech, A. P., James, R., Coggins, J. R., and Kleanthous, C. (1995) J. Biol. Chem. 270, 25827–25836 26. Le Sann, C., Gower, M. A., and Abell, A. D. (2004) Mini Rev. Med. Chem. 4, 747–756 27. Parker, E. J., Gonza´lez Bello, C., Coggins, J. R., Hawkins, A. R., and Abell, C. (2000) Bioorg. Med. Chem. Lett. 10, 231–234 28. Reilly, A., Morgan, P., Davis, K., Kelly, S. M., Greene, J., Rowe, A. J., Har- ding, S. E., Price, N. C., Coggins, J. R., and Kleanthous, C. (1994) J. Biol. Chem. 269, 5523–5526 29. Bu¨rgi, H. B., Dunitz, J. D., Lehn, J. M., and Wipff, G. (1974) Tetrahedron 30, 1563–1572 30. Fleming, I. (2010). in Molecular Orbitals and Organic Chemical Reac- tions, pp 214–215, John Wiley & Sons, Hoboken, NJ 31. Soares da Costa, T. P., Muscroft-Taylor, A. C., Dobson, R. C., Devenish, S. R., Jameson, G. B., and Gerrard, J. A. (2010) Biochimie 92, 837–845 32. Blom, N., and Sygusch, J. (1997) Nat. Struct. Biol. 4, 36–39 33. Turner, M. J., Smith, B. W., and Haslam, E. (1975) J. Chem. Soc. Perkin Trans. I 1, 52–55 34. Vaz, A. D., Butler, J. R., and Nugent, M. J. (1975) J. Am. Chem. Soc. 97, 5914–5915 35. Radisky, E. S., and Koshland, D. E., Jr. (2002) Proc. Natl. Acad. Sci. U.S.A. 99, 10316–10321 Multiple Roles for Active Site Histidine in Type I DHQDs FEBRUARY 4, 2011•VOLUME 286•NUMBER 5 JOURNAL OF BIOLOGICAL CHEMISTRY 3539
3M81
Crystal structure of Acetyl xylan esterase (TM0077) from THERMOTOGA MARITIMA at 2.50 A resolution (native apo structure)
Functional and structural characterization of a thermostable acetyl esterase from Thermotoga maritima Mark Levisson1,*, Gye Won Han2,3,*, Marc C. Deller2,3, Qingping Xu2,4, Peter Biely5, Sjon Hendriks1, Lynn F. Ten Eyck6,7, Claus Flensburg8, Pietro Roversi8, Mitchell D. Miller2,4, Daniel McMullan9, Frank von Delft2,3,‡, Andreas Kreusch10, Ashley M. Deacon2,4, John van der Oost1, Scott A. Lesley2,3,10, Marc-André Elsliger2,3, Servé W. M. Kengen1,†, and Ian A. Wilson2,3,† 1Laboratory of Microbiology, Wageningen University, 6703 HB, Wageningen, The Netherlands 2Joint Center for Structural Genomics, http://www.jcsg.org 3Department of Molecular Biology, The Scripps Research Institute, La Jolla, California 92037 4Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Stanford University, Menlo Park, California 92045 5Institute of Chemistry, Slovak Academy of Sciences, 845 38 Bratislava, Slovakia 6Department of Chemistry and Biochemistry, University of California at San Diego, La Jolla, California 92093-0505 7San Diego Supercomputer Center, University of California at San Diego, La Jolla, California 92093-0505 8Global Phasing Ltd. Sheraton House, Castle Park, Cambridge CB3 0AX, United Kingdom 9Protein Therapeutics Department, Genomics Institute of the Novartis Research Foundation, San Diego, California 92121 10Protein Sciences Department, Genomics Institute of the Novartis Research Foundation, San Diego, California 92121 Abstract TM0077 from Thermotoga maritima is a member of the carbohydrate esterase family 7 and is active on a variety of acetylated compounds, including cephalosporin C. TM0077 esterase activity is confined to short-chain acyl esters (C2-C3), and is optimal around 100°C and pH 7.5. The positional specificity of TM0077 was investigated using 4-nitrophenyl-β-D-xylopyranoside monoacetates as substrates in a β-xylosidase-coupled assay. TM0077 hydrolyzes acetate at positions 2, 3 and 4 with equal efficiency. No activity was detected on xylan or acetylated xylan, which implies that TM0077 is an acetyl esterase and not an acetyl xylan esterase as currently annotated. Selenomethionine-substituted and native structures of TM0077 were determined at 2.1 Å and 2.5 Å resolution, respectively, revealing a classic α/β-hydrolase fold. TM0077 assembles into a doughnut-shaped hexamer with small tunnels on either side leading to an inner cavity, which contains the six catalytic centers. Structures of TM0077 with covalently bound phenylmethylsulfonyl fluoride (PMSF) and paraoxon were determined to 2.4 Å and 2.1 Å, respectively, and confirmed that both inhibitors bind covalently to the catalytic serine (Ser188). Upon binding of inhibitor, the catalytic serine adopts an altered conformation, as observed in other esterase and lipases, and supports a previously proposed catalytic mechanism in which this Ser hydroxyl rotation prevents reversal of the reaction and allows access of a water molecule for completion of the reaction. †Correspondence to: Ian A. Wilson, Ph.D., Department of Molecular Biology, The Scripps Research Institute, La Jolla, California 92037; (858) 784-2939 Fax: (858) 784-2980; wilson@scripps.edu or Servé W. M. Kengen, Ph.D., Laboratory of Microbiology, Wageningen University, 6703 HB, Wageningen, The Netherlands; 31 317 483737, Fax: 31 317-483829; serve.kengen@wur.nl. *ML and GWH contributed equally to this work. ‡Current address: The Structural Genomics Consortium, Roosevelt Drive, Headington, Oxford OX3 7DQ, UK NIH Public Access Author Manuscript Proteins. Author manuscript; available in PMC 2013 June 01. Published in final edited form as: Proteins. 2012 June ; 80(6): 1545–1559. doi:10.1002/prot.24041. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Keywords Acetyl esterase; Thermotoga maritima; crystal structure; α/β hydrolase; inhibitor; serine rotation INTRODUCTION Thermotoga maritima is a hyperthermophilic bacterium that grows optimally at 80°C and is able to metabolize a variety of simple and complex carbohydrates, including glucose, sucrose, starch, cellulose, and xylan 1. Its carbohydrate utilization potential was confirmed by analysis of its sequenced genome 2. The xylan degrading pathway of T. maritima has been studied using microarrays 2–4, and several genes encoding transporters, xylanases, and a β-xylosidase have been identified. Among the enzymes with a differential expression pattern in the microarray was a predicted acetyl xylan esterase (locus tag TM0077, axeA) 3,5. Depending on the source, the xylan backbone may contain a varying degree of acetylated xylose residues. Therefore, in addition to xylanases and xylosidases, the complete degradation of xylan requires esterases/deacetylases 6. Presently, esterases and deacetylases that are active on carbohydrate substrates have been classified into 16 families by Henrissat and coworkers (Carbohydrate-Active enZymes Server (CAZy)) 7. According to this classification, the predicted acetyl xylan esterase from T. maritima would be a member of family 7 of the carbohydrate esterases (CE7). In addition to the acetyl xylan esterase activity, enzymes in the CE7 family are rather unusual in that they display a high specific activity towards the antibiotic cephalosporin C [(Fig. 1(a-b)] 8. Cephalosporins belong to the β-lactam class of antibiotics, which also includes penicillin, and affect bacterial cell growth by inhibiting the penicillin-binding-protein that cross-links peptide glycans required for cell wall formation 9. The production of deacetylated cephalosporins is of great interest because these compounds are valuable building blocks for the production of semi-synthetic β-lactam antibiotics10,11. To explore the catalytic capacity of the predicted acetyl xylan esterase from T. maritima and gain a better insight into the structure and function of the family 7 carbohydrate esterases, TM0077 was expressed and purified, and three-dimensional structures of the native enzyme and its complexes with phenylmethylsulfonyl fluoride (PMSF) and paraoxon inhibitors, were determined by x-ray crystallography. Furthermore, the enzyme was functionally characterized, and various biochemical properties including the positional specificity of the esterase were investigated. MATERIALS AND METHODS Gene cloning TM0077 was selected as part of the Joint Center for Structural Genomics (JCSG) effort on complete structural coverage of the T. maritima soluble proteome as a large-scale center for high-throughput structure determination funded under the NIHGMS Protein Structure Initiative (PSI) 12. The gene encoding TM0077 (GenBank: AAD35171.1, GI:4980565; SwissProt: Q9WXT2) was amplified by polymerase chain reaction (PCR) from genomic DNA using PfuTurbo DNA polymerase (Stratagene) and primers corresponding to the predicted 5′ and 3′ ends. The PCR product was cloned into plasmid pMH1, which encodes an expression and purification tag (MGSDKIHHHHHH) at the amino terminus of the protein. The cloning junctions were confirmed by DNA sequencing. Levisson et al. Page 2 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript TM0077-SeMet protein production and purification Protein production was performed in a selenomethionine-containing medium using the Escherichia coli methionine auxotrophic strain DL41. Expression was induced by the addition of 0.15% L-arabinose. At the end of fermentation, cells were harvested and subjected to one freeze/thaw cycle, and subsequently sonicated in Lysis Buffer [50 mM Tris pH 7.9, 50 mM NaCl, 1 mM MgCl2, 0.25 mM Tris(2-carboxyethyl)phosphine hydrochloride (TCEP), 1 mg/ml lysozyme] and the lysate was centrifuged at 3,400 × g for one hour. The soluble fraction was applied to nickel-chelating resin (GE Healthcare) pre-equilibrated with Equilibration Buffer [50 mM potassium phosphate pH 7.8, 300 mM NaCl, 10% (v/v) glycerol, 0.25 mM TCEP] containing 20 mM imidazole. The resin was washed with Equilibration Buffer containing 40 mM imidazole, and the protein was eluted with Elution Buffer [20 mM Tris pH 7.9, 300 mM imidazole, 10% (v/v) glycerol, 0.25 mM TCEP]. The eluate was buffer exchanged with Buffer Q [20 mM Tris pH 7.9, 5% (v/v) glycerol, 0.25 mM TCEP] containing 50 mM NaCl and applied to a RESOURCE Q column (GE Healthcare) pre-equilibrated with the same buffer. The protein was eluted using a linear gradient of 50 to 500 mM NaCl in Buffer Q and purified further with a HiLoad 16/60 Superdex 200 column (GE Healthcare), using Crystallization Buffer [20 mM Tris pH 7.9, 150 mM NaCl, 0.25 mM TCEP] as the mobile phase. For crystallization trials, the peak Superdex 200 fractions were concentrated to ~15 mg/mL by centrifugal ultrafiltration (Millipore). Molecular weight and oligomeric state of TM0077 were determined using a 1 cm × 30 cm Superdex 200 column (GE Healthcare) coupled with miniDAWN static light scattering (SEC/SLS) and Optilab differential refractive index detectors (Wyatt Technology). The mobile phase consisted of 20 mM Tris pH 8.0, 150 mM NaCl, and 0.02% (w/v) sodium azide. Native TM0077 production and purification For protein production, E. coli DL41 cells were grown in LB medium for 8 hours (an OD600 well above 2.0 was reached). Subsequently, the culture was induced by adding 0.15% L-arabinose and incubated another 16 hours at 37°C. Cells were harvested by centrifugation at 10,000 × g for 20 min. The cell pellet was resuspended in 30 ml of Lysis Buffer 2 [50 mM Tris-HCl pH 8.0, 50 mM NaCl, 10 mM imidazole, 0.25 mM TCEP]. The cells were disrupted by two passages through a French press at 110 MPa. The crude cell extract was treated with DNAse I at room temperature for 30 min and subsequently centrifuged at 43,000 × g for 30 min in order to remove cell debris. The supernatant was heated at 70°C for 25 min and then centrifuged to remove the precipitated proteins. The supernatant was filtered and loaded onto a nickel-chelating column packed with 20 ml of Ni- NTA His-Bind Resin (Novagen) and equilibrated in 50 mM Tris-HCl pH 8.0, 300 mM NaCl, 2% (v/v) glycerol, and 0.25 mM TCEP. The column was washed with 20 mM imidazole in the same buffer, and proteins were subsequently eluted with a linear gradient of 20–500 mM imidazole in the same buffer. Fractions containing esterase activity were pooled and loaded onto a HiPrep Desalting column (GE Healthcare) equilibrated with 20 mM Tris- HCl pH 8.0, 150 mM NaCl, and 0.25 mM TCEP. The homogeneity of the protein was checked by SDS-PAGE, and activity staining of the SDS-PAGE gel was performed using α- napthyl acetate, as described previously 13. The protein concentration was determined at 280 nm using a NanoDrop ND-1000 Spectrophotometer. Crystallization Crystals of selenomethionine-substituted TM0077 were obtained by hanging drop vapor diffusion against a 250 μl crystallization solution consisting of 20% (w/v) PEG-3000, 0.1 M HEPES pH 7.5, 0.2 M NaCl. Drops consisted of 0.5 μl protein and 0.5 μl crystallization solution. Native TM0077 was crystallized using nanodrop vapor diffusion techniques against a crystallization solution consisting of 0.2 M calcium acetate hydrate, 20% (w/v) Levisson et al. Page 3 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript PEG 3350, pH 7.3 at 20°C. Protein was concentrated to 22.8 mg/ml. Drops consisted of 100 nl protein and 100 nl of crystallization solution and a 60 μl reservoir of crystallization solution. Crystals of TM0077 in complex with inhibitors PMSF or paraoxon were obtained at 4°C in the same conditions with the same reagents as the native crystals. PMSF or paraoxon were added in a molar ratio of 1:3 (protein:inhibitor). Data collection For cryoprotection, the TM0077-SeMet crystal was transferred to crystallization solution supplemented with 15% (v/v) glycerol. The crystal was mounted in a cryoloop and subsequently flash-cooled in liquid nitrogen. X-ray data were collected at 100 K on beamline BL9-2 at the Stanford Synchrotron Radiation Lightsource (SSRL, Menlo Park, CA) using a Quantum 4 CCD detector (ADSC). A TM0077-SeMet MAD data set was collected to 2.1 Å resolution and the data were indexed in monoclinic space group P21, with unit cell parameters a = 152.6 Å, b = 131.0 Å, and c = 157.8 Å, and β=118.9°, and 12 molecules in the asymmetric unit. Data were indexed and integrated with DENZO and then scaled with SCALEPACK 14. Native TM0077, TM0077-PMSF complex (TM0077-PMS) and TM0077-paraoxon complex (TM0077-DEP) crystals were transferred to crystallization solution supplemented with 10% (v/v) ethylene glycol and flash-cooled to 100K. Data were collected at beamline 5.0.3 of the Advanced Light Source (ALS, Berkeley, CA) and processed with the HKL2000 package 14. The native data set was collected to 2.5 Å resolution, and TM0077-PMS and TM0077-DEP data sets were collected to 2.4 and 2.1 Å, respectively. All data were indexed in orthorhombic space group P212121, with unit cell parameters approximately a=103Å b=104Å c=221Å (See Table 1), and six molecules in the asymmetric unit. Data reduction and refinement statistics for TM0077-SeMet, TM0077-Native, TM0077-PMS and TM0077- DEP are summarized in Table I. Structure solution and refinement The TM0077-SeMet structure was solved by MAD phasing method using a two-wavelength MAD dataset. At the time of the initial data collection (2001), the structure determination of Se-MAD TM0077 posed a significant challenge to crystallographic programs, which were still under active development. As a result, modifications were made in various structure determination and refinement programs to achieve success. For initial phasing, SHELXD 15 was used to find candidate SeMet substructure sites. Attempts to complete phasing were unsuccessful due to the translational non-crystallographic symmetry (NCS) (not recognized initially). Self-consistent sets (partial sets) were found using the CCP4 program PROFESSS 16 and additional SeMet sites were found by SHELXD, and added to these partial sets. The SHARP 17 run did not complete initially; however, updates of SHARP and ARP/wARP eventually helped to resolve issues and an initial trace was obtained by ARP/ wARP. The structure was then refined with BUSTER 18 using tight NCS restraints to an Rcryst and Rfree of 18.6% and 22.3%, respectively. Model building was performed using O 19 and the structure was refined using Refmac5 20. Refinement statistics are summarized in Table I. The final model contains 12 protein molecules (chains A-L) in the asymmetric unit each consisting of residues 2-323. The MolProbity 21 Ramachandran plot analysis showed that 97.4% of all residues are in favored regions with a single outlier, Gln120 of chain B, which is supported by unambiguous electron density. Ramachandran outlier Gln120 of chain B of TM0077-SeMet is due to crystal packing with chain C. The backbone carbonyl oxygens of Gln120 and Gly119 of chain B makes hydrogen bonds with the backbone nitrogen of Gln140 of chain C (3.19 and 3.11 Å, respectively). Levisson et al. Page 4 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The native TM0077 structure, TM0077-PMS and TM0077-DEP structures were solved by molecular replacement using PHASER 22,23 with the TM0077-SeMet hexamer coordinates (pdb: 1vlq; A-F chains) as a search model. One hexamer was successfully located and the structure was further refined with Refmac5 20 using tight NCS restraints to an Rcryst and Rfree of 16.7% and 21.2% (native TM0077), 16.0% and 20.8% (TM0077-PMS) and 16.7% and 20.5% (TM0077-DEP), respectively. Iterative cycles of refinement and building were performed with Refmac5, Phenix 24,25 and Coot 26. All other crystallographic manipulations were carried out with the CCP4 package 16. Refinement statistics are summarized in Table I. The final model of native TM0077 contains residues 3-324 (chains A, B, C, D and F) and 3-323 (chain E) in the asymmetric unit. Analysis of main-chain torsion angles using MolProbity 21 showed that 97.8% of the residues are in favored regions of the Ramachandran plot with 0.2% outliers (Asn302 of chains B, C and D), which are supported by unambiguous electron density. The final model of TM0077-PMS contains residues 3-323 for all chains in the asymmetric unit with 97.5% of the residues in favored regions with 0.2% outliers (Asn302 of chains A, B, D and F). The final model of TM0077-DEP contains residues 0-324 for (chains A, B, C and F) and 0-323 (chains D and E) in the asymmetric unit, respectively, with 97.6% of the residues in favored region of the Ramachandran plot with 0.2% outliers (Asn302 of B, C, D and F chains). Ramachandran outlier Asn302 in the TM0077-Native, TM0077-PMS and TM0077-DEP structures is a neighbor to the catalytic triad residue His303 and may reflect a slightly different state for these structures compared to the Se-Met structure. Structure validation and deposition The quality of the crystal structure was analyzed using the JCSG Quality Control server (http://smb.slac.stanford.edu/jcsg/QC). This server processes the coordinates and data through a variety of validation tools including AutoDepInputTool 27 MolProbity 21, WHATIF 5.0 28, RESOLVE 29, MOLEMAN2 30 as well as several in-house scripts, and summarizes the results. Protein quaternary structure analysis were performed using the PISA server 30. Figures were prepared with PyMOL (DeLano Scientific) 31. RMSD values were calculated using the ProCKSI-Server 32. The structural data have been deposited in the RCSB Protein Data Bank (PDB) with accession codes 1vlq for TM0077-SeMet, 3m81 for TM0077-native, 3m83 for TM0077-DEP and 3m82 for TM0077-PMS. Enzyme assays Esterase activity was measured using p-nitrophenyl esters as described previously 13. Briefly, the standard assay consisted of activity measurements with 0.2 mM p-nitrophenyl acetate as substrate in 50 mM citrate-phosphate (pH 6) at 70°C. The p-nitrophenol liberated was measured continuously at 405 nm on a Hitachi U-2001 spectrophotometer with a temperature-controlled cuvette holder. Extinction coefficients of p-nitrophenol were determined prior to each measurement. Kinetic parameters were determined by direct fitting the data, obtained from multiple measurements, to the Michaelis–Menten curve (Tablecurve 2d, version 5.0). The effect of pH on esterase activity was studied in the pH range from 5 to 10. The buffers used were 50 mM citrate-phosphate (pH 5–8) and 50 mM CAPS (3-(cyclohexylamino) 1- propanesulphonic acid) (pH 9.5–10). The pH of the buffers was set at room temperature, and temperature corrections were made using their temperature coefficients: −0.0028 pH/°C for citrate-phosphate buffer and −0.018 pH/°C for CAPS buffer. The effect of temperature on esterase activity was studied in the range of 40–100°C using 0.2 mM p-nitrophenyl acetate as substrate. Enzyme thermostability was determined by incubating the enzyme in a 50 mM Tris-HCl, 150 mM NaCl (pH 7.8) buffer at 90°C and 100°C for various time intervals. Residual activity was determined in the standard assay. Levisson et al. Page 5 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Inhibition kinetics of PMSF and paraoxon were determined as described for the acetylcholinesterase from electric eel 33. All experiments were performed at 70°C in 50 mM citrate-phosphate (pH 6) buffer and 0.2 mM p-nitrophenyl acetate as substrate. The kinetic constants for the inhibition of TM0077 with PMSF and paraoxon were measured in the concentration range of 1.0–10.0 mM and 0.2–1.0 mM, respectively. Deacetylase activity was determined using high-performance liquid chromatography (HPLC) by measuring the amount of acetic acid released from the substrates cephalosporin C, 7-aminocephalosporanic acid, glucose-pentaacetate and acetylated xylan. Xylan was acetylated by the method described by Johnson 34. The reaction mixture contained 0.9 ml of substrate solution (dissolved in 50 mM Tris-HCl, pH 7.5) and 0.1 ml of enzyme solution, and was incubated at 37°C for various time intervals (0–10 min). The reaction was stopped by adding 0.2 ml of stop solution (100 mN H2SO4 and 30 mM crotonate) and placing the sample on ice. The conditions for HPLC were as follows: column, KC811 Shodex; detection, RI and UV detectors; solvent, 3 mN H2SO4; flow rate, 1.5 ml/min; temperature, 30°C; internal standard, crotonate. One unit of enzyme activity was defined as the amount of enzyme that releases one μmol of acetic acid per minute. Activity on xylan was measured quantitatively using DMSO-extracted xylan (1% polysaccharide solution in 0.1 M sodium phosphate buffer pH 6) at 60°C 35. Xylan will precipitate as a consequence of deacetylation, resulting in a rapid turbidity of the solution. Positional specificity assay The positional specificity of TM0077 was investigated using an enzyme-coupled assay on monoacetylated 4-nitrophenyl β-D-xylopyranosides (pNP-Xyl) as described 36. The β- xylosidase XloA (locus tag: TM0076) from T. maritima was cloned into the vector pET24d in frame with a C-terminal His6-tag. The enzyme was expressed and purified as described above for native TM0077. Activity of XloA was confirmed by measuring the release of p- nitrophenol at 405 nm from the substrate 4-nitrophenyl β-D-xylopyranoside. The enzyme-coupled assay was performed at 60°C in a total volume of 125 μl, which contained 0.1 M sodium phosphate (pH 6 or 7), 2-O-, 3-O-, or 4-O-acetyl pNP-Xyl, the β- xylosidase XloA, and TM0077. Stable 50x-concentrated stock solutions of the substrates were prepared in DMSO. The reaction was started by addition of 2.5 μl of a stock solution to a preheated reaction mixture consisting of phosphate buffer, auxiliary β-xylosidase XloA in excess, and TM0077. The reaction was terminated by addition of 800 μl of a 2% solution of Na2CO3. Liberated p-nitrophenol was determined at 405 nm against substrate and enzyme blanks. A short incubation time for activity determination was used to suppress acetyl migration on the xylopyranosyl-ring, which is significant at pH 6 or 7 37. The kinetic constants were determined at pH 7 and 60°C with reaction times of 2 or 5 minutes. RESULTS and DISCUSSION In silico analysis TM0077 consists of 325 amino acids with a calculated molecular mass of 37 kDa. Sequence analysis, using the SignalP 3.0 server, revealed that TM0077 has no predicted signal sequence and is, therefore, believed to be an intracellular enzyme. Analysis of the gene organization indicates that the TM0077 gene co-localizes with genes encoding a xylanase (TM0070) 38, ABC transporter components (TM0071-TM0075), and a β-xylosidase (TM0076) 39. BLAST-P analysis showed that TM0077 has highest similarity to putative acetyl esterases, acetyl xylan esterases and cephalosporin C deacetylases. Among the BLAST results, a predicted acetyl xylan esterase-related protein from T. maritima (locus tag: TM0435) was also identified. TM0077 was compared with other members of the CE7 Levisson et al. Page 6 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript family using structure-based, multi-sequence alignment and the putative catalytic triad, Ser188, Asp274, and His303, was identified from conservation throughout the analyzed sequences. The putative nucleophilic serine (Ser188) is located within a conserved pentapeptide consensus sequence, Gly-Xaa-Ser-Gln-Gly, typical of this family. Previously, a signature sequence motif, [RGQ]-(x:~70)-[GxSQG]-(x:~115)-[HE] (where x indicates any amino acid), had been suggested for the CE7 family based on an aminoacid alignment of 12 sequences 40. In an updated alignment consisting now of 50 sequences, we observed many sequences that have this signature motif, but it is not conserved throughout the entire family (See Supporting Information and Fig. S1 for the multi-sequence alignment). Overall structure The crystal structure of seleno-methionine incorporated TM0077 (TM0077-SeMet) was determined to 2.1 Å resolution by multi-wavelength anomalous dispersion (MAD) (Table I) with twelve molecules per asymmetric unit. A native apo structure (TM0077-Native) was determined in a different space group (see Methods) to 2.5 Å by molecular replacement, using TM0077-SeMet as a search model, with six molecules in the asymmetric unit (Table I). Each monomer of the native hexamer contained a calcium ion (see below) bound by Lys22, Glu26, and Asp25 via a bridging water molecule. Superposition of the TM0077- SeMet and the TM0077-Native structures gave a root-mean-square difference (RMSD) of 0.12 Å over 321 Cα atoms, which indicates that these structures are nearly identical as expected. In general, the TM0077 structure resembles the canonical α/β-hydrolase fold, which consists of a central, twisted, eight-stranded β-sheet surrounded by α-helices on both sides, with β2 antiparallel to the other strands. TM0077 deviates slightly from the canonical α/β- hydrolase fold at two locations: a three-helix insertion after strand β6 and an extension of the N-terminus (Fig. 2). Insertions after β6 or β7 are common for α/β-hydrolases and are proposed to help shape the substrate-binding site 41. The N-terminus is extended by two helices (αA-1 and αA-2) and an antiparallel β-strand (β-1) that aligns with the other eight β- strands (β1-β8) and extends the central β-sheet. This nine-stranded β-sheet is highly twisted, and β-1 and β8 at the extreme edges are rotated approximately 130° relative to each other. Helices αA and αB both contain a short 310-helix segment at their N-terminus. Helices αA-1, αA-2, αB, αC, αD, αD1, αD2, αD3, αE, and the 310-helix η2 are located on one side of the central β-sheet, and helices αA, αF and the 310-helix η1 are on the other side. A structural similarity search was performed using the program DALI 42. Monomer A of the TM0077-SeMet structure was used as a search model and similarity was found with cephalosporin C deacetylases, acetyl xylan esterases, acylamino-releasing enzymes, dipeptidyl peptidases and some esterases and lipases. TM0077 is structurally most similar to cephalosporin C deacetylase (CAH) from B. subtilis (PDB: 1ods) 40, acetyl xylan esterase (AXE) from B. pumilus (PDB: 3fvr and 2xlb), acetyl xylan esterase (AXE1) from Thermoanaerobacterium sp. JW/SL YS485 (PDB: 3fcy), and acylpeptide hydrolase/esterase apAPH from Aeropyrum pernix K1 (PDB: 1ve6) 43. The sequence identity between TM0077 and CAH is 41% and the two structures align with a Z-score of 46 and an RMSD of 1.5 Å over 312 Cα atoms. The sequence identity with apAPH is 17% with a Z-score of 23.3 and an RMSD of 2.3 Å over 230 Cα atoms. Superpositions of TM0077 with CAH and with apAPH are shown in Fig. 3. Quaternary structure The crystal structure of TM0077-SeMet contains two hexamers in the asymmetric unit that are related by a non-crystallographic two-fold axis. Each hexamer contains a dimer of trimers with a back-to-back arrangement (Fig. 4). The apo and the complex crystals Levisson et al. Page 7 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript contained one hexamer in the asymmetric unit. Crystallographic packing analysis using PISA (EBI) 44 indicated that the relevant physiological oligomeric state of TM0077 is a hexamer, which was confirmed by size exclusion chromatography coupled with static light scattering. Further analyses of the hexameric assembly indicated that two main interfaces play an essential role in complex formation. The first interface between subunit A and B (green and cyan in Fig. 4) (identical to C/D and E/F) is stabilized by seven hydrogen bonds on average and has a buried surface area of 1024 Å2 contributed by each chain. The second interface between A and F (green and purple in Fig. 4) (B/C and D/E) is stabilized by 17 hydrogen bonds on average with a buried surface area of 1079 Å2 contributed by each chain However, a multiple sequence alignment of TM0077 with other CE7 esterases showed that the residues involved in these two main interfaces are not conserved. Other secondary interfaces bury around 514 Å2 contributed by each chain. The hexamer has a total buried surface area of 18,860 Å2, which is approximately 30% of the total surface area. Approximately 3,143 Å2 per monomer is, therefore, buried upon complex formation. The TM0077 hexamer has a doughnut-shape when viewed from the side, with the six active sites located in the interior of the complex, where they line an oval-shaped cavity [Fig. 4(a)]. This cavity is accessible via two entrances, one on each side of the flat hexamer. Each of these entrances is approximately 13 Å wide and connects to a short tunnel or pore spanning approximately 10 Å to reach the inner cavity. Interestingly, in the TM0077-SeMet hexamer, the entrance to the internal cavity is blocked by three phenylalanine residues (Phe4), one for each of three monomers that compose half of the hexamer [Fig. 4(b)]. Residue Phe4 is located in the mobile N-terminus (high B-values), which may indicate some flexibility or multiple conformations. Calcium ions were identified, by the electron density and coordination geometry, supported by their presence in the crystallization reagents, in the native TM0077, TM0077-PMS and TM0077-DEP structures, but not in the TM0077-SeMet structure. The SeMet protein was crystallized without any calcium in the crystallization reagents. In each subunit of the hexamer, one calcium ion is located at the N-terminal region of helix αA-1, and is coordinated by the backbone carbonyl of Lys22 and the Glu26 carboxylate. The remainder of the calcium coordination sphere is filled with waters from a neighboring solvent channel present in all molecules in the asymmetric unit. The Asp25 carboxylate contributes to the calcium binding via one of the coordinating water molecules. Another calcium ion is bound in a crystal packing interface between chain A and chain C′ of a crystallographic symmetry- related hexamer. This calcium is coordinated by the carboxylates of GluA45 and AspA58 from one chain and the carboxylate from Glu C’45 (bidentate coordination) of the symmetry-related chain with three water molecules completing a capped-octahedral coordination sphere. An equivalent calcium binding site is also observed in the crystal packing interface between chains D and B′. No significant increase or reduction of activity of TM0077 was observed in the presence of calcium ions or EDTA. Therefore, it seems that these calcium ions are not important for activity. On the other hand, calcium may help stabilize the structure. No calcium was present in the B. subtilis CAH structure 40; however, Lys22, Glu26 and Ser25 are conserved and may also act as a calcium binding site. Enzyme activity The activity of TM0077 was investigated using p-nitrophenol esters with varying acyl-chain length, ranging from C2 to C18. TM0077 is only active on the short-chain p-nitrophenol esters of acetate and propionate and does not hydrolyze esters with acyl chains longer than four carbons. No significant difference was found in the catalytic efficiency (kcat/Km) for the hydrolysis of p-nitrophenyl with acyl chains containing 2 to 3 carbons (Table II) [Fig. 1(c)]. Levisson et al. Page 8 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The effect of temperature on activity was studied using p-nitrophenyl acetate as substrate. The esterase activity increased from 40°C upwards until 100°C [Fig. 5(a)]. An Arrhenius analysis resulted in linear plots in the temperature ranges of 40–60°C and 60–100°C [Fig. 5(a); inset], with calculated activation energies for the formation of the enzyme substrate- enzyme complex of 33.7 and 21.9 kJ/mol, respectively. The transition or break in linearity of the Arrhenius plot at 60°C (1000/T (K) = 3.0) could indicate some conformational change of the enzyme. TM0077 is fairly resistant to thermal inactivation. An approximate 50% transient increase in activity is seen during the first 10 to 20 minutes when the enzyme is incubated at 90°C. After 30 minutes, inactivation of function occurs by first order kinetics with a half-life of approximately 120 minutes [Fig. 5(b)]. A transient activation has also been observed for other thermophilic esterases, such from Sulfolobus shibatae 45, and it is believed that a high temperature is needed in order to obtain the optimal conformation for catalysis. TM0077 was not stable at 100°C, resulting in a half-life of less than 5 minutes. However, the optimum temperature and thermal stability of TM0077 are still considerably higher than those reported for other characterized CE7 esterases, including the Thermoanaerobacterium enzyme that has a temperature optimum of 80°C and a half-life of 1h at 75°C 46. The effect of pH on activity was measured in the pH range of 4.8 to 9.2 using the substrate p-nitrophenyl acetate. TM0077 displayed maximum activity at approximately pH 7.5 [Fig. 5(c)], which is comparable to other CE7 esterases, such as the acetyl xylan esterases from Thermoanaerobacterium sp. strain JW/SL-YS485 46. Positional specificity The positional specificity of TM0077 was tested on three monoacetates of 4-nitrophenyl β- D-xylopyranoside (pNP-Xyl). To determine the enzyme activity, the β-xylosidase XloA 39 (TM0076) from T. maritima is required as an auxiliary enzyme. This thermostable XloA enzyme was, therefore, cloned, heterologously expressed, purified to homogeneity, and its activity was confirmed by measuring release of p-nitrophenol from the substrate pNP-Xyl (data not shown). The β-xylosidase was not active on the three monoacetates of pNP-Xyl. In the XloA-coupled assay, TM0077 hydrolyzed acetate from positions 2, 3 and 4 of pNP-Xyl with similar catalytic efficiency. The results are summarized in Table II. In addition, TM0077 was investigated for its ability to remove acetyl groups from 7- aminocephalosporanic acid (7-ACA), cephalosporin C, glucose penta-acetate, N-acetyl-D- glucosamine, xylan and acetylated xylan. TM0077 has no activity for acetylated and non- acetylated xylan polymers, indicating that it is, indeed, an acetyl esterase and not an acetyl xylan esterase. As expected for an acetyl esterase, TM0077 displayed high activity on glucose penta-acetate with a turnover number of 2680 s−1. Like other members of CE7, TM0077 was also able to hydrolyze the acetyl groups from both cephalosporin C and 7- ACA with a turnover number of 376 s−1 and 1140 s−1, respectively. TM0077 was not able to hydrolyze the acetyl group from N-acetyl-D-glucosamine, indicating that it is specific for ester bonds and unable to hydrolyze amide bonds. Inhibitor assays and TM0077 structures complexed with PMSF and paraoxon PMSF and paraoxon [Fig. 1(d,e)] are competitive irreversible inhibitors of esterases. Inhibition proceeds by the formation of a reversible Michaelis complex, followed by an irreversible step and inhibition can, therefore, be characterized by two parameters: a dissociation constant and a binding rate constant. The inhibition kinetics for paraoxon and PMSF were investigated in the presence of p-nitrophenyl acetate, as described previously 47, and the dissociation and rate constants were 0.5 ± 0.1 mM and 0.13 ± 0.02 s−1 for paraoxon, and 1.1 ± 0.2 mM and 0.020 ± 0.001 s−1 for PMSF, respectively. The acetyl xylan esterase Levisson et al. Page 9 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript from Bacillus pumilus (BpAxe) has slightly reduced sensitivity to paraoxon (dissociation and rate constants of respectively 5.4 mM and 0.012 s−1), likely due to steric hindrance of two tyrosine residues (Tyr91 and Tyr206) that hamper the binding of paraoxon. Although these residues are essentially conserved in TM0077 (Tyr92 and Phe213), TM0077 is more sensitive to paraoxon than BpAxe48. In comparison to EST2 of Alicyclobacillus acidocaldarius 49 and EstA of T. maritima 47, the TM0077 dissociation constant is slightly higher, but the rate constant is comparable. No significant stimulation or reduction of activity of TM0077 was observed in the presence of divalent metal ions or ethylenediaminetetraacetic acid (EDTA). To obtain more information about inhibitor binding and any possible conformational changes during catalysis, TM0077 was co-crystallized with the inhibitors PMSF and paraoxon and the PMSF (TM0077-PMS) and paraoxon (TM0077-DEP) structures were determined to 2.4 Å and 2.1 Å, respectively (Table I). The electron density map of TM0077 with PMSF showed clear density for the PMSF covalent modification. The fluorine was cleaved from the PMSF molecule during the binding reaction and the phenylmethyl sulfonyl (PMS) moiety is covalently bound to the Oγ atom of Ser188. The native apo and PMS- bound structures superimpose well with RMSD’s of 0.09–0.11 Å over 320–321 Cα atoms. Electron density maps of the paraoxon-bound structure displayed clear density for a diethyl- phosphate moiety covalently bound to the Oγ atom of Ser188. This covalent modification indicates that the p-nitrophenol group of paraoxon was cleaved off during co-crystallization, and a tetrahedral product reminiscent of the first transition state was formed during carboxyl ester hydrolysis. The native apo and paraoxon-bound structures superimpose with RMSD’s of 0.12–0.32 Å over 320–322 Cα atoms. Attempts to obtain co-crystals of TM0077 with cephalosporin C, even at a low temperature of 4°C, were unsuccessful. Analysis of the active site TM0077 has a classic catalytic triad, consisting of Ser188 as the nucleophile, His303 as the proton acceptor/donor, and Asp274 as the acidic residue stabilizing the histidine (Fig. 6). The catalytic serine Ser188 is located within a conserved pentapeptide sequence, Gly-X-Ser- X-Gly (GGSQG), characteristic of esterases and lipases. The positions of Ser188, Asp274, and His303 are consistent with their expected locations in the canonical fold of the α/β- hydrolase family. Ser188 is located at the nucleophile elbow in a sharp turn between β5 and helix αC. The presence of three glycine residues (Gly186, Gly187, and Gly190) in close proximity to Ser188 prevents steric hindrance and facilitates access to the nucleophile elbow. Asp274 and His303 are located in loops between β7 and helix αE, and between β8 and helix αF, respectively. The oxyanion hole is formed by the backbone amide groups of Tyr92 and Gln189. The catalytic triad and oxyanion hole are located in a depression on the surface of TM0077. This ellipsoid pocket (S1), which is approximately 12 Å wide, extends 15 Å from the catalytic serine. A smaller pocket (S2), approximately 5 Å long, extends to the other side of the catalytic serine [Fig. 6(a)]. The volume of both pockets combined (S1 + S2) is 1082 Å3 (CASTp analysis; 50). The substrate-binding pocket is bordered by residues from helices αA and αF, and its base is formed by residues from β-strands 4, 5, 6, and their adjacent C-terminal loops. The overall pocket is hydrophobic, although it does have some polar residues (Gln88, Asp210, and Gln314), which may interact with the substrate. In the native apo structure, the Ser188 hydroxyl makes a hydrogen bond with the imidazole of His303 [Fig. 6(b)]. Extra density was observed near the side chain of Ser188 and was interpreted as a chloride ion based on electron density size and shape as well as the geometry of the interactions with surrounding residues. This chloride ion is bound at the entrance of the oxyanion hole, forming hydrogen bonds with the backbone amides of Tyr92 and Gln189. In the PMSF-bound structure, the phenyl ring of the inhibitor is located in the small active site groove surrounded by hydrophobic residues Tyr92, Trp124, Pro228, Ile276, Levisson et al. Page 10 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript and His303 [Fig. 6(c)]. The sulfonyl group of PMSF makes hydrogen bonds with the backbone amides of Tyr92 and Gln189. In the paraoxon-bound structure, the diethyl- phosphate (DEP) moiety is stabilized by hydrogen-bonding interactions with the oxyanion hole. One of the two ethyl arms of bound paraoxon points toward the larger pocket in the protein, while the other follows the groove of the small pocket. The two ethyl arms are stabilized by packing against Tyr92, Trp124, Pro228, Ile276, and His303 [Fig. 6(d)]. Two rotamers of the catalytic serine Although no large conformational changes were observed upon binding of PMSF or paraoxon, a different rotamer of the catalytic serine side chain was observed compared to native TM0077 [Fig. 7(a,b)]. Similar changes have been observed in several other esterases and have been shown to play a key role in the catalytic mechanism (see CONCLUSION for more details). In the native structure, the catalytic Ser188 Oγ is in the plane of the imidazole ring of His303, as most commonly observed in the resting state of esterases and lipases 51. The Ser188 Oγ forms a hydrogen bond (2.6 Å) with His303 Nε2. In the PMSF- and paraoxon-bound structures, the conformation of the catalytic serine changes; the Ser188 Oγ rotates about 110°, increasing the distance (3.1 Å and 2.8 Å for PMSF and paraoxon bound structures, respectively) to the His303 imidazole ring. In the TM0077-SeMet structure, the catalytic serine is also rotated over ~110°, with a distance to the imidazole ring of 3.0 Å [Fig. 7(c.)]. A probable explanation for this observation could be the protonation of His303, since TM0077-SeMet was crystallized at a lower pH (pH 4.2) compared to the native TM0077 (pH 7.3). Furthermore, extra electron density was identified in the TM0077-SeMet structure, suggesting a partially occupied acyl intermediate on Ser188. However, as this density is not sufficiently clear and interpretable to fit an acyl intermediate, water molecules were modeled instead. No rearrangements of any other residues in the active site were observed. CONCLUSION TM0077 from the hyperthermophilic bacterium T. maritima was predicted from its gene sequence to be an acetyl xylan esterase. We have expressed and purified TM0077 and experimentally demonstrated that it has ester-hydrolyzing activity. The TM0077 activity was restricted to short acyl chain esters (C2 and C3) when artificial p-nitrophenyl-esters were used as substrates. In addition, the enzyme has high specific activity on glucose penta- acetate. However, no activity was detected on xylan or acetylated xylan. Thus, TM0077 should be reclassified as an acetyl esterase, and not as an acetyl xylan esterase as currently annotated 52. Furthermore, the lack of any apparent signal sequence suggests that the protein is not secreted. Thus, the predicted intracellular location of TM0077 is compatible with a role other than the deacetylation of extracellular xylan. Based on these results, we conclude that the likely biological function of TM0077 is removal of the remaining acetyl groups from the short, end products of xylan degradation that are imported into the cytoplasm. The resulting deacetylated xylose oligomers are the substrates for a β-xylosidase. This role for TM0077 is in good agreement with the clustering of the TM0077 gene with other genes involved in xylose metabolism. However, it cannot be ruled out that TM0077 may also act on other small, acetylated compounds. TM0077 is the first esterase from the CE7 family to be tested for its positional specificity for the deacetylation of 4-nitrophenyl-β-D-xylopyranoside. TM0077 hydrolyzes acetate at the 2, 3 and 4 positions of 4-nitrophenyl-β-D-xylopyranoside with similar efficiency. Conversely, the CtAxe esterase from Clostridium thermocellum in the CE4 family shows a clear preference for hydrolyzing acetate at the 2 position 53, and Penicillium purpurogenum AXE II esterase, a member of the CE5 family, also has a preference for acetate at position 2 54. Levisson et al. Page 11 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript This lack of preference for a specific position of the acetate group correlates with the relative broad substrate specificity of the CE7 esterases. Esterases and deacetylases in the CE7 family are unusual in that they are active towards both acetylated xylo-oligosaccharides and the antibiotic cephalosporin C [Fig. 1(a,b)]. Therefore, TM0077 was investigated for activity towards the substrates 7-ACA and cephalosporin C. The activity of TM0077 on these substrates is approximately ten-fold higher than that of the acetyl xylan esterase from B. pumilus 55 or the acetyl esterase from Thermoanaerobacterium sp. strain JW/SL YS485 56. TM0077 has a higher hydrolytic activity on 7-ACA compared to cephalosporin C, as described for other CE7 esterases 40,55,56. Nonetheless, it is unlikely that both compounds are natural substrates, because the stability of these compounds at the optimal growth temperature (80°C) of T. maritima is very low. Crystal structures of TM0077 in complex with inhibitors PMSF and paraoxon revealed that, upon binding of PMSF or paraoxon, the reaction is trapped at the acylation step via the formation of a covalent tetrahedral reaction product. In the complexed structures, the negatively charged oxygen of the tetrahedral intermediate, derived from the substrate oxyanion, is stabilized by hydrogen bonds with the backbone amide groups of Tyr92 and Gln189. Comparison of the TM0077 complexed structures with the native structure shows that the catalytic serine (Ser188) Oγ rotates about 110°, thereby increasing the distance between Ser188 Oγ and His303 Nε2. Such a conformational change of the catalytic serine has been observed in several other esterases, including Fusarium solani cutinase 57, Penicillium purpurogenum acetyl xylan esterase 51, Rhodococcus sp. strain MB1 cocaine esterase 58, Bacillus subtilis lipase 59, Rhodococcus sp. strain H1 heroine esterase 60, and Aspergillus niger feruloyl esterase 61. The classical model for the catalytic mechanism of esterases consists of a sequential two-step hydrolysis. The first reaction involves nucleophilic attack by the catalytic serine on the substrate carbonyl carbon, resulting in an acyl-enzyme and the liberation of an alcohol. In the second reaction, a water molecule performs a nucleophilic attack on the acyl-enzyme, the acyl-enzyme bond breaks and the carboxylate is released 62. Although the catalytic mechanism of esterases is well established, it is unclear why the initially generated tetrahedral intermediate does not collapse back to the reactant complex during the nucleophilic attack of the substrate. A previously proposed mechanism that would prevent this collapse is the spatial reorganization of the catalytic residues during the initial catalytic step, causing the residues to separate and thereby drive the reaction forward 62–64. The apo and inhibitor bound structures of TM0077, presented herein, support this proposed mechanism. Moreover, in a recent study of the serine protease mechanism, it was suggested that subtle atomic motions of the catalytic serine and histidine residues during the catalytic cycle favor the forward reaction 65. Thus, rotation of Ser188 Oγ of TM0077 may be required to inhibit reversal of the reaction. In addition, such changes may facilitate the access of water to the catalytic histidine so that the second step of the reaction can go to completion. Deacetyl cephalosporins are valuable building blocks for the production of semisynthetic β- lactam antibiotics. These compounds are derived from cephalosporin C or 7- aminocephalosporanic acid via enzymatic or chemical processes 10. The thermostable TM0077 esterase may be valuable in the preparation of derivatives of β-lactam antibiotics. Recently, the substrate specificity of the acetyl xylan esterase from P. purpurogenum was engineered to accept a range of fatty acid esters of up to 14 carbons compared to its wild- type preference for acetate54. It might also be possible to engineer TM0077 and enable the (de)acetylation of cephalosporins at the C10 position with various acyl chains. Because of its high stability and activity on 7-ACA and cephalosporin C, TM0077 presents an attractive candidate for the production of new semi-synthetic antibiotics. Levisson et al. Page 12 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We gratefully acknowledge contributions from George Sheldrick for modifications of the SHELXD program, and for Global Phasing Ltd. that made significant improvements in the automation of autoSHARP. We also thank Victor Lamzin for updates of chain docking of the ARP/wARP program, and Gerard Bricogne and Eleanor Dodson for helpful discussion on phasing for the large TM0077-SeMet structure, and Willem J. van Berkel for valuable discussion on the catalytic mechanism of TM0077. Portions of this research were carried out at the Stanford Synchrotron Radiation Lightsource (SSRL) and the Advanced Light Source (ALS). The SSRL is a Directorate of SLAC National Accelerator Laboratory and an Office of Science User Facility operated for the U.S. Department of Energy Office of Science by Stanford University. The SSRL Structural Molecular Biology Program is supported by the DOE Office of Biological and Environmental Research, and by the National Institutes of Health, National Center for Research Resources, Biomedical Technology Program (P41RR001209), and the National Institute of General Medical Sciences. The ALS is supported by the Director, Office of Science, Office of Basic Energy Sciences, Materials Sciences Division, of the U.S. Department of Energy under Contract No. DE- AC02-05CH11231 at Lawrence Berkeley National Laboratory. Genomic DNA from Thermotoga maritima MSB8 (DSM3109) (ATCC #43589D-5) was obtained from the American Type Culture Collection (ATCC). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of General Medical Sciences or the National Institutes of Health. Grant sponsor: NIH Grant numbers U54 GM094586 and U54 GM074898 (Protein Structure Initiative); Grant sponsor: The Graduate School VLAG Wageningen, the Netherlands (ML). References 1. Huber, R.; Hannig, M. Thermotogales. In: Dworkin, M.; Falkow, S.; Rosenberg, E.; Schleifer, K-H.; Stackebrandt, E., editors. The Prokaryotes: An Evolving Electronic Resource for the Microbiological Community. Vol. 7. New-York: Springer-Verlag; 2006. p. 899-922. 2. Nelson KE, Clayton RA, Gill SR, Gwinn ML, Dodson RJ, Haft DH, Hickey EK, Peterson JD, Nelson WC, Ketchum KA, McDonald L, Utterback TR, Malek JA, Linher KD, Garrett MM, Stewart AM, Cotton MD, Pratt MS, Phillips CA, Richardson D, Heidelberg J, Sutton GG, Fleischmann RD, Eisen JA, White O, Salzberg SL, Smith HO, Venter JC, Fraser CM. Evidence for lateral gene transfer between Archaea and bacteria from genome sequence of Thermotoga maritima. Nature. 1999; 399:323–329. [PubMed: 10360571] 3. Chhabra SR, Shockley KR, Conners SB, Scott KL, Wolfinger RD, Kelly RM. Carbohydrate- induced differential gene expression patterns in the hyperthermophilic bacterium Thermotoga maritima. J Biol Chem. 2003; 278:7540–7552. [PubMed: 12475972] 4. Conners SB, Montero CI, Comfort DA, Shockley KR, Johnson MR, Chhabra SR, Kelly RM. An expression-driven approach to the prediction of carbohydrate transport and utilization regulons in the hyperthermophilic bacterium Thermotoga maritima. J Bacteriol. 2005; 187:7267–7282. [PubMed: 16237010] 5. VanFossen AL, Lewis DL, Nichols JD, Kelly RM. Polysaccharide degradation and synthesis by extremely thermophilic anaerobes. Ann N Y Acad Sci. 2008; 1125:322–337. [PubMed: 18378602] 6. Biely P, Mackenzie CR, Puls J, Schneider H. Cooperativity of esterases and xylanases in the enzymatic degradation of acetyl xylan. Bio-Technology. 1986; 4:731–733. 7. Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V, Henrissat B. The Carbohydrate- Active EnZymes database (CAZy): an expert resource for Glycogenomics. Nucleic Acids Res. 2009; 37:D233–D238. [PubMed: 18838391] 8. Topakas E, Paul C. Microbial xylanolytic carbohydrate esterases. Industrial Enzymes. 2007:83–97. 9. Weil J, Miramonti J, Ladisch MR. Cephalosporin-C: Mode of action and biosynthetic pathway. Enzyme Microb Technol. 1995; 17:85–87. 10. Barends TRM, Yoshida H, Dijkstra BW. Three-dimensional structures of enzymes useful for beta- lactam antibiotic production. Curr Opin Biotechnol. 2004; 15:356–363. [PubMed: 15358004] 11. Martínez-Martínez I, Montoro-García S, Lozada-Ramírez JD, Sánchez-Ferrer Á, García-Carmona F. A colorimetric assay for the determination of acetyl xylan esterase or cephalosporin C acetyl Levisson et al. Page 13 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript esterase activities using 7-amino cephalosporanic acid, cephalosporin C, or acetylated xylan as substrate. Anal Biochem. 2007; 369:210–217. [PubMed: 17651681] 12. Lesley SA, Kuhn P, Godzik A, Deacon AM, Mathews I, Kreusch A, Spraggon G, Klock HE, McMullan D, Shin T, Vincent J, Robb A, Brinen LS, Miller MD, McPhillips TM, Miller MA, Scheibe D, Canaves JM, Guda C, Jaroszewski L, Selby TL, Elsliger MA, Wooley J, Taylor SS, Hodgson KO, Wilson IA, Schultz PG, Stevens RC. Structural genomics of the Thermotoga maritima proteome implemented in a high-throughput structure determination pipeline. Proc Natl Acad Sci USA. 2002; 99:11664–11669. [PubMed: 12193646] 13. Levisson M, van der Oost J, Kengen SW. Characterization and structural modeling of a new type of thermostable esterase from Thermotoga maritima. FEBS Journal. 2007; 274:2832–2842. [PubMed: 17466017] 14. Otwinowski Z, Minor W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 1997; 276:307–326. 15. Schneider TR, Sheldrick GM. Substructure solution with SHELXD. Acta Crystallogr Sect D Biol Crystallogr. 2002; 58:1772–1779. [PubMed: 12351820] 16. Collaborative Computational Project Number 4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr Sect D Biol Crystallogr. 1994; 50:760–763. [PubMed: 15299374] 17. Bricogne G, Vonrhein C, Flensburg C, Schiltz M, Paciorek W. Generation, representation and flow of phase information in structure determination: recent developments in and around SHARP 2.0. Acta Crystallogr Sect D Biol Crystallogr. 2003; 59:2023–2030. [PubMed: 14573958] 18. Bricogne, G.; Blanc, E.; Brandl, M.; Flensburg, C.; Keller, P.; Paciorek, W.; Roversi, P.; Smart, O.; Vonrhein, CTW. BUSTER, version 2.8.0. Cambridge, United Kingdom: Global Phasing Ltd; 2009. 19. Jones TA, Zou JY, Cowan SW, Kjeldgaard M. Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr A Found Crystallogr. 1991; 47:110–119. 20. Winn MD, Murshudov GN, Papiz MZ. Macromolecular TLS refinement in REFMAC at moderate resolutions. Methods Enzymol. 2003; 374:300–321. [PubMed: 14696379] 21. Davis IW, Murray LW, Richardson JS, Richardson DC. MOLPROBITY: structure validation and all-atom contact analysis for nucleic acids and their complexes. Nucleic Acids Res. 2004; 32:W615–W619. [PubMed: 15215462] 22. McCoy AJ, Grosse-Kunstleve RW, Storoni LC, Read RJ. Likelihood-enhanced fast translation functions. Acta Crystallogr Sect D Biol Crystallogr. 2005; 61:458–464. [PubMed: 15805601] 23. Storoni LC, McCoy AJ, Read RJ. Likelihood-enhanced fast rotation functions. Acta Crystallogr Sect D Biol Crystallogr. 2004; 60:432–438. [PubMed: 14993666] 24. Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Terwilliger TC. PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr Sect D Biol Crystallogr. 2002; 58:1948– 1954. [PubMed: 12393927] 25. Murshudov GN, Vagin AA, Dodson EJ. Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr Sect D Biol Crystallogr. 1997; 53:240–255. [PubMed: 15299926] 26. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr Sect D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] 27. Yang H, Guranovic V, Dutta S, Feng Z, Berman HM, Westbrook JD. Automated and accurate deposition of structures solved by X-ray diffraction to the Protein Data Bank. Acta Crystallogr Sect D Biol Crystallogr. 2004; 60:1833–1839. [PubMed: 15388930] 28. Vriend G. WHAT IF: a molecular modeling and drug design program. J Mol Graph. 1990; 8:52– 56. [PubMed: 2268628] 29. Terwilliger TC. Automated side-chain model building and sequence assignment by template matching. Acta Crystallogr Sect D Biol Crystallogr. 2003; 59:45–49. [PubMed: 12499538] 30. Kleywegt GJ. Validation of protein crystal structures. Acta Crystallogr Sect D Biol Crystallogr. 2000; 56:249–265. [PubMed: 10713511] Levisson et al. Page 14 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 31. DeLano, WL. The Pymol molecular graphics system. DeLano Scientific; San Carlos, CA: 2002. 32. Barthel D, Hirst JD, Blazewicz J, Burke EK, Krasnogor N. ProCKSI: a decision support system for Protein (structure) Comparison, Knowledge, Similarity and Information. BMC Bioinformatics. 2007; 8:416. [PubMed: 17963510] 33. Forsberg A, Puu G. Kinetics for the inhibition of acetylcholinesterase from the electric eel by some organophosphates and carbamates. Eur J Biochem. 1984; 140:153–156. [PubMed: 6705793] 34. Johnson KG, Harrison BA, Schneider H, Mackenzie CR, Fontana JD. Xylan-hydrolyzing enzymes from Streptomyces spp. Enzyme Microb Technol. 1988; 10:403–409. 35. Biely P, Mackenzie CR, Schneider H. Production of acetyl xylan esterase by Trichoderma reesei and Schizophyllum commune. Can J Microbiol. 1988; 34:767–772. 36. Biely P, Mastihubova M, la Grange DC, van Zyl WH, Prior BA. Enzyme-coupled assay of acetylxylan esterases on monoacetylated 4-nitrophenyl beta-D-xylopyranosides. Anal Biochem. 2004; 332:109–115. [PubMed: 15301955] 37. Mastihubova M, Biely P. Lipase-catalysed preparation of acetates of 4-nitrophenyl beta-D- xylopyranoside and their use in kinetic studies of acetyl migration. Carbohydr Res. 2004; 339:1353–1360. [PubMed: 15113674] 38. Jiang ZQ, Kobayashi A, Ahsan MM, Lite L, Kitaoka M, Hayashi K. Characterization of a thermostable family 10 endo-xylanase (XynB) from Thermotoga maritima that cleaves p- nitrophenyl-beta-D-xyloside. J Biosci Bioeng. 2001; 92:423–428. [PubMed: 16233122] 39. Xue YM, Shao WL. Expression and characterization of a thermostable beta-xylosidase from the hyperthermophile, Thermotoga maritima. Biotechnol Lett. 2004; 26:1511–1515. [PubMed: 15604789] 40. Vincent F, Charnock SJ, Verschueren KHG, Turkenburg JP, Scott DJ, Offen WA, Roberts S, Pell G, Gilbert HJ, Davies GJ, Brannigan JA. Multifunctional xylooligosaccharide/cephalosporin C deacetylase revealed by the hexameric structure of the Bacillus subtilis enzyme at 1. 9 angstrom resolution. J Mol Biol. 2003; 330:593–606. [PubMed: 12842474] 41. Nardini M, Dijkstra BW. Alpha/beta hydrolase fold enzymes: the family keeps growing. Curr Opin Struct Biol. 1999; 9:732–737. [PubMed: 10607665] 42. Holm L, Kaariainen S, Rosenstrom P, Schenkel A. Searching protein structure databases with DaliLite v.3. Bioinformatics. 2008; 24:2780–2781. [PubMed: 18818215] 43. Bartlam M, Wang G, Yang H, Gao R, Zhao X, Xie G, Cao S, Feng Y, Rao Z. Crystal structure of an acylpeptide hydrolase/esterase from Aeropyrum pernix K1. Structure. 2004; 12:1481–1488. [PubMed: 15296741] 44. Krissinel E, Henrick K. Inference of macromolecular assemblies from crystalline state. J Mol Biol. 2007; 372:774–797. [PubMed: 17681537] 45. Huddleston S, Yallop CA, Charalambous BM. The identification and partial characterisation of a novel inducible extracellular thermostable esterase from the archaeon Sulfolobus shibatae. Biochem Biophys Res Commun. 1995; 216:495–500. [PubMed: 7488139] 46. Shao WL, Wiegel J. Purification and characterization of 2 thermostable acetyl xylan esterases from Thermoanaerobacterium sp strain JW/SL-YS485. Appl Environ Microbiol. 1995; 61:729–733. [PubMed: 7574610] 47. Levisson M, Sun L, Hendriks S, Swinkels P, Akveld T, Bultema JB, Barendregt A, van den Heuvel RH, Dijkstra BW, van der Oost J, Kengen SW. Crystal structure and biochemical properties of a novel thermostable esterase containing an immunoglobulin-like domain. J Mol Biol. 2009; 385:949–962. [PubMed: 19013466] 48. Montoro-García S, Gil-Ortiz F, García-Carmona F, Polo LM, Rubio VAS-F. The crystal structure of the cephalosporin deacetylating enzyme acetyl xylan esterase bound to paraoxon explains the low sensitivity of this serine hydrolase to organophosphate inactivation. Biochem J. 201110.1042/ BJ20101859 49. Febbraio F, D’Andrea SE, Mandrich L, Merone L, Rossi M, Nucci R, Manco G. Irreversible inhibition of the thermophilic esterase EST2 from Alicyclobacillus acidocaldarius. Extremophiles. 2008; 12:719–728. [PubMed: 18622571] Levisson et al. Page 15 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 50. Dundas J, Ouyang Z, Tseng J, Binkowski A, Turpaz Y, Liang J. CASTp: computed atlas of surface topography of proteins with structural and topographical mapping of functionally annotated residues. Nucleic Acids Res. 2006; 34:W116–W118. [PubMed: 16844972] 51. Ghosh D, Sawicki M, Lala P, Erman M, Pangborn W, Eyzaguirre J, Gutierrez R, Jornvall H, Thiel DJ. Multiple conformations of catalytic serine and histidine in acetylxylan esterase at 0.90 angstrom. J Biol Chem. 2001; 276:11159–11166. [PubMed: 11134051] 52. Williamson G, Kroon PA, Faulds CB. Hairy plant polysaccharides: a close shave with microbial esterases. Microbiology. 1998; 144:2011–2023. [PubMed: 9720023] 53. Biely P, Mastihubova M, Puchart V. The vicinal hydroxyl group is prerequisite for metal activation of Clostridium thermocellum acetylxylan esterase. Biochim Biophys Acta. 2007; 1770:565–570. [PubMed: 17261352] 54. Colombres M, Garate JA, Lagos CF, Araya-Secchi R, Norambuena P, Quiroz S, Larrondo L, Perez-Acle T, Eyzaguirre J. An eleven amino acid residue deletion expands the substrate specificity of acetyl xylan esterase II (AXE II) from Penicillium purpurogenum. J Comput Aided Mol Des. 2008; 22:19–28. [PubMed: 18060506] 55. Degrassi G, Kojic M, Ljubijankic G, Venturi V. The acetyl xylan esterase of Bacillus pumilus belongs to a family of esterases with broad substrate specificity. Microbiology. 2000; 146:1585– 1591. [PubMed: 10878123] 56. Lorenz WW, Wiegel J. Isolation, analysis, and expression of two genes from Thermoanaerobacterium sp. strain JW/SL YS485: a beta-xylosidase and a novel acetyl xylan esterase with cephalosporin C deacetylase activity. J Bacteriol. 1997; 179:5436–5441. [PubMed: 9286998] 57. Longhi S, Czjzek M, Lamzin V, Nicolas A, Cambillau C. Atomic resolution (1. 0 angstrom) crystal structure of Fusarium solani cutinase: Stereochemical analysis. J Mol Biol. 1997; 268:779–799. [PubMed: 9175860] 58. Larsen NA, Turner JM, Stevens J, Rosser SJ, Basran A, Lerner RA, Bruce NC, Wilson IA. Crystal structure of a bacterial cocaine esterase. Nat Struct Biol. 2002; 9:17–21. [PubMed: 11742345] 59. Kawasaki K, Kondo H, Suzuki M, Ohgiya S, Tsuda S. Alternate conformations observed in catalytic serine of Bacillus subtilis lipase determined at 1. 3 angstrom resolution. Acta Crystallogr Sect D Biol Crystallogr. 2002; 58:1168–1174. [PubMed: 12077437] 60. Zhu X, Larsen NA, Basran A, Bruce NC, Wilson IA. Observation of an arsenic adduct in an acetyl esterase crystal structure. J Biol Chem. 2003; 278:2008–2014. [PubMed: 12421810] 61. McAuley KE, Svendsen A, Patkar SA, Wilson KS. Structure of a feruloyl esterase from Aspergillus niger. Acta Crystallogr Sect D Biol Crystallogr. 2004; 60:878–887. [PubMed: 15103133] 62. Hedstrom L. Serine protease mechanism and specificity. Chem Rev. 2002; 102:4501–4523. [PubMed: 12475199] 63. Ash EL, Sudmeier JL, Day RM, Vincent M, Torchilin EV, Haddad KC, Bradshaw EM, Sanford DG, Bachovchin WW. Unusual 1H NMR chemical shifts support (His) Cε1-H···O=C H-bond: Proposal for reaction-driven ring flip mechanism in serine protease catalysis. Proc Natl Acad Sci USA. 2000; 97:10371–10376. [PubMed: 10984533] 64. Bizzozero SA, Dutler H. Stereochemical aspects of peptide hydrolysis catalyzed by serine proteases of the chymotrypsin type. Bioorg Chem. 1981; 10:46–62. 65. Radisky ES, Lee JM, Lu CJK, Koshland DE Jr. Insights into the serine protease mechanism from atomic resolution structures of trypsin reaction intermediates. Proc Natl Acad Sci USA. 2006; 103:6835–6840. [PubMed: 16636277] 66. Cruickshank DW. Remarks about protein structure precision. Acta Crystallogr Sect D Biol Crystallogr. 1999; 55:583–601. [PubMed: 10089455] 67. Diederichs K, Karplus PA. Improved R-factors for diffraction data analysis in macromolecular crystallography. Nat Struct Biol. 1997; 4:269–275. [PubMed: 9095194] 68. Weiss MS, Hilgenfeld R. On the use of the merging R factor as a quality indicator for X-ray data. J Appl Crystallogr. 1997; 30:203–205. 69. Weiss MS. Global indicators of X-ray data quality. J Appl Crystallogr. 2001; 34:130–135. Levisson et al. Page 16 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Substrates and inhibitors of the CE7 family of enzymes. Structures of (A) acetylated xylooligosaccharide, (B) cephalosporin C, (C) p-nitrophenyl-acetate, (D) phenylmethylsulfonyl fluoride (PMSF), and (E) paraoxon. Levisson et al. Page 17 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Overall fold and topology of TM0077. (A) Stereo view of a TM0077 protomer. The β- strands are labeled numerically (-1 to 8) with the core strands in red, α-helices are labeled alphabetically (A-2 to F) and 310-helices are labeled with an Eta (η1 and η2) with the core helices in cyan. The three-helix insertion after β6 is colored green and the N-terminal extension is colored sky blue. The figure was generated using Pymol 31. (B) Topology diagram of TM0077, with the helices displayed as cylinders and the strands displayed as arrows following the color and label scheme of (A). The location of residues forming the catalytic triad is also indicated. Levisson et al. Page 18 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Structural superposition of TM0077 with structurally related esterases. Superposition of TM0077 (yellow) with (A) the cephalosporin C deacetylase (CAH) from B. subtilis (PDB: 1ods; blue) 40 and (B) the α/β-hydrolase domain of the acylpeptide hydrolase/esterase apAPH from A. pernix K1 (PDB: 1ve6; grey) 43. Levisson et al. Page 19 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. TM0077 oligomeric assembly. (A) Surface representation of the biological unit of the TM0077-Native hexamer with each monomer in a different color (left). The “cross section” shows the entrances on either side of the assembly and the internal cavity (center), and a 90° rotated view of the TM0077-Native hexamer, with a close-up view of the open central hole (right). (B) Surface representation of the biological hexamer unit of CAH from B. subtilis 40 (left) and the TM0077-SeMet hexamer with a close-up view of the blocked central hole (right). Levisson et al. Page 20 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Effect of temperature and pH on esterase activity. (A) The esterase activity was studied using pNP-C2 as a substrate at temperatures ranging from 40–100°C. The inset shows the temperature dependence as an Arrhenius plot. (B) Thermal stability of TM0077 at 90°C. (C) The effect of pH on esterase activity studied using pNP-C2 as a substrate at pH values in the range of 4.8–9.2. Levisson et al. Page 21 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. TM0077 catalytic site. (A) Surface representation of the TM0077 catalytic site, with His303, Asp274 and the intermediate DEP-modified Ser188 shown as sticks. The two binding pockets are indicated with S1 and S2. (B) Apo TM0077 with a bound chloride ion (green sphere), (C) TM0077 with PMS-modified Ser188 and (D) TM0077 with DEP-modified Ser188. The catalytic residues are shown as sticks, with the hydrogen bonds shown as dashed lines. Carbon atoms are in green (apo), cyan (PMS) or blue (DEP), oxygen atoms in red, sulfur atoms in yellow and phosphate in orange. Electron density omit maps shown for inhibitor modified Ser188 contoured at 1σ show that the PMS and DEP are covalently bonded to Ser188 in (C) and (D), respectively. Distances are shown in Ångströms. Levisson et al. Page 22 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 7. Conformational change of Ser188 Oγ. The Oγ atom of the Ser188 is rotated ~110° between the native apo structure (cyan) and (A) the complexed PMS-modified Ser188 structure (pink), (B) the DEP-modified Ser188 structure (light blue) and (C) the SeMet structure (purple). The different hydrogen bonds made for the Ser Oγ in the native versus complexed structures are shown as dashed black lines with distances in Ångströms. Levisson et al. Page 23 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Levisson et al. Page 24 Table I Summary of crystal parameters, data collection, and refinement statistics TM0077-SeMet TM0077-Native TM0077-PMS TM0077-DEP Space group P 21 P 21 21 21 P 21 21 21 P 21 21 21 Unit cell parameters a=152.64Å b=130.95Å c=157.82Å β=118.90° a=103.46Å b=103.79Å c=221.02Å a=103.57Å b=104.50Å c=221.61Å a=103.80Å b=104.43Å c=221.64Å Data collection λ1 MAD-Se λ2 MAD-Se Wavelength (Å) 0.9791 0.9183 0.9765 0.9765 0.9765 Resolution range (Å) 29.6 – 2.10 29.6 – 2.10 48.8 – 2.50 49.0 – 2.40 49.0 – 2.12 No. observations 1,119,236 1,100,249 1,222,016 765,546 989,949 No. unique reflections 293,140 291,757 83,045 94,681 123,070 Completeness (%) 93.0 (61.8)a 92.6 (60.8) 100 (100) 100 (100) 89.8 (53.5) Mean I/σ(I) 9.1 (2.4)a 9.6 (2.2) 14.4 (2.9) 11.5 (3.4) 15.3 (2.2) Rmerge on I (%) 12.3 (52.5) a 11.9 (57.9) 20.7 (109.7) c 18.0 (67.4) 9.5 (51.9) Rmeas on I (%) 14.3 (62.2) a 13.9 (68.7) 21.4 (113.6) 19.2 (71.9) 10.2 (60.2) Rpim on I (%) 7.2 (32.7) a 7.1 (36.2) 5.5 (29.2) 6.7 (24.9) 3.5 (29.2) Highest resolution shell (Å) 2.15 – 2.10 2.15 – 2.10 2.56 – 2.50 2.46 – 2.40 2.18 – 2.12 Model and refinement statistics Resolution range (Å) 29.6 – 2.10 48.8 – 2.50 49.0–2.40 49.0 – 2.12 No. reflections (total) 293,097 b 83,045 94,680 122,994 No. reflections (test) 14,726 4,200 4,742 6,188 Completeness (% total) 92.8 100.0 100.0 89.8 Data set used in refinement λ1 MAD-Se Cutoff criteria |F| > 0 |F| > 0 |F| > 0 |F| > 0 Rcryst 0.186 0.167 0.160 0.167 Rfree 0.223 0.212 0.208 0.205 Stereochemical parameters Restraints (RMSD observed) Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Levisson et al. Page 25 TM0077-SeMet TM0077-Native TM0077-PMS TM0077-DEP Bond angle (°) 1.48 1.47 1.53 1.44 Bond length (Å) 0.018 0.017 0.017 0.015 Av. isotropic B-value (Å2) 27.9 24.7 19.4 19.6 ESU based on Rfree 0.17 0.25 0.22 0.18 Water molecules/other solvent molecules 2,464/1 507/24 946/17 987/23 PDB ID 1vlq 3m81 3m82 3m83 aHighest resolution shell ESU = Estimated overall coordinate error 16,66. Rmerge=ΣhklΣi|Ii(hkl)-<I(hkl)>|/ΣhklΣiIi(hkl), Rmeas(redundancy-independent Rmerge)=Σhkl[Nhkl/(Nhkl-1)]1/2Σi|Ii(hkl)-<I(hkl)>|/ΣhklΣiIi(hkl), and Rpim(precision-indicating Rmerge)=Σhkl[1/ (Nhkl-1)]1/2Σi|Ii(hkl)-<I(hkl)>|/ΣhklΣiIi(hkl) 67–69. Rcryst = Σ| |Fobs|-|Fcalc| |/Σ|Fobs| where Fcalc and Fobs are the calculated and observed structure factor amplitudes, respectively. Rfree = as for Rcryst, but for 5.0 % of the total reflections chosen at random and omitted from refinement. bTypically, the number of unique reflections used in refinement is slightly less than the total number that were integrated and scaled. Reflections are excluded due to systematic absences, negative intensities, and rounding errors in the resolution limits and cell parameters. cRmerge of the highest resolution shell is high due to high redundancy (14.7). However, the completeness and mean I/σ of the highest resolution shell are reasonable, and these data were included in the refinement. Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Levisson et al. Page 26 Table II Kinetic parameters for hydrolysis of various esters Ester Km (mM) kcat (s−1) kcat/Km (s−1 mM−1) pNP-Acetate 0.185 ± 0.026 57.5 ± 2.2 310.8 ± 45.3 pNP-Propionate 0.137 ± 0.013 41.3 ± 1.1 301.5 ± 29.7 2-O-acetyl pNP-Xyl 3.6 ± 0.5 76.1 ± 19.2 21.1 ± 6.1 3-O-acetyl pNP-Xyl 4.2 ± 0.4 70.1 ± 7.7 16.7 ± 2.4 4-O-acetyl pNP-Xyl 4.0 ± 0.1 78.6 ± 12.9 19.7 ± 3.3 Proteins. Author manuscript; available in PMC 2013 June 01.
3M82
Crystal structure of Acetyl xylan esterase (TM0077) from THERMOTOGA MARITIMA at 2.40 A resolution (PMSF inhibitor complex structure)
Functional and structural characterization of a thermostable acetyl esterase from Thermotoga maritima Mark Levisson1,*, Gye Won Han2,3,*, Marc C. Deller2,3, Qingping Xu2,4, Peter Biely5, Sjon Hendriks1, Lynn F. Ten Eyck6,7, Claus Flensburg8, Pietro Roversi8, Mitchell D. Miller2,4, Daniel McMullan9, Frank von Delft2,3,‡, Andreas Kreusch10, Ashley M. Deacon2,4, John van der Oost1, Scott A. Lesley2,3,10, Marc-André Elsliger2,3, Servé W. M. Kengen1,†, and Ian A. Wilson2,3,† 1Laboratory of Microbiology, Wageningen University, 6703 HB, Wageningen, The Netherlands 2Joint Center for Structural Genomics, http://www.jcsg.org 3Department of Molecular Biology, The Scripps Research Institute, La Jolla, California 92037 4Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Stanford University, Menlo Park, California 92045 5Institute of Chemistry, Slovak Academy of Sciences, 845 38 Bratislava, Slovakia 6Department of Chemistry and Biochemistry, University of California at San Diego, La Jolla, California 92093-0505 7San Diego Supercomputer Center, University of California at San Diego, La Jolla, California 92093-0505 8Global Phasing Ltd. Sheraton House, Castle Park, Cambridge CB3 0AX, United Kingdom 9Protein Therapeutics Department, Genomics Institute of the Novartis Research Foundation, San Diego, California 92121 10Protein Sciences Department, Genomics Institute of the Novartis Research Foundation, San Diego, California 92121 Abstract TM0077 from Thermotoga maritima is a member of the carbohydrate esterase family 7 and is active on a variety of acetylated compounds, including cephalosporin C. TM0077 esterase activity is confined to short-chain acyl esters (C2-C3), and is optimal around 100°C and pH 7.5. The positional specificity of TM0077 was investigated using 4-nitrophenyl-β-D-xylopyranoside monoacetates as substrates in a β-xylosidase-coupled assay. TM0077 hydrolyzes acetate at positions 2, 3 and 4 with equal efficiency. No activity was detected on xylan or acetylated xylan, which implies that TM0077 is an acetyl esterase and not an acetyl xylan esterase as currently annotated. Selenomethionine-substituted and native structures of TM0077 were determined at 2.1 Å and 2.5 Å resolution, respectively, revealing a classic α/β-hydrolase fold. TM0077 assembles into a doughnut-shaped hexamer with small tunnels on either side leading to an inner cavity, which contains the six catalytic centers. Structures of TM0077 with covalently bound phenylmethylsulfonyl fluoride (PMSF) and paraoxon were determined to 2.4 Å and 2.1 Å, respectively, and confirmed that both inhibitors bind covalently to the catalytic serine (Ser188). Upon binding of inhibitor, the catalytic serine adopts an altered conformation, as observed in other esterase and lipases, and supports a previously proposed catalytic mechanism in which this Ser hydroxyl rotation prevents reversal of the reaction and allows access of a water molecule for completion of the reaction. †Correspondence to: Ian A. Wilson, Ph.D., Department of Molecular Biology, The Scripps Research Institute, La Jolla, California 92037; (858) 784-2939 Fax: (858) 784-2980; wilson@scripps.edu or Servé W. M. Kengen, Ph.D., Laboratory of Microbiology, Wageningen University, 6703 HB, Wageningen, The Netherlands; 31 317 483737, Fax: 31 317-483829; serve.kengen@wur.nl. *ML and GWH contributed equally to this work. ‡Current address: The Structural Genomics Consortium, Roosevelt Drive, Headington, Oxford OX3 7DQ, UK NIH Public Access Author Manuscript Proteins. Author manuscript; available in PMC 2013 June 01. Published in final edited form as: Proteins. 2012 June ; 80(6): 1545–1559. doi:10.1002/prot.24041. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Keywords Acetyl esterase; Thermotoga maritima; crystal structure; α/β hydrolase; inhibitor; serine rotation INTRODUCTION Thermotoga maritima is a hyperthermophilic bacterium that grows optimally at 80°C and is able to metabolize a variety of simple and complex carbohydrates, including glucose, sucrose, starch, cellulose, and xylan 1. Its carbohydrate utilization potential was confirmed by analysis of its sequenced genome 2. The xylan degrading pathway of T. maritima has been studied using microarrays 2–4, and several genes encoding transporters, xylanases, and a β-xylosidase have been identified. Among the enzymes with a differential expression pattern in the microarray was a predicted acetyl xylan esterase (locus tag TM0077, axeA) 3,5. Depending on the source, the xylan backbone may contain a varying degree of acetylated xylose residues. Therefore, in addition to xylanases and xylosidases, the complete degradation of xylan requires esterases/deacetylases 6. Presently, esterases and deacetylases that are active on carbohydrate substrates have been classified into 16 families by Henrissat and coworkers (Carbohydrate-Active enZymes Server (CAZy)) 7. According to this classification, the predicted acetyl xylan esterase from T. maritima would be a member of family 7 of the carbohydrate esterases (CE7). In addition to the acetyl xylan esterase activity, enzymes in the CE7 family are rather unusual in that they display a high specific activity towards the antibiotic cephalosporin C [(Fig. 1(a-b)] 8. Cephalosporins belong to the β-lactam class of antibiotics, which also includes penicillin, and affect bacterial cell growth by inhibiting the penicillin-binding-protein that cross-links peptide glycans required for cell wall formation 9. The production of deacetylated cephalosporins is of great interest because these compounds are valuable building blocks for the production of semi-synthetic β-lactam antibiotics10,11. To explore the catalytic capacity of the predicted acetyl xylan esterase from T. maritima and gain a better insight into the structure and function of the family 7 carbohydrate esterases, TM0077 was expressed and purified, and three-dimensional structures of the native enzyme and its complexes with phenylmethylsulfonyl fluoride (PMSF) and paraoxon inhibitors, were determined by x-ray crystallography. Furthermore, the enzyme was functionally characterized, and various biochemical properties including the positional specificity of the esterase were investigated. MATERIALS AND METHODS Gene cloning TM0077 was selected as part of the Joint Center for Structural Genomics (JCSG) effort on complete structural coverage of the T. maritima soluble proteome as a large-scale center for high-throughput structure determination funded under the NIHGMS Protein Structure Initiative (PSI) 12. The gene encoding TM0077 (GenBank: AAD35171.1, GI:4980565; SwissProt: Q9WXT2) was amplified by polymerase chain reaction (PCR) from genomic DNA using PfuTurbo DNA polymerase (Stratagene) and primers corresponding to the predicted 5′ and 3′ ends. The PCR product was cloned into plasmid pMH1, which encodes an expression and purification tag (MGSDKIHHHHHH) at the amino terminus of the protein. The cloning junctions were confirmed by DNA sequencing. Levisson et al. Page 2 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript TM0077-SeMet protein production and purification Protein production was performed in a selenomethionine-containing medium using the Escherichia coli methionine auxotrophic strain DL41. Expression was induced by the addition of 0.15% L-arabinose. At the end of fermentation, cells were harvested and subjected to one freeze/thaw cycle, and subsequently sonicated in Lysis Buffer [50 mM Tris pH 7.9, 50 mM NaCl, 1 mM MgCl2, 0.25 mM Tris(2-carboxyethyl)phosphine hydrochloride (TCEP), 1 mg/ml lysozyme] and the lysate was centrifuged at 3,400 × g for one hour. The soluble fraction was applied to nickel-chelating resin (GE Healthcare) pre-equilibrated with Equilibration Buffer [50 mM potassium phosphate pH 7.8, 300 mM NaCl, 10% (v/v) glycerol, 0.25 mM TCEP] containing 20 mM imidazole. The resin was washed with Equilibration Buffer containing 40 mM imidazole, and the protein was eluted with Elution Buffer [20 mM Tris pH 7.9, 300 mM imidazole, 10% (v/v) glycerol, 0.25 mM TCEP]. The eluate was buffer exchanged with Buffer Q [20 mM Tris pH 7.9, 5% (v/v) glycerol, 0.25 mM TCEP] containing 50 mM NaCl and applied to a RESOURCE Q column (GE Healthcare) pre-equilibrated with the same buffer. The protein was eluted using a linear gradient of 50 to 500 mM NaCl in Buffer Q and purified further with a HiLoad 16/60 Superdex 200 column (GE Healthcare), using Crystallization Buffer [20 mM Tris pH 7.9, 150 mM NaCl, 0.25 mM TCEP] as the mobile phase. For crystallization trials, the peak Superdex 200 fractions were concentrated to ~15 mg/mL by centrifugal ultrafiltration (Millipore). Molecular weight and oligomeric state of TM0077 were determined using a 1 cm × 30 cm Superdex 200 column (GE Healthcare) coupled with miniDAWN static light scattering (SEC/SLS) and Optilab differential refractive index detectors (Wyatt Technology). The mobile phase consisted of 20 mM Tris pH 8.0, 150 mM NaCl, and 0.02% (w/v) sodium azide. Native TM0077 production and purification For protein production, E. coli DL41 cells were grown in LB medium for 8 hours (an OD600 well above 2.0 was reached). Subsequently, the culture was induced by adding 0.15% L-arabinose and incubated another 16 hours at 37°C. Cells were harvested by centrifugation at 10,000 × g for 20 min. The cell pellet was resuspended in 30 ml of Lysis Buffer 2 [50 mM Tris-HCl pH 8.0, 50 mM NaCl, 10 mM imidazole, 0.25 mM TCEP]. The cells were disrupted by two passages through a French press at 110 MPa. The crude cell extract was treated with DNAse I at room temperature for 30 min and subsequently centrifuged at 43,000 × g for 30 min in order to remove cell debris. The supernatant was heated at 70°C for 25 min and then centrifuged to remove the precipitated proteins. The supernatant was filtered and loaded onto a nickel-chelating column packed with 20 ml of Ni- NTA His-Bind Resin (Novagen) and equilibrated in 50 mM Tris-HCl pH 8.0, 300 mM NaCl, 2% (v/v) glycerol, and 0.25 mM TCEP. The column was washed with 20 mM imidazole in the same buffer, and proteins were subsequently eluted with a linear gradient of 20–500 mM imidazole in the same buffer. Fractions containing esterase activity were pooled and loaded onto a HiPrep Desalting column (GE Healthcare) equilibrated with 20 mM Tris- HCl pH 8.0, 150 mM NaCl, and 0.25 mM TCEP. The homogeneity of the protein was checked by SDS-PAGE, and activity staining of the SDS-PAGE gel was performed using α- napthyl acetate, as described previously 13. The protein concentration was determined at 280 nm using a NanoDrop ND-1000 Spectrophotometer. Crystallization Crystals of selenomethionine-substituted TM0077 were obtained by hanging drop vapor diffusion against a 250 μl crystallization solution consisting of 20% (w/v) PEG-3000, 0.1 M HEPES pH 7.5, 0.2 M NaCl. Drops consisted of 0.5 μl protein and 0.5 μl crystallization solution. Native TM0077 was crystallized using nanodrop vapor diffusion techniques against a crystallization solution consisting of 0.2 M calcium acetate hydrate, 20% (w/v) Levisson et al. Page 3 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript PEG 3350, pH 7.3 at 20°C. Protein was concentrated to 22.8 mg/ml. Drops consisted of 100 nl protein and 100 nl of crystallization solution and a 60 μl reservoir of crystallization solution. Crystals of TM0077 in complex with inhibitors PMSF or paraoxon were obtained at 4°C in the same conditions with the same reagents as the native crystals. PMSF or paraoxon were added in a molar ratio of 1:3 (protein:inhibitor). Data collection For cryoprotection, the TM0077-SeMet crystal was transferred to crystallization solution supplemented with 15% (v/v) glycerol. The crystal was mounted in a cryoloop and subsequently flash-cooled in liquid nitrogen. X-ray data were collected at 100 K on beamline BL9-2 at the Stanford Synchrotron Radiation Lightsource (SSRL, Menlo Park, CA) using a Quantum 4 CCD detector (ADSC). A TM0077-SeMet MAD data set was collected to 2.1 Å resolution and the data were indexed in monoclinic space group P21, with unit cell parameters a = 152.6 Å, b = 131.0 Å, and c = 157.8 Å, and β=118.9°, and 12 molecules in the asymmetric unit. Data were indexed and integrated with DENZO and then scaled with SCALEPACK 14. Native TM0077, TM0077-PMSF complex (TM0077-PMS) and TM0077-paraoxon complex (TM0077-DEP) crystals were transferred to crystallization solution supplemented with 10% (v/v) ethylene glycol and flash-cooled to 100K. Data were collected at beamline 5.0.3 of the Advanced Light Source (ALS, Berkeley, CA) and processed with the HKL2000 package 14. The native data set was collected to 2.5 Å resolution, and TM0077-PMS and TM0077-DEP data sets were collected to 2.4 and 2.1 Å, respectively. All data were indexed in orthorhombic space group P212121, with unit cell parameters approximately a=103Å b=104Å c=221Å (See Table 1), and six molecules in the asymmetric unit. Data reduction and refinement statistics for TM0077-SeMet, TM0077-Native, TM0077-PMS and TM0077- DEP are summarized in Table I. Structure solution and refinement The TM0077-SeMet structure was solved by MAD phasing method using a two-wavelength MAD dataset. At the time of the initial data collection (2001), the structure determination of Se-MAD TM0077 posed a significant challenge to crystallographic programs, which were still under active development. As a result, modifications were made in various structure determination and refinement programs to achieve success. For initial phasing, SHELXD 15 was used to find candidate SeMet substructure sites. Attempts to complete phasing were unsuccessful due to the translational non-crystallographic symmetry (NCS) (not recognized initially). Self-consistent sets (partial sets) were found using the CCP4 program PROFESSS 16 and additional SeMet sites were found by SHELXD, and added to these partial sets. The SHARP 17 run did not complete initially; however, updates of SHARP and ARP/wARP eventually helped to resolve issues and an initial trace was obtained by ARP/ wARP. The structure was then refined with BUSTER 18 using tight NCS restraints to an Rcryst and Rfree of 18.6% and 22.3%, respectively. Model building was performed using O 19 and the structure was refined using Refmac5 20. Refinement statistics are summarized in Table I. The final model contains 12 protein molecules (chains A-L) in the asymmetric unit each consisting of residues 2-323. The MolProbity 21 Ramachandran plot analysis showed that 97.4% of all residues are in favored regions with a single outlier, Gln120 of chain B, which is supported by unambiguous electron density. Ramachandran outlier Gln120 of chain B of TM0077-SeMet is due to crystal packing with chain C. The backbone carbonyl oxygens of Gln120 and Gly119 of chain B makes hydrogen bonds with the backbone nitrogen of Gln140 of chain C (3.19 and 3.11 Å, respectively). Levisson et al. Page 4 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The native TM0077 structure, TM0077-PMS and TM0077-DEP structures were solved by molecular replacement using PHASER 22,23 with the TM0077-SeMet hexamer coordinates (pdb: 1vlq; A-F chains) as a search model. One hexamer was successfully located and the structure was further refined with Refmac5 20 using tight NCS restraints to an Rcryst and Rfree of 16.7% and 21.2% (native TM0077), 16.0% and 20.8% (TM0077-PMS) and 16.7% and 20.5% (TM0077-DEP), respectively. Iterative cycles of refinement and building were performed with Refmac5, Phenix 24,25 and Coot 26. All other crystallographic manipulations were carried out with the CCP4 package 16. Refinement statistics are summarized in Table I. The final model of native TM0077 contains residues 3-324 (chains A, B, C, D and F) and 3-323 (chain E) in the asymmetric unit. Analysis of main-chain torsion angles using MolProbity 21 showed that 97.8% of the residues are in favored regions of the Ramachandran plot with 0.2% outliers (Asn302 of chains B, C and D), which are supported by unambiguous electron density. The final model of TM0077-PMS contains residues 3-323 for all chains in the asymmetric unit with 97.5% of the residues in favored regions with 0.2% outliers (Asn302 of chains A, B, D and F). The final model of TM0077-DEP contains residues 0-324 for (chains A, B, C and F) and 0-323 (chains D and E) in the asymmetric unit, respectively, with 97.6% of the residues in favored region of the Ramachandran plot with 0.2% outliers (Asn302 of B, C, D and F chains). Ramachandran outlier Asn302 in the TM0077-Native, TM0077-PMS and TM0077-DEP structures is a neighbor to the catalytic triad residue His303 and may reflect a slightly different state for these structures compared to the Se-Met structure. Structure validation and deposition The quality of the crystal structure was analyzed using the JCSG Quality Control server (http://smb.slac.stanford.edu/jcsg/QC). This server processes the coordinates and data through a variety of validation tools including AutoDepInputTool 27 MolProbity 21, WHATIF 5.0 28, RESOLVE 29, MOLEMAN2 30 as well as several in-house scripts, and summarizes the results. Protein quaternary structure analysis were performed using the PISA server 30. Figures were prepared with PyMOL (DeLano Scientific) 31. RMSD values were calculated using the ProCKSI-Server 32. The structural data have been deposited in the RCSB Protein Data Bank (PDB) with accession codes 1vlq for TM0077-SeMet, 3m81 for TM0077-native, 3m83 for TM0077-DEP and 3m82 for TM0077-PMS. Enzyme assays Esterase activity was measured using p-nitrophenyl esters as described previously 13. Briefly, the standard assay consisted of activity measurements with 0.2 mM p-nitrophenyl acetate as substrate in 50 mM citrate-phosphate (pH 6) at 70°C. The p-nitrophenol liberated was measured continuously at 405 nm on a Hitachi U-2001 spectrophotometer with a temperature-controlled cuvette holder. Extinction coefficients of p-nitrophenol were determined prior to each measurement. Kinetic parameters were determined by direct fitting the data, obtained from multiple measurements, to the Michaelis–Menten curve (Tablecurve 2d, version 5.0). The effect of pH on esterase activity was studied in the pH range from 5 to 10. The buffers used were 50 mM citrate-phosphate (pH 5–8) and 50 mM CAPS (3-(cyclohexylamino) 1- propanesulphonic acid) (pH 9.5–10). The pH of the buffers was set at room temperature, and temperature corrections were made using their temperature coefficients: −0.0028 pH/°C for citrate-phosphate buffer and −0.018 pH/°C for CAPS buffer. The effect of temperature on esterase activity was studied in the range of 40–100°C using 0.2 mM p-nitrophenyl acetate as substrate. Enzyme thermostability was determined by incubating the enzyme in a 50 mM Tris-HCl, 150 mM NaCl (pH 7.8) buffer at 90°C and 100°C for various time intervals. Residual activity was determined in the standard assay. Levisson et al. Page 5 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Inhibition kinetics of PMSF and paraoxon were determined as described for the acetylcholinesterase from electric eel 33. All experiments were performed at 70°C in 50 mM citrate-phosphate (pH 6) buffer and 0.2 mM p-nitrophenyl acetate as substrate. The kinetic constants for the inhibition of TM0077 with PMSF and paraoxon were measured in the concentration range of 1.0–10.0 mM and 0.2–1.0 mM, respectively. Deacetylase activity was determined using high-performance liquid chromatography (HPLC) by measuring the amount of acetic acid released from the substrates cephalosporin C, 7-aminocephalosporanic acid, glucose-pentaacetate and acetylated xylan. Xylan was acetylated by the method described by Johnson 34. The reaction mixture contained 0.9 ml of substrate solution (dissolved in 50 mM Tris-HCl, pH 7.5) and 0.1 ml of enzyme solution, and was incubated at 37°C for various time intervals (0–10 min). The reaction was stopped by adding 0.2 ml of stop solution (100 mN H2SO4 and 30 mM crotonate) and placing the sample on ice. The conditions for HPLC were as follows: column, KC811 Shodex; detection, RI and UV detectors; solvent, 3 mN H2SO4; flow rate, 1.5 ml/min; temperature, 30°C; internal standard, crotonate. One unit of enzyme activity was defined as the amount of enzyme that releases one μmol of acetic acid per minute. Activity on xylan was measured quantitatively using DMSO-extracted xylan (1% polysaccharide solution in 0.1 M sodium phosphate buffer pH 6) at 60°C 35. Xylan will precipitate as a consequence of deacetylation, resulting in a rapid turbidity of the solution. Positional specificity assay The positional specificity of TM0077 was investigated using an enzyme-coupled assay on monoacetylated 4-nitrophenyl β-D-xylopyranosides (pNP-Xyl) as described 36. The β- xylosidase XloA (locus tag: TM0076) from T. maritima was cloned into the vector pET24d in frame with a C-terminal His6-tag. The enzyme was expressed and purified as described above for native TM0077. Activity of XloA was confirmed by measuring the release of p- nitrophenol at 405 nm from the substrate 4-nitrophenyl β-D-xylopyranoside. The enzyme-coupled assay was performed at 60°C in a total volume of 125 μl, which contained 0.1 M sodium phosphate (pH 6 or 7), 2-O-, 3-O-, or 4-O-acetyl pNP-Xyl, the β- xylosidase XloA, and TM0077. Stable 50x-concentrated stock solutions of the substrates were prepared in DMSO. The reaction was started by addition of 2.5 μl of a stock solution to a preheated reaction mixture consisting of phosphate buffer, auxiliary β-xylosidase XloA in excess, and TM0077. The reaction was terminated by addition of 800 μl of a 2% solution of Na2CO3. Liberated p-nitrophenol was determined at 405 nm against substrate and enzyme blanks. A short incubation time for activity determination was used to suppress acetyl migration on the xylopyranosyl-ring, which is significant at pH 6 or 7 37. The kinetic constants were determined at pH 7 and 60°C with reaction times of 2 or 5 minutes. RESULTS and DISCUSSION In silico analysis TM0077 consists of 325 amino acids with a calculated molecular mass of 37 kDa. Sequence analysis, using the SignalP 3.0 server, revealed that TM0077 has no predicted signal sequence and is, therefore, believed to be an intracellular enzyme. Analysis of the gene organization indicates that the TM0077 gene co-localizes with genes encoding a xylanase (TM0070) 38, ABC transporter components (TM0071-TM0075), and a β-xylosidase (TM0076) 39. BLAST-P analysis showed that TM0077 has highest similarity to putative acetyl esterases, acetyl xylan esterases and cephalosporin C deacetylases. Among the BLAST results, a predicted acetyl xylan esterase-related protein from T. maritima (locus tag: TM0435) was also identified. TM0077 was compared with other members of the CE7 Levisson et al. Page 6 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript family using structure-based, multi-sequence alignment and the putative catalytic triad, Ser188, Asp274, and His303, was identified from conservation throughout the analyzed sequences. The putative nucleophilic serine (Ser188) is located within a conserved pentapeptide consensus sequence, Gly-Xaa-Ser-Gln-Gly, typical of this family. Previously, a signature sequence motif, [RGQ]-(x:~70)-[GxSQG]-(x:~115)-[HE] (where x indicates any amino acid), had been suggested for the CE7 family based on an aminoacid alignment of 12 sequences 40. In an updated alignment consisting now of 50 sequences, we observed many sequences that have this signature motif, but it is not conserved throughout the entire family (See Supporting Information and Fig. S1 for the multi-sequence alignment). Overall structure The crystal structure of seleno-methionine incorporated TM0077 (TM0077-SeMet) was determined to 2.1 Å resolution by multi-wavelength anomalous dispersion (MAD) (Table I) with twelve molecules per asymmetric unit. A native apo structure (TM0077-Native) was determined in a different space group (see Methods) to 2.5 Å by molecular replacement, using TM0077-SeMet as a search model, with six molecules in the asymmetric unit (Table I). Each monomer of the native hexamer contained a calcium ion (see below) bound by Lys22, Glu26, and Asp25 via a bridging water molecule. Superposition of the TM0077- SeMet and the TM0077-Native structures gave a root-mean-square difference (RMSD) of 0.12 Å over 321 Cα atoms, which indicates that these structures are nearly identical as expected. In general, the TM0077 structure resembles the canonical α/β-hydrolase fold, which consists of a central, twisted, eight-stranded β-sheet surrounded by α-helices on both sides, with β2 antiparallel to the other strands. TM0077 deviates slightly from the canonical α/β- hydrolase fold at two locations: a three-helix insertion after strand β6 and an extension of the N-terminus (Fig. 2). Insertions after β6 or β7 are common for α/β-hydrolases and are proposed to help shape the substrate-binding site 41. The N-terminus is extended by two helices (αA-1 and αA-2) and an antiparallel β-strand (β-1) that aligns with the other eight β- strands (β1-β8) and extends the central β-sheet. This nine-stranded β-sheet is highly twisted, and β-1 and β8 at the extreme edges are rotated approximately 130° relative to each other. Helices αA and αB both contain a short 310-helix segment at their N-terminus. Helices αA-1, αA-2, αB, αC, αD, αD1, αD2, αD3, αE, and the 310-helix η2 are located on one side of the central β-sheet, and helices αA, αF and the 310-helix η1 are on the other side. A structural similarity search was performed using the program DALI 42. Monomer A of the TM0077-SeMet structure was used as a search model and similarity was found with cephalosporin C deacetylases, acetyl xylan esterases, acylamino-releasing enzymes, dipeptidyl peptidases and some esterases and lipases. TM0077 is structurally most similar to cephalosporin C deacetylase (CAH) from B. subtilis (PDB: 1ods) 40, acetyl xylan esterase (AXE) from B. pumilus (PDB: 3fvr and 2xlb), acetyl xylan esterase (AXE1) from Thermoanaerobacterium sp. JW/SL YS485 (PDB: 3fcy), and acylpeptide hydrolase/esterase apAPH from Aeropyrum pernix K1 (PDB: 1ve6) 43. The sequence identity between TM0077 and CAH is 41% and the two structures align with a Z-score of 46 and an RMSD of 1.5 Å over 312 Cα atoms. The sequence identity with apAPH is 17% with a Z-score of 23.3 and an RMSD of 2.3 Å over 230 Cα atoms. Superpositions of TM0077 with CAH and with apAPH are shown in Fig. 3. Quaternary structure The crystal structure of TM0077-SeMet contains two hexamers in the asymmetric unit that are related by a non-crystallographic two-fold axis. Each hexamer contains a dimer of trimers with a back-to-back arrangement (Fig. 4). The apo and the complex crystals Levisson et al. Page 7 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript contained one hexamer in the asymmetric unit. Crystallographic packing analysis using PISA (EBI) 44 indicated that the relevant physiological oligomeric state of TM0077 is a hexamer, which was confirmed by size exclusion chromatography coupled with static light scattering. Further analyses of the hexameric assembly indicated that two main interfaces play an essential role in complex formation. The first interface between subunit A and B (green and cyan in Fig. 4) (identical to C/D and E/F) is stabilized by seven hydrogen bonds on average and has a buried surface area of 1024 Å2 contributed by each chain. The second interface between A and F (green and purple in Fig. 4) (B/C and D/E) is stabilized by 17 hydrogen bonds on average with a buried surface area of 1079 Å2 contributed by each chain However, a multiple sequence alignment of TM0077 with other CE7 esterases showed that the residues involved in these two main interfaces are not conserved. Other secondary interfaces bury around 514 Å2 contributed by each chain. The hexamer has a total buried surface area of 18,860 Å2, which is approximately 30% of the total surface area. Approximately 3,143 Å2 per monomer is, therefore, buried upon complex formation. The TM0077 hexamer has a doughnut-shape when viewed from the side, with the six active sites located in the interior of the complex, where they line an oval-shaped cavity [Fig. 4(a)]. This cavity is accessible via two entrances, one on each side of the flat hexamer. Each of these entrances is approximately 13 Å wide and connects to a short tunnel or pore spanning approximately 10 Å to reach the inner cavity. Interestingly, in the TM0077-SeMet hexamer, the entrance to the internal cavity is blocked by three phenylalanine residues (Phe4), one for each of three monomers that compose half of the hexamer [Fig. 4(b)]. Residue Phe4 is located in the mobile N-terminus (high B-values), which may indicate some flexibility or multiple conformations. Calcium ions were identified, by the electron density and coordination geometry, supported by their presence in the crystallization reagents, in the native TM0077, TM0077-PMS and TM0077-DEP structures, but not in the TM0077-SeMet structure. The SeMet protein was crystallized without any calcium in the crystallization reagents. In each subunit of the hexamer, one calcium ion is located at the N-terminal region of helix αA-1, and is coordinated by the backbone carbonyl of Lys22 and the Glu26 carboxylate. The remainder of the calcium coordination sphere is filled with waters from a neighboring solvent channel present in all molecules in the asymmetric unit. The Asp25 carboxylate contributes to the calcium binding via one of the coordinating water molecules. Another calcium ion is bound in a crystal packing interface between chain A and chain C′ of a crystallographic symmetry- related hexamer. This calcium is coordinated by the carboxylates of GluA45 and AspA58 from one chain and the carboxylate from Glu C’45 (bidentate coordination) of the symmetry-related chain with three water molecules completing a capped-octahedral coordination sphere. An equivalent calcium binding site is also observed in the crystal packing interface between chains D and B′. No significant increase or reduction of activity of TM0077 was observed in the presence of calcium ions or EDTA. Therefore, it seems that these calcium ions are not important for activity. On the other hand, calcium may help stabilize the structure. No calcium was present in the B. subtilis CAH structure 40; however, Lys22, Glu26 and Ser25 are conserved and may also act as a calcium binding site. Enzyme activity The activity of TM0077 was investigated using p-nitrophenol esters with varying acyl-chain length, ranging from C2 to C18. TM0077 is only active on the short-chain p-nitrophenol esters of acetate and propionate and does not hydrolyze esters with acyl chains longer than four carbons. No significant difference was found in the catalytic efficiency (kcat/Km) for the hydrolysis of p-nitrophenyl with acyl chains containing 2 to 3 carbons (Table II) [Fig. 1(c)]. Levisson et al. Page 8 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The effect of temperature on activity was studied using p-nitrophenyl acetate as substrate. The esterase activity increased from 40°C upwards until 100°C [Fig. 5(a)]. An Arrhenius analysis resulted in linear plots in the temperature ranges of 40–60°C and 60–100°C [Fig. 5(a); inset], with calculated activation energies for the formation of the enzyme substrate- enzyme complex of 33.7 and 21.9 kJ/mol, respectively. The transition or break in linearity of the Arrhenius plot at 60°C (1000/T (K) = 3.0) could indicate some conformational change of the enzyme. TM0077 is fairly resistant to thermal inactivation. An approximate 50% transient increase in activity is seen during the first 10 to 20 minutes when the enzyme is incubated at 90°C. After 30 minutes, inactivation of function occurs by first order kinetics with a half-life of approximately 120 minutes [Fig. 5(b)]. A transient activation has also been observed for other thermophilic esterases, such from Sulfolobus shibatae 45, and it is believed that a high temperature is needed in order to obtain the optimal conformation for catalysis. TM0077 was not stable at 100°C, resulting in a half-life of less than 5 minutes. However, the optimum temperature and thermal stability of TM0077 are still considerably higher than those reported for other characterized CE7 esterases, including the Thermoanaerobacterium enzyme that has a temperature optimum of 80°C and a half-life of 1h at 75°C 46. The effect of pH on activity was measured in the pH range of 4.8 to 9.2 using the substrate p-nitrophenyl acetate. TM0077 displayed maximum activity at approximately pH 7.5 [Fig. 5(c)], which is comparable to other CE7 esterases, such as the acetyl xylan esterases from Thermoanaerobacterium sp. strain JW/SL-YS485 46. Positional specificity The positional specificity of TM0077 was tested on three monoacetates of 4-nitrophenyl β- D-xylopyranoside (pNP-Xyl). To determine the enzyme activity, the β-xylosidase XloA 39 (TM0076) from T. maritima is required as an auxiliary enzyme. This thermostable XloA enzyme was, therefore, cloned, heterologously expressed, purified to homogeneity, and its activity was confirmed by measuring release of p-nitrophenol from the substrate pNP-Xyl (data not shown). The β-xylosidase was not active on the three monoacetates of pNP-Xyl. In the XloA-coupled assay, TM0077 hydrolyzed acetate from positions 2, 3 and 4 of pNP-Xyl with similar catalytic efficiency. The results are summarized in Table II. In addition, TM0077 was investigated for its ability to remove acetyl groups from 7- aminocephalosporanic acid (7-ACA), cephalosporin C, glucose penta-acetate, N-acetyl-D- glucosamine, xylan and acetylated xylan. TM0077 has no activity for acetylated and non- acetylated xylan polymers, indicating that it is, indeed, an acetyl esterase and not an acetyl xylan esterase. As expected for an acetyl esterase, TM0077 displayed high activity on glucose penta-acetate with a turnover number of 2680 s−1. Like other members of CE7, TM0077 was also able to hydrolyze the acetyl groups from both cephalosporin C and 7- ACA with a turnover number of 376 s−1 and 1140 s−1, respectively. TM0077 was not able to hydrolyze the acetyl group from N-acetyl-D-glucosamine, indicating that it is specific for ester bonds and unable to hydrolyze amide bonds. Inhibitor assays and TM0077 structures complexed with PMSF and paraoxon PMSF and paraoxon [Fig. 1(d,e)] are competitive irreversible inhibitors of esterases. Inhibition proceeds by the formation of a reversible Michaelis complex, followed by an irreversible step and inhibition can, therefore, be characterized by two parameters: a dissociation constant and a binding rate constant. The inhibition kinetics for paraoxon and PMSF were investigated in the presence of p-nitrophenyl acetate, as described previously 47, and the dissociation and rate constants were 0.5 ± 0.1 mM and 0.13 ± 0.02 s−1 for paraoxon, and 1.1 ± 0.2 mM and 0.020 ± 0.001 s−1 for PMSF, respectively. The acetyl xylan esterase Levisson et al. Page 9 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript from Bacillus pumilus (BpAxe) has slightly reduced sensitivity to paraoxon (dissociation and rate constants of respectively 5.4 mM and 0.012 s−1), likely due to steric hindrance of two tyrosine residues (Tyr91 and Tyr206) that hamper the binding of paraoxon. Although these residues are essentially conserved in TM0077 (Tyr92 and Phe213), TM0077 is more sensitive to paraoxon than BpAxe48. In comparison to EST2 of Alicyclobacillus acidocaldarius 49 and EstA of T. maritima 47, the TM0077 dissociation constant is slightly higher, but the rate constant is comparable. No significant stimulation or reduction of activity of TM0077 was observed in the presence of divalent metal ions or ethylenediaminetetraacetic acid (EDTA). To obtain more information about inhibitor binding and any possible conformational changes during catalysis, TM0077 was co-crystallized with the inhibitors PMSF and paraoxon and the PMSF (TM0077-PMS) and paraoxon (TM0077-DEP) structures were determined to 2.4 Å and 2.1 Å, respectively (Table I). The electron density map of TM0077 with PMSF showed clear density for the PMSF covalent modification. The fluorine was cleaved from the PMSF molecule during the binding reaction and the phenylmethyl sulfonyl (PMS) moiety is covalently bound to the Oγ atom of Ser188. The native apo and PMS- bound structures superimpose well with RMSD’s of 0.09–0.11 Å over 320–321 Cα atoms. Electron density maps of the paraoxon-bound structure displayed clear density for a diethyl- phosphate moiety covalently bound to the Oγ atom of Ser188. This covalent modification indicates that the p-nitrophenol group of paraoxon was cleaved off during co-crystallization, and a tetrahedral product reminiscent of the first transition state was formed during carboxyl ester hydrolysis. The native apo and paraoxon-bound structures superimpose with RMSD’s of 0.12–0.32 Å over 320–322 Cα atoms. Attempts to obtain co-crystals of TM0077 with cephalosporin C, even at a low temperature of 4°C, were unsuccessful. Analysis of the active site TM0077 has a classic catalytic triad, consisting of Ser188 as the nucleophile, His303 as the proton acceptor/donor, and Asp274 as the acidic residue stabilizing the histidine (Fig. 6). The catalytic serine Ser188 is located within a conserved pentapeptide sequence, Gly-X-Ser- X-Gly (GGSQG), characteristic of esterases and lipases. The positions of Ser188, Asp274, and His303 are consistent with their expected locations in the canonical fold of the α/β- hydrolase family. Ser188 is located at the nucleophile elbow in a sharp turn between β5 and helix αC. The presence of three glycine residues (Gly186, Gly187, and Gly190) in close proximity to Ser188 prevents steric hindrance and facilitates access to the nucleophile elbow. Asp274 and His303 are located in loops between β7 and helix αE, and between β8 and helix αF, respectively. The oxyanion hole is formed by the backbone amide groups of Tyr92 and Gln189. The catalytic triad and oxyanion hole are located in a depression on the surface of TM0077. This ellipsoid pocket (S1), which is approximately 12 Å wide, extends 15 Å from the catalytic serine. A smaller pocket (S2), approximately 5 Å long, extends to the other side of the catalytic serine [Fig. 6(a)]. The volume of both pockets combined (S1 + S2) is 1082 Å3 (CASTp analysis; 50). The substrate-binding pocket is bordered by residues from helices αA and αF, and its base is formed by residues from β-strands 4, 5, 6, and their adjacent C-terminal loops. The overall pocket is hydrophobic, although it does have some polar residues (Gln88, Asp210, and Gln314), which may interact with the substrate. In the native apo structure, the Ser188 hydroxyl makes a hydrogen bond with the imidazole of His303 [Fig. 6(b)]. Extra density was observed near the side chain of Ser188 and was interpreted as a chloride ion based on electron density size and shape as well as the geometry of the interactions with surrounding residues. This chloride ion is bound at the entrance of the oxyanion hole, forming hydrogen bonds with the backbone amides of Tyr92 and Gln189. In the PMSF-bound structure, the phenyl ring of the inhibitor is located in the small active site groove surrounded by hydrophobic residues Tyr92, Trp124, Pro228, Ile276, Levisson et al. Page 10 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript and His303 [Fig. 6(c)]. The sulfonyl group of PMSF makes hydrogen bonds with the backbone amides of Tyr92 and Gln189. In the paraoxon-bound structure, the diethyl- phosphate (DEP) moiety is stabilized by hydrogen-bonding interactions with the oxyanion hole. One of the two ethyl arms of bound paraoxon points toward the larger pocket in the protein, while the other follows the groove of the small pocket. The two ethyl arms are stabilized by packing against Tyr92, Trp124, Pro228, Ile276, and His303 [Fig. 6(d)]. Two rotamers of the catalytic serine Although no large conformational changes were observed upon binding of PMSF or paraoxon, a different rotamer of the catalytic serine side chain was observed compared to native TM0077 [Fig. 7(a,b)]. Similar changes have been observed in several other esterases and have been shown to play a key role in the catalytic mechanism (see CONCLUSION for more details). In the native structure, the catalytic Ser188 Oγ is in the plane of the imidazole ring of His303, as most commonly observed in the resting state of esterases and lipases 51. The Ser188 Oγ forms a hydrogen bond (2.6 Å) with His303 Nε2. In the PMSF- and paraoxon-bound structures, the conformation of the catalytic serine changes; the Ser188 Oγ rotates about 110°, increasing the distance (3.1 Å and 2.8 Å for PMSF and paraoxon bound structures, respectively) to the His303 imidazole ring. In the TM0077-SeMet structure, the catalytic serine is also rotated over ~110°, with a distance to the imidazole ring of 3.0 Å [Fig. 7(c.)]. A probable explanation for this observation could be the protonation of His303, since TM0077-SeMet was crystallized at a lower pH (pH 4.2) compared to the native TM0077 (pH 7.3). Furthermore, extra electron density was identified in the TM0077-SeMet structure, suggesting a partially occupied acyl intermediate on Ser188. However, as this density is not sufficiently clear and interpretable to fit an acyl intermediate, water molecules were modeled instead. No rearrangements of any other residues in the active site were observed. CONCLUSION TM0077 from the hyperthermophilic bacterium T. maritima was predicted from its gene sequence to be an acetyl xylan esterase. We have expressed and purified TM0077 and experimentally demonstrated that it has ester-hydrolyzing activity. The TM0077 activity was restricted to short acyl chain esters (C2 and C3) when artificial p-nitrophenyl-esters were used as substrates. In addition, the enzyme has high specific activity on glucose penta- acetate. However, no activity was detected on xylan or acetylated xylan. Thus, TM0077 should be reclassified as an acetyl esterase, and not as an acetyl xylan esterase as currently annotated 52. Furthermore, the lack of any apparent signal sequence suggests that the protein is not secreted. Thus, the predicted intracellular location of TM0077 is compatible with a role other than the deacetylation of extracellular xylan. Based on these results, we conclude that the likely biological function of TM0077 is removal of the remaining acetyl groups from the short, end products of xylan degradation that are imported into the cytoplasm. The resulting deacetylated xylose oligomers are the substrates for a β-xylosidase. This role for TM0077 is in good agreement with the clustering of the TM0077 gene with other genes involved in xylose metabolism. However, it cannot be ruled out that TM0077 may also act on other small, acetylated compounds. TM0077 is the first esterase from the CE7 family to be tested for its positional specificity for the deacetylation of 4-nitrophenyl-β-D-xylopyranoside. TM0077 hydrolyzes acetate at the 2, 3 and 4 positions of 4-nitrophenyl-β-D-xylopyranoside with similar efficiency. Conversely, the CtAxe esterase from Clostridium thermocellum in the CE4 family shows a clear preference for hydrolyzing acetate at the 2 position 53, and Penicillium purpurogenum AXE II esterase, a member of the CE5 family, also has a preference for acetate at position 2 54. Levisson et al. Page 11 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript This lack of preference for a specific position of the acetate group correlates with the relative broad substrate specificity of the CE7 esterases. Esterases and deacetylases in the CE7 family are unusual in that they are active towards both acetylated xylo-oligosaccharides and the antibiotic cephalosporin C [Fig. 1(a,b)]. Therefore, TM0077 was investigated for activity towards the substrates 7-ACA and cephalosporin C. The activity of TM0077 on these substrates is approximately ten-fold higher than that of the acetyl xylan esterase from B. pumilus 55 or the acetyl esterase from Thermoanaerobacterium sp. strain JW/SL YS485 56. TM0077 has a higher hydrolytic activity on 7-ACA compared to cephalosporin C, as described for other CE7 esterases 40,55,56. Nonetheless, it is unlikely that both compounds are natural substrates, because the stability of these compounds at the optimal growth temperature (80°C) of T. maritima is very low. Crystal structures of TM0077 in complex with inhibitors PMSF and paraoxon revealed that, upon binding of PMSF or paraoxon, the reaction is trapped at the acylation step via the formation of a covalent tetrahedral reaction product. In the complexed structures, the negatively charged oxygen of the tetrahedral intermediate, derived from the substrate oxyanion, is stabilized by hydrogen bonds with the backbone amide groups of Tyr92 and Gln189. Comparison of the TM0077 complexed structures with the native structure shows that the catalytic serine (Ser188) Oγ rotates about 110°, thereby increasing the distance between Ser188 Oγ and His303 Nε2. Such a conformational change of the catalytic serine has been observed in several other esterases, including Fusarium solani cutinase 57, Penicillium purpurogenum acetyl xylan esterase 51, Rhodococcus sp. strain MB1 cocaine esterase 58, Bacillus subtilis lipase 59, Rhodococcus sp. strain H1 heroine esterase 60, and Aspergillus niger feruloyl esterase 61. The classical model for the catalytic mechanism of esterases consists of a sequential two-step hydrolysis. The first reaction involves nucleophilic attack by the catalytic serine on the substrate carbonyl carbon, resulting in an acyl-enzyme and the liberation of an alcohol. In the second reaction, a water molecule performs a nucleophilic attack on the acyl-enzyme, the acyl-enzyme bond breaks and the carboxylate is released 62. Although the catalytic mechanism of esterases is well established, it is unclear why the initially generated tetrahedral intermediate does not collapse back to the reactant complex during the nucleophilic attack of the substrate. A previously proposed mechanism that would prevent this collapse is the spatial reorganization of the catalytic residues during the initial catalytic step, causing the residues to separate and thereby drive the reaction forward 62–64. The apo and inhibitor bound structures of TM0077, presented herein, support this proposed mechanism. Moreover, in a recent study of the serine protease mechanism, it was suggested that subtle atomic motions of the catalytic serine and histidine residues during the catalytic cycle favor the forward reaction 65. Thus, rotation of Ser188 Oγ of TM0077 may be required to inhibit reversal of the reaction. In addition, such changes may facilitate the access of water to the catalytic histidine so that the second step of the reaction can go to completion. Deacetyl cephalosporins are valuable building blocks for the production of semisynthetic β- lactam antibiotics. These compounds are derived from cephalosporin C or 7- aminocephalosporanic acid via enzymatic or chemical processes 10. The thermostable TM0077 esterase may be valuable in the preparation of derivatives of β-lactam antibiotics. Recently, the substrate specificity of the acetyl xylan esterase from P. purpurogenum was engineered to accept a range of fatty acid esters of up to 14 carbons compared to its wild- type preference for acetate54. It might also be possible to engineer TM0077 and enable the (de)acetylation of cephalosporins at the C10 position with various acyl chains. Because of its high stability and activity on 7-ACA and cephalosporin C, TM0077 presents an attractive candidate for the production of new semi-synthetic antibiotics. Levisson et al. Page 12 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We gratefully acknowledge contributions from George Sheldrick for modifications of the SHELXD program, and for Global Phasing Ltd. that made significant improvements in the automation of autoSHARP. We also thank Victor Lamzin for updates of chain docking of the ARP/wARP program, and Gerard Bricogne and Eleanor Dodson for helpful discussion on phasing for the large TM0077-SeMet structure, and Willem J. van Berkel for valuable discussion on the catalytic mechanism of TM0077. Portions of this research were carried out at the Stanford Synchrotron Radiation Lightsource (SSRL) and the Advanced Light Source (ALS). The SSRL is a Directorate of SLAC National Accelerator Laboratory and an Office of Science User Facility operated for the U.S. Department of Energy Office of Science by Stanford University. The SSRL Structural Molecular Biology Program is supported by the DOE Office of Biological and Environmental Research, and by the National Institutes of Health, National Center for Research Resources, Biomedical Technology Program (P41RR001209), and the National Institute of General Medical Sciences. The ALS is supported by the Director, Office of Science, Office of Basic Energy Sciences, Materials Sciences Division, of the U.S. Department of Energy under Contract No. DE- AC02-05CH11231 at Lawrence Berkeley National Laboratory. Genomic DNA from Thermotoga maritima MSB8 (DSM3109) (ATCC #43589D-5) was obtained from the American Type Culture Collection (ATCC). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of General Medical Sciences or the National Institutes of Health. Grant sponsor: NIH Grant numbers U54 GM094586 and U54 GM074898 (Protein Structure Initiative); Grant sponsor: The Graduate School VLAG Wageningen, the Netherlands (ML). References 1. Huber, R.; Hannig, M. Thermotogales. In: Dworkin, M.; Falkow, S.; Rosenberg, E.; Schleifer, K-H.; Stackebrandt, E., editors. The Prokaryotes: An Evolving Electronic Resource for the Microbiological Community. Vol. 7. New-York: Springer-Verlag; 2006. p. 899-922. 2. Nelson KE, Clayton RA, Gill SR, Gwinn ML, Dodson RJ, Haft DH, Hickey EK, Peterson JD, Nelson WC, Ketchum KA, McDonald L, Utterback TR, Malek JA, Linher KD, Garrett MM, Stewart AM, Cotton MD, Pratt MS, Phillips CA, Richardson D, Heidelberg J, Sutton GG, Fleischmann RD, Eisen JA, White O, Salzberg SL, Smith HO, Venter JC, Fraser CM. Evidence for lateral gene transfer between Archaea and bacteria from genome sequence of Thermotoga maritima. Nature. 1999; 399:323–329. [PubMed: 10360571] 3. Chhabra SR, Shockley KR, Conners SB, Scott KL, Wolfinger RD, Kelly RM. Carbohydrate- induced differential gene expression patterns in the hyperthermophilic bacterium Thermotoga maritima. J Biol Chem. 2003; 278:7540–7552. [PubMed: 12475972] 4. Conners SB, Montero CI, Comfort DA, Shockley KR, Johnson MR, Chhabra SR, Kelly RM. An expression-driven approach to the prediction of carbohydrate transport and utilization regulons in the hyperthermophilic bacterium Thermotoga maritima. J Bacteriol. 2005; 187:7267–7282. [PubMed: 16237010] 5. VanFossen AL, Lewis DL, Nichols JD, Kelly RM. Polysaccharide degradation and synthesis by extremely thermophilic anaerobes. Ann N Y Acad Sci. 2008; 1125:322–337. [PubMed: 18378602] 6. Biely P, Mackenzie CR, Puls J, Schneider H. Cooperativity of esterases and xylanases in the enzymatic degradation of acetyl xylan. Bio-Technology. 1986; 4:731–733. 7. Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V, Henrissat B. The Carbohydrate- Active EnZymes database (CAZy): an expert resource for Glycogenomics. Nucleic Acids Res. 2009; 37:D233–D238. [PubMed: 18838391] 8. Topakas E, Paul C. Microbial xylanolytic carbohydrate esterases. Industrial Enzymes. 2007:83–97. 9. Weil J, Miramonti J, Ladisch MR. Cephalosporin-C: Mode of action and biosynthetic pathway. Enzyme Microb Technol. 1995; 17:85–87. 10. Barends TRM, Yoshida H, Dijkstra BW. Three-dimensional structures of enzymes useful for beta- lactam antibiotic production. Curr Opin Biotechnol. 2004; 15:356–363. [PubMed: 15358004] 11. Martínez-Martínez I, Montoro-García S, Lozada-Ramírez JD, Sánchez-Ferrer Á, García-Carmona F. A colorimetric assay for the determination of acetyl xylan esterase or cephalosporin C acetyl Levisson et al. Page 13 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript esterase activities using 7-amino cephalosporanic acid, cephalosporin C, or acetylated xylan as substrate. Anal Biochem. 2007; 369:210–217. [PubMed: 17651681] 12. Lesley SA, Kuhn P, Godzik A, Deacon AM, Mathews I, Kreusch A, Spraggon G, Klock HE, McMullan D, Shin T, Vincent J, Robb A, Brinen LS, Miller MD, McPhillips TM, Miller MA, Scheibe D, Canaves JM, Guda C, Jaroszewski L, Selby TL, Elsliger MA, Wooley J, Taylor SS, Hodgson KO, Wilson IA, Schultz PG, Stevens RC. Structural genomics of the Thermotoga maritima proteome implemented in a high-throughput structure determination pipeline. Proc Natl Acad Sci USA. 2002; 99:11664–11669. [PubMed: 12193646] 13. Levisson M, van der Oost J, Kengen SW. Characterization and structural modeling of a new type of thermostable esterase from Thermotoga maritima. FEBS Journal. 2007; 274:2832–2842. [PubMed: 17466017] 14. Otwinowski Z, Minor W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 1997; 276:307–326. 15. Schneider TR, Sheldrick GM. Substructure solution with SHELXD. Acta Crystallogr Sect D Biol Crystallogr. 2002; 58:1772–1779. [PubMed: 12351820] 16. Collaborative Computational Project Number 4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr Sect D Biol Crystallogr. 1994; 50:760–763. [PubMed: 15299374] 17. Bricogne G, Vonrhein C, Flensburg C, Schiltz M, Paciorek W. Generation, representation and flow of phase information in structure determination: recent developments in and around SHARP 2.0. Acta Crystallogr Sect D Biol Crystallogr. 2003; 59:2023–2030. [PubMed: 14573958] 18. Bricogne, G.; Blanc, E.; Brandl, M.; Flensburg, C.; Keller, P.; Paciorek, W.; Roversi, P.; Smart, O.; Vonrhein, CTW. BUSTER, version 2.8.0. Cambridge, United Kingdom: Global Phasing Ltd; 2009. 19. Jones TA, Zou JY, Cowan SW, Kjeldgaard M. Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr A Found Crystallogr. 1991; 47:110–119. 20. Winn MD, Murshudov GN, Papiz MZ. Macromolecular TLS refinement in REFMAC at moderate resolutions. Methods Enzymol. 2003; 374:300–321. [PubMed: 14696379] 21. Davis IW, Murray LW, Richardson JS, Richardson DC. MOLPROBITY: structure validation and all-atom contact analysis for nucleic acids and their complexes. Nucleic Acids Res. 2004; 32:W615–W619. [PubMed: 15215462] 22. McCoy AJ, Grosse-Kunstleve RW, Storoni LC, Read RJ. Likelihood-enhanced fast translation functions. Acta Crystallogr Sect D Biol Crystallogr. 2005; 61:458–464. [PubMed: 15805601] 23. Storoni LC, McCoy AJ, Read RJ. Likelihood-enhanced fast rotation functions. Acta Crystallogr Sect D Biol Crystallogr. 2004; 60:432–438. [PubMed: 14993666] 24. Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Terwilliger TC. PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr Sect D Biol Crystallogr. 2002; 58:1948– 1954. [PubMed: 12393927] 25. Murshudov GN, Vagin AA, Dodson EJ. Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr Sect D Biol Crystallogr. 1997; 53:240–255. [PubMed: 15299926] 26. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr Sect D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] 27. Yang H, Guranovic V, Dutta S, Feng Z, Berman HM, Westbrook JD. Automated and accurate deposition of structures solved by X-ray diffraction to the Protein Data Bank. Acta Crystallogr Sect D Biol Crystallogr. 2004; 60:1833–1839. [PubMed: 15388930] 28. Vriend G. WHAT IF: a molecular modeling and drug design program. J Mol Graph. 1990; 8:52– 56. [PubMed: 2268628] 29. Terwilliger TC. Automated side-chain model building and sequence assignment by template matching. Acta Crystallogr Sect D Biol Crystallogr. 2003; 59:45–49. [PubMed: 12499538] 30. Kleywegt GJ. Validation of protein crystal structures. Acta Crystallogr Sect D Biol Crystallogr. 2000; 56:249–265. [PubMed: 10713511] Levisson et al. Page 14 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 31. DeLano, WL. The Pymol molecular graphics system. DeLano Scientific; San Carlos, CA: 2002. 32. Barthel D, Hirst JD, Blazewicz J, Burke EK, Krasnogor N. ProCKSI: a decision support system for Protein (structure) Comparison, Knowledge, Similarity and Information. BMC Bioinformatics. 2007; 8:416. [PubMed: 17963510] 33. Forsberg A, Puu G. Kinetics for the inhibition of acetylcholinesterase from the electric eel by some organophosphates and carbamates. Eur J Biochem. 1984; 140:153–156. [PubMed: 6705793] 34. Johnson KG, Harrison BA, Schneider H, Mackenzie CR, Fontana JD. Xylan-hydrolyzing enzymes from Streptomyces spp. Enzyme Microb Technol. 1988; 10:403–409. 35. Biely P, Mackenzie CR, Schneider H. Production of acetyl xylan esterase by Trichoderma reesei and Schizophyllum commune. Can J Microbiol. 1988; 34:767–772. 36. Biely P, Mastihubova M, la Grange DC, van Zyl WH, Prior BA. Enzyme-coupled assay of acetylxylan esterases on monoacetylated 4-nitrophenyl beta-D-xylopyranosides. Anal Biochem. 2004; 332:109–115. [PubMed: 15301955] 37. Mastihubova M, Biely P. Lipase-catalysed preparation of acetates of 4-nitrophenyl beta-D- xylopyranoside and their use in kinetic studies of acetyl migration. Carbohydr Res. 2004; 339:1353–1360. [PubMed: 15113674] 38. Jiang ZQ, Kobayashi A, Ahsan MM, Lite L, Kitaoka M, Hayashi K. Characterization of a thermostable family 10 endo-xylanase (XynB) from Thermotoga maritima that cleaves p- nitrophenyl-beta-D-xyloside. J Biosci Bioeng. 2001; 92:423–428. [PubMed: 16233122] 39. Xue YM, Shao WL. Expression and characterization of a thermostable beta-xylosidase from the hyperthermophile, Thermotoga maritima. Biotechnol Lett. 2004; 26:1511–1515. [PubMed: 15604789] 40. Vincent F, Charnock SJ, Verschueren KHG, Turkenburg JP, Scott DJ, Offen WA, Roberts S, Pell G, Gilbert HJ, Davies GJ, Brannigan JA. Multifunctional xylooligosaccharide/cephalosporin C deacetylase revealed by the hexameric structure of the Bacillus subtilis enzyme at 1. 9 angstrom resolution. J Mol Biol. 2003; 330:593–606. [PubMed: 12842474] 41. Nardini M, Dijkstra BW. Alpha/beta hydrolase fold enzymes: the family keeps growing. Curr Opin Struct Biol. 1999; 9:732–737. [PubMed: 10607665] 42. Holm L, Kaariainen S, Rosenstrom P, Schenkel A. Searching protein structure databases with DaliLite v.3. Bioinformatics. 2008; 24:2780–2781. [PubMed: 18818215] 43. Bartlam M, Wang G, Yang H, Gao R, Zhao X, Xie G, Cao S, Feng Y, Rao Z. Crystal structure of an acylpeptide hydrolase/esterase from Aeropyrum pernix K1. Structure. 2004; 12:1481–1488. [PubMed: 15296741] 44. Krissinel E, Henrick K. Inference of macromolecular assemblies from crystalline state. J Mol Biol. 2007; 372:774–797. [PubMed: 17681537] 45. Huddleston S, Yallop CA, Charalambous BM. The identification and partial characterisation of a novel inducible extracellular thermostable esterase from the archaeon Sulfolobus shibatae. Biochem Biophys Res Commun. 1995; 216:495–500. [PubMed: 7488139] 46. Shao WL, Wiegel J. Purification and characterization of 2 thermostable acetyl xylan esterases from Thermoanaerobacterium sp strain JW/SL-YS485. Appl Environ Microbiol. 1995; 61:729–733. [PubMed: 7574610] 47. Levisson M, Sun L, Hendriks S, Swinkels P, Akveld T, Bultema JB, Barendregt A, van den Heuvel RH, Dijkstra BW, van der Oost J, Kengen SW. Crystal structure and biochemical properties of a novel thermostable esterase containing an immunoglobulin-like domain. J Mol Biol. 2009; 385:949–962. [PubMed: 19013466] 48. Montoro-García S, Gil-Ortiz F, García-Carmona F, Polo LM, Rubio VAS-F. The crystal structure of the cephalosporin deacetylating enzyme acetyl xylan esterase bound to paraoxon explains the low sensitivity of this serine hydrolase to organophosphate inactivation. Biochem J. 201110.1042/ BJ20101859 49. Febbraio F, D’Andrea SE, Mandrich L, Merone L, Rossi M, Nucci R, Manco G. Irreversible inhibition of the thermophilic esterase EST2 from Alicyclobacillus acidocaldarius. Extremophiles. 2008; 12:719–728. [PubMed: 18622571] Levisson et al. Page 15 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 50. Dundas J, Ouyang Z, Tseng J, Binkowski A, Turpaz Y, Liang J. CASTp: computed atlas of surface topography of proteins with structural and topographical mapping of functionally annotated residues. Nucleic Acids Res. 2006; 34:W116–W118. [PubMed: 16844972] 51. Ghosh D, Sawicki M, Lala P, Erman M, Pangborn W, Eyzaguirre J, Gutierrez R, Jornvall H, Thiel DJ. Multiple conformations of catalytic serine and histidine in acetylxylan esterase at 0.90 angstrom. J Biol Chem. 2001; 276:11159–11166. [PubMed: 11134051] 52. Williamson G, Kroon PA, Faulds CB. Hairy plant polysaccharides: a close shave with microbial esterases. Microbiology. 1998; 144:2011–2023. [PubMed: 9720023] 53. Biely P, Mastihubova M, Puchart V. The vicinal hydroxyl group is prerequisite for metal activation of Clostridium thermocellum acetylxylan esterase. Biochim Biophys Acta. 2007; 1770:565–570. [PubMed: 17261352] 54. Colombres M, Garate JA, Lagos CF, Araya-Secchi R, Norambuena P, Quiroz S, Larrondo L, Perez-Acle T, Eyzaguirre J. An eleven amino acid residue deletion expands the substrate specificity of acetyl xylan esterase II (AXE II) from Penicillium purpurogenum. J Comput Aided Mol Des. 2008; 22:19–28. [PubMed: 18060506] 55. Degrassi G, Kojic M, Ljubijankic G, Venturi V. The acetyl xylan esterase of Bacillus pumilus belongs to a family of esterases with broad substrate specificity. Microbiology. 2000; 146:1585– 1591. [PubMed: 10878123] 56. Lorenz WW, Wiegel J. Isolation, analysis, and expression of two genes from Thermoanaerobacterium sp. strain JW/SL YS485: a beta-xylosidase and a novel acetyl xylan esterase with cephalosporin C deacetylase activity. J Bacteriol. 1997; 179:5436–5441. [PubMed: 9286998] 57. Longhi S, Czjzek M, Lamzin V, Nicolas A, Cambillau C. Atomic resolution (1. 0 angstrom) crystal structure of Fusarium solani cutinase: Stereochemical analysis. J Mol Biol. 1997; 268:779–799. [PubMed: 9175860] 58. Larsen NA, Turner JM, Stevens J, Rosser SJ, Basran A, Lerner RA, Bruce NC, Wilson IA. Crystal structure of a bacterial cocaine esterase. Nat Struct Biol. 2002; 9:17–21. [PubMed: 11742345] 59. Kawasaki K, Kondo H, Suzuki M, Ohgiya S, Tsuda S. Alternate conformations observed in catalytic serine of Bacillus subtilis lipase determined at 1. 3 angstrom resolution. Acta Crystallogr Sect D Biol Crystallogr. 2002; 58:1168–1174. [PubMed: 12077437] 60. Zhu X, Larsen NA, Basran A, Bruce NC, Wilson IA. Observation of an arsenic adduct in an acetyl esterase crystal structure. J Biol Chem. 2003; 278:2008–2014. [PubMed: 12421810] 61. McAuley KE, Svendsen A, Patkar SA, Wilson KS. Structure of a feruloyl esterase from Aspergillus niger. Acta Crystallogr Sect D Biol Crystallogr. 2004; 60:878–887. [PubMed: 15103133] 62. Hedstrom L. Serine protease mechanism and specificity. Chem Rev. 2002; 102:4501–4523. [PubMed: 12475199] 63. Ash EL, Sudmeier JL, Day RM, Vincent M, Torchilin EV, Haddad KC, Bradshaw EM, Sanford DG, Bachovchin WW. Unusual 1H NMR chemical shifts support (His) Cε1-H···O=C H-bond: Proposal for reaction-driven ring flip mechanism in serine protease catalysis. Proc Natl Acad Sci USA. 2000; 97:10371–10376. [PubMed: 10984533] 64. Bizzozero SA, Dutler H. Stereochemical aspects of peptide hydrolysis catalyzed by serine proteases of the chymotrypsin type. Bioorg Chem. 1981; 10:46–62. 65. Radisky ES, Lee JM, Lu CJK, Koshland DE Jr. Insights into the serine protease mechanism from atomic resolution structures of trypsin reaction intermediates. Proc Natl Acad Sci USA. 2006; 103:6835–6840. [PubMed: 16636277] 66. Cruickshank DW. Remarks about protein structure precision. Acta Crystallogr Sect D Biol Crystallogr. 1999; 55:583–601. [PubMed: 10089455] 67. Diederichs K, Karplus PA. Improved R-factors for diffraction data analysis in macromolecular crystallography. Nat Struct Biol. 1997; 4:269–275. [PubMed: 9095194] 68. Weiss MS, Hilgenfeld R. On the use of the merging R factor as a quality indicator for X-ray data. J Appl Crystallogr. 1997; 30:203–205. 69. Weiss MS. Global indicators of X-ray data quality. J Appl Crystallogr. 2001; 34:130–135. Levisson et al. Page 16 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Substrates and inhibitors of the CE7 family of enzymes. Structures of (A) acetylated xylooligosaccharide, (B) cephalosporin C, (C) p-nitrophenyl-acetate, (D) phenylmethylsulfonyl fluoride (PMSF), and (E) paraoxon. Levisson et al. Page 17 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Overall fold and topology of TM0077. (A) Stereo view of a TM0077 protomer. The β- strands are labeled numerically (-1 to 8) with the core strands in red, α-helices are labeled alphabetically (A-2 to F) and 310-helices are labeled with an Eta (η1 and η2) with the core helices in cyan. The three-helix insertion after β6 is colored green and the N-terminal extension is colored sky blue. The figure was generated using Pymol 31. (B) Topology diagram of TM0077, with the helices displayed as cylinders and the strands displayed as arrows following the color and label scheme of (A). The location of residues forming the catalytic triad is also indicated. Levisson et al. Page 18 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Structural superposition of TM0077 with structurally related esterases. Superposition of TM0077 (yellow) with (A) the cephalosporin C deacetylase (CAH) from B. subtilis (PDB: 1ods; blue) 40 and (B) the α/β-hydrolase domain of the acylpeptide hydrolase/esterase apAPH from A. pernix K1 (PDB: 1ve6; grey) 43. Levisson et al. Page 19 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. TM0077 oligomeric assembly. (A) Surface representation of the biological unit of the TM0077-Native hexamer with each monomer in a different color (left). The “cross section” shows the entrances on either side of the assembly and the internal cavity (center), and a 90° rotated view of the TM0077-Native hexamer, with a close-up view of the open central hole (right). (B) Surface representation of the biological hexamer unit of CAH from B. subtilis 40 (left) and the TM0077-SeMet hexamer with a close-up view of the blocked central hole (right). Levisson et al. Page 20 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Effect of temperature and pH on esterase activity. (A) The esterase activity was studied using pNP-C2 as a substrate at temperatures ranging from 40–100°C. The inset shows the temperature dependence as an Arrhenius plot. (B) Thermal stability of TM0077 at 90°C. (C) The effect of pH on esterase activity studied using pNP-C2 as a substrate at pH values in the range of 4.8–9.2. Levisson et al. Page 21 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. TM0077 catalytic site. (A) Surface representation of the TM0077 catalytic site, with His303, Asp274 and the intermediate DEP-modified Ser188 shown as sticks. The two binding pockets are indicated with S1 and S2. (B) Apo TM0077 with a bound chloride ion (green sphere), (C) TM0077 with PMS-modified Ser188 and (D) TM0077 with DEP-modified Ser188. The catalytic residues are shown as sticks, with the hydrogen bonds shown as dashed lines. Carbon atoms are in green (apo), cyan (PMS) or blue (DEP), oxygen atoms in red, sulfur atoms in yellow and phosphate in orange. Electron density omit maps shown for inhibitor modified Ser188 contoured at 1σ show that the PMS and DEP are covalently bonded to Ser188 in (C) and (D), respectively. Distances are shown in Ångströms. Levisson et al. Page 22 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 7. Conformational change of Ser188 Oγ. The Oγ atom of the Ser188 is rotated ~110° between the native apo structure (cyan) and (A) the complexed PMS-modified Ser188 structure (pink), (B) the DEP-modified Ser188 structure (light blue) and (C) the SeMet structure (purple). The different hydrogen bonds made for the Ser Oγ in the native versus complexed structures are shown as dashed black lines with distances in Ångströms. Levisson et al. Page 23 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Levisson et al. Page 24 Table I Summary of crystal parameters, data collection, and refinement statistics TM0077-SeMet TM0077-Native TM0077-PMS TM0077-DEP Space group P 21 P 21 21 21 P 21 21 21 P 21 21 21 Unit cell parameters a=152.64Å b=130.95Å c=157.82Å β=118.90° a=103.46Å b=103.79Å c=221.02Å a=103.57Å b=104.50Å c=221.61Å a=103.80Å b=104.43Å c=221.64Å Data collection λ1 MAD-Se λ2 MAD-Se Wavelength (Å) 0.9791 0.9183 0.9765 0.9765 0.9765 Resolution range (Å) 29.6 – 2.10 29.6 – 2.10 48.8 – 2.50 49.0 – 2.40 49.0 – 2.12 No. observations 1,119,236 1,100,249 1,222,016 765,546 989,949 No. unique reflections 293,140 291,757 83,045 94,681 123,070 Completeness (%) 93.0 (61.8)a 92.6 (60.8) 100 (100) 100 (100) 89.8 (53.5) Mean I/σ(I) 9.1 (2.4)a 9.6 (2.2) 14.4 (2.9) 11.5 (3.4) 15.3 (2.2) Rmerge on I (%) 12.3 (52.5) a 11.9 (57.9) 20.7 (109.7) c 18.0 (67.4) 9.5 (51.9) Rmeas on I (%) 14.3 (62.2) a 13.9 (68.7) 21.4 (113.6) 19.2 (71.9) 10.2 (60.2) Rpim on I (%) 7.2 (32.7) a 7.1 (36.2) 5.5 (29.2) 6.7 (24.9) 3.5 (29.2) Highest resolution shell (Å) 2.15 – 2.10 2.15 – 2.10 2.56 – 2.50 2.46 – 2.40 2.18 – 2.12 Model and refinement statistics Resolution range (Å) 29.6 – 2.10 48.8 – 2.50 49.0–2.40 49.0 – 2.12 No. reflections (total) 293,097 b 83,045 94,680 122,994 No. reflections (test) 14,726 4,200 4,742 6,188 Completeness (% total) 92.8 100.0 100.0 89.8 Data set used in refinement λ1 MAD-Se Cutoff criteria |F| > 0 |F| > 0 |F| > 0 |F| > 0 Rcryst 0.186 0.167 0.160 0.167 Rfree 0.223 0.212 0.208 0.205 Stereochemical parameters Restraints (RMSD observed) Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Levisson et al. Page 25 TM0077-SeMet TM0077-Native TM0077-PMS TM0077-DEP Bond angle (°) 1.48 1.47 1.53 1.44 Bond length (Å) 0.018 0.017 0.017 0.015 Av. isotropic B-value (Å2) 27.9 24.7 19.4 19.6 ESU based on Rfree 0.17 0.25 0.22 0.18 Water molecules/other solvent molecules 2,464/1 507/24 946/17 987/23 PDB ID 1vlq 3m81 3m82 3m83 aHighest resolution shell ESU = Estimated overall coordinate error 16,66. Rmerge=ΣhklΣi|Ii(hkl)-<I(hkl)>|/ΣhklΣiIi(hkl), Rmeas(redundancy-independent Rmerge)=Σhkl[Nhkl/(Nhkl-1)]1/2Σi|Ii(hkl)-<I(hkl)>|/ΣhklΣiIi(hkl), and Rpim(precision-indicating Rmerge)=Σhkl[1/ (Nhkl-1)]1/2Σi|Ii(hkl)-<I(hkl)>|/ΣhklΣiIi(hkl) 67–69. Rcryst = Σ| |Fobs|-|Fcalc| |/Σ|Fobs| where Fcalc and Fobs are the calculated and observed structure factor amplitudes, respectively. Rfree = as for Rcryst, but for 5.0 % of the total reflections chosen at random and omitted from refinement. bTypically, the number of unique reflections used in refinement is slightly less than the total number that were integrated and scaled. Reflections are excluded due to systematic absences, negative intensities, and rounding errors in the resolution limits and cell parameters. cRmerge of the highest resolution shell is high due to high redundancy (14.7). However, the completeness and mean I/σ of the highest resolution shell are reasonable, and these data were included in the refinement. Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Levisson et al. Page 26 Table II Kinetic parameters for hydrolysis of various esters Ester Km (mM) kcat (s−1) kcat/Km (s−1 mM−1) pNP-Acetate 0.185 ± 0.026 57.5 ± 2.2 310.8 ± 45.3 pNP-Propionate 0.137 ± 0.013 41.3 ± 1.1 301.5 ± 29.7 2-O-acetyl pNP-Xyl 3.6 ± 0.5 76.1 ± 19.2 21.1 ± 6.1 3-O-acetyl pNP-Xyl 4.2 ± 0.4 70.1 ± 7.7 16.7 ± 2.4 4-O-acetyl pNP-Xyl 4.0 ± 0.1 78.6 ± 12.9 19.7 ± 3.3 Proteins. Author manuscript; available in PMC 2013 June 01.
3M83
Crystal structure of Acetyl xylan esterase (TM0077) from THERMOTOGA MARITIMA at 2.12 A resolution (paraoxon inhibitor complex structure)
Functional and structural characterization of a thermostable acetyl esterase from Thermotoga maritima Mark Levisson1,*, Gye Won Han2,3,*, Marc C. Deller2,3, Qingping Xu2,4, Peter Biely5, Sjon Hendriks1, Lynn F. Ten Eyck6,7, Claus Flensburg8, Pietro Roversi8, Mitchell D. Miller2,4, Daniel McMullan9, Frank von Delft2,3,‡, Andreas Kreusch10, Ashley M. Deacon2,4, John van der Oost1, Scott A. Lesley2,3,10, Marc-André Elsliger2,3, Servé W. M. Kengen1,†, and Ian A. Wilson2,3,† 1Laboratory of Microbiology, Wageningen University, 6703 HB, Wageningen, The Netherlands 2Joint Center for Structural Genomics, http://www.jcsg.org 3Department of Molecular Biology, The Scripps Research Institute, La Jolla, California 92037 4Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Stanford University, Menlo Park, California 92045 5Institute of Chemistry, Slovak Academy of Sciences, 845 38 Bratislava, Slovakia 6Department of Chemistry and Biochemistry, University of California at San Diego, La Jolla, California 92093-0505 7San Diego Supercomputer Center, University of California at San Diego, La Jolla, California 92093-0505 8Global Phasing Ltd. Sheraton House, Castle Park, Cambridge CB3 0AX, United Kingdom 9Protein Therapeutics Department, Genomics Institute of the Novartis Research Foundation, San Diego, California 92121 10Protein Sciences Department, Genomics Institute of the Novartis Research Foundation, San Diego, California 92121 Abstract TM0077 from Thermotoga maritima is a member of the carbohydrate esterase family 7 and is active on a variety of acetylated compounds, including cephalosporin C. TM0077 esterase activity is confined to short-chain acyl esters (C2-C3), and is optimal around 100°C and pH 7.5. The positional specificity of TM0077 was investigated using 4-nitrophenyl-β-D-xylopyranoside monoacetates as substrates in a β-xylosidase-coupled assay. TM0077 hydrolyzes acetate at positions 2, 3 and 4 with equal efficiency. No activity was detected on xylan or acetylated xylan, which implies that TM0077 is an acetyl esterase and not an acetyl xylan esterase as currently annotated. Selenomethionine-substituted and native structures of TM0077 were determined at 2.1 Å and 2.5 Å resolution, respectively, revealing a classic α/β-hydrolase fold. TM0077 assembles into a doughnut-shaped hexamer with small tunnels on either side leading to an inner cavity, which contains the six catalytic centers. Structures of TM0077 with covalently bound phenylmethylsulfonyl fluoride (PMSF) and paraoxon were determined to 2.4 Å and 2.1 Å, respectively, and confirmed that both inhibitors bind covalently to the catalytic serine (Ser188). Upon binding of inhibitor, the catalytic serine adopts an altered conformation, as observed in other esterase and lipases, and supports a previously proposed catalytic mechanism in which this Ser hydroxyl rotation prevents reversal of the reaction and allows access of a water molecule for completion of the reaction. †Correspondence to: Ian A. Wilson, Ph.D., Department of Molecular Biology, The Scripps Research Institute, La Jolla, California 92037; (858) 784-2939 Fax: (858) 784-2980; wilson@scripps.edu or Servé W. M. Kengen, Ph.D., Laboratory of Microbiology, Wageningen University, 6703 HB, Wageningen, The Netherlands; 31 317 483737, Fax: 31 317-483829; serve.kengen@wur.nl. *ML and GWH contributed equally to this work. ‡Current address: The Structural Genomics Consortium, Roosevelt Drive, Headington, Oxford OX3 7DQ, UK NIH Public Access Author Manuscript Proteins. Author manuscript; available in PMC 2013 June 01. Published in final edited form as: Proteins. 2012 June ; 80(6): 1545–1559. doi:10.1002/prot.24041. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Keywords Acetyl esterase; Thermotoga maritima; crystal structure; α/β hydrolase; inhibitor; serine rotation INTRODUCTION Thermotoga maritima is a hyperthermophilic bacterium that grows optimally at 80°C and is able to metabolize a variety of simple and complex carbohydrates, including glucose, sucrose, starch, cellulose, and xylan 1. Its carbohydrate utilization potential was confirmed by analysis of its sequenced genome 2. The xylan degrading pathway of T. maritima has been studied using microarrays 2–4, and several genes encoding transporters, xylanases, and a β-xylosidase have been identified. Among the enzymes with a differential expression pattern in the microarray was a predicted acetyl xylan esterase (locus tag TM0077, axeA) 3,5. Depending on the source, the xylan backbone may contain a varying degree of acetylated xylose residues. Therefore, in addition to xylanases and xylosidases, the complete degradation of xylan requires esterases/deacetylases 6. Presently, esterases and deacetylases that are active on carbohydrate substrates have been classified into 16 families by Henrissat and coworkers (Carbohydrate-Active enZymes Server (CAZy)) 7. According to this classification, the predicted acetyl xylan esterase from T. maritima would be a member of family 7 of the carbohydrate esterases (CE7). In addition to the acetyl xylan esterase activity, enzymes in the CE7 family are rather unusual in that they display a high specific activity towards the antibiotic cephalosporin C [(Fig. 1(a-b)] 8. Cephalosporins belong to the β-lactam class of antibiotics, which also includes penicillin, and affect bacterial cell growth by inhibiting the penicillin-binding-protein that cross-links peptide glycans required for cell wall formation 9. The production of deacetylated cephalosporins is of great interest because these compounds are valuable building blocks for the production of semi-synthetic β-lactam antibiotics10,11. To explore the catalytic capacity of the predicted acetyl xylan esterase from T. maritima and gain a better insight into the structure and function of the family 7 carbohydrate esterases, TM0077 was expressed and purified, and three-dimensional structures of the native enzyme and its complexes with phenylmethylsulfonyl fluoride (PMSF) and paraoxon inhibitors, were determined by x-ray crystallography. Furthermore, the enzyme was functionally characterized, and various biochemical properties including the positional specificity of the esterase were investigated. MATERIALS AND METHODS Gene cloning TM0077 was selected as part of the Joint Center for Structural Genomics (JCSG) effort on complete structural coverage of the T. maritima soluble proteome as a large-scale center for high-throughput structure determination funded under the NIHGMS Protein Structure Initiative (PSI) 12. The gene encoding TM0077 (GenBank: AAD35171.1, GI:4980565; SwissProt: Q9WXT2) was amplified by polymerase chain reaction (PCR) from genomic DNA using PfuTurbo DNA polymerase (Stratagene) and primers corresponding to the predicted 5′ and 3′ ends. The PCR product was cloned into plasmid pMH1, which encodes an expression and purification tag (MGSDKIHHHHHH) at the amino terminus of the protein. The cloning junctions were confirmed by DNA sequencing. Levisson et al. Page 2 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript TM0077-SeMet protein production and purification Protein production was performed in a selenomethionine-containing medium using the Escherichia coli methionine auxotrophic strain DL41. Expression was induced by the addition of 0.15% L-arabinose. At the end of fermentation, cells were harvested and subjected to one freeze/thaw cycle, and subsequently sonicated in Lysis Buffer [50 mM Tris pH 7.9, 50 mM NaCl, 1 mM MgCl2, 0.25 mM Tris(2-carboxyethyl)phosphine hydrochloride (TCEP), 1 mg/ml lysozyme] and the lysate was centrifuged at 3,400 × g for one hour. The soluble fraction was applied to nickel-chelating resin (GE Healthcare) pre-equilibrated with Equilibration Buffer [50 mM potassium phosphate pH 7.8, 300 mM NaCl, 10% (v/v) glycerol, 0.25 mM TCEP] containing 20 mM imidazole. The resin was washed with Equilibration Buffer containing 40 mM imidazole, and the protein was eluted with Elution Buffer [20 mM Tris pH 7.9, 300 mM imidazole, 10% (v/v) glycerol, 0.25 mM TCEP]. The eluate was buffer exchanged with Buffer Q [20 mM Tris pH 7.9, 5% (v/v) glycerol, 0.25 mM TCEP] containing 50 mM NaCl and applied to a RESOURCE Q column (GE Healthcare) pre-equilibrated with the same buffer. The protein was eluted using a linear gradient of 50 to 500 mM NaCl in Buffer Q and purified further with a HiLoad 16/60 Superdex 200 column (GE Healthcare), using Crystallization Buffer [20 mM Tris pH 7.9, 150 mM NaCl, 0.25 mM TCEP] as the mobile phase. For crystallization trials, the peak Superdex 200 fractions were concentrated to ~15 mg/mL by centrifugal ultrafiltration (Millipore). Molecular weight and oligomeric state of TM0077 were determined using a 1 cm × 30 cm Superdex 200 column (GE Healthcare) coupled with miniDAWN static light scattering (SEC/SLS) and Optilab differential refractive index detectors (Wyatt Technology). The mobile phase consisted of 20 mM Tris pH 8.0, 150 mM NaCl, and 0.02% (w/v) sodium azide. Native TM0077 production and purification For protein production, E. coli DL41 cells were grown in LB medium for 8 hours (an OD600 well above 2.0 was reached). Subsequently, the culture was induced by adding 0.15% L-arabinose and incubated another 16 hours at 37°C. Cells were harvested by centrifugation at 10,000 × g for 20 min. The cell pellet was resuspended in 30 ml of Lysis Buffer 2 [50 mM Tris-HCl pH 8.0, 50 mM NaCl, 10 mM imidazole, 0.25 mM TCEP]. The cells were disrupted by two passages through a French press at 110 MPa. The crude cell extract was treated with DNAse I at room temperature for 30 min and subsequently centrifuged at 43,000 × g for 30 min in order to remove cell debris. The supernatant was heated at 70°C for 25 min and then centrifuged to remove the precipitated proteins. The supernatant was filtered and loaded onto a nickel-chelating column packed with 20 ml of Ni- NTA His-Bind Resin (Novagen) and equilibrated in 50 mM Tris-HCl pH 8.0, 300 mM NaCl, 2% (v/v) glycerol, and 0.25 mM TCEP. The column was washed with 20 mM imidazole in the same buffer, and proteins were subsequently eluted with a linear gradient of 20–500 mM imidazole in the same buffer. Fractions containing esterase activity were pooled and loaded onto a HiPrep Desalting column (GE Healthcare) equilibrated with 20 mM Tris- HCl pH 8.0, 150 mM NaCl, and 0.25 mM TCEP. The homogeneity of the protein was checked by SDS-PAGE, and activity staining of the SDS-PAGE gel was performed using α- napthyl acetate, as described previously 13. The protein concentration was determined at 280 nm using a NanoDrop ND-1000 Spectrophotometer. Crystallization Crystals of selenomethionine-substituted TM0077 were obtained by hanging drop vapor diffusion against a 250 μl crystallization solution consisting of 20% (w/v) PEG-3000, 0.1 M HEPES pH 7.5, 0.2 M NaCl. Drops consisted of 0.5 μl protein and 0.5 μl crystallization solution. Native TM0077 was crystallized using nanodrop vapor diffusion techniques against a crystallization solution consisting of 0.2 M calcium acetate hydrate, 20% (w/v) Levisson et al. Page 3 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript PEG 3350, pH 7.3 at 20°C. Protein was concentrated to 22.8 mg/ml. Drops consisted of 100 nl protein and 100 nl of crystallization solution and a 60 μl reservoir of crystallization solution. Crystals of TM0077 in complex with inhibitors PMSF or paraoxon were obtained at 4°C in the same conditions with the same reagents as the native crystals. PMSF or paraoxon were added in a molar ratio of 1:3 (protein:inhibitor). Data collection For cryoprotection, the TM0077-SeMet crystal was transferred to crystallization solution supplemented with 15% (v/v) glycerol. The crystal was mounted in a cryoloop and subsequently flash-cooled in liquid nitrogen. X-ray data were collected at 100 K on beamline BL9-2 at the Stanford Synchrotron Radiation Lightsource (SSRL, Menlo Park, CA) using a Quantum 4 CCD detector (ADSC). A TM0077-SeMet MAD data set was collected to 2.1 Å resolution and the data were indexed in monoclinic space group P21, with unit cell parameters a = 152.6 Å, b = 131.0 Å, and c = 157.8 Å, and β=118.9°, and 12 molecules in the asymmetric unit. Data were indexed and integrated with DENZO and then scaled with SCALEPACK 14. Native TM0077, TM0077-PMSF complex (TM0077-PMS) and TM0077-paraoxon complex (TM0077-DEP) crystals were transferred to crystallization solution supplemented with 10% (v/v) ethylene glycol and flash-cooled to 100K. Data were collected at beamline 5.0.3 of the Advanced Light Source (ALS, Berkeley, CA) and processed with the HKL2000 package 14. The native data set was collected to 2.5 Å resolution, and TM0077-PMS and TM0077-DEP data sets were collected to 2.4 and 2.1 Å, respectively. All data were indexed in orthorhombic space group P212121, with unit cell parameters approximately a=103Å b=104Å c=221Å (See Table 1), and six molecules in the asymmetric unit. Data reduction and refinement statistics for TM0077-SeMet, TM0077-Native, TM0077-PMS and TM0077- DEP are summarized in Table I. Structure solution and refinement The TM0077-SeMet structure was solved by MAD phasing method using a two-wavelength MAD dataset. At the time of the initial data collection (2001), the structure determination of Se-MAD TM0077 posed a significant challenge to crystallographic programs, which were still under active development. As a result, modifications were made in various structure determination and refinement programs to achieve success. For initial phasing, SHELXD 15 was used to find candidate SeMet substructure sites. Attempts to complete phasing were unsuccessful due to the translational non-crystallographic symmetry (NCS) (not recognized initially). Self-consistent sets (partial sets) were found using the CCP4 program PROFESSS 16 and additional SeMet sites were found by SHELXD, and added to these partial sets. The SHARP 17 run did not complete initially; however, updates of SHARP and ARP/wARP eventually helped to resolve issues and an initial trace was obtained by ARP/ wARP. The structure was then refined with BUSTER 18 using tight NCS restraints to an Rcryst and Rfree of 18.6% and 22.3%, respectively. Model building was performed using O 19 and the structure was refined using Refmac5 20. Refinement statistics are summarized in Table I. The final model contains 12 protein molecules (chains A-L) in the asymmetric unit each consisting of residues 2-323. The MolProbity 21 Ramachandran plot analysis showed that 97.4% of all residues are in favored regions with a single outlier, Gln120 of chain B, which is supported by unambiguous electron density. Ramachandran outlier Gln120 of chain B of TM0077-SeMet is due to crystal packing with chain C. The backbone carbonyl oxygens of Gln120 and Gly119 of chain B makes hydrogen bonds with the backbone nitrogen of Gln140 of chain C (3.19 and 3.11 Å, respectively). Levisson et al. Page 4 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The native TM0077 structure, TM0077-PMS and TM0077-DEP structures were solved by molecular replacement using PHASER 22,23 with the TM0077-SeMet hexamer coordinates (pdb: 1vlq; A-F chains) as a search model. One hexamer was successfully located and the structure was further refined with Refmac5 20 using tight NCS restraints to an Rcryst and Rfree of 16.7% and 21.2% (native TM0077), 16.0% and 20.8% (TM0077-PMS) and 16.7% and 20.5% (TM0077-DEP), respectively. Iterative cycles of refinement and building were performed with Refmac5, Phenix 24,25 and Coot 26. All other crystallographic manipulations were carried out with the CCP4 package 16. Refinement statistics are summarized in Table I. The final model of native TM0077 contains residues 3-324 (chains A, B, C, D and F) and 3-323 (chain E) in the asymmetric unit. Analysis of main-chain torsion angles using MolProbity 21 showed that 97.8% of the residues are in favored regions of the Ramachandran plot with 0.2% outliers (Asn302 of chains B, C and D), which are supported by unambiguous electron density. The final model of TM0077-PMS contains residues 3-323 for all chains in the asymmetric unit with 97.5% of the residues in favored regions with 0.2% outliers (Asn302 of chains A, B, D and F). The final model of TM0077-DEP contains residues 0-324 for (chains A, B, C and F) and 0-323 (chains D and E) in the asymmetric unit, respectively, with 97.6% of the residues in favored region of the Ramachandran plot with 0.2% outliers (Asn302 of B, C, D and F chains). Ramachandran outlier Asn302 in the TM0077-Native, TM0077-PMS and TM0077-DEP structures is a neighbor to the catalytic triad residue His303 and may reflect a slightly different state for these structures compared to the Se-Met structure. Structure validation and deposition The quality of the crystal structure was analyzed using the JCSG Quality Control server (http://smb.slac.stanford.edu/jcsg/QC). This server processes the coordinates and data through a variety of validation tools including AutoDepInputTool 27 MolProbity 21, WHATIF 5.0 28, RESOLVE 29, MOLEMAN2 30 as well as several in-house scripts, and summarizes the results. Protein quaternary structure analysis were performed using the PISA server 30. Figures were prepared with PyMOL (DeLano Scientific) 31. RMSD values were calculated using the ProCKSI-Server 32. The structural data have been deposited in the RCSB Protein Data Bank (PDB) with accession codes 1vlq for TM0077-SeMet, 3m81 for TM0077-native, 3m83 for TM0077-DEP and 3m82 for TM0077-PMS. Enzyme assays Esterase activity was measured using p-nitrophenyl esters as described previously 13. Briefly, the standard assay consisted of activity measurements with 0.2 mM p-nitrophenyl acetate as substrate in 50 mM citrate-phosphate (pH 6) at 70°C. The p-nitrophenol liberated was measured continuously at 405 nm on a Hitachi U-2001 spectrophotometer with a temperature-controlled cuvette holder. Extinction coefficients of p-nitrophenol were determined prior to each measurement. Kinetic parameters were determined by direct fitting the data, obtained from multiple measurements, to the Michaelis–Menten curve (Tablecurve 2d, version 5.0). The effect of pH on esterase activity was studied in the pH range from 5 to 10. The buffers used were 50 mM citrate-phosphate (pH 5–8) and 50 mM CAPS (3-(cyclohexylamino) 1- propanesulphonic acid) (pH 9.5–10). The pH of the buffers was set at room temperature, and temperature corrections were made using their temperature coefficients: −0.0028 pH/°C for citrate-phosphate buffer and −0.018 pH/°C for CAPS buffer. The effect of temperature on esterase activity was studied in the range of 40–100°C using 0.2 mM p-nitrophenyl acetate as substrate. Enzyme thermostability was determined by incubating the enzyme in a 50 mM Tris-HCl, 150 mM NaCl (pH 7.8) buffer at 90°C and 100°C for various time intervals. Residual activity was determined in the standard assay. Levisson et al. Page 5 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Inhibition kinetics of PMSF and paraoxon were determined as described for the acetylcholinesterase from electric eel 33. All experiments were performed at 70°C in 50 mM citrate-phosphate (pH 6) buffer and 0.2 mM p-nitrophenyl acetate as substrate. The kinetic constants for the inhibition of TM0077 with PMSF and paraoxon were measured in the concentration range of 1.0–10.0 mM and 0.2–1.0 mM, respectively. Deacetylase activity was determined using high-performance liquid chromatography (HPLC) by measuring the amount of acetic acid released from the substrates cephalosporin C, 7-aminocephalosporanic acid, glucose-pentaacetate and acetylated xylan. Xylan was acetylated by the method described by Johnson 34. The reaction mixture contained 0.9 ml of substrate solution (dissolved in 50 mM Tris-HCl, pH 7.5) and 0.1 ml of enzyme solution, and was incubated at 37°C for various time intervals (0–10 min). The reaction was stopped by adding 0.2 ml of stop solution (100 mN H2SO4 and 30 mM crotonate) and placing the sample on ice. The conditions for HPLC were as follows: column, KC811 Shodex; detection, RI and UV detectors; solvent, 3 mN H2SO4; flow rate, 1.5 ml/min; temperature, 30°C; internal standard, crotonate. One unit of enzyme activity was defined as the amount of enzyme that releases one μmol of acetic acid per minute. Activity on xylan was measured quantitatively using DMSO-extracted xylan (1% polysaccharide solution in 0.1 M sodium phosphate buffer pH 6) at 60°C 35. Xylan will precipitate as a consequence of deacetylation, resulting in a rapid turbidity of the solution. Positional specificity assay The positional specificity of TM0077 was investigated using an enzyme-coupled assay on monoacetylated 4-nitrophenyl β-D-xylopyranosides (pNP-Xyl) as described 36. The β- xylosidase XloA (locus tag: TM0076) from T. maritima was cloned into the vector pET24d in frame with a C-terminal His6-tag. The enzyme was expressed and purified as described above for native TM0077. Activity of XloA was confirmed by measuring the release of p- nitrophenol at 405 nm from the substrate 4-nitrophenyl β-D-xylopyranoside. The enzyme-coupled assay was performed at 60°C in a total volume of 125 μl, which contained 0.1 M sodium phosphate (pH 6 or 7), 2-O-, 3-O-, or 4-O-acetyl pNP-Xyl, the β- xylosidase XloA, and TM0077. Stable 50x-concentrated stock solutions of the substrates were prepared in DMSO. The reaction was started by addition of 2.5 μl of a stock solution to a preheated reaction mixture consisting of phosphate buffer, auxiliary β-xylosidase XloA in excess, and TM0077. The reaction was terminated by addition of 800 μl of a 2% solution of Na2CO3. Liberated p-nitrophenol was determined at 405 nm against substrate and enzyme blanks. A short incubation time for activity determination was used to suppress acetyl migration on the xylopyranosyl-ring, which is significant at pH 6 or 7 37. The kinetic constants were determined at pH 7 and 60°C with reaction times of 2 or 5 minutes. RESULTS and DISCUSSION In silico analysis TM0077 consists of 325 amino acids with a calculated molecular mass of 37 kDa. Sequence analysis, using the SignalP 3.0 server, revealed that TM0077 has no predicted signal sequence and is, therefore, believed to be an intracellular enzyme. Analysis of the gene organization indicates that the TM0077 gene co-localizes with genes encoding a xylanase (TM0070) 38, ABC transporter components (TM0071-TM0075), and a β-xylosidase (TM0076) 39. BLAST-P analysis showed that TM0077 has highest similarity to putative acetyl esterases, acetyl xylan esterases and cephalosporin C deacetylases. Among the BLAST results, a predicted acetyl xylan esterase-related protein from T. maritima (locus tag: TM0435) was also identified. TM0077 was compared with other members of the CE7 Levisson et al. Page 6 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript family using structure-based, multi-sequence alignment and the putative catalytic triad, Ser188, Asp274, and His303, was identified from conservation throughout the analyzed sequences. The putative nucleophilic serine (Ser188) is located within a conserved pentapeptide consensus sequence, Gly-Xaa-Ser-Gln-Gly, typical of this family. Previously, a signature sequence motif, [RGQ]-(x:~70)-[GxSQG]-(x:~115)-[HE] (where x indicates any amino acid), had been suggested for the CE7 family based on an aminoacid alignment of 12 sequences 40. In an updated alignment consisting now of 50 sequences, we observed many sequences that have this signature motif, but it is not conserved throughout the entire family (See Supporting Information and Fig. S1 for the multi-sequence alignment). Overall structure The crystal structure of seleno-methionine incorporated TM0077 (TM0077-SeMet) was determined to 2.1 Å resolution by multi-wavelength anomalous dispersion (MAD) (Table I) with twelve molecules per asymmetric unit. A native apo structure (TM0077-Native) was determined in a different space group (see Methods) to 2.5 Å by molecular replacement, using TM0077-SeMet as a search model, with six molecules in the asymmetric unit (Table I). Each monomer of the native hexamer contained a calcium ion (see below) bound by Lys22, Glu26, and Asp25 via a bridging water molecule. Superposition of the TM0077- SeMet and the TM0077-Native structures gave a root-mean-square difference (RMSD) of 0.12 Å over 321 Cα atoms, which indicates that these structures are nearly identical as expected. In general, the TM0077 structure resembles the canonical α/β-hydrolase fold, which consists of a central, twisted, eight-stranded β-sheet surrounded by α-helices on both sides, with β2 antiparallel to the other strands. TM0077 deviates slightly from the canonical α/β- hydrolase fold at two locations: a three-helix insertion after strand β6 and an extension of the N-terminus (Fig. 2). Insertions after β6 or β7 are common for α/β-hydrolases and are proposed to help shape the substrate-binding site 41. The N-terminus is extended by two helices (αA-1 and αA-2) and an antiparallel β-strand (β-1) that aligns with the other eight β- strands (β1-β8) and extends the central β-sheet. This nine-stranded β-sheet is highly twisted, and β-1 and β8 at the extreme edges are rotated approximately 130° relative to each other. Helices αA and αB both contain a short 310-helix segment at their N-terminus. Helices αA-1, αA-2, αB, αC, αD, αD1, αD2, αD3, αE, and the 310-helix η2 are located on one side of the central β-sheet, and helices αA, αF and the 310-helix η1 are on the other side. A structural similarity search was performed using the program DALI 42. Monomer A of the TM0077-SeMet structure was used as a search model and similarity was found with cephalosporin C deacetylases, acetyl xylan esterases, acylamino-releasing enzymes, dipeptidyl peptidases and some esterases and lipases. TM0077 is structurally most similar to cephalosporin C deacetylase (CAH) from B. subtilis (PDB: 1ods) 40, acetyl xylan esterase (AXE) from B. pumilus (PDB: 3fvr and 2xlb), acetyl xylan esterase (AXE1) from Thermoanaerobacterium sp. JW/SL YS485 (PDB: 3fcy), and acylpeptide hydrolase/esterase apAPH from Aeropyrum pernix K1 (PDB: 1ve6) 43. The sequence identity between TM0077 and CAH is 41% and the two structures align with a Z-score of 46 and an RMSD of 1.5 Å over 312 Cα atoms. The sequence identity with apAPH is 17% with a Z-score of 23.3 and an RMSD of 2.3 Å over 230 Cα atoms. Superpositions of TM0077 with CAH and with apAPH are shown in Fig. 3. Quaternary structure The crystal structure of TM0077-SeMet contains two hexamers in the asymmetric unit that are related by a non-crystallographic two-fold axis. Each hexamer contains a dimer of trimers with a back-to-back arrangement (Fig. 4). The apo and the complex crystals Levisson et al. Page 7 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript contained one hexamer in the asymmetric unit. Crystallographic packing analysis using PISA (EBI) 44 indicated that the relevant physiological oligomeric state of TM0077 is a hexamer, which was confirmed by size exclusion chromatography coupled with static light scattering. Further analyses of the hexameric assembly indicated that two main interfaces play an essential role in complex formation. The first interface between subunit A and B (green and cyan in Fig. 4) (identical to C/D and E/F) is stabilized by seven hydrogen bonds on average and has a buried surface area of 1024 Å2 contributed by each chain. The second interface between A and F (green and purple in Fig. 4) (B/C and D/E) is stabilized by 17 hydrogen bonds on average with a buried surface area of 1079 Å2 contributed by each chain However, a multiple sequence alignment of TM0077 with other CE7 esterases showed that the residues involved in these two main interfaces are not conserved. Other secondary interfaces bury around 514 Å2 contributed by each chain. The hexamer has a total buried surface area of 18,860 Å2, which is approximately 30% of the total surface area. Approximately 3,143 Å2 per monomer is, therefore, buried upon complex formation. The TM0077 hexamer has a doughnut-shape when viewed from the side, with the six active sites located in the interior of the complex, where they line an oval-shaped cavity [Fig. 4(a)]. This cavity is accessible via two entrances, one on each side of the flat hexamer. Each of these entrances is approximately 13 Å wide and connects to a short tunnel or pore spanning approximately 10 Å to reach the inner cavity. Interestingly, in the TM0077-SeMet hexamer, the entrance to the internal cavity is blocked by three phenylalanine residues (Phe4), one for each of three monomers that compose half of the hexamer [Fig. 4(b)]. Residue Phe4 is located in the mobile N-terminus (high B-values), which may indicate some flexibility or multiple conformations. Calcium ions were identified, by the electron density and coordination geometry, supported by their presence in the crystallization reagents, in the native TM0077, TM0077-PMS and TM0077-DEP structures, but not in the TM0077-SeMet structure. The SeMet protein was crystallized without any calcium in the crystallization reagents. In each subunit of the hexamer, one calcium ion is located at the N-terminal region of helix αA-1, and is coordinated by the backbone carbonyl of Lys22 and the Glu26 carboxylate. The remainder of the calcium coordination sphere is filled with waters from a neighboring solvent channel present in all molecules in the asymmetric unit. The Asp25 carboxylate contributes to the calcium binding via one of the coordinating water molecules. Another calcium ion is bound in a crystal packing interface between chain A and chain C′ of a crystallographic symmetry- related hexamer. This calcium is coordinated by the carboxylates of GluA45 and AspA58 from one chain and the carboxylate from Glu C’45 (bidentate coordination) of the symmetry-related chain with three water molecules completing a capped-octahedral coordination sphere. An equivalent calcium binding site is also observed in the crystal packing interface between chains D and B′. No significant increase or reduction of activity of TM0077 was observed in the presence of calcium ions or EDTA. Therefore, it seems that these calcium ions are not important for activity. On the other hand, calcium may help stabilize the structure. No calcium was present in the B. subtilis CAH structure 40; however, Lys22, Glu26 and Ser25 are conserved and may also act as a calcium binding site. Enzyme activity The activity of TM0077 was investigated using p-nitrophenol esters with varying acyl-chain length, ranging from C2 to C18. TM0077 is only active on the short-chain p-nitrophenol esters of acetate and propionate and does not hydrolyze esters with acyl chains longer than four carbons. No significant difference was found in the catalytic efficiency (kcat/Km) for the hydrolysis of p-nitrophenyl with acyl chains containing 2 to 3 carbons (Table II) [Fig. 1(c)]. Levisson et al. Page 8 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript The effect of temperature on activity was studied using p-nitrophenyl acetate as substrate. The esterase activity increased from 40°C upwards until 100°C [Fig. 5(a)]. An Arrhenius analysis resulted in linear plots in the temperature ranges of 40–60°C and 60–100°C [Fig. 5(a); inset], with calculated activation energies for the formation of the enzyme substrate- enzyme complex of 33.7 and 21.9 kJ/mol, respectively. The transition or break in linearity of the Arrhenius plot at 60°C (1000/T (K) = 3.0) could indicate some conformational change of the enzyme. TM0077 is fairly resistant to thermal inactivation. An approximate 50% transient increase in activity is seen during the first 10 to 20 minutes when the enzyme is incubated at 90°C. After 30 minutes, inactivation of function occurs by first order kinetics with a half-life of approximately 120 minutes [Fig. 5(b)]. A transient activation has also been observed for other thermophilic esterases, such from Sulfolobus shibatae 45, and it is believed that a high temperature is needed in order to obtain the optimal conformation for catalysis. TM0077 was not stable at 100°C, resulting in a half-life of less than 5 minutes. However, the optimum temperature and thermal stability of TM0077 are still considerably higher than those reported for other characterized CE7 esterases, including the Thermoanaerobacterium enzyme that has a temperature optimum of 80°C and a half-life of 1h at 75°C 46. The effect of pH on activity was measured in the pH range of 4.8 to 9.2 using the substrate p-nitrophenyl acetate. TM0077 displayed maximum activity at approximately pH 7.5 [Fig. 5(c)], which is comparable to other CE7 esterases, such as the acetyl xylan esterases from Thermoanaerobacterium sp. strain JW/SL-YS485 46. Positional specificity The positional specificity of TM0077 was tested on three monoacetates of 4-nitrophenyl β- D-xylopyranoside (pNP-Xyl). To determine the enzyme activity, the β-xylosidase XloA 39 (TM0076) from T. maritima is required as an auxiliary enzyme. This thermostable XloA enzyme was, therefore, cloned, heterologously expressed, purified to homogeneity, and its activity was confirmed by measuring release of p-nitrophenol from the substrate pNP-Xyl (data not shown). The β-xylosidase was not active on the three monoacetates of pNP-Xyl. In the XloA-coupled assay, TM0077 hydrolyzed acetate from positions 2, 3 and 4 of pNP-Xyl with similar catalytic efficiency. The results are summarized in Table II. In addition, TM0077 was investigated for its ability to remove acetyl groups from 7- aminocephalosporanic acid (7-ACA), cephalosporin C, glucose penta-acetate, N-acetyl-D- glucosamine, xylan and acetylated xylan. TM0077 has no activity for acetylated and non- acetylated xylan polymers, indicating that it is, indeed, an acetyl esterase and not an acetyl xylan esterase. As expected for an acetyl esterase, TM0077 displayed high activity on glucose penta-acetate with a turnover number of 2680 s−1. Like other members of CE7, TM0077 was also able to hydrolyze the acetyl groups from both cephalosporin C and 7- ACA with a turnover number of 376 s−1 and 1140 s−1, respectively. TM0077 was not able to hydrolyze the acetyl group from N-acetyl-D-glucosamine, indicating that it is specific for ester bonds and unable to hydrolyze amide bonds. Inhibitor assays and TM0077 structures complexed with PMSF and paraoxon PMSF and paraoxon [Fig. 1(d,e)] are competitive irreversible inhibitors of esterases. Inhibition proceeds by the formation of a reversible Michaelis complex, followed by an irreversible step and inhibition can, therefore, be characterized by two parameters: a dissociation constant and a binding rate constant. The inhibition kinetics for paraoxon and PMSF were investigated in the presence of p-nitrophenyl acetate, as described previously 47, and the dissociation and rate constants were 0.5 ± 0.1 mM and 0.13 ± 0.02 s−1 for paraoxon, and 1.1 ± 0.2 mM and 0.020 ± 0.001 s−1 for PMSF, respectively. The acetyl xylan esterase Levisson et al. Page 9 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript from Bacillus pumilus (BpAxe) has slightly reduced sensitivity to paraoxon (dissociation and rate constants of respectively 5.4 mM and 0.012 s−1), likely due to steric hindrance of two tyrosine residues (Tyr91 and Tyr206) that hamper the binding of paraoxon. Although these residues are essentially conserved in TM0077 (Tyr92 and Phe213), TM0077 is more sensitive to paraoxon than BpAxe48. In comparison to EST2 of Alicyclobacillus acidocaldarius 49 and EstA of T. maritima 47, the TM0077 dissociation constant is slightly higher, but the rate constant is comparable. No significant stimulation or reduction of activity of TM0077 was observed in the presence of divalent metal ions or ethylenediaminetetraacetic acid (EDTA). To obtain more information about inhibitor binding and any possible conformational changes during catalysis, TM0077 was co-crystallized with the inhibitors PMSF and paraoxon and the PMSF (TM0077-PMS) and paraoxon (TM0077-DEP) structures were determined to 2.4 Å and 2.1 Å, respectively (Table I). The electron density map of TM0077 with PMSF showed clear density for the PMSF covalent modification. The fluorine was cleaved from the PMSF molecule during the binding reaction and the phenylmethyl sulfonyl (PMS) moiety is covalently bound to the Oγ atom of Ser188. The native apo and PMS- bound structures superimpose well with RMSD’s of 0.09–0.11 Å over 320–321 Cα atoms. Electron density maps of the paraoxon-bound structure displayed clear density for a diethyl- phosphate moiety covalently bound to the Oγ atom of Ser188. This covalent modification indicates that the p-nitrophenol group of paraoxon was cleaved off during co-crystallization, and a tetrahedral product reminiscent of the first transition state was formed during carboxyl ester hydrolysis. The native apo and paraoxon-bound structures superimpose with RMSD’s of 0.12–0.32 Å over 320–322 Cα atoms. Attempts to obtain co-crystals of TM0077 with cephalosporin C, even at a low temperature of 4°C, were unsuccessful. Analysis of the active site TM0077 has a classic catalytic triad, consisting of Ser188 as the nucleophile, His303 as the proton acceptor/donor, and Asp274 as the acidic residue stabilizing the histidine (Fig. 6). The catalytic serine Ser188 is located within a conserved pentapeptide sequence, Gly-X-Ser- X-Gly (GGSQG), characteristic of esterases and lipases. The positions of Ser188, Asp274, and His303 are consistent with their expected locations in the canonical fold of the α/β- hydrolase family. Ser188 is located at the nucleophile elbow in a sharp turn between β5 and helix αC. The presence of three glycine residues (Gly186, Gly187, and Gly190) in close proximity to Ser188 prevents steric hindrance and facilitates access to the nucleophile elbow. Asp274 and His303 are located in loops between β7 and helix αE, and between β8 and helix αF, respectively. The oxyanion hole is formed by the backbone amide groups of Tyr92 and Gln189. The catalytic triad and oxyanion hole are located in a depression on the surface of TM0077. This ellipsoid pocket (S1), which is approximately 12 Å wide, extends 15 Å from the catalytic serine. A smaller pocket (S2), approximately 5 Å long, extends to the other side of the catalytic serine [Fig. 6(a)]. The volume of both pockets combined (S1 + S2) is 1082 Å3 (CASTp analysis; 50). The substrate-binding pocket is bordered by residues from helices αA and αF, and its base is formed by residues from β-strands 4, 5, 6, and their adjacent C-terminal loops. The overall pocket is hydrophobic, although it does have some polar residues (Gln88, Asp210, and Gln314), which may interact with the substrate. In the native apo structure, the Ser188 hydroxyl makes a hydrogen bond with the imidazole of His303 [Fig. 6(b)]. Extra density was observed near the side chain of Ser188 and was interpreted as a chloride ion based on electron density size and shape as well as the geometry of the interactions with surrounding residues. This chloride ion is bound at the entrance of the oxyanion hole, forming hydrogen bonds with the backbone amides of Tyr92 and Gln189. In the PMSF-bound structure, the phenyl ring of the inhibitor is located in the small active site groove surrounded by hydrophobic residues Tyr92, Trp124, Pro228, Ile276, Levisson et al. Page 10 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript and His303 [Fig. 6(c)]. The sulfonyl group of PMSF makes hydrogen bonds with the backbone amides of Tyr92 and Gln189. In the paraoxon-bound structure, the diethyl- phosphate (DEP) moiety is stabilized by hydrogen-bonding interactions with the oxyanion hole. One of the two ethyl arms of bound paraoxon points toward the larger pocket in the protein, while the other follows the groove of the small pocket. The two ethyl arms are stabilized by packing against Tyr92, Trp124, Pro228, Ile276, and His303 [Fig. 6(d)]. Two rotamers of the catalytic serine Although no large conformational changes were observed upon binding of PMSF or paraoxon, a different rotamer of the catalytic serine side chain was observed compared to native TM0077 [Fig. 7(a,b)]. Similar changes have been observed in several other esterases and have been shown to play a key role in the catalytic mechanism (see CONCLUSION for more details). In the native structure, the catalytic Ser188 Oγ is in the plane of the imidazole ring of His303, as most commonly observed in the resting state of esterases and lipases 51. The Ser188 Oγ forms a hydrogen bond (2.6 Å) with His303 Nε2. In the PMSF- and paraoxon-bound structures, the conformation of the catalytic serine changes; the Ser188 Oγ rotates about 110°, increasing the distance (3.1 Å and 2.8 Å for PMSF and paraoxon bound structures, respectively) to the His303 imidazole ring. In the TM0077-SeMet structure, the catalytic serine is also rotated over ~110°, with a distance to the imidazole ring of 3.0 Å [Fig. 7(c.)]. A probable explanation for this observation could be the protonation of His303, since TM0077-SeMet was crystallized at a lower pH (pH 4.2) compared to the native TM0077 (pH 7.3). Furthermore, extra electron density was identified in the TM0077-SeMet structure, suggesting a partially occupied acyl intermediate on Ser188. However, as this density is not sufficiently clear and interpretable to fit an acyl intermediate, water molecules were modeled instead. No rearrangements of any other residues in the active site were observed. CONCLUSION TM0077 from the hyperthermophilic bacterium T. maritima was predicted from its gene sequence to be an acetyl xylan esterase. We have expressed and purified TM0077 and experimentally demonstrated that it has ester-hydrolyzing activity. The TM0077 activity was restricted to short acyl chain esters (C2 and C3) when artificial p-nitrophenyl-esters were used as substrates. In addition, the enzyme has high specific activity on glucose penta- acetate. However, no activity was detected on xylan or acetylated xylan. Thus, TM0077 should be reclassified as an acetyl esterase, and not as an acetyl xylan esterase as currently annotated 52. Furthermore, the lack of any apparent signal sequence suggests that the protein is not secreted. Thus, the predicted intracellular location of TM0077 is compatible with a role other than the deacetylation of extracellular xylan. Based on these results, we conclude that the likely biological function of TM0077 is removal of the remaining acetyl groups from the short, end products of xylan degradation that are imported into the cytoplasm. The resulting deacetylated xylose oligomers are the substrates for a β-xylosidase. This role for TM0077 is in good agreement with the clustering of the TM0077 gene with other genes involved in xylose metabolism. However, it cannot be ruled out that TM0077 may also act on other small, acetylated compounds. TM0077 is the first esterase from the CE7 family to be tested for its positional specificity for the deacetylation of 4-nitrophenyl-β-D-xylopyranoside. TM0077 hydrolyzes acetate at the 2, 3 and 4 positions of 4-nitrophenyl-β-D-xylopyranoside with similar efficiency. Conversely, the CtAxe esterase from Clostridium thermocellum in the CE4 family shows a clear preference for hydrolyzing acetate at the 2 position 53, and Penicillium purpurogenum AXE II esterase, a member of the CE5 family, also has a preference for acetate at position 2 54. Levisson et al. Page 11 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript This lack of preference for a specific position of the acetate group correlates with the relative broad substrate specificity of the CE7 esterases. Esterases and deacetylases in the CE7 family are unusual in that they are active towards both acetylated xylo-oligosaccharides and the antibiotic cephalosporin C [Fig. 1(a,b)]. Therefore, TM0077 was investigated for activity towards the substrates 7-ACA and cephalosporin C. The activity of TM0077 on these substrates is approximately ten-fold higher than that of the acetyl xylan esterase from B. pumilus 55 or the acetyl esterase from Thermoanaerobacterium sp. strain JW/SL YS485 56. TM0077 has a higher hydrolytic activity on 7-ACA compared to cephalosporin C, as described for other CE7 esterases 40,55,56. Nonetheless, it is unlikely that both compounds are natural substrates, because the stability of these compounds at the optimal growth temperature (80°C) of T. maritima is very low. Crystal structures of TM0077 in complex with inhibitors PMSF and paraoxon revealed that, upon binding of PMSF or paraoxon, the reaction is trapped at the acylation step via the formation of a covalent tetrahedral reaction product. In the complexed structures, the negatively charged oxygen of the tetrahedral intermediate, derived from the substrate oxyanion, is stabilized by hydrogen bonds with the backbone amide groups of Tyr92 and Gln189. Comparison of the TM0077 complexed structures with the native structure shows that the catalytic serine (Ser188) Oγ rotates about 110°, thereby increasing the distance between Ser188 Oγ and His303 Nε2. Such a conformational change of the catalytic serine has been observed in several other esterases, including Fusarium solani cutinase 57, Penicillium purpurogenum acetyl xylan esterase 51, Rhodococcus sp. strain MB1 cocaine esterase 58, Bacillus subtilis lipase 59, Rhodococcus sp. strain H1 heroine esterase 60, and Aspergillus niger feruloyl esterase 61. The classical model for the catalytic mechanism of esterases consists of a sequential two-step hydrolysis. The first reaction involves nucleophilic attack by the catalytic serine on the substrate carbonyl carbon, resulting in an acyl-enzyme and the liberation of an alcohol. In the second reaction, a water molecule performs a nucleophilic attack on the acyl-enzyme, the acyl-enzyme bond breaks and the carboxylate is released 62. Although the catalytic mechanism of esterases is well established, it is unclear why the initially generated tetrahedral intermediate does not collapse back to the reactant complex during the nucleophilic attack of the substrate. A previously proposed mechanism that would prevent this collapse is the spatial reorganization of the catalytic residues during the initial catalytic step, causing the residues to separate and thereby drive the reaction forward 62–64. The apo and inhibitor bound structures of TM0077, presented herein, support this proposed mechanism. Moreover, in a recent study of the serine protease mechanism, it was suggested that subtle atomic motions of the catalytic serine and histidine residues during the catalytic cycle favor the forward reaction 65. Thus, rotation of Ser188 Oγ of TM0077 may be required to inhibit reversal of the reaction. In addition, such changes may facilitate the access of water to the catalytic histidine so that the second step of the reaction can go to completion. Deacetyl cephalosporins are valuable building blocks for the production of semisynthetic β- lactam antibiotics. These compounds are derived from cephalosporin C or 7- aminocephalosporanic acid via enzymatic or chemical processes 10. The thermostable TM0077 esterase may be valuable in the preparation of derivatives of β-lactam antibiotics. Recently, the substrate specificity of the acetyl xylan esterase from P. purpurogenum was engineered to accept a range of fatty acid esters of up to 14 carbons compared to its wild- type preference for acetate54. It might also be possible to engineer TM0077 and enable the (de)acetylation of cephalosporins at the C10 position with various acyl chains. Because of its high stability and activity on 7-ACA and cephalosporin C, TM0077 presents an attractive candidate for the production of new semi-synthetic antibiotics. Levisson et al. Page 12 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Supplementary Material Refer to Web version on PubMed Central for supplementary material. Acknowledgments We gratefully acknowledge contributions from George Sheldrick for modifications of the SHELXD program, and for Global Phasing Ltd. that made significant improvements in the automation of autoSHARP. We also thank Victor Lamzin for updates of chain docking of the ARP/wARP program, and Gerard Bricogne and Eleanor Dodson for helpful discussion on phasing for the large TM0077-SeMet structure, and Willem J. van Berkel for valuable discussion on the catalytic mechanism of TM0077. Portions of this research were carried out at the Stanford Synchrotron Radiation Lightsource (SSRL) and the Advanced Light Source (ALS). The SSRL is a Directorate of SLAC National Accelerator Laboratory and an Office of Science User Facility operated for the U.S. Department of Energy Office of Science by Stanford University. The SSRL Structural Molecular Biology Program is supported by the DOE Office of Biological and Environmental Research, and by the National Institutes of Health, National Center for Research Resources, Biomedical Technology Program (P41RR001209), and the National Institute of General Medical Sciences. The ALS is supported by the Director, Office of Science, Office of Basic Energy Sciences, Materials Sciences Division, of the U.S. Department of Energy under Contract No. DE- AC02-05CH11231 at Lawrence Berkeley National Laboratory. Genomic DNA from Thermotoga maritima MSB8 (DSM3109) (ATCC #43589D-5) was obtained from the American Type Culture Collection (ATCC). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of General Medical Sciences or the National Institutes of Health. Grant sponsor: NIH Grant numbers U54 GM094586 and U54 GM074898 (Protein Structure Initiative); Grant sponsor: The Graduate School VLAG Wageningen, the Netherlands (ML). References 1. Huber, R.; Hannig, M. Thermotogales. In: Dworkin, M.; Falkow, S.; Rosenberg, E.; Schleifer, K-H.; Stackebrandt, E., editors. The Prokaryotes: An Evolving Electronic Resource for the Microbiological Community. Vol. 7. New-York: Springer-Verlag; 2006. p. 899-922. 2. Nelson KE, Clayton RA, Gill SR, Gwinn ML, Dodson RJ, Haft DH, Hickey EK, Peterson JD, Nelson WC, Ketchum KA, McDonald L, Utterback TR, Malek JA, Linher KD, Garrett MM, Stewart AM, Cotton MD, Pratt MS, Phillips CA, Richardson D, Heidelberg J, Sutton GG, Fleischmann RD, Eisen JA, White O, Salzberg SL, Smith HO, Venter JC, Fraser CM. Evidence for lateral gene transfer between Archaea and bacteria from genome sequence of Thermotoga maritima. Nature. 1999; 399:323–329. [PubMed: 10360571] 3. Chhabra SR, Shockley KR, Conners SB, Scott KL, Wolfinger RD, Kelly RM. Carbohydrate- induced differential gene expression patterns in the hyperthermophilic bacterium Thermotoga maritima. J Biol Chem. 2003; 278:7540–7552. [PubMed: 12475972] 4. Conners SB, Montero CI, Comfort DA, Shockley KR, Johnson MR, Chhabra SR, Kelly RM. An expression-driven approach to the prediction of carbohydrate transport and utilization regulons in the hyperthermophilic bacterium Thermotoga maritima. J Bacteriol. 2005; 187:7267–7282. [PubMed: 16237010] 5. VanFossen AL, Lewis DL, Nichols JD, Kelly RM. Polysaccharide degradation and synthesis by extremely thermophilic anaerobes. Ann N Y Acad Sci. 2008; 1125:322–337. [PubMed: 18378602] 6. Biely P, Mackenzie CR, Puls J, Schneider H. Cooperativity of esterases and xylanases in the enzymatic degradation of acetyl xylan. Bio-Technology. 1986; 4:731–733. 7. Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V, Henrissat B. The Carbohydrate- Active EnZymes database (CAZy): an expert resource for Glycogenomics. Nucleic Acids Res. 2009; 37:D233–D238. [PubMed: 18838391] 8. Topakas E, Paul C. Microbial xylanolytic carbohydrate esterases. Industrial Enzymes. 2007:83–97. 9. Weil J, Miramonti J, Ladisch MR. Cephalosporin-C: Mode of action and biosynthetic pathway. Enzyme Microb Technol. 1995; 17:85–87. 10. Barends TRM, Yoshida H, Dijkstra BW. Three-dimensional structures of enzymes useful for beta- lactam antibiotic production. Curr Opin Biotechnol. 2004; 15:356–363. [PubMed: 15358004] 11. Martínez-Martínez I, Montoro-García S, Lozada-Ramírez JD, Sánchez-Ferrer Á, García-Carmona F. A colorimetric assay for the determination of acetyl xylan esterase or cephalosporin C acetyl Levisson et al. Page 13 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript esterase activities using 7-amino cephalosporanic acid, cephalosporin C, or acetylated xylan as substrate. Anal Biochem. 2007; 369:210–217. [PubMed: 17651681] 12. Lesley SA, Kuhn P, Godzik A, Deacon AM, Mathews I, Kreusch A, Spraggon G, Klock HE, McMullan D, Shin T, Vincent J, Robb A, Brinen LS, Miller MD, McPhillips TM, Miller MA, Scheibe D, Canaves JM, Guda C, Jaroszewski L, Selby TL, Elsliger MA, Wooley J, Taylor SS, Hodgson KO, Wilson IA, Schultz PG, Stevens RC. Structural genomics of the Thermotoga maritima proteome implemented in a high-throughput structure determination pipeline. Proc Natl Acad Sci USA. 2002; 99:11664–11669. [PubMed: 12193646] 13. Levisson M, van der Oost J, Kengen SW. Characterization and structural modeling of a new type of thermostable esterase from Thermotoga maritima. FEBS Journal. 2007; 274:2832–2842. [PubMed: 17466017] 14. Otwinowski Z, Minor W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 1997; 276:307–326. 15. Schneider TR, Sheldrick GM. Substructure solution with SHELXD. Acta Crystallogr Sect D Biol Crystallogr. 2002; 58:1772–1779. [PubMed: 12351820] 16. Collaborative Computational Project Number 4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr Sect D Biol Crystallogr. 1994; 50:760–763. [PubMed: 15299374] 17. Bricogne G, Vonrhein C, Flensburg C, Schiltz M, Paciorek W. Generation, representation and flow of phase information in structure determination: recent developments in and around SHARP 2.0. Acta Crystallogr Sect D Biol Crystallogr. 2003; 59:2023–2030. [PubMed: 14573958] 18. Bricogne, G.; Blanc, E.; Brandl, M.; Flensburg, C.; Keller, P.; Paciorek, W.; Roversi, P.; Smart, O.; Vonrhein, CTW. BUSTER, version 2.8.0. Cambridge, United Kingdom: Global Phasing Ltd; 2009. 19. Jones TA, Zou JY, Cowan SW, Kjeldgaard M. Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr A Found Crystallogr. 1991; 47:110–119. 20. Winn MD, Murshudov GN, Papiz MZ. Macromolecular TLS refinement in REFMAC at moderate resolutions. Methods Enzymol. 2003; 374:300–321. [PubMed: 14696379] 21. Davis IW, Murray LW, Richardson JS, Richardson DC. MOLPROBITY: structure validation and all-atom contact analysis for nucleic acids and their complexes. Nucleic Acids Res. 2004; 32:W615–W619. [PubMed: 15215462] 22. McCoy AJ, Grosse-Kunstleve RW, Storoni LC, Read RJ. Likelihood-enhanced fast translation functions. Acta Crystallogr Sect D Biol Crystallogr. 2005; 61:458–464. [PubMed: 15805601] 23. Storoni LC, McCoy AJ, Read RJ. Likelihood-enhanced fast rotation functions. Acta Crystallogr Sect D Biol Crystallogr. 2004; 60:432–438. [PubMed: 14993666] 24. Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Terwilliger TC. PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr Sect D Biol Crystallogr. 2002; 58:1948– 1954. [PubMed: 12393927] 25. Murshudov GN, Vagin AA, Dodson EJ. Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr Sect D Biol Crystallogr. 1997; 53:240–255. [PubMed: 15299926] 26. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr Sect D Biol Crystallogr. 2004; 60:2126–2132. [PubMed: 15572765] 27. Yang H, Guranovic V, Dutta S, Feng Z, Berman HM, Westbrook JD. Automated and accurate deposition of structures solved by X-ray diffraction to the Protein Data Bank. Acta Crystallogr Sect D Biol Crystallogr. 2004; 60:1833–1839. [PubMed: 15388930] 28. Vriend G. WHAT IF: a molecular modeling and drug design program. J Mol Graph. 1990; 8:52– 56. [PubMed: 2268628] 29. Terwilliger TC. Automated side-chain model building and sequence assignment by template matching. Acta Crystallogr Sect D Biol Crystallogr. 2003; 59:45–49. [PubMed: 12499538] 30. Kleywegt GJ. Validation of protein crystal structures. Acta Crystallogr Sect D Biol Crystallogr. 2000; 56:249–265. [PubMed: 10713511] Levisson et al. Page 14 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 31. DeLano, WL. The Pymol molecular graphics system. DeLano Scientific; San Carlos, CA: 2002. 32. Barthel D, Hirst JD, Blazewicz J, Burke EK, Krasnogor N. ProCKSI: a decision support system for Protein (structure) Comparison, Knowledge, Similarity and Information. BMC Bioinformatics. 2007; 8:416. [PubMed: 17963510] 33. Forsberg A, Puu G. Kinetics for the inhibition of acetylcholinesterase from the electric eel by some organophosphates and carbamates. Eur J Biochem. 1984; 140:153–156. [PubMed: 6705793] 34. Johnson KG, Harrison BA, Schneider H, Mackenzie CR, Fontana JD. Xylan-hydrolyzing enzymes from Streptomyces spp. Enzyme Microb Technol. 1988; 10:403–409. 35. Biely P, Mackenzie CR, Schneider H. Production of acetyl xylan esterase by Trichoderma reesei and Schizophyllum commune. Can J Microbiol. 1988; 34:767–772. 36. Biely P, Mastihubova M, la Grange DC, van Zyl WH, Prior BA. Enzyme-coupled assay of acetylxylan esterases on monoacetylated 4-nitrophenyl beta-D-xylopyranosides. Anal Biochem. 2004; 332:109–115. [PubMed: 15301955] 37. Mastihubova M, Biely P. Lipase-catalysed preparation of acetates of 4-nitrophenyl beta-D- xylopyranoside and their use in kinetic studies of acetyl migration. Carbohydr Res. 2004; 339:1353–1360. [PubMed: 15113674] 38. Jiang ZQ, Kobayashi A, Ahsan MM, Lite L, Kitaoka M, Hayashi K. Characterization of a thermostable family 10 endo-xylanase (XynB) from Thermotoga maritima that cleaves p- nitrophenyl-beta-D-xyloside. J Biosci Bioeng. 2001; 92:423–428. [PubMed: 16233122] 39. Xue YM, Shao WL. Expression and characterization of a thermostable beta-xylosidase from the hyperthermophile, Thermotoga maritima. Biotechnol Lett. 2004; 26:1511–1515. [PubMed: 15604789] 40. Vincent F, Charnock SJ, Verschueren KHG, Turkenburg JP, Scott DJ, Offen WA, Roberts S, Pell G, Gilbert HJ, Davies GJ, Brannigan JA. Multifunctional xylooligosaccharide/cephalosporin C deacetylase revealed by the hexameric structure of the Bacillus subtilis enzyme at 1. 9 angstrom resolution. J Mol Biol. 2003; 330:593–606. [PubMed: 12842474] 41. Nardini M, Dijkstra BW. Alpha/beta hydrolase fold enzymes: the family keeps growing. Curr Opin Struct Biol. 1999; 9:732–737. [PubMed: 10607665] 42. Holm L, Kaariainen S, Rosenstrom P, Schenkel A. Searching protein structure databases with DaliLite v.3. Bioinformatics. 2008; 24:2780–2781. [PubMed: 18818215] 43. Bartlam M, Wang G, Yang H, Gao R, Zhao X, Xie G, Cao S, Feng Y, Rao Z. Crystal structure of an acylpeptide hydrolase/esterase from Aeropyrum pernix K1. Structure. 2004; 12:1481–1488. [PubMed: 15296741] 44. Krissinel E, Henrick K. Inference of macromolecular assemblies from crystalline state. J Mol Biol. 2007; 372:774–797. [PubMed: 17681537] 45. Huddleston S, Yallop CA, Charalambous BM. The identification and partial characterisation of a novel inducible extracellular thermostable esterase from the archaeon Sulfolobus shibatae. Biochem Biophys Res Commun. 1995; 216:495–500. [PubMed: 7488139] 46. Shao WL, Wiegel J. Purification and characterization of 2 thermostable acetyl xylan esterases from Thermoanaerobacterium sp strain JW/SL-YS485. Appl Environ Microbiol. 1995; 61:729–733. [PubMed: 7574610] 47. Levisson M, Sun L, Hendriks S, Swinkels P, Akveld T, Bultema JB, Barendregt A, van den Heuvel RH, Dijkstra BW, van der Oost J, Kengen SW. Crystal structure and biochemical properties of a novel thermostable esterase containing an immunoglobulin-like domain. J Mol Biol. 2009; 385:949–962. [PubMed: 19013466] 48. Montoro-García S, Gil-Ortiz F, García-Carmona F, Polo LM, Rubio VAS-F. The crystal structure of the cephalosporin deacetylating enzyme acetyl xylan esterase bound to paraoxon explains the low sensitivity of this serine hydrolase to organophosphate inactivation. Biochem J. 201110.1042/ BJ20101859 49. Febbraio F, D’Andrea SE, Mandrich L, Merone L, Rossi M, Nucci R, Manco G. Irreversible inhibition of the thermophilic esterase EST2 from Alicyclobacillus acidocaldarius. Extremophiles. 2008; 12:719–728. [PubMed: 18622571] Levisson et al. Page 15 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 50. Dundas J, Ouyang Z, Tseng J, Binkowski A, Turpaz Y, Liang J. CASTp: computed atlas of surface topography of proteins with structural and topographical mapping of functionally annotated residues. Nucleic Acids Res. 2006; 34:W116–W118. [PubMed: 16844972] 51. Ghosh D, Sawicki M, Lala P, Erman M, Pangborn W, Eyzaguirre J, Gutierrez R, Jornvall H, Thiel DJ. Multiple conformations of catalytic serine and histidine in acetylxylan esterase at 0.90 angstrom. J Biol Chem. 2001; 276:11159–11166. [PubMed: 11134051] 52. Williamson G, Kroon PA, Faulds CB. Hairy plant polysaccharides: a close shave with microbial esterases. Microbiology. 1998; 144:2011–2023. [PubMed: 9720023] 53. Biely P, Mastihubova M, Puchart V. The vicinal hydroxyl group is prerequisite for metal activation of Clostridium thermocellum acetylxylan esterase. Biochim Biophys Acta. 2007; 1770:565–570. [PubMed: 17261352] 54. Colombres M, Garate JA, Lagos CF, Araya-Secchi R, Norambuena P, Quiroz S, Larrondo L, Perez-Acle T, Eyzaguirre J. An eleven amino acid residue deletion expands the substrate specificity of acetyl xylan esterase II (AXE II) from Penicillium purpurogenum. J Comput Aided Mol Des. 2008; 22:19–28. [PubMed: 18060506] 55. Degrassi G, Kojic M, Ljubijankic G, Venturi V. The acetyl xylan esterase of Bacillus pumilus belongs to a family of esterases with broad substrate specificity. Microbiology. 2000; 146:1585– 1591. [PubMed: 10878123] 56. Lorenz WW, Wiegel J. Isolation, analysis, and expression of two genes from Thermoanaerobacterium sp. strain JW/SL YS485: a beta-xylosidase and a novel acetyl xylan esterase with cephalosporin C deacetylase activity. J Bacteriol. 1997; 179:5436–5441. [PubMed: 9286998] 57. Longhi S, Czjzek M, Lamzin V, Nicolas A, Cambillau C. Atomic resolution (1. 0 angstrom) crystal structure of Fusarium solani cutinase: Stereochemical analysis. J Mol Biol. 1997; 268:779–799. [PubMed: 9175860] 58. Larsen NA, Turner JM, Stevens J, Rosser SJ, Basran A, Lerner RA, Bruce NC, Wilson IA. Crystal structure of a bacterial cocaine esterase. Nat Struct Biol. 2002; 9:17–21. [PubMed: 11742345] 59. Kawasaki K, Kondo H, Suzuki M, Ohgiya S, Tsuda S. Alternate conformations observed in catalytic serine of Bacillus subtilis lipase determined at 1. 3 angstrom resolution. Acta Crystallogr Sect D Biol Crystallogr. 2002; 58:1168–1174. [PubMed: 12077437] 60. Zhu X, Larsen NA, Basran A, Bruce NC, Wilson IA. Observation of an arsenic adduct in an acetyl esterase crystal structure. J Biol Chem. 2003; 278:2008–2014. [PubMed: 12421810] 61. McAuley KE, Svendsen A, Patkar SA, Wilson KS. Structure of a feruloyl esterase from Aspergillus niger. Acta Crystallogr Sect D Biol Crystallogr. 2004; 60:878–887. [PubMed: 15103133] 62. Hedstrom L. Serine protease mechanism and specificity. Chem Rev. 2002; 102:4501–4523. [PubMed: 12475199] 63. Ash EL, Sudmeier JL, Day RM, Vincent M, Torchilin EV, Haddad KC, Bradshaw EM, Sanford DG, Bachovchin WW. Unusual 1H NMR chemical shifts support (His) Cε1-H···O=C H-bond: Proposal for reaction-driven ring flip mechanism in serine protease catalysis. Proc Natl Acad Sci USA. 2000; 97:10371–10376. [PubMed: 10984533] 64. Bizzozero SA, Dutler H. Stereochemical aspects of peptide hydrolysis catalyzed by serine proteases of the chymotrypsin type. Bioorg Chem. 1981; 10:46–62. 65. Radisky ES, Lee JM, Lu CJK, Koshland DE Jr. Insights into the serine protease mechanism from atomic resolution structures of trypsin reaction intermediates. Proc Natl Acad Sci USA. 2006; 103:6835–6840. [PubMed: 16636277] 66. Cruickshank DW. Remarks about protein structure precision. Acta Crystallogr Sect D Biol Crystallogr. 1999; 55:583–601. [PubMed: 10089455] 67. Diederichs K, Karplus PA. Improved R-factors for diffraction data analysis in macromolecular crystallography. Nat Struct Biol. 1997; 4:269–275. [PubMed: 9095194] 68. Weiss MS, Hilgenfeld R. On the use of the merging R factor as a quality indicator for X-ray data. J Appl Crystallogr. 1997; 30:203–205. 69. Weiss MS. Global indicators of X-ray data quality. J Appl Crystallogr. 2001; 34:130–135. Levisson et al. Page 16 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Substrates and inhibitors of the CE7 family of enzymes. Structures of (A) acetylated xylooligosaccharide, (B) cephalosporin C, (C) p-nitrophenyl-acetate, (D) phenylmethylsulfonyl fluoride (PMSF), and (E) paraoxon. Levisson et al. Page 17 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 2. Overall fold and topology of TM0077. (A) Stereo view of a TM0077 protomer. The β- strands are labeled numerically (-1 to 8) with the core strands in red, α-helices are labeled alphabetically (A-2 to F) and 310-helices are labeled with an Eta (η1 and η2) with the core helices in cyan. The three-helix insertion after β6 is colored green and the N-terminal extension is colored sky blue. The figure was generated using Pymol 31. (B) Topology diagram of TM0077, with the helices displayed as cylinders and the strands displayed as arrows following the color and label scheme of (A). The location of residues forming the catalytic triad is also indicated. Levisson et al. Page 18 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 3. Structural superposition of TM0077 with structurally related esterases. Superposition of TM0077 (yellow) with (A) the cephalosporin C deacetylase (CAH) from B. subtilis (PDB: 1ods; blue) 40 and (B) the α/β-hydrolase domain of the acylpeptide hydrolase/esterase apAPH from A. pernix K1 (PDB: 1ve6; grey) 43. Levisson et al. Page 19 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 4. TM0077 oligomeric assembly. (A) Surface representation of the biological unit of the TM0077-Native hexamer with each monomer in a different color (left). The “cross section” shows the entrances on either side of the assembly and the internal cavity (center), and a 90° rotated view of the TM0077-Native hexamer, with a close-up view of the open central hole (right). (B) Surface representation of the biological hexamer unit of CAH from B. subtilis 40 (left) and the TM0077-SeMet hexamer with a close-up view of the blocked central hole (right). Levisson et al. Page 20 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 5. Effect of temperature and pH on esterase activity. (A) The esterase activity was studied using pNP-C2 as a substrate at temperatures ranging from 40–100°C. The inset shows the temperature dependence as an Arrhenius plot. (B) Thermal stability of TM0077 at 90°C. (C) The effect of pH on esterase activity studied using pNP-C2 as a substrate at pH values in the range of 4.8–9.2. Levisson et al. Page 21 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 6. TM0077 catalytic site. (A) Surface representation of the TM0077 catalytic site, with His303, Asp274 and the intermediate DEP-modified Ser188 shown as sticks. The two binding pockets are indicated with S1 and S2. (B) Apo TM0077 with a bound chloride ion (green sphere), (C) TM0077 with PMS-modified Ser188 and (D) TM0077 with DEP-modified Ser188. The catalytic residues are shown as sticks, with the hydrogen bonds shown as dashed lines. Carbon atoms are in green (apo), cyan (PMS) or blue (DEP), oxygen atoms in red, sulfur atoms in yellow and phosphate in orange. Electron density omit maps shown for inhibitor modified Ser188 contoured at 1σ show that the PMS and DEP are covalently bonded to Ser188 in (C) and (D), respectively. Distances are shown in Ångströms. Levisson et al. Page 22 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 7. Conformational change of Ser188 Oγ. The Oγ atom of the Ser188 is rotated ~110° between the native apo structure (cyan) and (A) the complexed PMS-modified Ser188 structure (pink), (B) the DEP-modified Ser188 structure (light blue) and (C) the SeMet structure (purple). The different hydrogen bonds made for the Ser Oγ in the native versus complexed structures are shown as dashed black lines with distances in Ångströms. Levisson et al. Page 23 Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Levisson et al. Page 24 Table I Summary of crystal parameters, data collection, and refinement statistics TM0077-SeMet TM0077-Native TM0077-PMS TM0077-DEP Space group P 21 P 21 21 21 P 21 21 21 P 21 21 21 Unit cell parameters a=152.64Å b=130.95Å c=157.82Å β=118.90° a=103.46Å b=103.79Å c=221.02Å a=103.57Å b=104.50Å c=221.61Å a=103.80Å b=104.43Å c=221.64Å Data collection λ1 MAD-Se λ2 MAD-Se Wavelength (Å) 0.9791 0.9183 0.9765 0.9765 0.9765 Resolution range (Å) 29.6 – 2.10 29.6 – 2.10 48.8 – 2.50 49.0 – 2.40 49.0 – 2.12 No. observations 1,119,236 1,100,249 1,222,016 765,546 989,949 No. unique reflections 293,140 291,757 83,045 94,681 123,070 Completeness (%) 93.0 (61.8)a 92.6 (60.8) 100 (100) 100 (100) 89.8 (53.5) Mean I/σ(I) 9.1 (2.4)a 9.6 (2.2) 14.4 (2.9) 11.5 (3.4) 15.3 (2.2) Rmerge on I (%) 12.3 (52.5) a 11.9 (57.9) 20.7 (109.7) c 18.0 (67.4) 9.5 (51.9) Rmeas on I (%) 14.3 (62.2) a 13.9 (68.7) 21.4 (113.6) 19.2 (71.9) 10.2 (60.2) Rpim on I (%) 7.2 (32.7) a 7.1 (36.2) 5.5 (29.2) 6.7 (24.9) 3.5 (29.2) Highest resolution shell (Å) 2.15 – 2.10 2.15 – 2.10 2.56 – 2.50 2.46 – 2.40 2.18 – 2.12 Model and refinement statistics Resolution range (Å) 29.6 – 2.10 48.8 – 2.50 49.0–2.40 49.0 – 2.12 No. reflections (total) 293,097 b 83,045 94,680 122,994 No. reflections (test) 14,726 4,200 4,742 6,188 Completeness (% total) 92.8 100.0 100.0 89.8 Data set used in refinement λ1 MAD-Se Cutoff criteria |F| > 0 |F| > 0 |F| > 0 |F| > 0 Rcryst 0.186 0.167 0.160 0.167 Rfree 0.223 0.212 0.208 0.205 Stereochemical parameters Restraints (RMSD observed) Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Levisson et al. Page 25 TM0077-SeMet TM0077-Native TM0077-PMS TM0077-DEP Bond angle (°) 1.48 1.47 1.53 1.44 Bond length (Å) 0.018 0.017 0.017 0.015 Av. isotropic B-value (Å2) 27.9 24.7 19.4 19.6 ESU based on Rfree 0.17 0.25 0.22 0.18 Water molecules/other solvent molecules 2,464/1 507/24 946/17 987/23 PDB ID 1vlq 3m81 3m82 3m83 aHighest resolution shell ESU = Estimated overall coordinate error 16,66. Rmerge=ΣhklΣi|Ii(hkl)-<I(hkl)>|/ΣhklΣiIi(hkl), Rmeas(redundancy-independent Rmerge)=Σhkl[Nhkl/(Nhkl-1)]1/2Σi|Ii(hkl)-<I(hkl)>|/ΣhklΣiIi(hkl), and Rpim(precision-indicating Rmerge)=Σhkl[1/ (Nhkl-1)]1/2Σi|Ii(hkl)-<I(hkl)>|/ΣhklΣiIi(hkl) 67–69. Rcryst = Σ| |Fobs|-|Fcalc| |/Σ|Fobs| where Fcalc and Fobs are the calculated and observed structure factor amplitudes, respectively. Rfree = as for Rcryst, but for 5.0 % of the total reflections chosen at random and omitted from refinement. bTypically, the number of unique reflections used in refinement is slightly less than the total number that were integrated and scaled. Reflections are excluded due to systematic absences, negative intensities, and rounding errors in the resolution limits and cell parameters. cRmerge of the highest resolution shell is high due to high redundancy (14.7). However, the completeness and mean I/σ of the highest resolution shell are reasonable, and these data were included in the refinement. Proteins. Author manuscript; available in PMC 2013 June 01. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript Levisson et al. Page 26 Table II Kinetic parameters for hydrolysis of various esters Ester Km (mM) kcat (s−1) kcat/Km (s−1 mM−1) pNP-Acetate 0.185 ± 0.026 57.5 ± 2.2 310.8 ± 45.3 pNP-Propionate 0.137 ± 0.013 41.3 ± 1.1 301.5 ± 29.7 2-O-acetyl pNP-Xyl 3.6 ± 0.5 76.1 ± 19.2 21.1 ± 6.1 3-O-acetyl pNP-Xyl 4.2 ± 0.4 70.1 ± 7.7 16.7 ± 2.4 4-O-acetyl pNP-Xyl 4.0 ± 0.1 78.6 ± 12.9 19.7 ± 3.3 Proteins. Author manuscript; available in PMC 2013 June 01.
3M85
Archaeoglobus fulgidus exosome y70a with RNA bound to the active site
Quantitative analysis of processive RNA degradation by the archaeal RNA exosome Sophia Hartung1, Theresa Niederberger1, Marianne Hartung2, Achim Tresch1 and Karl-Peter Hopfner1,* 1Center for Integrated Protein Sciences and Munich Center for Advanced Photonics at the Gene Center, Department of Biochemistry, Ludwig-Maximilians-University Munich, Feodor-Lynen-Strasse 25, 81377 Munich and 2General Electric - Global Research, Freisinger Landstrasse 50, 85748 Munich, Germany Received February 9, 2010; Revised March 18, 2010; Accepted March 21, 2010 ABSTRACT RNA exosomes are large multisubunit assemblies involved in controlled RNA processing. The archaeal exosome possesses a heterohexameric processing chamber with three RNase-PH-like active sites, capped by Rrp4- or Csl4-type subunits containing RNA-binding domains. RNA degradation by RNA exosomes has not been studied in a quan- titative manner because of the complex kinetics involved, and exosome features contributing to effi- cient RNA degradation remain unclear. Here we derive a quantitative kinetic model for degradation of a model substrate by the archaeal exosome. Markov Chain Monte Carlo methods for parameter estimation allow for the comparison of reaction kinetics between different exosome variants and substrates. We show that long substrates are degraded in a processive and short RNA in a more distributive manner and that the cap proteins influ- ence degradation speed. Our results, supported by small angle X-ray scattering, suggest that the Rrp4-type cap efficiently recruits RNA but prevents fast RNA degradation of longer RNAs by molecular friction, likely by RNA contacts to its unique KH-domain. We also show that formation of the RNase-PH like ring with entrapped RNA is not required for high catalytic efficiency, suggesting that the exosome chamber evolved for controlled processivity, rather than for catalytic chemistry in RNA decay. INTRODUCTION The eukaryotic and archaeal RNA exosomes and their distant relative, the bacterial degradosome, are large multiprotein assemblies that function as central cellular RNA processing and degradation machineries. The RNA exosome was originally found in yeast as an essen- tial protein complex with 30 ! 50 exonuclease activity. First, identified for the 30-processing of the yeast 5.8S ribo- somal RNA (1), the yeast RNA exosome subsequently turned out to be important for the trimming and degrad- ation of the 30-end of several nuclear RNA precursors (2). In addition, the exosome was shown to be also active in the cytoplasm by controlling mRNA turnover (3), and by its implication in various mRNA surveillance pathways like the non-sense-mediated and the non-stop decay pathways (4–7). Due to its involvement in all the different RNA processing and surveillance pathways the exosome is apparently one of the central exonucleases of a yeast cell [for reviews see for instance (8,9)]. Structural homologues of the yeast exosome were sub- sequently identified in humans, previously known as the PM-Scl (polymyositis–scleroderma overlap syndrome) complex, and in archaea (10–12). A variety of structural studies revealed a conserved architecture of exosome like complexes (13–18): exosomes consist of nine conserved core subunits, six RNase PH type subunits and three subunits with S1 and KH or zinc-ribbon domains. The six RNase-PH like domains form a ring, arranged as trimers of pseudo-dimers. In archaea, the ring is formed by three (archaeal)aRrp41:aRrp42 dimers, while human and yeast exosomes contain six different RNase PH type subunits. The archaeal exosome possesses a central chamber within the RNase PH ring which contains three phos- phorolytic active sites. The actual active site is located in the aRrp41 subunits, but the whole aRrp41:aRrp42 dimer is involved in positioning the RNA. These sites degrade single-stranded RNA (ssRNA) in a phosphate dependent manner in 30 ! 50 direction. They also catalyse the reverse reaction of adding nucleoside diphosphates to the 30-end of RNA (13), liberating inorganic phosphate. In archaea, this *To whom correspondence should be addressed. Tel: +49 89 2180 76953; Fax: +49 89 2180 76999; Email: hopfner@lmb.uni-muenchen.de 5166–5176 Nucleic Acids Research, 2010, Vol. 38, No. 15 Published online 14 April 2010 doi:10.1093/nar/gkq238  The Author(s) 2010. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/ by-nc/2.5), which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. activity has been attributed to formation of poly-A-rich tails on RNA (19). A proposed RNA entry pore at one side of the chamber restricts entry to mostly unstructured ssRNA, providing an explanation for controlled RNA deg- radation. Furthermore, three Csl4 or Rrp4 type putative RNA recognition subunits are located on top of the (Rrp41:Rrp42)3 ring and frame the proposed RNA entry pore. Current models suggest that these domains recognize RNA substrates and help to funnel them into the process- ing chamber. Although the human exosome is structurally related to the archaeal complex, including S1 and KH domain con- taining subunits (Csl4, Rrp4 and Rrp40), it has lost phosphorolytic activity (14). Instead, it gained additional ectopic subunits: the hydrolytic RNase Rrp44 (20,21) has both exonucleolytic and endonucleolytic activities, located in RNB and PIN domain subunits, respectively (14,20,22–24). A second hydrolytic RNase (Rrp6) was identified as transient part of the nuclear complex (10). Recent results indicate that despite the ectoptic placement of the nuclease active sites, RNA is still threaded through the nuclease deficient RNase PH type ring (25). A variety of groups have biochemically observed processive RNA degradation, in particular for the archaeal exosome. From structural studies, it was proposed that RNA is channelled through an entry pore between the S1 domains of aCsl4 or aRrp4 trimers into the processing chamber, where the 30-end of the RNA is pos- itioned in one of the three phosphorolytic active sites, subsequently degrading RNA base-per-base (16,26,27). The presence of the cap proteins Csl4 and Rrp4 in general increases the degradation efficiency of the exosome, but it is unclear how they do so. For instance, if the cap proteins recruit RNA, one would expect an increase in the general binding affinity. However, once RNA has entered the processing chamber, high affinity binding to the ectopic domains should slow down processive degradation. Another mechanism that is not yet understood is why processivity depends on the length of the RNA molecules (28). To address these questions and to develop means to quantitatively analyse processive degradation, we performed quantitative high-resolution RNase degradation activity assays with different variants of the Archaeoglobus fulgidus exosome. We evaluated dif- ferent kinetic models and developed a Markov Chain Monte Carlo (MCMC) analysis to fit the model to the data and derive appropriate rate constants of individual RNA degradation steps. Our data identify different struc- tural contributors to processivity, suggesting that ectopic RNA-binding domains, the entry pore and the active site are different contributors to processive degradation. The methods should be easily applicable also to other processive enzymes, including the hydrolytic nucleases of the eukaryotic exosomes. MATERIALS AND METHODS Protein expression and purification The Archaeaoglobus fulgidus RNA exosome with different Rrp4 and Csl4 caps, derivatives and mutants were expressed and purified as described (29). Site directed mu- tations were introduced using the QuickChange Site-directed Mutagenesis Kit (Stratagene) and verified by sequencing. Oligonucleotide sequences are provided in the Supplementary Table S2. Crystallization and structure determination An amount of 120 mM Csl4-exosome (Csl4:Rrp41:Rrp42)3 or its Y70ARrp42 mutant (= 27 g/l) were incubated with 400 mM RNA (3.3-fold excess, 6-mer CCCCUC) for 10 min on ice. Protein:RNA complexes were crystallized by sitting drop vapour diffusion technique by mixing 1 ml protein and 1 ml of reservoir solution (0.1 M NaAcetate, pH 4.6, 30% 3-Methyl-1,5-pentadiol (MPD), 100 mM NaCl) at 20C. Datasets were recorded at the ID-14-2 beamline (ESRF, Grenoble, France) to 2.4 A˚ (wild-type exosome) and at the PX I beamline (SLS, Villigen, Switzerland) to 3.0 A˚ (Y70ARrp42 mutant) and processed with X-ray Detector Software (XDS) (30). A model of the apo-Csl4-exosome complex (29) was used as a search model for molecular replacement using PHASER (31). Refinement to 2.4 A˚ and 3.0 A˚ , respectively was performed with CNS (32) and PHENIX (33). In the additional electron density RNA nucleotides were positioned using COOT (34). Refinement of the complete complexes was followed by iterative cycles of manual model completion with COOT and positional and B-factor refinement with CNS (Supplementary Table S1). Small angle X-ray scattering For small angle X-ray scattering (SAXS) studies, the (Rrp41:Rrp42:Rrp4)3 complex was purified as described above. To purify the exosome with endogenously bound Escherichia coli RNA the protocol was modified as follows: RNA was not washed offwith high salt, and in all buffers the salt concentration was 250 mM or lower. After the Ni-NTA affinity chromatography, the complex was loaded on an anion exchange column to remove unbound nucleic acids and the procedure was repeated to assure the total removal of free RNA. Not until only one distinct peak was eluted, the fractions were pooled, concentrated and flash frozen. The apo-complex was measured at 5, 10 and 15 mg/ml and the RNA complex was concentrated to an absorption at 280 nm of A280 = 55 and measured in a 1: 0, 1: 1 and 1: 2 dilution to evaluate the concentration dependency of scattering. Both complexes did not show concentration dependent aggregation and were not affected by long exposure to high-energy X-rays. SAXS data collection was performed in 20 mM Tris pH 7.4 and 200 mM NaCl buffer at the SIBYLS beamline (Advanced Light Source, Berkeley, CA, USA) (35). The radius of gyration was calculated using the Guinier plot in the linear region (constraint: s  Rg <1.3) and the calculation of the pair distribution function was done with GNOM within PRIMUS (36). Ab initio modelling of the solution structures was done with GASBORp (37) and more than 10 identically calculated models were aligned and averaged using DAMAVER and SUPCOMB (38). For analysis of the Nucleic Acids Research, 2010, Vol. 38, No. 15 5167 bound RNA, the protein was separated from the RNA by running the complex on a denaturing 6 M urea and 20% polyacrylamide gel and elution of the RNA from the gel. The pelleted RNA was sent to Vertis Biotechnologie AG, where the sample was poly(A)-tailed using poly(A) poly- merase followed by ligation of an RNA adapter to the 50-phosphate of the small RNAs. First-strand cDNA syn- thesis was then performed using an oligo(dT)-adapter primer and M-MLV H- reverse transcriptase. The result- ing cDNAs were PCR-amplified to 20–30 ng/ml in 19 cycles using standard Taq DNA polymerase. We cloned the cDNA products with EcoRV into pET21 vectors, transformed and amplified the plasmids and isolated and sequenced single clones. Cross-linking Site-specific crosslinking of the K37CRrp41:D143CRrp42 mutant was performed with a HBVS (1,6-Hexane-bis- vinylsulfon) crosslinker. The crosslinking reaction was performed with a 100-fold excess of crosslinker under oxygen-free conditions in a glove-box. We removed crosslinked protein from non-crosslinked protein complexes using a Superose 6 size-exclusion column, equilibrated with a running buffer containing 4 M guanidinium chloride. Protein from the peak correspond- ing to a crosslinked Rrp41/Rrp42 dimer was refolded in 50 mM Tris (pH 7.4), 200 mM NaCl, 500 mM arginine and 5% glycerol by dilution. The refolded protein was again applied onto a Superose 6 column in 50 mM Tris (pH 7.4) and 200 mM NaCl. Only the correctly refolded protein, verified by the formation of a hexamer in size exclusion chromatography, was used for further experiments. As a control sample, the same complex without crosslinker was partly denatured, purified and refolded in the same way as the crosslinked protein. RNase activity assays We carried out RNase activity assays using 32P-labelled poly(rA)-oligoribonucleotides with different lengths as substrate (26). RNA was incubated with [g-32P] ATP (Hartmann Analytics) and T4 polynucleotide kinase (NEB) for 45 min at 37C and purified by using MicroSpin G-25 columns (GE Healthcare). For each reaction, the protein (30 nM for the Csl4- and the Rrp4-capped wild-type exosome and the interface mutant; 60 nM for the cap-less exosome and the single site mutants R65ERrp41 and Y70ARrp42; 120 nM for the crosslinked cap-less exosome) was incubated with RNA in buffer containing 20 mM Tris (pH 7.8), 60 mM KCl, 10 mM MgCl2, 10% glycerol, 2 mM DTT, 0.1% PEG 8000, 10 mM NaH2PO4 (pH 7.8) and 0.8 U/ml RNasin (Promega) at 50C. Different time points were taken and the reaction was stopped by adding one volume of loading dye [0.75 g/l bromphenol blue, 0.75 g/l xylene cyanol, 25% (v/v) glycerol, 50% formamide]. The reaction products were resolved on a 20% polyacrylamide/6 M urea sequencing gel running at 50C and were analysed by phosphorimaging (GE Healthcare). The gel bands were quantified using the ImageQuant Software (GE Healthcare) and data analysis, simulation and fitting was done with MatLab (Mathworks). Models and kinetic data analysis Kinetic models are shown in Figure 3A. They are described by four parameters: association rate ka,i, dissoci- ation rate kd,i, cleavage rate kc,i and polymerization rate kp,i, one for each RNA of length i = 4, 5, . . ., 30. The cor- responding set of differential equations that quantitatively describe RNA degradation is shown in the supplement Data (Chapter 1). Since the reaction takes place in an excess of inorganic phosphate (10 mM phosphate compared to only 3.6 mM ADP at the time all RNA mol- ecules are totally degraded), we may assume no polymer- ization takes place, i.e. kp,i = 0 for all i. Consistently, we saw no synthesis of longer RNAs in our reactions. To obtain empirical estimates of the posterior parameter dis- tribution, we implemented a MCMC approach based on the Metropolis–Hastings algorithm. The key ingredients are the likelihood function, the prior, and the proposal distribution. The likelihood function penalizes the estima- tion error produced by a given model. More precisely, it penalizes the residuals, i.e. the deviation of the measured RNA amounts at each time point from the amounts that have been predicted from the current parameter set. We assume that the residuals are independent realizations of Gaussian distributions with zero mean. Since the vari- ances of these Gaussians are not known a priori, we assume a two-parameter error model with an additive and a multiplicative error component which has been proposed (39) in the context of spot quantification on arrays. We initialize the error model very conservatively (presuming large measurement errors). During the MCMC run, the error model is updated continuously by replacing it with an empirical estimate derived from the residuals that occurred in the Markov chain so far. The prior encodes prior knowledge/assumptions on the distribution of the parameters. It is sensible to require the kinetic parameters to vary smoothly with the RNA length i. This is made explicit by penalizing the difference of two successive kinetic parameters kx,i+1 and kx,i using a Gaussian prior on these differences. We em- phasize that this does not impose any restrictions on the absolute level of the parameter values. The comparison of the parameter levels obtained by different experiments is virtually unaffected by our prior choice and therefore practically unbiased. The proposal distribution generates a new parameter set as a candidate for the next MCMC step that is based on the current parameter set. We simply use a multivariate log-normal distribution with fixed diagonal covariance matrix, which is centred at the current parameter set. It turns out that the parameters of the model as stated above are not identifiable. We therefore fixed kd,i to one global constant kd, whereas the association param- eters ka,i are sampled individually. The parameters kc,i are set equal to one length-independent parameter kc. The details of this approach and its justification through extensive simulations are given in the supplementary Data (Chapters 2–4). 5168 Nucleic Acids Research, 2010, Vol. 38, No. 15 RESULTS RNA is not degraded with constant velocity Despite intense structural and biochemical research on RNA exosomes, a kinetic model, quantitative analysis of processive RNA degradation and a biochemical identifi- cation of elements that contribute to processive degrad- ation have not been studied, due to the complex kinetics involved. To address these issues, we performed RNase assays with 50-radioactively labelled 30-mer oligo(A) RNAs and the A. fulgidus Csl4- (Csl4:Rr41:Rrp42)3 and Rrp4-exosomes (Rrp4:Rr41:Rrp42)3. The reaction products and their time evolution were resolved on a denaturing sequencing gel and quantified by phosphorimaging (Figure 1), controls are shown in the supplemental material (Supplementary Figures S4 and S5). Several characteristic features of substrate degrad- ation by exosomes are revealed: First, RNA is not degraded at a constant speed, but the degradation of substrate has several phases and is distinct in different isoforms. In the Csl4 exosome (Figure 1A), after a slower first processing step, longer RNAs (>12–13 nt) are degraded very fast, seen by the low amount of intermediates in this range; shorter RNAs (<12–13 nt) are degraded slower and accumulate first before they are further degraded. On the contrary, the first processing step is faster in the Rrp4- than in the Csl4-exosome (Figure 1B). However, oligo-rA substrates >24 nt are degraded slower, intermediate substrates (24–13 nts) faster, and RNAs <13 nt slower again. This result is astonishing, considering homooligomeric se- quences are used and the effect is consequently not sequence dependent. In addition, the unexpected slow-fast-slow kinetics of the Rrp4 isoform reveals a quite complex length dependency of RNA degradation speeds. Second, the final degradation product is a 3-mer. Further degradation is extremely slow, comparable to spontaneous background hydrolytic cleavage under the present conditions. We hypothesized that features of the active site might interact specifically with the fourth base at the 30-end. Previous structural analysis with the Sulfolobus solfataricus exosome has shown that at least 4 nt are stably bound in the phosphoropytic active sites (26), but in the case of the Pyrococcus furiosus exosome some nucleotides were recognized (16). To get direct structural information for the A. fulgidus exosome:RNA interaction, used in this study, we crystallized our Csl4-exosome with a 6-mer RNA molecule (Figure 2; Supplementary Table S1). Four nu- cleotides from the 30-end are clearly visible in the unbiased Fo–Fc electron density, with weaker density for the two additional nucleotides. Interestingly, the side chain of Y70Rrp42 shows p-stacking with the fourth base (counting from the active site) and this seems to be a conserved feature among archaeal exosomes (16,26). This interaction specifically stabilizes the first 4 nt, while RNA positions+5 and+6 behind Y70Rrp42 appear not to be specifically recognized. To test the role of Y70Rrp42, we determined the co-crystal structure of the Csl4- exosome-Y70A mutant with a CCCCUC oligonucleotide. In fact, we only see clear electron density for 4 nt in the active site and the electron density at position+4 is weaker and less defined compared to the wild-type. Thus, the 3-mer as degradation end-product is likely the cause of inefficient recognition of RNA’s with <4 nt at the active site. Figure 1. Visualization of RNase activity of the archaeal exosome on denaturing polyacrylamide gels: the input (I) is a 30-mer polyA RNA radioactively labelled at the 50-end that is degraded from the 30-end to a final product (FP) of a 3-mer. Time points were taken in increasing intervals [in minutes: 0:10; 0:20; 0:30; 0:40; 0:50; 1:00; 1:10; 1:20; 1:40; 2:00; 2:20; 2:40; 3:00; 3:30; 4:00; 4:30; 5:00; 5:50; 6:00; 6:30; 7:00; 7:30; 8:00; 9:00; 10:00; 12:00; 14:00; 16:00; 18:00; 20:00; 25:00; 30:00; 35:00; 40:00; (B) ends at 8:00 min]. RNA degradation does clearly not occur with constant speed and the (Csl4:Rrp41:Rrp42)3 exosome (A) degrades RNA with a different time dependency than the (Rrp4:Rrp41:Rrp42)3 exosome (B). Nucleic Acids Research, 2010, Vol. 38, No. 15 5169 A kinetic model for processive RNA degradation by exosomes To obtain a comprehensive picture of the exosomal RNA decay, we need to analyse the reaction speeds in a quan- titative manner. The amount of RNA as function of time of intermediate i of an rA n-mer may be described by several rate constants (Figure 3A): an association rate constant ka,i of the 30-end of RNA to the active site; a corresponding dissociation rate constant kd,i; a rate of for- mation of intermediate i by cleavage of intermediate i+1, Figure 3. Three different models to describe the kinetics of RNA degradation by the exosome were tested: (A) scheme for the general kinetic model, which includes cleavage and polymerization rates kc and kp as well as association and dissociation rates ka and kd for all RNAs from 30–4 nt. (B–D) Quantified concentrations of RNA intermediates from Figure 1A, along with least square fits to different kinetic models. (B) Strict processivity considers only 27 different cleavage rates kc,30 –kc,4. (C) cleavage-and-polymerization considers 27 different cleavage rates kc,30 –kc,4, 27 different polymerization rates kp,30–kp,4 and one initial association rate ka,30 (=55 rates). With models (C) and (B), no reasonable fit could be obtained. (D) By including association, dissociation and cleavage and making rational simplifications (see text) we can convincingly fit the data with a model con- taining 28 different rate constants. Figure 2. Crystal structure of 6-mer RNA bound to the active site of the archaeal exosome. Rrp41 is shown in light and Rrp42 in dark green. The 2Fo–Fc electron density is contoured at 1.0s and only shown for the RNA and the side chain of Y70Rrp41. (A) In the wild-type exosome Y70 is stacking with the fourth base of the bound RNA, and only weak density can be seen for the fifth and sixth base. (B) Electron density for the fourth base of the RNA is much weaker in the Y70ARrp41 mutant compared to the wild-type and no density can be detected at this contour level for additional nucleotides. 5170 Nucleic Acids Research, 2010, Vol. 38, No. 15 kc,i+1; and by adenylation (polymerization) of intermedi- ate i1, kp,i-1; a rate of disappearance of intermediate i by cleavage of i, kc,i and by adenylation of i, kp,i. The system kinetics is then given by a set of differential equations (Supplementary Data). However, it is possible that this more general model can be further simplified. For instance, we likely can neglect adenylation (kp,i = 0) because our reaction conditions contain 10 mM phosphate compared to only 3.6 mM ADP at the time all RNA molecules are totally degraded, strongly shifting the reversible reaction towards degradation. In addition, the exosome might be strictly processive, i.e. association and dissociation rates of RNA intermediates are negligible compared to the cleavage rates (ka,i = kd,i = 0). Furthermore, all cleavage rates may be independent of the length of RNA, because they could be a local active site property (kc,i = kc,j). Hence we analysed three simplified models (Figure 3B–D). Once initial values for the rate constants, enzyme concentration and RNA substrate concentration (rA 30-mer) are provided, this corresponding set of differ- ential equations can be used to calculate the concentra- tions of all RNA intermediates over time. We then minimized the resulting least square deviations between the calculated and experimental concentrations of reaction intermediates by optimizing the rate constants using the ‘fminsearch’ parameter optimization procedure as implemented in Matlab. Using the ‘strict processivity’ model with 27 independ- ent variables (ka,i = kd,i = kp,i = 0) (Figure 3B), we obtained no reasonable fit of the experimental data. A second model including the adenylation reaction (55 inde- pendent variables, ka,i = kd,i = 0) could also not properly interpret the data (Figure 3C). Thus, simply adding more parameter does not automatically lead to reasonable fits and the RNA degradation activity cannot be convincingly explained by strict processivity. Consequently, we added association and dissociation of RNA intermediates to the equations and used the following alternative simplifica- tions: (i) adenylation is omitted (kp,i = 0; see above); (ii) the same cleavage rate is used for all RNA molecules (kc,i = kc,j), i.e. cleavage rate is a local active site property and not dependent on RNA length. We estimated starting values for kc and validated this simpli- fication from analysis of the initial exponential decay of RNAs substrates with different initial lengths (data not shown). (iii) Due to our experimental approach, we cannot experimentally distinguish between bound and free RNA since the gel bands represent the sum of free and exosome-bound RNA intermediates of length i. For that reason, we cannot reconstruct dissociation-, associ- ation- and cleavage rate constants independently of each other. Consequently, we do not treat the association and the dissociation rate constants independently, but analyse the ratio of ka,i/kd,i by setting kd,i’s to a constant low value, leaving ka,i free to vary. Variation of the value for kd did not result in significant changes in the analysis (Supplementary Data). These three reasonable simplifica- tions leave essentially one free parameter per intermediate plus one overall cleavage rate constant. Although this model has less degrees of freedom than the second model (55 versus 28), it can convincingly interpret the ex- perimental data for the both Csl4 and Rrp4 exosome variants and most mutants (Figure 3D). MCMC analysis of degradation To address the problem of multidimensional parameter fitting and to assess the variance in parameter esti- mation, we established MCMC simulations. Because of the difficulty in determination of separate values for the single rate constants, we defined an RNA length- dependent quantity vi vi ¼ kc,i Km,i ð1Þ with Km,i the Michaelis–Menten constant Km,i ¼ kc,i+kd,i ka,i ð2Þ vi is called ‘catalytic efficiency’ or ‘specificity constant’, as it is a measure of the velocity of RNA intermediate i deg- radation by the exosome. We are now in a position to test exosome features important for vi. We observe that for the Csl4 exosome, vi is highly dependent on the RNA length: the initial RNA processing step, likely determined by the initial association of RNA with the exosome, is generally slow. Once RNA is bound, vi is large and relatively constant for RNA lengths >13 nt. vi then progressively decreases for RNA lengths <13 nt until the final 3-mer appears (Figures 4B and 5A). This length dependency may be explained by the exosome structure: RNA molecules longer than 13 nt might still reach through the ‘neck’, and this topological interaction will induce a higher ‘local concentration’ of RNA at the active site with increased vi. Short RNAs, on the contrary, will lose this contact and due to their smaller size more easily diffuse out of the processing chamber, therefore decreasing vi. To test this idea, we analysed the Y70A mutant of the Csl4-exosome. The length profile of the catalytic efficiency has a similar shape than for the wild-type, although the catalytic efficiency is lower for all RNA intermediates (Figure 5B). For RNAs >13 nt, the difference in vi is 2- to 3-fold (about one log unit). However, the drop in vi for RNAs <13 nt is progressively more pronounced compared to the wild-type and towards short RNAs (<8 nt), the mutant is 20- to 150-fold (three to five log units) slower than the wild-type. This is consistent with the idea that for long RNAs the neck provides additional interaction and thus overcomes in part the destabilizing effect of Y70A. For shorter RNAs, the active site becomes the sole attachment, leading to a rapid drop of catalytic efficiency in the Y70A mutant. We also analysed the ‘neck’ mutant R65E, which has been shown to severely reduce exosome activity (16,26,27). This mutant exhibited a substantially delayed onset of degradation, presumably because RNA is unable to effi- ciently enter the active site (data not shown). A likely reason is the formation of non-productive RNA:protein complexes with RNA trapped on the outside of the exosome (29). At present, our model cannot deal with Nucleic Acids Research, 2010, Vol. 38, No. 15 5171 this scenario and we could not convincingly include—as only variant—the R65E mutant in the analysis. However, the data of the analysis of R65E are provided in the sup- plementary Figure S15 and following. Role of exosome ring formation and ring dynamics for RNA binding Although the initial binding of RNA appears slow, it seems unlikely that the 30-end directly finds its way through the small hole in the neck. It is perhaps more likely that the ring structure ‘breathes’—as observed e.g. in hexameric helicases—and allows some lateral entry at the neck. To explore this idea we analysed a crosslinked exosome, where the ring is rigidified by three site specific crosslinks, and a mutant that disrupts the ring structure into Rrp41:Rrp42 pairs. We compared these isoforms with the corresponding wild-type, the cap-less hexameric (Rrp41:Rrp42)3 ring (Figure 5C). From the structural analyses, it was observed that the Rrp41 and Rrp42 subunits possess two interfaces. One interface is larger, and characterized by contiguous b-sheets between Rrp41 and Rrp42 (40). The other interface is smaller, presumably more dynamic, and was chosen for the crosslinking and mutagenesis analysis (Supplementary Figure S2). (Rrp41:Rrp42)3 exhibit a biphasic length dependence of vi, similar to the Csl4-exosome. However, the catalytic efficiency of (Rrp41:Rrp42)3 is 5- to 10-fold higher Figure 5. Comparisons of the catalytic efficiency vi of different exosome variants versus RNA lengths: (A) differences in the cap proteins influence catalytic activity. This is shown by comparison of vi from the cap-less exosome (Rrp41:Rrp42)3 in magenta, the Csl4 capped exosome (Csl4:Rrp41:Rrp42)3 in red and the Rrp4 capped exosome (Rrp4:Rrp41:Rrp42)3 in blue. (B) Tyr70Rrp42 close to the active site is especially important to efficiently degrade small RNAs. The wild-typ Csl4 exosome is shown in red and the Y70ARrp42 mutant in green. (C) The role of the ring architecture and dynamics for cata- lytic activity is shown by comparing wild-type cap-less exosome (Rrp41:Rrp42)3 in magenta with the dimeric and open interface mutant (Rrp41:Rrp42)1 and a rigidified crosslinked variant that is less dynamic in yellow. A total of 1000 parameter sets have been randomly drawn from the stationary phase of the Markov chain. Thus for each RNA length and each timepoint, we obtained 1000 estimates whose distribution is displayed by boxplots. Figure 4. Catalytic efficiency vi for all RNA intermediates present during the degradation of a 30-mer RNA by the Csl4-Rrp41-Rrp42 exosome was determined with MCMC simulations. (A) shows the traceplot and (B) the final parameter set (burnin = 150 000). It can be seen that the MCMC chains vary in convergence speed as well as in variability. The boxplots in (B) illustrate the main advantage of the MCMC approach: it not only offers a set of parameters that best describe the measured data, but it also yields a posterior distribution for each catalytic efficiency parameter and thus provides a more com- prehensive summary of the data. 5172 Nucleic Acids Research, 2010, Vol. 38, No. 15 across the RNA spectrum than that of the Csl4-exosome, indicating that the Csl4 cap subunits do not substantially promote degradation of this model substrate. In addition, vi drops for RNAs <13 nt even for the (Rrp41:Rrp42)3 particle, indicating that the neck not e.g. the S1 domains of caps are responsible for the higher catalytic efficiency on longer RNAs. To explore the effect of the hexameric ring formation, we mutated Lys51 to Glu, located in the ‘smaller’ interface between alternating Rrp41 and Rrp42 pairs. This resulted in stable Rrp41:Rrp42 dimers that do not assemble into hexamers anymore (Supplementary Figure S3). The Rrp41:Rrp42 dimers exhibit catalytic efficiencies that are only slightly lower than the (Rrp41:Rrp42)3 particle for RNAs >13 nt, and almost identical to the corresponding hexamers for RNAs <13 nt. As a result, the drop around 13 nt from a faster to a slower degradation is much less pronounced in the Rrp41:Rrp42 dimer, further supporting the idea that encapsulation in the neck is responsible for higher degradation speeds. The relatively high activity of the dimers is possibly also a result of the effective ‘tripli- cation’ of active sites, i.e. only one RNA molecule can be degraded by a (Rrp41:Rrp42)3, while three RNA mol- ecules can be degraded by three Rrp41:Rrp42 dimers. In addition, while RNA probably dissociates faster from the dimers, this effect could be compensated by a faster asso- ciation of RNA to the readily accessible active sites in the open dimers. The opposite is observed, when the ring structure is crosslinked. We introduced cysteines on the outside of the RNase-PH ring and crosslinked the three Rrp41:Rrp42 dimers via a thiol specific bifunctional crosslinker. This procedure resulted in a hexameric RNase PH ring with wild-type-like size and shape accord- ing to gel filtration (Supplementary Figure S3). Comparison of ni between the crosslinked isoform with the (Rrp41:Rrp42)3 hexamer, revealed a dramatically reduced ni (500- to 2000-fold) indicating that rigidifying the exosome by the crosslink severely affects catalytic ef- ficiency. We cannot formally rule out that the crosslinking affects activity by other means, but considering that the hexamer disrupting mutation at the same interface does not severely reduce activity, a plausible scenario is that the rigidified exosome does not allow efficient association with RNA anymore. Thus, taken together, the self-compartmentalization of exosomes is probably not an evolution for high activity, but rather for controlled RNA degradation. Effect of the cap structures To learn about the role of the cap proteins in exosome activity, we compared the rate constants of the cap-less exosome with the Csl4 and the Rrp4 exosome. The initial rates for degradation of the 30-mer rA are similar for the Csl4-exosome and the cap-less version, but consid- erably faster for the Rrp4-exosome (Figure 5A). This in- dicates that cap proteins can influence recognition and recruitment of RNA substrates and that this step is more efficient for the Rrp4 exosome. However, while RNA degradation for medium and short RNAs is quite comparable between the Csl4 and Rrp4 exosomes, there is an interesting difference for long RNA molecules (>24 nt). The Rrp4 exosome is quite slower for RNAs >24 nt, faster for RNA between 24 and 13 nt, and then progressively slower for RNAs <13 nt. This remarkable length depen- dency is clearly evident in degradation profiles (Figure 1). The most likely explanation is that long RNAs might still have contacts with Rrp4, where a more specific binding site holds them partially back from rapid degradation. In principle, this could be viewed as molecular friction. When RNAs are shorter, they loose contact to Rrp4 and degrad- ation speed is increased. The Csl4 protein and the Rrp4 protein differ in their domain structures. While Csl4 contains a Zn-ribbon domain, Rrp4 possesses a KH-domain, which is a typical RNA-binding domain and could recognize the oligo-rA. Such a binding could be responsible for the faster first degradation step, because it more efficiently sequesters RNAs on the exosome surface, but may subsequently slow down degradation until RNAs are too short to maintain simultaneous contacts at the KH domain and active site. However, the Rrp4 isoform is more efficient for smaller RNA species than the Csl4 and capless isoforms. Since these shorter RNAs cannot form dual contacts with the active site and outside the caps, the Rrp4 could also influ- ence the dynamics or other properties of the RNase-PH ring, for instance to help in loading of RNA into the ring structure. SAXS structure of the Rrp4 exosome with endogenously purified bacterial RNA To explore the role of the Rrp4 cap in efficiently recruiting RNAs further, we performed SAXS studies with a nuclease deficient nine-subunit Rrp4 exosome bound to RNA: we had noticed that this nuclease deficient Rrp4-exosome (D180A in Rrp41) very efficiently co-purifies with E. coli RNA. To determine the kind of RNA that binds to the exosome we run it on a denaturing gel together with RNAs with known sizes and could estimate the size of the RNA to be between 55 and 65 nt (Supplementary Figure S1). Cloning and sequencing of bound RNA molecules revealed a set of much shorter in- homogeneous mixed sequences (Supplementary Table S3). It is possible that the bound RNAs are a mixture of various mRNAs from E. coli, although the isolated RNA is larger than the identified sequences and it is possible that highly structured RNAs such as tRNAs are underrepresented due to inefficient amplification and cloning. Comparison of the SAXS structure of apo-Rrp4-exosome with the RNA bound complex shows an increase in the radius of gyration from 39.6 A˚ to 46.8 A˚ when RNA is bound and the corresponding pair distribu- tion functions contains longer vectors (Figure 6A), likely because additional scattering elements from RNA protrude from the compact protein core. The resulting ab initio model of the complex overlaid with the crystal structure of the apo-complex clearly indicates additional mass from the bound RNA (Figure 6B and C). This clear additional mass is distributed in the centre of the cap structure on top of the neck region but also protrudes Nucleic Acids Research, 2010, Vol. 38, No. 15 5173 away from the complex. When looking at the overlay with the crystal structure it appears that the RNA is bound at the KH and the S1-domains. The SAXS analysis supports the model that RNA binds near the KH-domain on the outside of the caps and reveals a low-resolution image of trapped exosome–RNA complexes. DISCUSSION RNA exosomes are large, self-compartmentalized nucle- ases, implicated in processive, controlled degradation of a large variety of RNAs. While the archaeal exosome possesses three phosphorolytic active sites within the com- partment, the eukaryotic exosomes apparently have lost this activity but adopted hydrolytic RNase subunits that are bound at the outside of the evolutionary conserved core. Nevertheless, recent data suggest that RNA is still threaded through the eukaryotic core exosome before it is degraded in ectopic hydrolytic active sites, suggesting that the core particle retained critical ‘structural’ functions re- garding RNA degradation such as increased processivity or controlled RNA degradation (25). To be able to quantitatively address RNA exosome activities, we derived a kinetic model for the complex RNA degradation of the archaeal RNA exosome using Markov Chain Monte Carlo analysis. The kinetic model gives a realistic assessment of the velocity of the exosome and mutant variants during processive degradation of a rA 30-mer oligonucleotide. The considerable effort we had to put into the MCMC simulation pays offeventually. We are now able to derive a realistic joint posterior distribu- tion of kinetic parameters, enabling us to quantify the relation of different parameters in either the same or in distinct exosome mutants. This would have been impos- sible with a conventional least squares fit of the data, which produces very unstable parameter estimates (see Supplementary Data for a comparison), although the obtained fits are very good (Figure 3D). With this in hand, we find several interesting and unex- pected features of RNA degradation activities. First, kinetic evaluation of RNA degradation of exosomes needs to include association and dissociation rate con- stants. Thus the kinetics cannot be treated as strictly processive, at least for RNA species in the assessed Figure 6. SAXS structure of the Rrp4 exosome with endogenously purified bacterial RNA. (A) SAXS data of the Rrp4 exosome (green) and the Rrp4 exosome with RNA (orange) (curves show the scattering intensity I(q) as a function of the scattering angle 2y and X-ray wavelength , where q = (4sin/y)) and the pair-distribution function describing intramolecular distances; in the presence of RNA longer distances occur and the radius of gyration increases. (B) Average of 10 independent ab initio models for the apo exosome and the RNA-bound complex superimposed with the crystal structure. The additional density for the RNA is clearly visible. 5174 Nucleic Acids Research, 2010, Vol. 38, No. 15 length range. This does not necessarily imply that RNA dissociates and rebinds completely from the exosome. Longer RNAs may be retained within the neck as well as cap domains, while binding and dissociating from the active site in the processing chamber. The association and dissociation constants can thus be understood as ‘ratcheting’ constants that influence the rate of translation along the RNA to and from the active site. For short substrates that are unable to simultaneously bind neck and active site this connection is lost, the dissociation in- creases, degradation speeds drop and the exosome changes from being fast and processive to a slower distributive enzyme. Our results also show how neck region and active site features contribute to exosome activity. Although we could not quantitatively address the importance of Arg65 in the neck with the simplified model in hand, this residue appears to be important for loading RNA into the processing chamber, but not for efficient degrad- ation once RNA is bound. This conclusion is derived from the observation that while the initial degradation is sub- stantially delayed, the appearance of smaller RNA species is qualitatively similar to the wild-type Csl4 exosome. Taken together with the observation that crosslinking severely reduces processing and the RNase PH ring needs to breath or display some conformational dynamics, it is unlikely that RNA is simply threaded into the processing chamber like a yarn through the eye of a needle. Rather, we propose that initial RNA binding includes some lateral entry near the neck. We are also in the position now to address the influence of the cap proteins Rrp4 and Csl4. These proteins possess a variety of domains with unclear function in exosome activity. While eukaryotic exosomes have defined heterotrimeric caps, the stoichiometry of cap proteins in archaeal exosomes is not defined in vitro and perhaps variable in vivo. For the archaeal exosome, the Csl4 capped isoform displays similar degradation kinetics to that of the capless variant, and the function of this type of cap remains to be shown. However, the Rrp4 isoform substantially differs from the other two variants and our analysis suggests that Rrp4 more efficiently recruits RNA to the exosome. In fact, RNA from the heterologous expression in E. coli is very tightly bound to the Rrp4 exosome. It must be noted that the gene coding for Rrp4 is in the same operon as genes for Rrp41 and Rrp42, indicating that this cap is perhaps a ‘default’ isoform of the exosome, while the Csl4 cap, located else- where in the genome, might be differentially regulated. The cap structures, however, also influence the degrad- ation of short RNAs. This is to some extent surprising, since short RNAs (<13 nt) cannot bind to the caps and the active site at the same time. However, the Rrp4 subunit more intimately interacts with the RNase PH ring than the Csl4 protein and might influence also the dynamics of the RNase PH type ring. Likewise, binding of RNA to the KH domains, consistent with the lateral density of RNA in the SAXS models, may position it better for loading into the processing chamber. In sum, we present here a robust method to analyse the complex degradation kinetics of a partially processive degradation enzyme in a quantitative manner, with esti- mates of the posterior distribution of the model param- eters. We applied this analysis to degradation of RNA by several isoforms and variants of the archaeal exosome and could reveal a variety of features that are important for catalytic efficiency. The objective of this manuscript is to derive a general method that can now be used to unravel the biochemistry of exosomes in a more quantitative manner. The method can now form a basis for compre- hensive analysis of different substrates, other RNA se- quences, as well as mutants of this system or related systems such as the eukaryotic exosome. ACCESSION NUMBERS 3M7N, 3M85. SUPPLEMENTARY DATA Supplementary Data are available at NAR Online. ACKNOWLEDGEMENTS The authors thank Christian Luginsland for help in protein purification, Katharina Bu¨ ttner for exosome con- structs and Katja Lammens and Gregor Witte for helpful discussions. The authors thank the staffof the European Synchrotron Radiation Facility (beamline 14–2) and the Swiss Light Source (beamline PX I) for help with diffrac- tion data collection and Michal Hammel from the Advanced Light Source (SIBYLS beamline) for help with scattering data collection. FUNDING Deutsche Forschungsgemeinschaft (HO2489/3 and SFB646); Center for Integrated Protein Science Munich. Funding for open access charge: Deutsche Forschungsgemeinschaft. Conflict of interest statement. None declared. REFERENCES 1. Mitchell,P., Petfalski,E., Shevchenko,A., Mann,M. and Tollervey,D. (1997) The exosome: a conserved eukaryotic RNA processing complex containing multiple 30–>50 exoribonucleases. Cell, 91, 457–466. 2. Allmang,C., Kufel,J., Chanfreau,G., Mitchell,P., Petfalski,E. and Tollervey,D. (1999) Functions of the exosome in rRNA, snoRNA and snRNA synthesis. EMBO J., 18, 5399–5410. 3. Anderson,J.S. and Parker,R.P. (1998) The 30 to 50 degradation of yeast mRNAs is a general mechanism for mRNA turnover that requires the SKI2 DEVH box protein and 30 to 50 exonucleases of the exosome complex. EMBO J., 17, 1497–1506. 4. Isken,O. and Maquat,L.E. (2007) Quality control of eukaryotic mRNA: safeguarding cells from abnormal mRNA function. Genes Dev., 21, 1833–1856. 5. Thompson,D.M. and Parker,R. (2007) Cytoplasmic decay of intergenic transcripts in Saccharomyces cerevisiae. Mol. Cell. Biol., 27, 92–101. 6. Mitchell,P. and Tollervey,D. (2003) An NMD pathway in yeast involving accelerated deadenylation and exosome-mediated 30–>50 degradation. Mol. Cell., 11, 1405–1413. Nucleic Acids Research, 2010, Vol. 38, No. 15 5175 7. Vasudevan,S., Peltz,S.W. and Wilusz,C.J. (2002) Non-stop decay– a new mRNA surveillance pathway. Bioessays, 24, 785–788. 8. Houseley,J., LaCava,J. and Tollervey,D. (2006) RNA-quality control by the exosome. Nat. Rev. Mol. Cell. Biol., 7, 529–539. 9. Schmid,M. and Jensen,T.H. (2008) The exosome: a multipurpose RNA-decay machine. Trends Biochem. Sci., 33, 501–510. 10. Allmang,C., Petfalski,E., Podtelejnikov,A., Mann,M., Tollervey,D. and Mitchell,P. (1999) The yeast exosome and human PM-Scl are related complexes of 30 –> 50 exonucleases. Genes Dev., 13, 2148–2158. 11. Koonin,E.V., Wolf,Y.I. and Aravind,L. (2001) Prediction of the archaeal exosome and its connections with the proteasome and the translation and transcription machineries by a comparative-genomic approach. Genome Res., 11, 240–252. 12. Evguenieva-Hackenberg,E., Walter,P., Hochleitner,E., Lottspeich,F. and Klug,G. (2003) An exosome-like complex in Sulfolobus solfataricus. EMBO Rep., 4, 889–893. 13. Bu¨ ttner,K., Wenig,K. and Hopfner,K.P. (2006) The exosome: a macromolecular cage for controlled RNA degradation. Mol. Microbiol., 61, 1372–1379. 14. Liu,Q., Greimann,J.C. and Lima,C.D. (2006) Reconstitution, activities, and structure of the eukaryotic RNA exosome. Cell, 127, 1223–1237. 15. Lorentzen,E., Basquin,J., Tomecki,R., Dziembowski,A. and Conti,E. (2008) Structure of the active subunit of the yeast exosome core, Rrp44: diverse modes of substrate recruitment in the RNase II nuclease family. Mol. Cell., 29, 717–728. 16. Navarro,M.V., Oliveira,C.C., Zanchin,N.I. and Guimaraes,B.G. (2008) Insights into the mechanism of progressive RNA degradation by the archaeal exosome. J. Biol. Chem., 283, 14120–14131. 17. Ramos,C.R., Oliveira,C.L., Torriani,I.L. and Oliveira,C.C. (2006) The Pyrococcus exosome complex: structural and functional characterization. J. Biol. Chem., 281, 6751–6759. 18. Wang,H.W., Wang,J., Ding,F., Callahan,K., Bratkowski,M.A., Butler,J.S., Nogales,E. and Ke,A. (2007) Architecture of the yeast Rrp44 exosome complex suggests routes of RNA recruitment for 30 end processing. Proc. Natl Acad. Sci. USA, 104, 16844–16849. 19. Portnoy,V., Evguenieva-Hackenberg,E., Klein,F., Walter,P., Lorentzen,E., Klug,G. and Schuster,G. (2005) RNA polyadenylation in Archaea: not observed in Haloferax while the exosome polynucleotidylates RNA in Sulfolobus. EMBO Rep., 6, 1188–1193. 20. Dziembowski,A., Lorentzen,E., Conti,E. and Seraphin,B. (2007) A single subunit, Dis3, is essentially responsible for yeast exosome core activity. Nat. Struct. Mol. Biol., 14, 15–22. 21. Schneider,C., Anderson,J.T. and Tollervey,D. (2007) The exosome subunit Rrp44 plays a direct role in RNA substrate recognition. Mol. Cell., 27, 324–331. 22. Liu,Q., Greimann,J.C. and Lima,C.D. (2007) Reconstruction, activities, and structure of the eukaryotic RNA exosome. Erratum. Cell, 131, 188–189. 23. Lebreton,A., Tomecki,R., Dziembowski,A. and Seraphin,B. (2008) Endonucleolytic RNA cleavage by a eukaryotic exosome. Nature, 456, 993–996. 24. Schaeffer,D., Tsanova,B., Barbas,A., Reis,F.P., Dastidar,E.G., Sanchez-Rotunno,M., Arraiano,C.M. and van Hoof,A. (2009) The exosome contains domains with specific endoribonuclease, exoribonuclease and cytoplasmic mRNA decay activities. Nat. Struct. Mol. Biol., 16, 56–62. 25. Bonneau,F., Basquin,J., Ebert,J., Lorentzen,E. and Conti,E. (2009) The yeast exosome functions as a macromolecular cage to channel RNA substrates for degradation. Cell, 139, 547–559. 26. Lorentzen,E. and Conti,E. (2005) Structural basis of 30 end RNA recognition and exoribonucleolytic cleavage by an exosome RNase PH core. Mol. Cell., 20, 473–481. 27. Lorentzen,E., Dziembowski,A., Lindner,D., Seraphin,B. and Conti,E. (2007) RNA channelling by the archaeal exosome. EMBO Rep., 8, 470–476. 28. Walter,P., Klein,F., Lorentzen,E., Ilchmann,A., Klug,G. and Evguenieva-Hackenberg,E. (2006) Characterization of native and reconstituted exosome complexes from the hyperthermophilic archaeon Sulfolobus solfataricus. Mol. Microbiol., 62, 1076–1089. 29. Bu¨ ttner,K., Wenig,K. and Hopfner,K.P. (2005) Structural framework for the mechanism of archaeal exosomes in RNA processing. Mol. Cell., 20, 461–471. 30. Kabsch,W. (1993) Automatic data processing of rotation diffraction data from crystals of initially unkonwn symmetry and cell constants. J. Appl. Cryst., 26, 795–800. 31. CCP4. (1994) The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D Biol. Crystallogr., 50, 760–763. 32. Brunger,A.T., Adams,P.D., Clore,G.M., DeLano,W.L., Gros,P., Grosse-Kunstleve,R.W., Jiang,J.S., Kuszewski,J., Nilges,M., Pannu,N.S. et al. (1998) Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr. D Biol. Crystallogr., 54, 905–921. 33. Afonine,P.V., Grosse-Kunstleve,R.W. and Adams,P.D. (2005) A robust bulk-solvent correction and anisotropic scaling procedure. Acta Crystallogr. D Biol. Crystallogr., 61, 850–855. 34. Emsley,P. and Cowtan,K. (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr., 60, 2126–2132. 35. Hura,G.L., Menon,A.L., Hammel,M., Rambo,R.P., Poole,F.L., 2nd, Tsutakawa,S.E., Jenney,F.E. Jr, Classen,S., Frankel,K.A., Hopkins,R.C. et al. (2009) Robust, high-throughput solution structural analyses by small angle X-ray scattering (SAXS). Nat. Methods, 6, 606–612. 36. Konarev,P.V., Volkov,V.V., Sokolova,A.V., Koch,M.H.J. and Svergun,D.I. (2003) PRIMUS: a Windows PC-based system for small-angle scattering data analysis. J. Appl. Cryst., 36, 1277–1282. 37. Svergun,D.I., Petoukhov,M.V. and Koch,M.H. (2001) Determination of domain structure of proteins from X-ray solution scattering. Biophys. J., 80, 2946–2953. 38. Volkov,V.V. and Svergun,D.I. (2003) Uniqueness of ab initio shape determination in small-angle scattering. J. Appl. Cryst., 36, 860–864. 39. Rocke,D.M. and Durbin,B. (2001) A model for measurement error for gene expression arrays. J. Comput. Biol., 8, 557–569. 40. Lorentzen,E., Walter,P., Fribourg,S., Evguenieva-Hackenberg,E., Klug,G. and Conti,E. (2005) The archaeal exosome core is a hexameric ring structure with three catalytic subunits. Nat. Struct. Mol. Biol., 12, 575–581. 5176 Nucleic Acids Research, 2010, Vol. 38, No. 15
3M89
Structure of TubZ-GTP-g-S
Plasmid protein TubR uses a distinct mode of HTH- DNA binding and recruits the prokaryotic tubulin homolog TubZ to effect DNA partition Lisheng Ni, Weijun Xu, Muthiah Kumaraswami, and Maria A. Schumacher1 Department of Biochemistry and Molecular Biology, University of Texas, M. D. Anderson Cancer Center, Unit 1000, Houston, TX 77030 Edited by Robert T. Sauer, Massachusetts Institute of Technology, Cambridge, MA, and approved May 11, 2010 (received for review March 22, 2010) The segregation of plasmid DNA typically requires three elements: a DNA centromere site, an NTPase, and a centromere-binding protein. Because of their simplicity, plasmid partition systems represent tractable models to study the molecular basis of DNA segregation. Unlike eukaryotes, which utilize the GTPase tubulin to segregate DNA, the most common plasmid-encoded NTPases contain Walker-box and actin-like folds. Recently, a plasmid stabi- lity cassette on Bacillus thuringiensis pBtoxis encoding a putative FtsZ/tubulin-like NTPase called TubZ and DNA-binding protein called TubR has been described. How these proteins collaborate to impart plasmid stability, however, is unknown. Here we show that the TubR structure consists of an intertwined dimer with a winged helix-turn-helix (HTH) motif. Strikingly, however, the TubR recognition helices mediate dimerization, making canonical HTH– DNA interactions impossible. Mutagenesis data indicate that a basic patch, encompassing the two wing regions and the N termini of the recognition helices, mediates DNA binding, which indicates an unusual HTH–DNA interaction mode in which the N termini of the recognition helices insert into a single DNA groove and the wings into adjacent DNA grooves. The TubZ structure shows that it is as similar structurally to eukaryotic tubulin as it is to bacterial FtsZ. TubZ forms polymers with guanine nucleotide-binding characteristics and polymer dynamics similar to tubulin. Finally, we show that the exposed TubZ C-terminal region interacts with TubR-DNA, linking the TubR-bound pBtoxis to TubZ polymerization. The combined data suggest a mechanism for TubZ-polymer pow- ered plasmid movement. T he cytoskeletons of eukaryotic cells are constructed of three primary elements: actin, tubulin, and intermediate filaments. Although it had long been presumed that the proteins forming these elements were absent in prokaryotes, it is now known that prokaryotes contain structural homologs to all three components. These prokaryotic proteins appear to carry out distinct functions compared to their eukaryotic counterparts; however, their roles are similar enough to indicate a likely common ancestor. The best known prokaryotic homolog of tubulin is FtsZ. Both FtsZ and tubulin form long filamentous structures by head to tail associa- tion in a manner regulated by GTP, which binds between adjacent subunits (1–4). However, unlike tubulin, FtsZ does not function in DNA segregation but rather cell division. Specifically, it forms a cytokinetic ring called the Z ring at midcell, which mediates septation (5, 6). Recently, however, prokaryotic proteins encoded on large plasmids harbored in bacilli showing 15–20% sequence similarity to both FtsZ and tubulin have been identified and dubbed TubZ (7–12). Studies showed that the Bacillus thuringien- sis TubZ protein from the pBtoxis plasmid is essential for plasmid DNA segregation. DNA segregation of most low copy number plasmids is carried out by specific partition (par) systems. These systems require only three elements: a centromere DNA site, a centromere-binding protein, and a partition NTPase (13, 14). Partition systems have been classified into two main types on the basis of the kind of NTPase present (15). Type I systems contain NTPases with deviant Walker A-type ATPase folds, whereas type II systems uti- lize actin-like NTPases. Interestingly, both types of NTPases form polymers in NTP-dependent manners that are implicated to play a role in plasmid DNA separation (16–19). The recent discovery of TubZ NTPases has led to the designation of “type III” par sys- tems (13, 14). The best studied of these systems is that found on the pBtoxis plasmid in B. thuringiensis. This plasmid stability sys- tem is represented by an operon encoding two proteins: ORF156 (TubZ) and ORF157 (TubR) (7–9, 11). TubR is a 11.6 kDa DNA- binding protein that shows no sequence homology to any known protein. Studies showed that TubR binds a 48-bp centromere con- taining four repeat sites in the pBtoxis plasmid and also autore- gulates tubRZ transcription (8, 9). TubZ is a 54.4 kDa protein that can assemble into filaments in a GTP-dependent manner (12). Both proteins were found to be required for plasmid stability (9). However, how the TubR and TubZ proteins work together to effect pBtoxis plasmid segregation is not known. To gain insight into the molecular mechanism utilized by these proteins in DNA segregation, we carried out structural and biochemical studies on the pBtoxis TubR and TubZ proteins. The TubR structure reveals that it employs a helix-turn-helix (HTH) motif in a previously un- described manner to bind DNA. TubZ contains a tubulin/FtsZ fold but has structural distinctions from these proteins indicating that it forms distinct protofilaments. TubR binds the flexible C-terminal region of TubZ, thus attaching the TubZ filament to the pBtoxis plasmid, providing a mechanism for plasmid move- ment and, ultimately, segregation. Results and Discussion Overall Structure of pBtoxis TubR. The crystal structure of the 107- residue pBtoxis TubR protein was solved to 2.0-Å resolution by selenomethionine multiple wavelength anomalous diffraction (MAD) methods (Table S1). The structure contains two TubR molecules in the crystallographic asymmetric unit and consists of residues 6–102 of one subunit and 4–100 of the second subunit, and has Rwork∕Rfree ¼ 23.8%∕27.0%. The TubR structure forms a highly intertwined dimer with dimensions 30 × 30 × 60 Å3 (Fig. 1A). Each TubR subunit has the topology β1-α1-α2-α3-α4-β2- β3-α5, which is similar to winged HTH motifs found in a number of DNA-binding proteins in both prokaryotes and eukaryotes (20). In TubR, α3-α4 forms the HTH motif and the loop between Author contributions: M.A.S. designed research; L.N., W.X., M.K., and M.A.S. performed research; M.A.S. analyzed data; and M.A.S. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Freely available online through the PNAS open access option. Data deposition: The atomic coordinates and structure factor amplitudes for the WT TubR (C2), WT TubR (I222), TubR(S63W), TubZ, and TubZ-GTP-γ-S structures have been deposited with the Protein Data Bank, www.pdb.org (PDB ID codes 3M8E, 3M9A, 3M8F, 3M8K, and 3M89). 1To whom correspondence may be addressed. E-mail: maschuma@mdanderson.org. This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1003817107/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1003817107 PNAS ∣June 29, 2010 ∣vol. 107 ∣no. 26 ∣11763–11768 BIOCHEMISTRY β2 and β3, the wing. Indeed, each TubR subunit shows the stron- gest structural similarity to members of the ArsR family of prokaryotic winged helix transcription regulators, in particular the Staphylococcus aureus CzrA protein (21, 22). Superposition of one subunit of TubR onto that of CzrA results in a rmsd of 2.7 Å. This similarity includes the core regions of the winged HTH, because the loops and N-terminal regions of the proteins are structurally distinct. For example, TubR contains a β-strand in its N-terminal region compared to a long helix in the CzrA struc- ture (Fig. 1B). This structural similarity initially suggested that TubR may be a member of the ArsR family of proteins. However, the arrangement of the TubR dimer was found to be strikingly different from the dimer organization exhibited by the ArsR proteins (Fig. 1C). ArsR family members are involved in metal-regulated tran- scription processes whereby they act as repressors in their apo forms and are induced off their DNA sites upon metal binding (22). The specific dimer structures of the ArsR proteins are cri- tical for creation of their metal-binding motifs. Not only does TubR form a very different dimer from the ArsR proteins, it also does not harbor any of their metal-binding signatures nor does it contain any other characterized metal-binding motif. Consistent with this, we find that the addition of metals has no effect on TubR DNA binding. Dimerization of the ArsR proteins is imparted by residues from the N-terminal regions, α1 and α5, which, impor- tantly, leaves its recognition helices exposed for DNA interaction (22). By contrast, the TubR dimer is formed primarily by contacts between its twofold related “recognition helices” α4 and α4·. This interaction results in a near complete burial of these helices, leav- ing only the N-terminal residues exposed to solvent. Whereas the α4 and α4· interaction creates the dimer core, the dimer is further stabilized by interactions between the twofold related β1 strands, which swap to form an antiparallel β-sheet. Residues from α1 and α5 interact with β1 to further seal the top of the dimer. The dimer interface formed by these interactions is predominantly hydropho- bic and buries a large 1;200 Å2 of subunit surface from solvent. TubR Forms a “Recognition Helix Dimer”: Implications for DNA Bind- ing. Gel filtration studies on TubR confirmed that it is a dimer in solution. However, the finding from the structure that the TubR α4 recognition helices are buried in the dimer core has important implications in terms of its DNA-binding mechanism. Indeed, it suggests that, although TubR contains a structurally canonical HTH, it is not utilized for DNA binding in a manner typical of HTH proteins. A second crystal form (I222) of TubR, which was solved to 2.5-Å resolution, revealed the same TubR dimer. The presence of the identical dimer in two different crystal forms and its large buried surface area supports that the dimer observed in the crystal structures is physiologically relevant. However, to test this, we mutated residues within the recognition helices that the structure indicates are critical for dimerization and assayed the ability of the mutant proteins to dimerize via gel filtration. Specifically, we mutated Ser-63 and Ala-67 individually to tryp- tophan and arginine. The structure shows that residues occupying positions 63 and 67 must be small and largely hydrophobic to permit the proper packing of the α4 helices in the dimer (Fig. S1A). Hence, the in- troduction of the bulky side chain of tryptophan and, in particu- lar, the large as well as charged side chain of arginine would be predicted to be highly disruptive to dimerization. Gel filtration analyses on purified mutant proteins clearly showed that the ar- ginine mutants exist primarily as monomers in solution (>80%), whereas the tryptophan mutants were able to maintain the di- meric state (Fig. 2A and Fig. S1B). However, all mutant proteins showed reduced or loss of DNA-binding activity as ascertained by fluorescence polarization (FP) studies, which examined TubR protein binding to its centromere site (Fig. 2B and Fig. S1 C and D) (9). The fact that the monomeric mutants were severely impaired in DNA binding was not surprising. However, the finding that the tryptophan mutants, which were largely dimeric, displayed reduced DNA-binding activity suggested that their oli- gomer structures might be altered. To address this issue, the struc- ture of S63W TubR was solved to 2.8-Å resolution, resulting in Rwork∕Rfree values of 20.4%∕26.6% (Table S1). The subunit struc- ture of S63W TubR is essentially identical to that of WT TubR as revealed by their superimposition (rmsd of 1.2 Å) (Fig. 2 C and D). However, this single subunit overlay shows that the S63W TubR dimer, although the same as the WT in general arrange- ment, is forced into a more open oligomer conformation in which one subunit is rotated 20° away from its dimer mate compared to WT. This rotation is required to accommodate the bulky S63W side chains (Fig. 2D). The N-terminal β1–β1′-sheet interaction appears to play a key role in holding the TubR subunits together. In addition, the tight stacking of the twofold related Trp63 indole groups (3.5 Å) provides a compensatory interaction that, com- bined with the β1–β1′ interaction, apparently permits retention of the dimer state, indicating why the TubR tryptophan mutants were able to maintain the dimer state, albeit an altered dimer state relative to WT (Fig. 2 C and D). By contrast, the TubR arginine mutations, which introduced both bulk and charge with- in the predominantly hydrophobic dimer interface, were highly destabilizing for dimerization. In addition, the finding that the S63W TubR mutant forms an altered dimer explains the severe effect on DNA binding because a correct dimer orientation is likely essential for binding to its palindromic DNA sites (9). TubR-DNA Model. Because all but the N-terminal residues of the TubR recognition helices are buried in the dimer interface, TubR must use a different mode of DNA binding than the ArsR or other HTH containing proteins (23). Examination of surface electro- statics of TubR reveals that one face of the protein is electro- negative, whereas the other is strongly electropositive (Fig. 3A). Notably, the positive region is composed of one large and contig- uous basic patch. Basic residues in this region correspond to Arg- 74, Arg-77, and Lys-79, in the wing and Lys-43, which is on α3, the helix preceding α4 in the HTH motif. These residues were mutated singly to alanine to examine their roles in DNA binding (Fig. 3B). FP experiments showed that mutation of the basic wing residues resulted in either a complete (R74A and R77A) or nearly com- plete (K79A) abrogation of DNA binding, indicating that the wings play a major role in TubR DNA binding. Residue Lys-43 is surface exposed and located at the center of the basic region on the TubR dimer (Fig. 3A). The K43A mutant also showed no binding to TubR, supporting the notion that the continuous basic patch of TubR represents its DNA-binding surface. To gain insight into the structural mechanism of DNA binding, a DNA duplex was docked onto the basic patch of the TubR di- Fig. 1. B. thuringiensis pBtoxis TubR structure. (A) One TubR subunit is red and the other cyan. Secondary structural elements and N and C termini are labeled. (B) Superimposition of one subunit of TubR (Red) onto a S. aureus CzrA subunit (Yellow). Regions with different structures are labeled. (C) Same superimposition as B showing the location of the other subunit in the TubR and CzrA dimers after one subunit is overlaid. A–C are in the same orienta- tion to highlight differences. Figs. 1 A–C, 2 C and D, 3 A, C, and E, 4B, and 5B were made by using PyMOL (31). 11764 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1003817107 Ni et al. mer by using the location of the mutations that affected DNA binding as a guide (Fig. 3C). This model revealed that the wings are positioned to interact with successive minor grooves, with either the bases or the phosphate backbone depending on the ability of the DNA to deform. In the model TubR interacts with a minimum of 14 bp of DNA. However, the centromere bound by Fig. 2. The TubR “recognition helix” dimer. (A) Gel fil- tration studies on TubR mutants S63W and S63R show- ing that the S63W mutant remains dimeric whereas S63R is >80% monomer. (B) Fluorescence polarization studies examining the ability of WT TubR, S63R, and S63W TubR mutants to bind iteronic DNA. Fluorescence polarization units (millipolarization) and TubR concen- tration (nM) are along the y and x axes, respectively. The Kd of WT TubR for the centromere DNA is 8  2 nm. (C) Superimposition of WT TubR (Green) onto the TubR S63W mutant structure (Tan). (D) Close-up of the site of the S63W mutation in the expanded TubR S63W dimer showing stacking interactions between the twofold related tryptophans. Fig. 3. TubR-DNA binding. (A) Electrostatic surface potential of the TubR dimer. Blue and red represent electropositive and electronegative regions, respec- tively. (Left) The electronegative side of the TubR dimer, and the side on the right is the electropositive side. Labeled on the left side are the locations of the mutated residues. (B) FP binding isotherms showing the DNA binding of WT TubR and the K43A, R74A, R77A, and K79A mutants. Fluorescence polarization units (millipolarization) and TubR concentration (nM) are along the y and x axes, respectively. (C) TubR-DNA model showing TubR electrostatic potential. (D) Stoichiometry of TubR (subunit) binding showing titration curve of TubR into the 48-mer iteron resulting in a molar ratio of TubR subunit to DNA of eight (or four) dimers. (E) Left: Ribbon diagram of the TubR-DNA model with the recognition helices colored yellow. Right: Ribbon diagram showing a canonical HTH–DNA interaction (the λ repressor-DNA complex) with the recognition helices colored yellow (32). Ni et al. PNAS ∣ June 29, 2010 ∣ vol. 107 ∣ no. 26 ∣ 11765 BIOCHEMISTRY TubR consists of four 12-bp sites with the consensus T(T/A)(T/A) (C/A)(G/A)GTTTA(A/C)(A/C) (9). Thus, we used FP to ascer- tain the binding stoichiometry of TubR for its 48-bp centromere. As shown in Fig. 3D, eight TubR subunits or four TubR dimers bind the 48-mer centromere, consistent with a dimer of TubR binding each palindrome. Thus, either TubR distorts its DNA or the TubR dimers bind with some degree of overlap on their DNA sites perhaps imparting cooperativity, as observed in other centromere-binding protein–DNA interactions (13, 14). In addi- tion to the insertion of the wings, a striking outcome of the mod- eling was the finding that the N termini of the recognition helices, which interact with each other in a parallel, coiled-coil-like man- ner, are in position to insert into a single major groove. Structures of HTH proteins bound to DNA have thus far shown that the recognition helices insert singly into successive major grooves by using residues in the first few turns or the central portion of the recognition helix to contact the DNA bases (Fig. 3E). Thus, in the TubR-DNA model, the HTH–DNA interaction is drama- tically different from any displayed previously by a HTH protein. TubZ Binds TubR-DNA. A characteristic feature exhibited by partition centromere-binding proteins is the ability to bind their partner NTPase (13, 14). To determine if TubR binds TubZ, we utilized a FP assay and found that full length (FL) TubZ bound avidly to the TubR-centromere complex (Fig. 4A). However, unlike other partition systems in which the NTPase must be com- plexed with nucleotide to bind its centromere-binding protein, the interaction of TubR with TubZ did not require the presence of GTP-Mg2þ. Previous studies have shown that the C-terminal regions of tubulin and FtsZ mediate key binding events with their target proteins (24–26). We noted that the terminal region of TubZ, consisting of residues 407–484, is the most divergent region between TubZ proteins and between TubZ and tubulin/FtsZ pro- teins, suggesting that it may be similarly utilized and bind TubR. To test this hypothesis, we constructed various TubZ truncations, TubZ(1-407), TubZ(1-442), TubZ(1-460), and TubZ(1-470), and examined the ability of each protein to bind TubR-DNA. TubZ (1-407) showed no binding to TubR, whereas the remaining trun- cation mutants bound weakly to TubR-DNA (Fig. 4A). Thus, the data indicate that the last 14 amino acids of TubZ are critical for the ability of TubZ to form a tight interaction with TubR but that residues 408–470 also play an important role in this interaction. These data demonstrate that TubR acts as a partition partner for TubZ, linking it to pBtoxis plasmid DNA. Although TubZ has been shown to form polymers in a GTP-dependent manner, the TubZ protein displays limited sequence similarity to tubu- lin/FtsZ, suggesting potential differences in TubZ and tubulin/ FtsZ structures (7–9). To gain insight into TubZ function, we next determined structures of B. thuringiensis pBtoxis TubZ. Structure of TubZ. Crystallization of FLTubZ was not successful, in either its apo form or bound to guanine nucleotides. We noted that FL TubZ degraded over time whereby C-terminal residues were proteolyzed. Therefore, truncated TubZ proteins were utilized in crystallization trials, and crystals were obtained of apo TubZ (1-428) and the structure solved by MAD (Table S2). The model consists of residues 1–79 and 91–404 and has an Rwork∕Rfree of 21.4∕24.9% to 2.3-Å resolution (Fig. 4B). No discernible oligomer- ization of apo TubZ(1-428) was observed in the crystal packing, and gel filtration analyses confirmed that it is monomeric (Fig. S2). The overall TubZ structure can be divided into two main domains: an N domain (residues 25–235) and a C domain (resi- dues 258–377). These domains are connected by a long, core helix, H7. The TubZ N domain has a Rossman fold and consists of six parallel β-strands with topology 3-2-1-4-5-6. The resulting β-sheet is sandwiched by five α-helices, with two helices on one side and three on the other. The C domain consists of four β-strands with the topology 1-4-2-3. The C-domain β-strands are arranged nearly perpendicular to those in the N domain. In addition to these main protein domains, there are two helices: one at the N terminus, H0, and a long helix at the C terminus, H11 (Fig. 4B). Database searches showed that TubZ indeed belongs to the tubulin/FtsZ fa- mily of proteins and is similar to both eukaryotic and prokaryotic members of the family; TubZ can be optimally superimposed with rmsds of 3.4 Å onto both bovine α tubulin and Pseudomonas aeruginosa FtsZ (1, 5). Whereas the two-domain architectures of tubulin, FtsZ, and TubZ are similar in overall structure, the extreme N- and C-terminal regions of these proteins are very di- vergent (Fig. S3 A–D). N-Terminal and C-Terminal Differences in TubZ, FtsZ, and Tubulin: Im- plications for Polymer Formation and Target Protein Binding. Tubulin proteins do not contain significant N-terminal extensions, whereas FtsZ proteins from different organisms show structural variability within their N-terminal regions. For instance, in the Escherichia coli FtsZ structure the N-terminal residues are disor- dered, whereas Methanococcus jannaschii FtsZ has an extra N-terminal helix, H0, which is flexibly attached to the body of the protein and has been captured in multiple orientations (5). Although H0 is not conserved in FtsZ proteins, one M. jannaschii FtsZ structure revealed a semicontinuous polymer in the crystal, thought to closely represent in vivo protofilaments, which utilizes H0 in subunit-subunit contacts (4). This finding suggests that the flexibly attached H0 is stabilized in a specific orientation by protofilament formation, at least in the M. jannaschii protein. The TubZ H0 helix extends in the opposite direction compared to that of the protofilament stabilized FtsZ H0 helix. Moreover, in TubZ, H0 is not flexibly attached to the N domain but is tightly anchored to the C domain through numerous interactions with the core helix and C-domain residues. The large number of inter- actions involving H0, and the fact that it covers what would other- wise be a surface exposed hydrophobic patch, indicate that the TubZ H0 does not undergo conformational changes during protofilament formation and is important for the general fold of TubZ (Fig. S3 A and C). Data suggest that FtsZ and tubulin form protofilaments with similar longitudinal contacts (4). However, the TubZ structure reveals key differences, primarily in its C-domain and C-terminal regions, which suggest that it forms protofilaments distinct from those formed by tubulin/FtsZ. A notable difference is the struc- ture of loop 7 (L7). This loop inserts into the adjacent subunit providing the key catalytic residues required for GTP hydrolysis. In tubulin/FtsZ proteins, L7 has the consensus GXXNXDXAD. In TubZ, the loop is very divergent in conformation compared Fig. 4. TubZ interacts with TubR-DNA and contains a tubulin/FtsZ fold. (A) FP assay measuring binding of FL TubZ, TubZ(1-470), TubZ(1-460), TubZ(1-442), and TubZ(1-407) to TubR-DNA. Below is the control (TubZ titrated into DNA alone). Millipolarization units and TubZ concentration (nM) are along the y and x axes, respectively. (B) TubZ(1-428) structure. The N domain or GTP-bind- ing domain is colored salmon and the C domain purple. The interdomain helix, H7, is red. TubZ also contains an N-terminal helix, H0 (Yellow), and a C-terminal helix, H11 (White). 11766 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1003817107 Ni et al. to that in FtsZ/tubulin and consists of the sequence 256- DNVTYDPSD-266. In addition to the N-terminal region, the extreme C-terminal extensions of tubulin, FtsZ, and TubZ are structurally divergent (Fig. S3B). In FtsZ, the C-terminal region forms a small, two-stranded β-sheet and continues into an ex- tended region that is involved in binding adaptor proteins such as FtsA and ZipA (6, 25). By contrast, the C-terminal regions of tubulin proteins consist of a two-helix bundle followed by an extended region. Like FtsZ, however, these regions interact with numerous target proteins such as the microtubule-associated pro- teins (MAPs) (24). Consistent with this, the C-terminal regions of tubulin have been shown to face the outside of the microtubule. A characteristic feature of the extreme C-terminal extensions of tubulin proteins is their highly acidic nature (3). This acidic region has been shown to be critical for binding to several MAPs that harbor a substantial basic character, such as tau, MAP2, and MAP4 (24, 27). The TubZ C-terminal region is also helical, but it contains a single, long helix. Notably, the TubZ-tubulin overlay shows that the long C-terminal helix of TubZ would dramatically clash with the adjacent subunit in a polymer, providing support for the no- tion that TubZ forms protofilaments different from tubulin/FtsZ (Figs. S3B and S4). Interestingly, and in contrast to tubulin pro- teins, the flexible C-terminal region of TubZ that follows H11 is highly basic, in particular the last 14 residues. We have shown that these residues play a central role in TubR binding (Fig. 4A). TubR uses its electropositive face for DNA binding, leaving exposed its opposite face for TubZ interaction. Notably, this exposed face is strongly electronegative and hence would complement the basic C-terminal tail of TubZ (Fig. 3A). Tubulin/FtsZ protofilaments combine to form higher-order structures. In tubulin, the protofilaments interact in a parallel manner to form microtubules. Central to microtubule formation are lateral contacts between protofilaments from the so-called M loop, between H10 and S9. In tubulin, this loop is composed of 13 residues (1–3). The corresponding loop is much shorter in FtsZ proteins, consistent with the fact that FtsZ does not form tubulin microtubule-like structures (5, 28–30). In TubZ, the M loop is even shorter than in FtsZ, spanning only four residues. In fact, the TubZ/tubulin overlay shows that the side of the molecule con- taining the M loop is the most divergent between these proteins. These combined findings suggest that TubZ not only forms pro- tofilaments with distinct longitudinal contacts compared to FtsZ and tubulin, but it also does not form tubulin-like microtubule structures. TubZ Interactions with Guanine Nucleotides. Consistent with TubZ being a member of the tubulin/FtsZ family, our isothermal titra- tion calorimetry (ITC) studies showed that TubZ binds guanine nucleotides with high affinity; Kds for GTP-γ-S and GDP were ∼0.69 and 26 μM, respectively (Fig. 5A). We next determined the structure of the TubZ-GTP-γ-S complex by soaking GTP-γ-S into preformed TubZ(1-428) crystals. The TubZ-GTP-γ-S struc- ture contains TubZ residues 1–79 and 91–404, one GTP-γ-S, and has Rwork∕Rfree ¼ 21.8%∕25.5% (Fig. 5B and Table S2). The structure shows that TubZ binds GTP-γ-S in the same GTP binding pocket as tubulin/FtsZ (1–5). Comparison of the apo and GTP-γ-S bound TubZ structures indicated that, like FtsZ, guanine nucleotide binding does not lead to significant conforma- tional changes (5). The phosphate binding pocket is formed by two of the most highly conserved regions between TubZ and tubulin/FtsZ called loops 1 and 4 (L1 and L4) (1–3). L1 contacts the GTP-γ-S α- and β-phosphate groups via the Gln-32 and Lys-33 amide nitrogens. The L1 region of FtsZ and tubulin contain the sequences GQ(A/G)G and GQCG, respectively, whereas in TubZ the motif is 31-GQKG-34. However, the alanine/glycine and cysteine residues in FtsZ and tubulin do not contact the bound nucleotide; the TubZ Lys-33 side chain makes stacking interac- tions with the guanine base (Fig. 5B). L4 represents the so-called signature motif [GGGTG(T/S)G], which serves as an identifier of tubulin/FtsZ family members. Like FtsZ and tubulin, the L4 region of TubZ-GTP-γ-S makes phosphate interactions via its glycine amide nitrogens. Whereas L1 and L4 residues of the N domain mediate phosphate contacts, the GTP-γ-S guanine moiety is specified from residues in the core helix, H5, and C domain. In this regard, an important motif is loop 6 (L6). In FtsZ and tubulin, L6 has the consensus (F/Y)XXX(N/D) and the conserved (F/Y) residue functions in guanine base stacking. This region in TubZ, 236-WKXXXN-241, is in an altered conformation compared to FtsZ and tubulin structures. Despite the presence of the trypto- phan, which might be expected to interact with the guanine, the side chain of Lys-237 instead stacks with the guanine ring. Hence, in the TubZ-GTP-γ-S structure, the guanine base does not interact with aromatic residues as in tubulin/FtsZ but is sand- wiched between the aliphatic portions of two lysine side chains, Lys-33 and Lys-237. Finally, two asparagine residues, Asn-213 and Asn-241, from L6 effectively read the guanine N2/N3 and N1/O6 atoms, respectively, providing high specificity in TubZ’s in- teraction with guanine nucleotides. pBtoxis DNA Segregation: TubR Plasmid Partition Model. Our data show that TubR binds to the flexible, C-terminal, basic region of TubZ. The flexibility and location of the TubZ C-terminal extension suggest that it is not required for polymerization and thus may be exposed on the surface of TubZ filaments. Indeed, negative stain EM images show that TubZ(1-407) forms polymers in a GTP-dependent manner similar to the FL protein (Fig. S5). Recent data suggesting that TubZ filaments are stabilized by Fig. 5. TubZ-guanine nucleotide interactions. (A) ITC binding isotherms showing TubZ-GDP (Left) and TubZ-GTP-γ-S interaction (Right). (B) Left: Overall structure of the TubZ-GTP-γ-S complex. β-strands are colored magen- ta and helices cyan, and the GTP-γ-S molecule is shown as cpk. Right: Close-up view of the GTP binding pocket with the initial Fo-Fc electron density map (Blue Mesh), contoured at 4.5σ, and calculated before the GTP-γ-S was included in refinement. Ni et al. PNAS ∣ June 29, 2010 ∣ vol. 107 ∣ no. 26 ∣ 11767 BIOCHEMISTRY the presence of a GTP cap and undergo treadmilling are consis- tent with the notion that TubZ displays tubulin-like polymer dy- namics (12). Thus, on the basis of the combined data, we suggest a model for TubR/TubZ mediated pBtoxis plasmid segregation shown in Fig. 6. In this model, multiple TubR dimers first bind to the iteronic DNA on the pBtoxis plasmid leading to the crea- tion of a high local concentration of TubR, which can recruit a TubZ polymer, likely by interactions between the acidic TubR dimer face and the basic C-terminal TubZ region. Importantly, this interaction serves to attach the pBtoxis plasmid to the TubZ polymer, which undergoes treadmilling, adding subunits at the þ end and losing subunits at the −end. The bound TubR-pBtoxis can be handed off from the −end to the molecules in the growing þ end, leading to the transport of the pBtoxis plasmid to the cell pole. Interestingly, it has been shown that once TubZ polymers reach and interact with the cell pole, they bend around the curved pole and continue growing in the other direction (7). The force of the interaction with the membrane likely causes the release of TubR-pBtoxis, the net result being transport of pBtoxis to the cell pole. Of course, this model is simplified and many questions re- main. For example, how directionality is achieved and how the replicated plasmids are driven to opposite cell poles is not clear. However, given the large size of the pBtoxis plasmid (8), it may be that only one TubR-pBtoxis “tram” can be bound at once by the rapidly treadmilling TubZ polymer and that, once one such a tram is unloaded after reaching the cell pole, another engages when the now reversed polymer treadmills toward the opposite cell pole. Materials and Methods Summary Detailed methods are provided in SI Materials and Methods. Briefly, the tubR and tubZ genes were codon optimized (for E. coli expression), subcloned into pET15b, expressed, and pur- ified. WT TubR crystals were grown with NaCl and phosphate. TubR S63W was crystallized with PEG and ethylene glycol and TubZ with sodium formate. Detailed assay conditions for FP, ITC, electron microscopy, and gel filtration are provided in SI Materials and Methods. ACKNOWLEDGMENTS. This work was supported by the Burroughs Wellcome Career Development Award 992863 and National Institutes of Health Grant GM074815 (to M.A.S.). 1. Nogales E, Wolf SG, Downing KH (1998) Structure of the αβ tubulin dimer by electron crystallography. Nature 391:199–203. 2. Nogales E, Whittaker M, Milligan RA, Downing KH (1999) High-resolution model of the microtubule. Cell 96:79–88. 3. Nogales E (2000) Structural insights into microtubule function. Annu Rev Biochem 69:277–302. 4. Oliva MA, Cordell SC, Löwe J (2004) Structural insights into FtsZ protofilament formation. Nat Struct Mol Biol 11:1243–1250. 5. Oliva MA, Trambaiolo D, Löwe J (2007) Structural insights into the conformational variability of FtsZ. J Mol Biol 373:1229–1242. 6. Margolin W (2005) FtsZ and the division of prokaryotic cells and organelles. Nat Rev Mol Cell Biol 6:862–871. 7. Larsen RA, et al. (2007) Treadmilling of a prokaryotic tubulin-like protein, TubZ, required for plasmid stability in Bacillus thuringiensis. Genes Dev 21:1340–1352. 8. Tang M, Bideshi DK, Park H-W, Federici BA (2006) Minireplicon from pBtoxis of Bacillus thuringiensis subsp. israelensis. App Environ Microbiol 72:6948–6954. 9. Tang M, Bideshi DK, Park H-W, Federici BA (2007) Iteron-binding ORF157 and FtsZ-like ORF156 proteins encoded by pBtoxis play a role in its replication in Bacillus thuringien- sis subsp. israelensis. J Bacteriol 189:8053–8058. 10. Anand SP, Akhtar P, Tinsley E, Watkins SC, Khan SA (2008) GTP-dependent polymer- ization of the tubulin-like RepX replication protein encoded by the pXO1 plasmid of Bacillus anthracis. Mol Microbiol 67:881–890. 11. Berry C, et al. (2002) Complete sequence and organization of pBtoxis, the toxin- coding plasmid of Bacillus thuringiensis subsp israelensis. Appl Environ Microbiol 68:5082–5095. 12. Chen Y, Erickson HP (2008) In vitro assembly studies of FtsZ/tubulin-like proteins (TubZ) from Bacillus plasmids: Evidence for a capping mechanism. J Biol Chem 283:8102–8109. 13. Hayes F, Barillà D (2006) The bacterial segrosome: A dynamic nucleoprotein machine for DNA trafficking and segregation. Nat Rev Microbiol 4:133–143. 14. Schumacher MA (2008) Structural biology of plasmid partition: Uncovering the molecular mechanisms of DNA segregation. Biochem J 412:1–18. 15. Gerdes K, Møller-Jensen J, Bugge Jensen R (2000) Plasmid and chromosome partitioning: Surprises from phylogeny. Mol Microbiol 37:455–466. 16. Møller-Jensen J, et al. (2003) Bacterial mitosis: ParM of plasmid R1 moves plasmid DNA by an actin-like insertional polymerization mechanism. Mol Cell 12:1477–1487. 17. Popp D, et al. (2008) Molecular structure of the ParM polymer and the mechanism leading to its nucleotide-driven dynamic instability. EMBO J 27:570–579. 18. Salje J, Löwe J (2008) Bacterial actin: Architecture of the ParMRC DNA partitioning complex. EMBO J 27:2230–2238. 19. Dunham TD, Xu W, Funnell BE, Schumacher MA (2009) Structural basis for ADP- mediated transcriptional regulation by P1 and P7 ParA. EMBO J 28:1792–1802. 20. Gajiwala KS, Burley SK (2000) Winged helix proteins. Curr Opin Struct Biol 10:110–116. 21. Eicken C, et al. (2003) A metal-ligand-mediated intersubunit allosteric switch in related SmtB/ArsR zinc sensor proteins. J Mol Biol 333:683–695. 22. Pennella M, Giedroc DP (2005) Structural determinants of metal selectivity in prokaryotic metal-responsive transcriptional regulators. Biometals 18:413–428. 23. Arunkumar AI, Campanello GC, Giedroc DP (2009) Solution structure of a paradigm ArsR family zinc sensor in the DNA-bound state. Proc Natl Acad Sci USA 106:18177–18182. 24. Downing KH (2000) Structural basis for the interaction of tubulin with proteins and drugs that affect microtubule dynamics. Annu Rev Cell Dev Biol 16:89–111. 25. Adams DW, Errington J (2009) Bacterial cell division: Assembly, maintenance and disassembly of the Z ring. Nat Rev Microbiol 7:642–653. 26. Errington J, Daniel RA, Scheffers DJ (2003) Cytokinesis in bacteria. Microbiol Mol Biol Rev 67:52–65. 27. Chau MF, et al. (1998) The microtubule-associated protein tau cross-links to two distinct sites on each alpha and beta tubulin monomer via separate domains. Biochemistry 37:17692–17703. 28. Bi EF, Lutkenhaus J (1991) FtsZ ring structure associated with division in Escherichia coli. Nature 354:161–164. 29. Erickson HP, Taylor D, Taylor KA, Bramhill D (1996) Bacterial cell division protein FtsZ assembles into protofilament sheets and minirings, structural homologs of tubulin polymers. Proc Natl Acad Sci USA 93:519–523. 30. Osawa M, Anderson DE, Erickson HP (2008) Reconstitution of contractile FtsZ rings in liposomes. Science 320:792–794. 31. Delano WL (2002) The PyMOL Molecular Graphics System (DeLano Scientific, San Carlos, CA). 32. Pabo CO, Lewis M (1982) The operator-binding domain of λ repressor: Structure and DNA recognition. Nature 298:443–447. Fig. 6. pBtoxis DNA partition model. In the first step, TubR, which is bound to its centromere on one of the replicated pBtoxis plasmids, contacts the TubZ C-terminal region (indicated by lines pointing from the TubZ “circles”) in a treadmilling TubZ polymer. TubZ subunits are lost from the −end and are added to the þ end. TubR is pulled along the growing polymer by its TubR-TubZ interaction until it reaches the cell pole and is knocked off when it comes into contact with the membrane at the cell pole. TubZ reverses direction and may pick up the other TubR-pBtoxis complex and deliver it similarly to the opposite cell pole. 11768 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1003817107 Ni et al.
3M8B
Crystal structure of spin-labeled BtuB V10R1 in the apo state
Conformational Exchange in a Membrane Transport Protein Is Altered in Protein Crystals Daniel M. Freed,† Peter S. Horanyi,‡ Michael C. Wiener,‡ and David S. Cafiso†‡* †Departments of Chemistry and ‡Molecular Physiology and Biological Physics, and the Biophysics Program, University of Virginia, Charlottesville, Virginia ABSTRACT Successful macromolecular crystallography requires solution conditions that may alter the conformational sampling of a macromolecule. Here, site-directed spin labeling is used to examine a conformational equilibrium within BtuB, the Escherichia coli outer membrane transporter for vitamin B12. Electron paramagnetic resonance (EPR) spectra from a spin label placed within the N-terminal energy coupling motif (Ton box) of BtuB indicate that this segment is in equilibrium between folded and unfolded forms. In bilayers, substrate binding shifts this equilibrium toward the unfolded form; however, EPR spectra from this same spin-labeled mutant indicate that this unfolding transition is blocked in protein crystals. Moreover, crystal structures of this spin-labeled mutant are consistent with the EPR result. When the free energy difference between substates is estimated from the EPR spectra, the crystal environment is found to alter this energy by 3 kcal/mol when compared to the bilayer state. Approximately half of this energy change is due to solutes or osmolytes in the crystallization buffer, and the remainder is contributed by the crystal lattice. These data provide a quantitative measure of how a conformational equilibrium in BtuB is modified in the crystal environment, and suggest that more-compact, less-hydrated substates will be favored in protein crystals. INTRODUCTION Proteins are dynamic and structurally heterogeneous. They exhibit collective and uncoupled motions over a wide range of timescales (1,2) and they may assume numerous discrete structural substates that are in equilibrium. This motion and sampling of structural states is important and appears to be critical to enzyme activity and allosteric regulation (3–5). In protein crystals, dynamics and structural heterogeneity are present, and at sufficiently high resolution, may be well represented by the use of multiple conformations of the side-chain (6) and backbone atoms (7). There are indications that protein crystallization and the conditions used for crystallization may alter protein dy- namics and conformational sampling. Molecular dynamics simulations of proteins in a crystal lattice have been performed for both soluble (8–10), and membrane (11) proteins. These computational efforts suggest that confor- mational sampling may be altered by protein crystallization. In addition, experiments employing precipitants or osmo- lytes similar to those used in protein crystallization demon- strate that these solutes may have a significant effect on exchange between long-lived conformational substates; for example, osmolytes have been found to alter conformational substates involved in enzymatic activity (12,13) and ion conduction (14). In BtuB, the outer membrane Escherichia coli vitamin B12 (CNCbl) transporter, an electron paramagnetic reso- nance (EPR)-based method termed site-directed spin labeling (SDSL), has been used to investigate the dynamics and structural transitions in an N-terminal energy coupling segment termed the Ton box (15). The Ton box couples BtuB to the inner membrane protein TonB, which provides energy for transport (16–18). SDSL provides strong evidence that the Ton box undergoes a vitamin B12-depen- dent unfolding (19,20), as depicted in Fig. 1. This event moves the Ton box 20–30 A˚ into the periplasmic space, where it may act as a trigger to initiate BtuB-TonB interac- tions (21). In contrast, the Ton box remains folded within the transporter in crystal structures of BtuB either in the pres- ence or absence of substrate. While there are small shifts in the conformation of the Ton box upon substrate binding, no evidence is seen for the substrate-dependent unfolding observed spectroscopically (22). The discrepancy between the spectroscopic and crystallo- graphic result might have several origins. EPR spectroscopy of membrane-associated BtuB demonstrates that there is an equilibrium between folded and unfolded substates of the Ton box, and that this equilibrium is shifted toward the more folded state by the osmolytes used in the BtuB crystal- lization (23–25). Osmolytes, such as polyethylene glycols (PEGs), are believed to be excluded from hydrated protein surfaces (26,27), thereby raising the energy of the protein and reducing its solubility. As a result, the presence of os- molytes will favor conformational substates that are less hydrated (12,28,29). The packing of the protein within the crystal lattice might also account for the difference between the spectroscopic and crystallographic result. Although protein-protein contacts within the unit cell should not interfere sterically with the unfolding of the Ton box, the contributions that the lattice might make to the Ton box equilibrium are not known. Submitted April 26, 2010, and accepted for publication June 14, 2010. *Correspondence: cafiso@virginia.edu Editor: David D. Thomas.  2010 by the Biophysical Society 0006-3495/10/09/1604/7 $2.00 doi: 10.1016/j.bpj.2010.06.026 1604 Biophysical Journal Volume 99 September 2010 1604–1610 To determine how the Ton box equilibrium is modified within the protein crystal compared to the bilayer state, we generated a spin-labeled mutant of BtuB where the nitro- xide side chain R1 (Fig. 1 c) is incorporated into the Ton box at position 10. EPR spectroscopy was then carried out in parallel with x-ray diffraction and structure determination of this labeled BtuB mutant (V10R1) using the same protein crystals. The EPR spectra obtained from the protein crystal indicate that the substrate-dependent Ton box transition is blocked. This spectroscopic result is consistent with crystal structures of this BtuB mutant, which indicate that the Ton box remains folded and the R1 side chain buried with or without substrate. By comparing EPR spectra of BtuB- V10R1 in bilayers with spectra from the protein crystal, we estimate the free energy change induced by the crystal environment on this conformational equilibrium, and dissect the energetic contributions made by solute and the crystal lattice. MATERIALS AND METHODS Mutagenesis, expression, and purification The V10C mutation was introduced into btuB using a QuikChange site- directed mutagenesis kit (Stratagene, La Jolla, CA), and was subsequently verified by nucleotide sequencing. Expression and purification of BtuB for the formation of protein crystals was performed as described previously (22), and BtuB was reconstituted into vesicle bilayers by following a procedure described elsewhere (30). Spin labeling For spin labeling, the first round of purification was paused before initiation of the salt gradient. The Q-Sepharose slurry (Amersham, Piscataway, NJ) bound with BtuB was transferred to a conical tube and reacted with 1 mL of 45 mM S-(1-oxy-2,2,5,6-tetramethylpyrroline-3-methyl)methane- thiosulfonate (MTSL; Toronto Research Chemicals, Ontario, Canada) for 4 h at room temperature. Crystallization and crystallographic data collection Purified BtuB (11 mg/mL in 30 mM Tris pH 8.0, 20 mM C8E4) was crys- tallized by mixing 1 mL of BtuB and 1 mL of reservoir buffer in an EasyXtal hanging-drop tray (Qiagen, Germantown, MD), containing 200 mL of total reservoir buffer for each crystallization condition, and followed by incuba- tion at 290 K. The reservoir buffer consisted of 200–550 mM magnesium acetate, 5.0–7.5% PEG3350, and 20 mM Bis Tris at pH 6.6. Crystals were visible after 1–2 days, and grew to ~200 mm in the longest dimension after 1–2 weeks. For crystals to be incubated with substrate, 1 mL of soaking buffer (150 mM calcium chloride, 2.5% PEG3350, 20 mM Bis Tris at pH 6.6, and 10 mM C8E4) was added to each well, followed by incubation overnight. The crystals were subsequently transferred into soaking buffer containing 1 mM cyanocobalamin and 20% glycerol, and allowed to incu- bate for at least 4 h. For x-ray diffraction, apo and Ca2þB12-soaked crystals were transferred to cryo-buffer (150 mM magnesium acetate or calcium chloride, 2.5% PEG3350, 20 mM Bis Tris at pH 6.6, 10 mM C8E4, and 20% glycerol) for 1–2 min before loop mounting and cryocooling by inser- tion into liquid nitrogen. Diffraction data were taken at 90 K at the 22ID beamline at the Advanced Photon Source (Argonne National Laboratory, Argonne, IL). See Table 1 for more details. Structure determination Indexing, integration, and scaling of the diffraction data was performed using HKL2000 (31). The structures were solved with PHASER (32) maximum likelihood molecular replacement method, using PDB deposi- tions 1NQE and 1NQH (22) as search models for the apo- and Ca2þB12- bound data, respectively. To reduce model bias, V10 was deleted from the apo-search model, and the entire Ton box was deleted from the Ca2þ- CNCbl-bound search model. Model building was done in COOT (33), and unrestrained TLS (34) refinement was performed using REFMAC (35) and PHENIX was used to refine the occupancy of the spin label (36). The spin-labeled residue V10R1 was manually built in COOT. Anom- alous difference Fourier maps were calculated to accurately position bound cobalt and calcium using Sfall and fast-Fourier transform (37). Completed structures were evaluated and validated with MolProbity (38). Electron paramagnetic resonance Apo- and Ca2þ B12-soaked crystals were incubated for at least 4 h in cryobuffer and soaking buffer (with 1 mM cyanocobalamin and 20% glyc- erol), respectively. Crystals were then transferred to a 0.60 ID  0.84 OD round capillary with a syringe (Hamilton Syringe, Bonaduz, Switzerland) for EPR spectroscopy, which was performed on an X-band EMX spectrom- eter (Bruker Biospin, Billerica, MA) equipped with a dielectric resonator. All EPR spectra were recorded with a 100 G magnetic field sweep at 2.0 mW incident power at a temperature of 298 K. The phasing, normaliza- tion, and subtraction of EPR spectra was performed using LabVIEW soft- ware provided by Dr. Christian Altenbach (University of California, Los Angeles, California). FIGURE 1 BtuB in the (a) apo form where the Ton box position is highlighted (PDB ID: 1NQE). (b) Vitamin B12 bound form of BtuB showing the state of the Ton box as determined by EPR spectra and pulse EPR distance measurements (based upon PDB ID 1NQH and spectroscopic restraints obtained for the Ton box in bilayers (21)). This unfolding event places the Ton box as much as 30 A˚ into the periplasmic space. (c) The structure of the spin-labeled R1 side chain and dihedral angles that define the rotamers of R1. Biophysical Journal 99(5) 1604–1610 Membrane Protein Conformational Exchange 1605 To determine the free energies and free energy changes between Ton box states, the population of each Ton box conformation was determined by spectral subtraction and quantitation of the spectral components as described previously (24). For BtuB V10R1, the EPR spectra are linear combinations of the spectra resulting from the folded and unfolded Ton box conformations. As a result, the fraction of spins in each population may be estimated by determining the contribution that each conformation makes to the total spectrum. The label at position 10 was chosen for these measurements, because EPR spectra for V10R1 yield dramatically different lineshapes for the folded and unfolded forms of the Ton box. As a result, it is easy to simulate both the folded and unfolded lineshape. In this case, the mobile lineshape was simulated (using Redfield theory (39)) and subtracted from the composite spectrum until a spectrum corresponding to the purely folded Ton box conformation was obtained. Double integration of the first derivative EPR spectra yields numbers that are proportional to spin number and was used to estimate the populations of folded and unfolded Ton box. RESULTS The Ton box exhibits a substrate-dependent unfolding in bilayers but not in protein crystals The label at position 10 was chosen for these experiments for two reasons. First, the incorporation of R1 at some sites may perturb the Ton box fold; however, the incorporation of R1 at position 10 does not appear to be highly perturbing (20). Second, the spectra from BtuB-V10R1 are particularly good at revealing different conformational substates of the Ton box, and these states are easily quantitated from the EPR spectra of V10R1. Shown in Fig. 2 a are EPR spectra for BtuB-V10R1 with and without substrate in lipid bilayers composed of POPC. Spectra for BtuB-V10R1 in bilayers have been reported previously (20), and in the absence of substrate the spectrum is dominated by a broad component resulting from an immobile spin-labeled side chain that is near the rigid-limit of nitroxide motion at X-band (tc 30–50 ns). This broad component results from a label that is in strong tertiary contact with other side chains in BtuB. In the presence of substrate, the spectrum changes dramatically and is domi- nated by a narrow high-amplitude component arising from a motionally averaged nitroxide attached to a disordered backbone segment. A careful examination of the EPR line- shapes in Fig. 2 a indicates that in each case (with or without substrate), both immobile and mobile components can be distinguished. These components represent folded and unfolded substates of the Ton box in equilibrium (40), and the populations of these substates may be estimated from the EPR spectra using spectral subtraction (see Methods). This estimate shows that in the presence of substrate TABLE 1 Data collection and refinement for BtuB-V10R1 Structure BtuB-V10R1 apo BtuB-V10R1 þCa2þB12 Data collection Beamline APS-22ID APS-22ID Wavelength (A˚ ) 1.000 1.000 Temperature (K) 90 90 Reflections observed 311,539 294,094 Unique reflections 32,472 32,358 Resolution range (A˚ )* 50–2.40 (2.49–2.40) 50–2.45 (2.54–2.45) Space group P3121 P3121 Cell dimensions a ¼ b ¼ 81.3 A˚ , c ¼ 226.6 A˚ a ¼ b ¼ 82.1 A˚ , c ¼ 224.5 A˚ a ¼ b ¼ 90, g ¼ 120 a ¼ b ¼ 90, g ¼ 120 Rsym (%) 9.1 (38.3) 12.1 (45.8) Redundancy 9.6 9.1 Refinement Resolution range (A˚ ) 44.1–2.44 (2.50–2.44) 44.0–2.44 (2.51–2.44) Reflections used 30,769 30,642 Completeness (%) 97.6 (79.3) 96.6 (67.3) Rcryst (%)y 22.1 22.9 Rfree (%)z 24.8 27.5 Root-mean-square deviations Bond lengths (A˚ ) 0.021 0.019 Bond angles () 1.839 2.037 Number of atoms Protein 4605 4865 Water 113 76 Other C8E4 (7), Mg (4) CNCbl (1), Ca2þ (3), C8E4 (6) PDB accession code 3M8B 3M8D *Highest resolution shell data shown in parentheses. yRcryst ¼ SkFobsj-jFcalck / SjFobsj, where Fobs and Fcalc are the observed and calculated structure factor amplitudes, respectively. zRfree is Rcryst calculated using 5% of the data which is randomly chosen and omitted from the refinement. FIGURE 2 EPR spectra for V10R1 with (red traces) and without (blue traces) substrate when BtuB is incorporated into (a) POPC bilayers, or (b) in the protein crystal. The inset below is a 10 vertical expansion showing a small signal from unfolded Ton box. The dashed vertical lines indicate the positions of signals resulting from immobilized (i) and mobile (m) nitroxide side chain, corresponding to folded and unfolded Ton box, respectively. Biophysical Journal 99(5) 1604–1610 1606 Freed et al. ~50% of the Ton box is unfolded, and the free energy differ- ence (DG) between these two states is approximately zero. Fig. 2 b shows an analogous pair of spectra obtained for V10R1 in protein crystals in cryo buffer (see Methods) with and without substrate. In the protein crystal, each spectrum reflects a nitroxide near the rigid limit of motion at X-band. The substrate-induced transition, which is clearly seen in bilayers (Fig. 2 a), is absent. A careful examination of the EPR spectrum for crystallized BtuB in the presence of vitamin B12 (Fig. 2 b) reveals a very minor mobile component (arrow in Fig. 2). This component matches the lineshape obtained for V10R1 in the unfolded state and appears to represent a small fraction of unfolded Ton box in the presence of substrate. Quantitation of this minor component by spectral subtraction indicates that it repre- sents <0.5% of the total spin signal from V10R1, and that the folded form of the Ton box is stabilized by at least 3 kcal/mol for BtuB bound to substrate in the protein crystal. Because the energy difference between the folded and unfolded states of the Ton box is close to zero in bilayers, the free energy difference between these two protein sub- states is altered (a DDG) by ~3 kcal/mol for BtuB-V10R1 in the protein crystal. Structures from crystals of BtuB-V10R1 show no evidence for a substrate-dependent unfolding Protein crystals of BtuB-V10R1 in the absence and presence of substrate diffracted to 2.4 A˚ and the refinement details are given in Table 1. In both cases, the Ton box is resolved and folded within the protein interior, and several extracellular loops become resolved in presence of ligand, as seen previ- ously for wild-type (wt) BtuB (22). Fig. 3, a and b, displays periplasmic views of BtuB-V10R1, where the position of V10R1 in the protein interior as well as the configuration of the Ton box is shown. The label is sitting at the bottom of a pocket facing the periplasmic surface; and as expected, it is interacting with a number of side chains, including R219 and R255. As a result, conversion between label ro- tamers should be highly restricted, consistent with the rigid limit spectra seen by EPR (Fig. 2 b). The angles for c1 and c2 (Fig. 1 c) for R1 typically assume a limited set of rotameric states on protein surface sites, where the rotamers allow for an interaction between Sd and HCa (41). Here V10R1 is found to have c1 and c2 angles of 56 and 69 in the apo form and 49 and 60 in the CNCbl-bound form, which are both in a {p, p} configu- ration using the conventions of Lovell et al. (42). The entire set of spin-label dihedral angles for V10R1 is given in Table 2. The Sd-HCa distance for the R1 side chain is ~4.5 A˚ , which is longer than that typically seen for R1 on helix surface sites. Although this rotamer is energetically allowed, it has not previously been observed in crystal structures (41), presumably due to the sterically restricted environment surrounding V10R1. Fig. 3 c compares the Ton box for the V10R1 mutant with and without CNCbl. The R1 side chain and the Ton box to which it is attached remain folded into the protein interior upon the addition of substrate, consistent with a lack of change in the EPR spectra shown in Fig. 2 b for FIGURE 3 (a) Periplasmic view of the structure and electron density (1s) showing the placement of the spin-labeled side chain V10R1 and residues that closely interact with the label in the apo form (PDB ID: 3M8B) of BtuB. (Magenta) Backbone of the Ton box. (Beige) N-terminal fold. (b) Periplasmic view of BtuB-V10R1 similar to that shown in panel a, except with van der Waals surfaces rendered for the atoms. The label, V10R1, is at the base of a periplasmic pocket in close tertiary contact with a number of atoms. (c) A comparison of the Ton box of BtuB- V10R1 with and without substrate. A side view of the crystal structure of the Ca2þ-B12 bound form of V10R1 (PDB ID: 3M8D) is shown with B12 bound, and the Ton box (magenta). This structure was aligned with the apo form of BtuB-V10R1 where only the Ton box is rendered (blue). Biophysical Journal 99(5) 1604–1610 Membrane Protein Conformational Exchange 1607 the protein crystal. Substrate addition to BtuB-V10R1 produces a change in the position of residue 7, as seen previ- ously for wt protein (22). However, residue 6, which is resolved in the wt structure, is not resolved for BtuB- V10R1 once substrate is bound. A B-factor analysis of the Ton box backbone and side-chain atoms indicates that when compared to wt BtuB structures, the BtuB-V10R1 has higher B-factors for the Ton box N-terminal to position 10, and a larger difference between apo- and ligand-bound forms. Nonetheless, the fold of the Ton box in BtuB- V10R1 is virtually identical to that seen in the wt structure (the root-mean-square deviation is 1.2 A˚ and 1.5 A˚ for the apo and CNCbl-bound forms of the Ton box, respectively, when V10R1 and wt are compared). Hence, minimal struc- tural changes in the Ton box are induced by this particular label. Both the crystal lattice and solutes shift the equilibrium between Ton box substates To determine whether the crystal lattice makes a contribu- tion to the free energy change when bilayer and crystal forms of BtuB are compared, EPR spectra from V10R1 were compared for the protein crystal and the protein solu- bilized into the cryo buffer. The two spectra for BtuB- V10R1 (in the CNCbl bound form) are compared in Fig. 4, a and b, and are clearly different. In particular, the spectrum from solubilized protein (Fig. 4 b) yields a mobile component with much higher amplitude than that for the protein crystal (Fig. 4 a). This mobile signal has a lineshape identical to that seen for the unfolded state in the bilayer (Fig. 2 a). Quantitation of the two components in this spectrum indicates that the mobile population makes up ~8 5 2% of the total spins. This fraction of unfolded Ton box corresponds to a change in free energy (a DDG for this transition relative to the bilayer reconstituted BtuB) of ~1.5 5 0.2 kcal/mol, indicating that solutes and the crystal lattice make roughly equal contributions to the change in conformational energy that is seen in the protein crystal. The lineshapes for the immobilized component in the absence of substrate for the bilayer reconstituted and crys- tallized BtuB-V10R1 are shown in Fig. 4, c and d, respec- tively. In this case the mobile component was subtracted from the bilayer BtuB-V10R1 (Fig. 2 a) to yield the immo- bile component in Fig. 4 c. Both these lineshapes result from immobile spin labels near the rigid limit of nitroxide motion. However, the hyperfine extrema in Fig. 4 c are not as distinct as in Fig. 4 d, and components representing the g-tensor anisotropy in the central (mI ¼ 0) resonance of BtuB-V10R1 are better resolved in the protein crystal (Fig. 4 d). This difference provides an indication that addi- tional motional modes are available for V10R1 in the bilayer environment. DISCUSSION In this work, SDSL was used to examine a conformational equilibrium in the Escherichia coli outer membrane trans- porter, BtuB, both in lipid bilayers and in protein crystals. The results indicate that the equilibrium between folded and unfolded forms of the Ton box is shifted by ~3 kcal/mol when the protein is taken from the bilayer phase to the protein crystal phase. This has the effect of stabilizing the folded form of the Ton box in the protein crystal, and it provides an expla- nation for the observation that the Ton box is resolved both in the absence and presenceofsubstrate in crystal structures(22), but is seen to unfold in bilayers. It should be emphasized that protein crystallography does not provide an incorrect structure for BtuB. The conditions of the protein crystal alter the equi- librium distribution of substates, compared to the distribution found by EPR, to favor the more-compact-ordered conformer. Osmolytes, such as PEGs, are well known to modify protein behavior (43), and previous work has demonstrated that osmolytes stabilize a folded form of the Ton box (15,24,25) and stabilize more-compact, less-hydrated conformations of the extracellular ligand-binding loops (44). The data obtained here indicate that solutes and the crystal lattice contribute almost equally to the energy change seen in the Ton box equilibrium. While the action of PEGs and other osmolytes is reasonably well understood, FIGURE 4 EPR spectra from BtuB-V10R1 with bound ligand in (a) the protein crystal, (b) in the crystallization buffer at a protein concentration too dilute to form crystals, and in the apo state in (c) lipid bilayers and (d) the protein crystal. The symbols i and m indicate immobilized and mobile components in the spectra for panel b. The spectrum in panel c is identical to the spectrum in Fig. 2 a, except that the small mobile component seen in Fig. 2 a has been subtracted. All spectra are 100 Gauss scans. TABLE 2 Summary of R1 side-chain dihedral angles and rotamer designation Mutant Rotamer c1 c2 c3 c4 c5 V10R1 apo {p,p} 56 69 83 67 67 V10R1 Ca2þ and B12 {p,p} 49 60 70 93 46 Biophysical Journal 99(5) 1604–1610 1608 Freed et al. it is not presently known how the protein lattice in the crystal couples to the Ton box equilibrium and stabilizes its folded form in BtuB. Previous work has shown that there is an interaction between charged residues near the Ton box and the BtuB b-barrel (40), and that eliminating this interac- tion unfolds the Ton box. Conceivably, a change in the dynamics or structure of the BtuB b-barrel when the protein is in the crystal lattice might alter the energy of this interac- tion and account for the effect of the lattice upon the Ton box. The structural biology of membrane proteins is far from mature, and it is not known whether the effects seen here on protein conformational sampling apply to a wider range of membrane proteins. Bacteriorhodopsin is perhaps the best-studied membrane protein, and there are indications that the kinetics of the photocycle are modified in the three-dimensional crystalline lattice when compared to the native membrane (45). Solution NMR can provide high- resolution structural data on membrane proteins in micelles, allowing comparisons to be made with crystal structures. NMR spectroscopy is often found to resolve portions of proteins that are not resolved by crystallography (as seen for DsbB (46)), presumably because NMR is better at exam- ining structures that are inherently dynamic. In outer membrane porins, such as OmpA, the strands of the b-barrel are shorter in the NMR-derived structures than in the crystal structures (47); however, it is not clear whether this differ- ence is a result of crystallization conditions or the micellar environment used for NMR. Even in a reconstituted bilayer environment (which is a much better approximation to the native environment than the protein crystal), protein confor- mational sampling may be modulated relative to the native environment (48–51); however, many of these effects appear to be due to the fraction of acidic lipids selected for the reconstitution, which in turn control local ion concentrations and pH. As for BtuB, the Ton box equilibrium does not appear to be modulated by lipid composition (Q. Xu and D. S. Cafiso, unpublished); this equilibrium is maintained within range of reconstituted bilayers as well as intact outer membrane preparations (19). Changes in the equilibrium distribution of protein confor- mational substates are thought to underlie protein signaling events (52) and allostery (53). Because of its fast timescale, EPR spectroscopy is particularly well suited to detect these conformational substates and to measure conformational equilibria in proteins. In BtuB, SDSL demonstrates that both the folded and unfolded states of the Ton box are sampled and that substrate binding shifts the equilibrium to the more disordered state. Furthermore, colicin E3, which is also a ligand for BtuB, shifts the Ton box equilibrium to favor the folded, ordered state of the Ton box (54). These are precisely the types of changes that are proposed to underlie protein signaling, and in the present case, they may function to regulate coupling between BtuB and inner membrane protein TonB. We thank Dr. Robert Nakamoto for careful reading of this manuscript. This work was supported by National Institutes of Health grants Nos. NIGMS 035215 to D.S.C. and NIGMS 079800 to M.C.W. Use of the Advanced Photon Source was supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under contract No. DE- AC02-06CH11357. Data were collected at Southeast Regional Collabora- tive Access Team (SER-CAT) 22-ID beamline at the Advanced Photon Source, Argon National Laboratory, Argonne, IL. Supporting institutions may be found at http://www.ser-cat.org/members.html. REFERENCES 1. Frauenfelder, H., S. G. Sligar, and P. G. Wolynes. 1991. The energy landscapes and motions of proteins. Science. 254:1598–1603. 2. Henzler-Wildman, K., and D. Kern. 2007. Dynamic personalities of proteins. Nature. 450:964–972. 3. Henzler-Wildman, K. A., M. Lei, ., D. Kern. 2007. A hierarchy of timescales in protein dynamics is linked to enzyme catalysis. Nature. 450:913–916. 4. Bahar, I., C. Chennubhotla, and D. Tobi. 2007. Intrinsic dynamics of enzymes in the unbound state and relation to allosteric regulation. Curr. Opin. Struct. Biol. 17:633–640. 5. Swain, J. F., and L. M. Gierasch. 2006. The changing landscape of protein allostery. Curr. Opin. Struct. Biol. 16:102–108. 6. Dunbrack, Jr., R. L. 2002. Rotamer libraries in the 21st century. Curr. Opin. Struct. Biol. 12:431–440. 7. Davis, I. W., W. B. Arendall, 3rd, ., J. S. Richardson. 2006. The backrub motion: how protein backbone shrugs when a sidechain dances. Structure. 14:265–274. 8. Walser, R., P. H. Hu¨nenberger, and W. F. van Gunsteren. 2002. Molecular dynamics simulations of a double unit cell in a protein crystal: volume relaxation at constant pressure and correlation of motions between the two unit cells. Proteins. 48:327–340. 9. Walser, R., P. H. Hu¨nenberger, and W. F. van Gunsteren. 2001. Comparison of different schemes to treat long-range electrostatic inter- actions in molecular dynamics simulations of a protein crystal. Proteins. 43:509–519. 10. Meinhold, L., and J. C. Smith. 2005. Fluctuations and correlations in crystalline protein dynamics: a simulation analysis of staphylococcal nuclease. Biophys. J. 88:2554–2563. 11. Bond, P. J., J. D. Faraldo-Go´mez, ., M. S. Sansom. 2006. Membrane protein dynamics and detergent interactions within a crystal: a simula- tion study of OmpA. Proc. Natl. Acad. Sci. USA. 103:9518–9523. 12. Colombo, M. F., D. C. Rau, and V. A. Parsegian. 1992. Protein solvation in allosteric regulation: a water effect on hemoglobin. Science. 256:1335–1336. 13. Parsegian, V. A., R. P. Rand, and D. C. Rau. 1995. Macromolecules and water: probing with osmotic stress. Methods Enzymol. 259:43–94. 14. Vodyanoy, I., S. M. Bezrukov, and V. A. Parsegian. 1993. Probing alamethicin channels with water-soluble polymers. Size-modulated osmotic action. Biophys. J. 65:2097–2105. 15. Kim, M., G. E. Fanucci, and D. S. Cafiso. 2007. Substrate-dependent transmembrane signaling in TonB-dependent transporters is not conserved. Proc. Natl. Acad. Sci. USA. 104:11975–11980. 16. Postle, K., and R. J. Kadner. 2003. Touch and go: tying TonB to trans- port. Mol. Microbiol. 49:869–882. 17. Wiener, M. C. 2005. TonB-dependent outer membrane transport: going for Baroque? Curr. Opin. Struct. Biol. 15:394–400. 18. Schauer, K., D. A. Rodionov, and H. de Reuse. 2008. New substrates for TonB-dependent transport: do we only see the ‘tip of the iceberg’? Trends Biochem. Sci. 33:330–338. 19. Merianos, H. J., N. Cadieux, ., D. S. Cafiso. 2000. Substrate-induced exposure of an energy-coupling motif of a membrane transporter. Nat. Struct. Biol. 7:205–209. Biophysical Journal 99(5) 1604–1610 Membrane Protein Conformational Exchange 1609 20. Fanucci, G. E., K. A. Coggshall, ., D. S. Cafiso. 2003. Substrate- induced conformational changes of the periplasmic N-terminus of an outer-membrane transporter by site-directed spin labeling. Biochem- istry. 42:1391–1400. 21. Xu, Q., J. F. Ellena, ., D. S. Cafiso. 2006. Substrate-dependent unfold- ing of the energy coupling motif of a membrane transport protein determined by double electron-electron resonance. Biochemistry. 45:10847–10854. 22. Chimento, D. P., A. K. Mohanty, ., M. C. Wiener. 2003. Substrate- induced transmembrane signaling in the cobalamin transporter BtuB. Nat. Struct. Biol. 10:394–401. 23. Fanucci, G. E., J. Y. Lee, and D. S. Cafiso. 2003. Spectroscopic evidence that osmolytes used in crystallization buffers inhibit a confor- mation change in a membrane protein. Biochemistry. 42:13106–13112. 24. Kim, M., Q. Xu, ., D. S. Cafiso. 2006. Solutes modify a conforma- tional transition in a membrane transport protein. Biophys. J. 90: 2922–2929. 25. Flores Jime´nez, R. H., M. A. Do Cao, ., D. S. Cafiso. 2010. Osmolytes modulate conformational exchange in solvent-exposed regions of membrane proteins. Protein Sci. 19:269–278. 26. Arakawa, T., and S. N. Timasheff. 1985. The stabilization of proteins by osmolytes. Biophys. J. 47:411–414. 27. Arakawa, T., and S. N. Timasheff. 1985. Mechanism of poly(ethylene glycol) interaction with proteins. Biochemistry. 24:6756–6762. 28. Timasheff, S. N. 2002. Protein hydration, thermodynamic binding, and preferential hydration. Biochemistry. 41:13473–13482. 29. Zimmerberg, J., F. Bezanilla, and V. A. Parsegian. 1990. Solute inac- cessible aqueous volume changes during opening of the potassium channel of the squid giant axon. Biophys. J. 57:1049–1064. 30. Fanucci, G. E., N. Cadieux, ., D. S. Cafiso. 2002. Structure and dynamics of the b-barrel of the membrane transporter BtuB by site- directed spin labeling. Biochemistry. 41:11543–11551. 31. Otwinowski, Z., and W. Minor. 1997. Processing of x-ray diffraction data collected in oscillation mode. Methods Enzymol. 276:307–326. 32. McCoy, A. J., R. W. Grosse-Kunstleve, ., R. J. Read. 2007. PHASER crystallographic software. J. Appl. Cryst. 40:658–674. 33. Emsley, P., and K. Cowtan. 2004. COOT: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 60: 2126–2132. 34. Adams, P. D., R. W. Grosse-Kunstleve, ., T. C. Terwilliger. 2002. PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr. D Biol. Crystallogr. 58:1948–1954. 35. Murshudov, G. N., A. A. Vagin, and E. J. Dodson. 1997. Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr. D Biol. Crystallogr. 53:240–255. 36. Winn, M. D., M. N. Isupov, and G. N. Murshudov. 2001. Use of TLS parameters to model anisotropic displacements in macromolecular refinement. Acta Crystallogr. D Biol. Crystallogr. 57:122–133. 37. Potterton, E., P. Briggs, ., E. Dodson. 2003. A graphical user interface to the CCP4 program suite. Acta Crystallogr. D Biol. Crystallogr. 59:1131–1137. 38. Davis, I. W., A. Leaver-Fay, ., D. C. Richardson. 2007. MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res. 35(Web Server issue):W375–W383. 39. Stone, T. J., T. Buckman, ., H. M. McConnell. 1965. Spin-labeled biomolecules. Proc. Natl. Acad. Sci. USA. 54:1010–1017. 40. Lukasik, S. M., K. W. Ho, and D. S. Cafiso. 2007. Molecular basis for substrate-dependent transmembrane signaling in an outer-membrane transporter. J. Mol. Biol. 370:807–811. 41. Guo, Z., D. Cascio, ., W. L. Hubbell. 2008. Structural determinants of nitroxide motion in spin-labeled proteins: solvent-exposed sites in helix B of T4 lysozyme. Protein Sci. 17:228–239. 42. Lovell, S. C., J. M. Word, ., D. C. Richardson. 2000. The penultimate rotamer library. Proteins. 40:389–408. 43. Auton, M., D. W. Bolen, and J. Ro¨sgen. 2008. Structural thermody- namics of protein preferential solvation: osmolyte solvation of proteins, amino acids, and peptides. Proteins. 73:802–813. 44. Kim, M., Q. Xu, ., D. S. Cafiso. 2008. Solutes alter the conformation of the ligand binding loops in outer membrane transporters. Biochem- istry. 47:670–679. 45. Efremov, R., V. I. Gordeliy, ., G. Bu¨ldt. 2006. Time-resolved microspectroscopy on a single crystal of bacteriorhodopsin reveals lattice-induced differences in the photocycle kinetics. Biophys. J. 91: 1441–1451. 46. Zhou, Y., T. Cierpicki, ., J. H. Bushweller. 2008. NMR solution structure of the integral membrane enzyme DsbB: functional insights into DsbB-catalyzed disulfide bond formation. Mol. Cell. 31:896–908. 47. Tamm, L. K., F. Abildgaard, ., J. H. Bushweller. 2003. Structure, dynamics and function of the outer membrane protein A (OmpA) and influenza hemagglutinin fusion domain in detergent micelles by solution NMR. FEBS Lett. 555:139–143. 48. Brown, M. F. 1994. Modulation of rhodopsin function by properties of the membrane bilayer. Chem. Phys. Lipids. 73:159–180. 49. Turnheim, K., J. Gruber, ., V. Ruiz-Gutie´rrez. 1999. Membrane phos- pholipid composition affects function of potassium channels from rabbit colon epithelium. Am. J. Physiol. 277:C83–C90. 50. Baenziger, J. E., S. E. Ryan, ., C. J. daCosta. 2008. Lipid composition alters drug action at the nicotinic acetylcholine receptor. Mol. Pharma- col. 73:880–890. 51. Charalambous, K., D. Miller, ., P. J. Booth. 2008. Lipid bilayer composition influences small multidrug transporters. BMC Biochem. 9:31. 52. Smock, R. G., and L. M. Gierasch. 2009. Sending signals dynamically. Science. 324:198–203. 53. Hilser, V. J. 2010. Biochemistry. An ensemble view of allostery. Science. 327:653–654. 54. Fanucci, G. E., N. Cadieux, ., D. S. Cafiso. 2003. Competing ligands stabilize alternate conformations of the energy coupling motif of a TonB-dependent outer membrane transporter. Proc. Natl. Acad. Sci. USA. 100:11382–11387. Biophysical Journal 99(5) 1604–1610 1610 Freed et al.
3M8D
Crystal structure of spin-labeled BtuB V10R1 with bound calcium and cyanocobalamin
Conformational Exchange in a Membrane Transport Protein Is Altered in Protein Crystals Daniel M. Freed,† Peter S. Horanyi,‡ Michael C. Wiener,‡ and David S. Cafiso†‡* †Departments of Chemistry and ‡Molecular Physiology and Biological Physics, and the Biophysics Program, University of Virginia, Charlottesville, Virginia ABSTRACT Successful macromolecular crystallography requires solution conditions that may alter the conformational sampling of a macromolecule. Here, site-directed spin labeling is used to examine a conformational equilibrium within BtuB, the Escherichia coli outer membrane transporter for vitamin B12. Electron paramagnetic resonance (EPR) spectra from a spin label placed within the N-terminal energy coupling motif (Ton box) of BtuB indicate that this segment is in equilibrium between folded and unfolded forms. In bilayers, substrate binding shifts this equilibrium toward the unfolded form; however, EPR spectra from this same spin-labeled mutant indicate that this unfolding transition is blocked in protein crystals. Moreover, crystal structures of this spin-labeled mutant are consistent with the EPR result. When the free energy difference between substates is estimated from the EPR spectra, the crystal environment is found to alter this energy by 3 kcal/mol when compared to the bilayer state. Approximately half of this energy change is due to solutes or osmolytes in the crystallization buffer, and the remainder is contributed by the crystal lattice. These data provide a quantitative measure of how a conformational equilibrium in BtuB is modified in the crystal environment, and suggest that more-compact, less-hydrated substates will be favored in protein crystals. INTRODUCTION Proteins are dynamic and structurally heterogeneous. They exhibit collective and uncoupled motions over a wide range of timescales (1,2) and they may assume numerous discrete structural substates that are in equilibrium. This motion and sampling of structural states is important and appears to be critical to enzyme activity and allosteric regulation (3–5). In protein crystals, dynamics and structural heterogeneity are present, and at sufficiently high resolution, may be well represented by the use of multiple conformations of the side-chain (6) and backbone atoms (7). There are indications that protein crystallization and the conditions used for crystallization may alter protein dy- namics and conformational sampling. Molecular dynamics simulations of proteins in a crystal lattice have been performed for both soluble (8–10), and membrane (11) proteins. These computational efforts suggest that confor- mational sampling may be altered by protein crystallization. In addition, experiments employing precipitants or osmo- lytes similar to those used in protein crystallization demon- strate that these solutes may have a significant effect on exchange between long-lived conformational substates; for example, osmolytes have been found to alter conformational substates involved in enzymatic activity (12,13) and ion conduction (14). In BtuB, the outer membrane Escherichia coli vitamin B12 (CNCbl) transporter, an electron paramagnetic reso- nance (EPR)-based method termed site-directed spin labeling (SDSL), has been used to investigate the dynamics and structural transitions in an N-terminal energy coupling segment termed the Ton box (15). The Ton box couples BtuB to the inner membrane protein TonB, which provides energy for transport (16–18). SDSL provides strong evidence that the Ton box undergoes a vitamin B12-depen- dent unfolding (19,20), as depicted in Fig. 1. This event moves the Ton box 20–30 A˚ into the periplasmic space, where it may act as a trigger to initiate BtuB-TonB interac- tions (21). In contrast, the Ton box remains folded within the transporter in crystal structures of BtuB either in the pres- ence or absence of substrate. While there are small shifts in the conformation of the Ton box upon substrate binding, no evidence is seen for the substrate-dependent unfolding observed spectroscopically (22). The discrepancy between the spectroscopic and crystallo- graphic result might have several origins. EPR spectroscopy of membrane-associated BtuB demonstrates that there is an equilibrium between folded and unfolded substates of the Ton box, and that this equilibrium is shifted toward the more folded state by the osmolytes used in the BtuB crystal- lization (23–25). Osmolytes, such as polyethylene glycols (PEGs), are believed to be excluded from hydrated protein surfaces (26,27), thereby raising the energy of the protein and reducing its solubility. As a result, the presence of os- molytes will favor conformational substates that are less hydrated (12,28,29). The packing of the protein within the crystal lattice might also account for the difference between the spectroscopic and crystallographic result. Although protein-protein contacts within the unit cell should not interfere sterically with the unfolding of the Ton box, the contributions that the lattice might make to the Ton box equilibrium are not known. Submitted April 26, 2010, and accepted for publication June 14, 2010. *Correspondence: cafiso@virginia.edu Editor: David D. Thomas.  2010 by the Biophysical Society 0006-3495/10/09/1604/7 $2.00 doi: 10.1016/j.bpj.2010.06.026 1604 Biophysical Journal Volume 99 September 2010 1604–1610 To determine how the Ton box equilibrium is modified within the protein crystal compared to the bilayer state, we generated a spin-labeled mutant of BtuB where the nitro- xide side chain R1 (Fig. 1 c) is incorporated into the Ton box at position 10. EPR spectroscopy was then carried out in parallel with x-ray diffraction and structure determination of this labeled BtuB mutant (V10R1) using the same protein crystals. The EPR spectra obtained from the protein crystal indicate that the substrate-dependent Ton box transition is blocked. This spectroscopic result is consistent with crystal structures of this BtuB mutant, which indicate that the Ton box remains folded and the R1 side chain buried with or without substrate. By comparing EPR spectra of BtuB- V10R1 in bilayers with spectra from the protein crystal, we estimate the free energy change induced by the crystal environment on this conformational equilibrium, and dissect the energetic contributions made by solute and the crystal lattice. MATERIALS AND METHODS Mutagenesis, expression, and purification The V10C mutation was introduced into btuB using a QuikChange site- directed mutagenesis kit (Stratagene, La Jolla, CA), and was subsequently verified by nucleotide sequencing. Expression and purification of BtuB for the formation of protein crystals was performed as described previously (22), and BtuB was reconstituted into vesicle bilayers by following a procedure described elsewhere (30). Spin labeling For spin labeling, the first round of purification was paused before initiation of the salt gradient. The Q-Sepharose slurry (Amersham, Piscataway, NJ) bound with BtuB was transferred to a conical tube and reacted with 1 mL of 45 mM S-(1-oxy-2,2,5,6-tetramethylpyrroline-3-methyl)methane- thiosulfonate (MTSL; Toronto Research Chemicals, Ontario, Canada) for 4 h at room temperature. Crystallization and crystallographic data collection Purified BtuB (11 mg/mL in 30 mM Tris pH 8.0, 20 mM C8E4) was crys- tallized by mixing 1 mL of BtuB and 1 mL of reservoir buffer in an EasyXtal hanging-drop tray (Qiagen, Germantown, MD), containing 200 mL of total reservoir buffer for each crystallization condition, and followed by incuba- tion at 290 K. The reservoir buffer consisted of 200–550 mM magnesium acetate, 5.0–7.5% PEG3350, and 20 mM Bis Tris at pH 6.6. Crystals were visible after 1–2 days, and grew to ~200 mm in the longest dimension after 1–2 weeks. For crystals to be incubated with substrate, 1 mL of soaking buffer (150 mM calcium chloride, 2.5% PEG3350, 20 mM Bis Tris at pH 6.6, and 10 mM C8E4) was added to each well, followed by incubation overnight. The crystals were subsequently transferred into soaking buffer containing 1 mM cyanocobalamin and 20% glycerol, and allowed to incu- bate for at least 4 h. For x-ray diffraction, apo and Ca2þB12-soaked crystals were transferred to cryo-buffer (150 mM magnesium acetate or calcium chloride, 2.5% PEG3350, 20 mM Bis Tris at pH 6.6, 10 mM C8E4, and 20% glycerol) for 1–2 min before loop mounting and cryocooling by inser- tion into liquid nitrogen. Diffraction data were taken at 90 K at the 22ID beamline at the Advanced Photon Source (Argonne National Laboratory, Argonne, IL). See Table 1 for more details. Structure determination Indexing, integration, and scaling of the diffraction data was performed using HKL2000 (31). The structures were solved with PHASER (32) maximum likelihood molecular replacement method, using PDB deposi- tions 1NQE and 1NQH (22) as search models for the apo- and Ca2þB12- bound data, respectively. To reduce model bias, V10 was deleted from the apo-search model, and the entire Ton box was deleted from the Ca2þ- CNCbl-bound search model. Model building was done in COOT (33), and unrestrained TLS (34) refinement was performed using REFMAC (35) and PHENIX was used to refine the occupancy of the spin label (36). The spin-labeled residue V10R1 was manually built in COOT. Anom- alous difference Fourier maps were calculated to accurately position bound cobalt and calcium using Sfall and fast-Fourier transform (37). Completed structures were evaluated and validated with MolProbity (38). Electron paramagnetic resonance Apo- and Ca2þ B12-soaked crystals were incubated for at least 4 h in cryobuffer and soaking buffer (with 1 mM cyanocobalamin and 20% glyc- erol), respectively. Crystals were then transferred to a 0.60 ID  0.84 OD round capillary with a syringe (Hamilton Syringe, Bonaduz, Switzerland) for EPR spectroscopy, which was performed on an X-band EMX spectrom- eter (Bruker Biospin, Billerica, MA) equipped with a dielectric resonator. All EPR spectra were recorded with a 100 G magnetic field sweep at 2.0 mW incident power at a temperature of 298 K. The phasing, normaliza- tion, and subtraction of EPR spectra was performed using LabVIEW soft- ware provided by Dr. Christian Altenbach (University of California, Los Angeles, California). FIGURE 1 BtuB in the (a) apo form where the Ton box position is highlighted (PDB ID: 1NQE). (b) Vitamin B12 bound form of BtuB showing the state of the Ton box as determined by EPR spectra and pulse EPR distance measurements (based upon PDB ID 1NQH and spectroscopic restraints obtained for the Ton box in bilayers (21)). This unfolding event places the Ton box as much as 30 A˚ into the periplasmic space. (c) The structure of the spin-labeled R1 side chain and dihedral angles that define the rotamers of R1. Biophysical Journal 99(5) 1604–1610 Membrane Protein Conformational Exchange 1605 To determine the free energies and free energy changes between Ton box states, the population of each Ton box conformation was determined by spectral subtraction and quantitation of the spectral components as described previously (24). For BtuB V10R1, the EPR spectra are linear combinations of the spectra resulting from the folded and unfolded Ton box conformations. As a result, the fraction of spins in each population may be estimated by determining the contribution that each conformation makes to the total spectrum. The label at position 10 was chosen for these measurements, because EPR spectra for V10R1 yield dramatically different lineshapes for the folded and unfolded forms of the Ton box. As a result, it is easy to simulate both the folded and unfolded lineshape. In this case, the mobile lineshape was simulated (using Redfield theory (39)) and subtracted from the composite spectrum until a spectrum corresponding to the purely folded Ton box conformation was obtained. Double integration of the first derivative EPR spectra yields numbers that are proportional to spin number and was used to estimate the populations of folded and unfolded Ton box. RESULTS The Ton box exhibits a substrate-dependent unfolding in bilayers but not in protein crystals The label at position 10 was chosen for these experiments for two reasons. First, the incorporation of R1 at some sites may perturb the Ton box fold; however, the incorporation of R1 at position 10 does not appear to be highly perturbing (20). Second, the spectra from BtuB-V10R1 are particularly good at revealing different conformational substates of the Ton box, and these states are easily quantitated from the EPR spectra of V10R1. Shown in Fig. 2 a are EPR spectra for BtuB-V10R1 with and without substrate in lipid bilayers composed of POPC. Spectra for BtuB-V10R1 in bilayers have been reported previously (20), and in the absence of substrate the spectrum is dominated by a broad component resulting from an immobile spin-labeled side chain that is near the rigid-limit of nitroxide motion at X-band (tc 30–50 ns). This broad component results from a label that is in strong tertiary contact with other side chains in BtuB. In the presence of substrate, the spectrum changes dramatically and is domi- nated by a narrow high-amplitude component arising from a motionally averaged nitroxide attached to a disordered backbone segment. A careful examination of the EPR line- shapes in Fig. 2 a indicates that in each case (with or without substrate), both immobile and mobile components can be distinguished. These components represent folded and unfolded substates of the Ton box in equilibrium (40), and the populations of these substates may be estimated from the EPR spectra using spectral subtraction (see Methods). This estimate shows that in the presence of substrate TABLE 1 Data collection and refinement for BtuB-V10R1 Structure BtuB-V10R1 apo BtuB-V10R1 þCa2þB12 Data collection Beamline APS-22ID APS-22ID Wavelength (A˚ ) 1.000 1.000 Temperature (K) 90 90 Reflections observed 311,539 294,094 Unique reflections 32,472 32,358 Resolution range (A˚ )* 50–2.40 (2.49–2.40) 50–2.45 (2.54–2.45) Space group P3121 P3121 Cell dimensions a ¼ b ¼ 81.3 A˚ , c ¼ 226.6 A˚ a ¼ b ¼ 82.1 A˚ , c ¼ 224.5 A˚ a ¼ b ¼ 90, g ¼ 120 a ¼ b ¼ 90, g ¼ 120 Rsym (%) 9.1 (38.3) 12.1 (45.8) Redundancy 9.6 9.1 Refinement Resolution range (A˚ ) 44.1–2.44 (2.50–2.44) 44.0–2.44 (2.51–2.44) Reflections used 30,769 30,642 Completeness (%) 97.6 (79.3) 96.6 (67.3) Rcryst (%)y 22.1 22.9 Rfree (%)z 24.8 27.5 Root-mean-square deviations Bond lengths (A˚ ) 0.021 0.019 Bond angles () 1.839 2.037 Number of atoms Protein 4605 4865 Water 113 76 Other C8E4 (7), Mg (4) CNCbl (1), Ca2þ (3), C8E4 (6) PDB accession code 3M8B 3M8D *Highest resolution shell data shown in parentheses. yRcryst ¼ SkFobsj-jFcalck / SjFobsj, where Fobs and Fcalc are the observed and calculated structure factor amplitudes, respectively. zRfree is Rcryst calculated using 5% of the data which is randomly chosen and omitted from the refinement. FIGURE 2 EPR spectra for V10R1 with (red traces) and without (blue traces) substrate when BtuB is incorporated into (a) POPC bilayers, or (b) in the protein crystal. The inset below is a 10 vertical expansion showing a small signal from unfolded Ton box. The dashed vertical lines indicate the positions of signals resulting from immobilized (i) and mobile (m) nitroxide side chain, corresponding to folded and unfolded Ton box, respectively. Biophysical Journal 99(5) 1604–1610 1606 Freed et al. ~50% of the Ton box is unfolded, and the free energy differ- ence (DG) between these two states is approximately zero. Fig. 2 b shows an analogous pair of spectra obtained for V10R1 in protein crystals in cryo buffer (see Methods) with and without substrate. In the protein crystal, each spectrum reflects a nitroxide near the rigid limit of motion at X-band. The substrate-induced transition, which is clearly seen in bilayers (Fig. 2 a), is absent. A careful examination of the EPR spectrum for crystallized BtuB in the presence of vitamin B12 (Fig. 2 b) reveals a very minor mobile component (arrow in Fig. 2). This component matches the lineshape obtained for V10R1 in the unfolded state and appears to represent a small fraction of unfolded Ton box in the presence of substrate. Quantitation of this minor component by spectral subtraction indicates that it repre- sents <0.5% of the total spin signal from V10R1, and that the folded form of the Ton box is stabilized by at least 3 kcal/mol for BtuB bound to substrate in the protein crystal. Because the energy difference between the folded and unfolded states of the Ton box is close to zero in bilayers, the free energy difference between these two protein sub- states is altered (a DDG) by ~3 kcal/mol for BtuB-V10R1 in the protein crystal. Structures from crystals of BtuB-V10R1 show no evidence for a substrate-dependent unfolding Protein crystals of BtuB-V10R1 in the absence and presence of substrate diffracted to 2.4 A˚ and the refinement details are given in Table 1. In both cases, the Ton box is resolved and folded within the protein interior, and several extracellular loops become resolved in presence of ligand, as seen previ- ously for wild-type (wt) BtuB (22). Fig. 3, a and b, displays periplasmic views of BtuB-V10R1, where the position of V10R1 in the protein interior as well as the configuration of the Ton box is shown. The label is sitting at the bottom of a pocket facing the periplasmic surface; and as expected, it is interacting with a number of side chains, including R219 and R255. As a result, conversion between label ro- tamers should be highly restricted, consistent with the rigid limit spectra seen by EPR (Fig. 2 b). The angles for c1 and c2 (Fig. 1 c) for R1 typically assume a limited set of rotameric states on protein surface sites, where the rotamers allow for an interaction between Sd and HCa (41). Here V10R1 is found to have c1 and c2 angles of 56 and 69 in the apo form and 49 and 60 in the CNCbl-bound form, which are both in a {p, p} configu- ration using the conventions of Lovell et al. (42). The entire set of spin-label dihedral angles for V10R1 is given in Table 2. The Sd-HCa distance for the R1 side chain is ~4.5 A˚ , which is longer than that typically seen for R1 on helix surface sites. Although this rotamer is energetically allowed, it has not previously been observed in crystal structures (41), presumably due to the sterically restricted environment surrounding V10R1. Fig. 3 c compares the Ton box for the V10R1 mutant with and without CNCbl. The R1 side chain and the Ton box to which it is attached remain folded into the protein interior upon the addition of substrate, consistent with a lack of change in the EPR spectra shown in Fig. 2 b for FIGURE 3 (a) Periplasmic view of the structure and electron density (1s) showing the placement of the spin-labeled side chain V10R1 and residues that closely interact with the label in the apo form (PDB ID: 3M8B) of BtuB. (Magenta) Backbone of the Ton box. (Beige) N-terminal fold. (b) Periplasmic view of BtuB-V10R1 similar to that shown in panel a, except with van der Waals surfaces rendered for the atoms. The label, V10R1, is at the base of a periplasmic pocket in close tertiary contact with a number of atoms. (c) A comparison of the Ton box of BtuB- V10R1 with and without substrate. A side view of the crystal structure of the Ca2þ-B12 bound form of V10R1 (PDB ID: 3M8D) is shown with B12 bound, and the Ton box (magenta). This structure was aligned with the apo form of BtuB-V10R1 where only the Ton box is rendered (blue). Biophysical Journal 99(5) 1604–1610 Membrane Protein Conformational Exchange 1607 the protein crystal. Substrate addition to BtuB-V10R1 produces a change in the position of residue 7, as seen previ- ously for wt protein (22). However, residue 6, which is resolved in the wt structure, is not resolved for BtuB- V10R1 once substrate is bound. A B-factor analysis of the Ton box backbone and side-chain atoms indicates that when compared to wt BtuB structures, the BtuB-V10R1 has higher B-factors for the Ton box N-terminal to position 10, and a larger difference between apo- and ligand-bound forms. Nonetheless, the fold of the Ton box in BtuB- V10R1 is virtually identical to that seen in the wt structure (the root-mean-square deviation is 1.2 A˚ and 1.5 A˚ for the apo and CNCbl-bound forms of the Ton box, respectively, when V10R1 and wt are compared). Hence, minimal struc- tural changes in the Ton box are induced by this particular label. Both the crystal lattice and solutes shift the equilibrium between Ton box substates To determine whether the crystal lattice makes a contribu- tion to the free energy change when bilayer and crystal forms of BtuB are compared, EPR spectra from V10R1 were compared for the protein crystal and the protein solu- bilized into the cryo buffer. The two spectra for BtuB- V10R1 (in the CNCbl bound form) are compared in Fig. 4, a and b, and are clearly different. In particular, the spectrum from solubilized protein (Fig. 4 b) yields a mobile component with much higher amplitude than that for the protein crystal (Fig. 4 a). This mobile signal has a lineshape identical to that seen for the unfolded state in the bilayer (Fig. 2 a). Quantitation of the two components in this spectrum indicates that the mobile population makes up ~8 5 2% of the total spins. This fraction of unfolded Ton box corresponds to a change in free energy (a DDG for this transition relative to the bilayer reconstituted BtuB) of ~1.5 5 0.2 kcal/mol, indicating that solutes and the crystal lattice make roughly equal contributions to the change in conformational energy that is seen in the protein crystal. The lineshapes for the immobilized component in the absence of substrate for the bilayer reconstituted and crys- tallized BtuB-V10R1 are shown in Fig. 4, c and d, respec- tively. In this case the mobile component was subtracted from the bilayer BtuB-V10R1 (Fig. 2 a) to yield the immo- bile component in Fig. 4 c. Both these lineshapes result from immobile spin labels near the rigid limit of nitroxide motion. However, the hyperfine extrema in Fig. 4 c are not as distinct as in Fig. 4 d, and components representing the g-tensor anisotropy in the central (mI ¼ 0) resonance of BtuB-V10R1 are better resolved in the protein crystal (Fig. 4 d). This difference provides an indication that addi- tional motional modes are available for V10R1 in the bilayer environment. DISCUSSION In this work, SDSL was used to examine a conformational equilibrium in the Escherichia coli outer membrane trans- porter, BtuB, both in lipid bilayers and in protein crystals. The results indicate that the equilibrium between folded and unfolded forms of the Ton box is shifted by ~3 kcal/mol when the protein is taken from the bilayer phase to the protein crystal phase. This has the effect of stabilizing the folded form of the Ton box in the protein crystal, and it provides an expla- nation for the observation that the Ton box is resolved both in the absence and presenceofsubstrate in crystal structures(22), but is seen to unfold in bilayers. It should be emphasized that protein crystallography does not provide an incorrect structure for BtuB. The conditions of the protein crystal alter the equi- librium distribution of substates, compared to the distribution found by EPR, to favor the more-compact-ordered conformer. Osmolytes, such as PEGs, are well known to modify protein behavior (43), and previous work has demonstrated that osmolytes stabilize a folded form of the Ton box (15,24,25) and stabilize more-compact, less-hydrated conformations of the extracellular ligand-binding loops (44). The data obtained here indicate that solutes and the crystal lattice contribute almost equally to the energy change seen in the Ton box equilibrium. While the action of PEGs and other osmolytes is reasonably well understood, FIGURE 4 EPR spectra from BtuB-V10R1 with bound ligand in (a) the protein crystal, (b) in the crystallization buffer at a protein concentration too dilute to form crystals, and in the apo state in (c) lipid bilayers and (d) the protein crystal. The symbols i and m indicate immobilized and mobile components in the spectra for panel b. The spectrum in panel c is identical to the spectrum in Fig. 2 a, except that the small mobile component seen in Fig. 2 a has been subtracted. All spectra are 100 Gauss scans. TABLE 2 Summary of R1 side-chain dihedral angles and rotamer designation Mutant Rotamer c1 c2 c3 c4 c5 V10R1 apo {p,p} 56 69 83 67 67 V10R1 Ca2þ and B12 {p,p} 49 60 70 93 46 Biophysical Journal 99(5) 1604–1610 1608 Freed et al. it is not presently known how the protein lattice in the crystal couples to the Ton box equilibrium and stabilizes its folded form in BtuB. Previous work has shown that there is an interaction between charged residues near the Ton box and the BtuB b-barrel (40), and that eliminating this interac- tion unfolds the Ton box. Conceivably, a change in the dynamics or structure of the BtuB b-barrel when the protein is in the crystal lattice might alter the energy of this interac- tion and account for the effect of the lattice upon the Ton box. The structural biology of membrane proteins is far from mature, and it is not known whether the effects seen here on protein conformational sampling apply to a wider range of membrane proteins. Bacteriorhodopsin is perhaps the best-studied membrane protein, and there are indications that the kinetics of the photocycle are modified in the three-dimensional crystalline lattice when compared to the native membrane (45). Solution NMR can provide high- resolution structural data on membrane proteins in micelles, allowing comparisons to be made with crystal structures. NMR spectroscopy is often found to resolve portions of proteins that are not resolved by crystallography (as seen for DsbB (46)), presumably because NMR is better at exam- ining structures that are inherently dynamic. In outer membrane porins, such as OmpA, the strands of the b-barrel are shorter in the NMR-derived structures than in the crystal structures (47); however, it is not clear whether this differ- ence is a result of crystallization conditions or the micellar environment used for NMR. Even in a reconstituted bilayer environment (which is a much better approximation to the native environment than the protein crystal), protein confor- mational sampling may be modulated relative to the native environment (48–51); however, many of these effects appear to be due to the fraction of acidic lipids selected for the reconstitution, which in turn control local ion concentrations and pH. As for BtuB, the Ton box equilibrium does not appear to be modulated by lipid composition (Q. Xu and D. S. Cafiso, unpublished); this equilibrium is maintained within range of reconstituted bilayers as well as intact outer membrane preparations (19). Changes in the equilibrium distribution of protein confor- mational substates are thought to underlie protein signaling events (52) and allostery (53). Because of its fast timescale, EPR spectroscopy is particularly well suited to detect these conformational substates and to measure conformational equilibria in proteins. In BtuB, SDSL demonstrates that both the folded and unfolded states of the Ton box are sampled and that substrate binding shifts the equilibrium to the more disordered state. Furthermore, colicin E3, which is also a ligand for BtuB, shifts the Ton box equilibrium to favor the folded, ordered state of the Ton box (54). These are precisely the types of changes that are proposed to underlie protein signaling, and in the present case, they may function to regulate coupling between BtuB and inner membrane protein TonB. We thank Dr. Robert Nakamoto for careful reading of this manuscript. This work was supported by National Institutes of Health grants Nos. NIGMS 035215 to D.S.C. and NIGMS 079800 to M.C.W. Use of the Advanced Photon Source was supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under contract No. DE- AC02-06CH11357. Data were collected at Southeast Regional Collabora- tive Access Team (SER-CAT) 22-ID beamline at the Advanced Photon Source, Argon National Laboratory, Argonne, IL. Supporting institutions may be found at http://www.ser-cat.org/members.html. REFERENCES 1. Frauenfelder, H., S. G. Sligar, and P. G. Wolynes. 1991. The energy landscapes and motions of proteins. Science. 254:1598–1603. 2. Henzler-Wildman, K., and D. Kern. 2007. Dynamic personalities of proteins. Nature. 450:964–972. 3. Henzler-Wildman, K. A., M. Lei, ., D. Kern. 2007. A hierarchy of timescales in protein dynamics is linked to enzyme catalysis. Nature. 450:913–916. 4. Bahar, I., C. Chennubhotla, and D. Tobi. 2007. Intrinsic dynamics of enzymes in the unbound state and relation to allosteric regulation. Curr. Opin. Struct. Biol. 17:633–640. 5. Swain, J. F., and L. M. Gierasch. 2006. The changing landscape of protein allostery. Curr. Opin. Struct. Biol. 16:102–108. 6. Dunbrack, Jr., R. L. 2002. Rotamer libraries in the 21st century. Curr. Opin. Struct. Biol. 12:431–440. 7. Davis, I. W., W. B. Arendall, 3rd, ., J. S. Richardson. 2006. The backrub motion: how protein backbone shrugs when a sidechain dances. Structure. 14:265–274. 8. Walser, R., P. H. Hu¨nenberger, and W. F. van Gunsteren. 2002. Molecular dynamics simulations of a double unit cell in a protein crystal: volume relaxation at constant pressure and correlation of motions between the two unit cells. Proteins. 48:327–340. 9. Walser, R., P. H. Hu¨nenberger, and W. F. van Gunsteren. 2001. Comparison of different schemes to treat long-range electrostatic inter- actions in molecular dynamics simulations of a protein crystal. Proteins. 43:509–519. 10. Meinhold, L., and J. C. Smith. 2005. Fluctuations and correlations in crystalline protein dynamics: a simulation analysis of staphylococcal nuclease. Biophys. J. 88:2554–2563. 11. Bond, P. J., J. D. Faraldo-Go´mez, ., M. S. Sansom. 2006. Membrane protein dynamics and detergent interactions within a crystal: a simula- tion study of OmpA. Proc. Natl. Acad. Sci. USA. 103:9518–9523. 12. Colombo, M. F., D. C. Rau, and V. A. Parsegian. 1992. Protein solvation in allosteric regulation: a water effect on hemoglobin. Science. 256:1335–1336. 13. Parsegian, V. A., R. P. Rand, and D. C. Rau. 1995. Macromolecules and water: probing with osmotic stress. Methods Enzymol. 259:43–94. 14. Vodyanoy, I., S. M. Bezrukov, and V. A. Parsegian. 1993. Probing alamethicin channels with water-soluble polymers. Size-modulated osmotic action. Biophys. J. 65:2097–2105. 15. Kim, M., G. E. Fanucci, and D. S. Cafiso. 2007. Substrate-dependent transmembrane signaling in TonB-dependent transporters is not conserved. Proc. Natl. Acad. Sci. USA. 104:11975–11980. 16. Postle, K., and R. J. Kadner. 2003. Touch and go: tying TonB to trans- port. Mol. Microbiol. 49:869–882. 17. Wiener, M. C. 2005. TonB-dependent outer membrane transport: going for Baroque? Curr. Opin. Struct. Biol. 15:394–400. 18. Schauer, K., D. A. Rodionov, and H. de Reuse. 2008. New substrates for TonB-dependent transport: do we only see the ‘tip of the iceberg’? Trends Biochem. Sci. 33:330–338. 19. Merianos, H. J., N. Cadieux, ., D. S. Cafiso. 2000. Substrate-induced exposure of an energy-coupling motif of a membrane transporter. Nat. Struct. Biol. 7:205–209. Biophysical Journal 99(5) 1604–1610 Membrane Protein Conformational Exchange 1609 20. Fanucci, G. E., K. A. Coggshall, ., D. S. Cafiso. 2003. Substrate- induced conformational changes of the periplasmic N-terminus of an outer-membrane transporter by site-directed spin labeling. Biochem- istry. 42:1391–1400. 21. Xu, Q., J. F. Ellena, ., D. S. Cafiso. 2006. Substrate-dependent unfold- ing of the energy coupling motif of a membrane transport protein determined by double electron-electron resonance. Biochemistry. 45:10847–10854. 22. Chimento, D. P., A. K. Mohanty, ., M. C. Wiener. 2003. Substrate- induced transmembrane signaling in the cobalamin transporter BtuB. Nat. Struct. Biol. 10:394–401. 23. Fanucci, G. E., J. Y. Lee, and D. S. Cafiso. 2003. Spectroscopic evidence that osmolytes used in crystallization buffers inhibit a confor- mation change in a membrane protein. Biochemistry. 42:13106–13112. 24. Kim, M., Q. Xu, ., D. S. Cafiso. 2006. Solutes modify a conforma- tional transition in a membrane transport protein. Biophys. J. 90: 2922–2929. 25. Flores Jime´nez, R. H., M. A. Do Cao, ., D. S. Cafiso. 2010. Osmolytes modulate conformational exchange in solvent-exposed regions of membrane proteins. Protein Sci. 19:269–278. 26. Arakawa, T., and S. N. Timasheff. 1985. The stabilization of proteins by osmolytes. Biophys. J. 47:411–414. 27. Arakawa, T., and S. N. Timasheff. 1985. Mechanism of poly(ethylene glycol) interaction with proteins. Biochemistry. 24:6756–6762. 28. Timasheff, S. N. 2002. Protein hydration, thermodynamic binding, and preferential hydration. Biochemistry. 41:13473–13482. 29. Zimmerberg, J., F. Bezanilla, and V. A. Parsegian. 1990. Solute inac- cessible aqueous volume changes during opening of the potassium channel of the squid giant axon. Biophys. J. 57:1049–1064. 30. Fanucci, G. E., N. Cadieux, ., D. S. Cafiso. 2002. Structure and dynamics of the b-barrel of the membrane transporter BtuB by site- directed spin labeling. Biochemistry. 41:11543–11551. 31. Otwinowski, Z., and W. Minor. 1997. Processing of x-ray diffraction data collected in oscillation mode. Methods Enzymol. 276:307–326. 32. McCoy, A. J., R. W. Grosse-Kunstleve, ., R. J. Read. 2007. PHASER crystallographic software. J. Appl. Cryst. 40:658–674. 33. Emsley, P., and K. Cowtan. 2004. COOT: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 60: 2126–2132. 34. Adams, P. D., R. W. Grosse-Kunstleve, ., T. C. Terwilliger. 2002. PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr. D Biol. Crystallogr. 58:1948–1954. 35. Murshudov, G. N., A. A. Vagin, and E. J. Dodson. 1997. Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr. D Biol. Crystallogr. 53:240–255. 36. Winn, M. D., M. N. Isupov, and G. N. Murshudov. 2001. Use of TLS parameters to model anisotropic displacements in macromolecular refinement. Acta Crystallogr. D Biol. Crystallogr. 57:122–133. 37. Potterton, E., P. Briggs, ., E. Dodson. 2003. A graphical user interface to the CCP4 program suite. Acta Crystallogr. D Biol. Crystallogr. 59:1131–1137. 38. Davis, I. W., A. Leaver-Fay, ., D. C. Richardson. 2007. MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res. 35(Web Server issue):W375–W383. 39. Stone, T. J., T. Buckman, ., H. M. McConnell. 1965. Spin-labeled biomolecules. Proc. Natl. Acad. Sci. USA. 54:1010–1017. 40. Lukasik, S. M., K. W. Ho, and D. S. Cafiso. 2007. Molecular basis for substrate-dependent transmembrane signaling in an outer-membrane transporter. J. Mol. Biol. 370:807–811. 41. Guo, Z., D. Cascio, ., W. L. Hubbell. 2008. Structural determinants of nitroxide motion in spin-labeled proteins: solvent-exposed sites in helix B of T4 lysozyme. Protein Sci. 17:228–239. 42. Lovell, S. C., J. M. Word, ., D. C. Richardson. 2000. The penultimate rotamer library. Proteins. 40:389–408. 43. Auton, M., D. W. Bolen, and J. Ro¨sgen. 2008. Structural thermody- namics of protein preferential solvation: osmolyte solvation of proteins, amino acids, and peptides. Proteins. 73:802–813. 44. Kim, M., Q. Xu, ., D. S. Cafiso. 2008. Solutes alter the conformation of the ligand binding loops in outer membrane transporters. Biochem- istry. 47:670–679. 45. Efremov, R., V. I. Gordeliy, ., G. Bu¨ldt. 2006. Time-resolved microspectroscopy on a single crystal of bacteriorhodopsin reveals lattice-induced differences in the photocycle kinetics. Biophys. J. 91: 1441–1451. 46. Zhou, Y., T. Cierpicki, ., J. H. Bushweller. 2008. NMR solution structure of the integral membrane enzyme DsbB: functional insights into DsbB-catalyzed disulfide bond formation. Mol. Cell. 31:896–908. 47. Tamm, L. K., F. Abildgaard, ., J. H. Bushweller. 2003. Structure, dynamics and function of the outer membrane protein A (OmpA) and influenza hemagglutinin fusion domain in detergent micelles by solution NMR. FEBS Lett. 555:139–143. 48. Brown, M. F. 1994. Modulation of rhodopsin function by properties of the membrane bilayer. Chem. Phys. Lipids. 73:159–180. 49. Turnheim, K., J. Gruber, ., V. Ruiz-Gutie´rrez. 1999. Membrane phos- pholipid composition affects function of potassium channels from rabbit colon epithelium. Am. J. Physiol. 277:C83–C90. 50. Baenziger, J. E., S. E. Ryan, ., C. J. daCosta. 2008. Lipid composition alters drug action at the nicotinic acetylcholine receptor. Mol. Pharma- col. 73:880–890. 51. Charalambous, K., D. Miller, ., P. J. Booth. 2008. Lipid bilayer composition influences small multidrug transporters. BMC Biochem. 9:31. 52. Smock, R. G., and L. M. Gierasch. 2009. Sending signals dynamically. Science. 324:198–203. 53. Hilser, V. J. 2010. Biochemistry. An ensemble view of allostery. Science. 327:653–654. 54. Fanucci, G. E., N. Cadieux, ., D. S. Cafiso. 2003. Competing ligands stabilize alternate conformations of the energy coupling motif of a TonB-dependent outer membrane transporter. Proc. Natl. Acad. Sci. USA. 100:11382–11387. Biophysical Journal 99(5) 1604–1610 1610 Freed et al.
3M8E
Protein structure of Type III plasmid segregation TubR
Plasmid protein TubR uses a distinct mode of HTH- DNA binding and recruits the prokaryotic tubulin homolog TubZ to effect DNA partition Lisheng Ni, Weijun Xu, Muthiah Kumaraswami, and Maria A. Schumacher1 Department of Biochemistry and Molecular Biology, University of Texas, M. D. Anderson Cancer Center, Unit 1000, Houston, TX 77030 Edited by Robert T. Sauer, Massachusetts Institute of Technology, Cambridge, MA, and approved May 11, 2010 (received for review March 22, 2010) The segregation of plasmid DNA typically requires three elements: a DNA centromere site, an NTPase, and a centromere-binding protein. Because of their simplicity, plasmid partition systems represent tractable models to study the molecular basis of DNA segregation. Unlike eukaryotes, which utilize the GTPase tubulin to segregate DNA, the most common plasmid-encoded NTPases contain Walker-box and actin-like folds. Recently, a plasmid stabi- lity cassette on Bacillus thuringiensis pBtoxis encoding a putative FtsZ/tubulin-like NTPase called TubZ and DNA-binding protein called TubR has been described. How these proteins collaborate to impart plasmid stability, however, is unknown. Here we show that the TubR structure consists of an intertwined dimer with a winged helix-turn-helix (HTH) motif. Strikingly, however, the TubR recognition helices mediate dimerization, making canonical HTH– DNA interactions impossible. Mutagenesis data indicate that a basic patch, encompassing the two wing regions and the N termini of the recognition helices, mediates DNA binding, which indicates an unusual HTH–DNA interaction mode in which the N termini of the recognition helices insert into a single DNA groove and the wings into adjacent DNA grooves. The TubZ structure shows that it is as similar structurally to eukaryotic tubulin as it is to bacterial FtsZ. TubZ forms polymers with guanine nucleotide-binding characteristics and polymer dynamics similar to tubulin. Finally, we show that the exposed TubZ C-terminal region interacts with TubR-DNA, linking the TubR-bound pBtoxis to TubZ polymerization. The combined data suggest a mechanism for TubZ-polymer pow- ered plasmid movement. T he cytoskeletons of eukaryotic cells are constructed of three primary elements: actin, tubulin, and intermediate filaments. Although it had long been presumed that the proteins forming these elements were absent in prokaryotes, it is now known that prokaryotes contain structural homologs to all three components. These prokaryotic proteins appear to carry out distinct functions compared to their eukaryotic counterparts; however, their roles are similar enough to indicate a likely common ancestor. The best known prokaryotic homolog of tubulin is FtsZ. Both FtsZ and tubulin form long filamentous structures by head to tail associa- tion in a manner regulated by GTP, which binds between adjacent subunits (1–4). However, unlike tubulin, FtsZ does not function in DNA segregation but rather cell division. Specifically, it forms a cytokinetic ring called the Z ring at midcell, which mediates septation (5, 6). Recently, however, prokaryotic proteins encoded on large plasmids harbored in bacilli showing 15–20% sequence similarity to both FtsZ and tubulin have been identified and dubbed TubZ (7–12). Studies showed that the Bacillus thuringien- sis TubZ protein from the pBtoxis plasmid is essential for plasmid DNA segregation. DNA segregation of most low copy number plasmids is carried out by specific partition (par) systems. These systems require only three elements: a centromere DNA site, a centromere-binding protein, and a partition NTPase (13, 14). Partition systems have been classified into two main types on the basis of the kind of NTPase present (15). Type I systems contain NTPases with deviant Walker A-type ATPase folds, whereas type II systems uti- lize actin-like NTPases. Interestingly, both types of NTPases form polymers in NTP-dependent manners that are implicated to play a role in plasmid DNA separation (16–19). The recent discovery of TubZ NTPases has led to the designation of “type III” par sys- tems (13, 14). The best studied of these systems is that found on the pBtoxis plasmid in B. thuringiensis. This plasmid stability sys- tem is represented by an operon encoding two proteins: ORF156 (TubZ) and ORF157 (TubR) (7–9, 11). TubR is a 11.6 kDa DNA- binding protein that shows no sequence homology to any known protein. Studies showed that TubR binds a 48-bp centromere con- taining four repeat sites in the pBtoxis plasmid and also autore- gulates tubRZ transcription (8, 9). TubZ is a 54.4 kDa protein that can assemble into filaments in a GTP-dependent manner (12). Both proteins were found to be required for plasmid stability (9). However, how the TubR and TubZ proteins work together to effect pBtoxis plasmid segregation is not known. To gain insight into the molecular mechanism utilized by these proteins in DNA segregation, we carried out structural and biochemical studies on the pBtoxis TubR and TubZ proteins. The TubR structure reveals that it employs a helix-turn-helix (HTH) motif in a previously un- described manner to bind DNA. TubZ contains a tubulin/FtsZ fold but has structural distinctions from these proteins indicating that it forms distinct protofilaments. TubR binds the flexible C-terminal region of TubZ, thus attaching the TubZ filament to the pBtoxis plasmid, providing a mechanism for plasmid move- ment and, ultimately, segregation. Results and Discussion Overall Structure of pBtoxis TubR. The crystal structure of the 107- residue pBtoxis TubR protein was solved to 2.0-Å resolution by selenomethionine multiple wavelength anomalous diffraction (MAD) methods (Table S1). The structure contains two TubR molecules in the crystallographic asymmetric unit and consists of residues 6–102 of one subunit and 4–100 of the second subunit, and has Rwork∕Rfree ¼ 23.8%∕27.0%. The TubR structure forms a highly intertwined dimer with dimensions 30 × 30 × 60 Å3 (Fig. 1A). Each TubR subunit has the topology β1-α1-α2-α3-α4-β2- β3-α5, which is similar to winged HTH motifs found in a number of DNA-binding proteins in both prokaryotes and eukaryotes (20). In TubR, α3-α4 forms the HTH motif and the loop between Author contributions: M.A.S. designed research; L.N., W.X., M.K., and M.A.S. performed research; M.A.S. analyzed data; and M.A.S. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Freely available online through the PNAS open access option. Data deposition: The atomic coordinates and structure factor amplitudes for the WT TubR (C2), WT TubR (I222), TubR(S63W), TubZ, and TubZ-GTP-γ-S structures have been deposited with the Protein Data Bank, www.pdb.org (PDB ID codes 3M8E, 3M9A, 3M8F, 3M8K, and 3M89). 1To whom correspondence may be addressed. E-mail: maschuma@mdanderson.org. This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1003817107/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1003817107 PNAS ∣June 29, 2010 ∣vol. 107 ∣no. 26 ∣11763–11768 BIOCHEMISTRY β2 and β3, the wing. Indeed, each TubR subunit shows the stron- gest structural similarity to members of the ArsR family of prokaryotic winged helix transcription regulators, in particular the Staphylococcus aureus CzrA protein (21, 22). Superposition of one subunit of TubR onto that of CzrA results in a rmsd of 2.7 Å. This similarity includes the core regions of the winged HTH, because the loops and N-terminal regions of the proteins are structurally distinct. For example, TubR contains a β-strand in its N-terminal region compared to a long helix in the CzrA struc- ture (Fig. 1B). This structural similarity initially suggested that TubR may be a member of the ArsR family of proteins. However, the arrangement of the TubR dimer was found to be strikingly different from the dimer organization exhibited by the ArsR proteins (Fig. 1C). ArsR family members are involved in metal-regulated tran- scription processes whereby they act as repressors in their apo forms and are induced off their DNA sites upon metal binding (22). The specific dimer structures of the ArsR proteins are cri- tical for creation of their metal-binding motifs. Not only does TubR form a very different dimer from the ArsR proteins, it also does not harbor any of their metal-binding signatures nor does it contain any other characterized metal-binding motif. Consistent with this, we find that the addition of metals has no effect on TubR DNA binding. Dimerization of the ArsR proteins is imparted by residues from the N-terminal regions, α1 and α5, which, impor- tantly, leaves its recognition helices exposed for DNA interaction (22). By contrast, the TubR dimer is formed primarily by contacts between its twofold related “recognition helices” α4 and α4·. This interaction results in a near complete burial of these helices, leav- ing only the N-terminal residues exposed to solvent. Whereas the α4 and α4· interaction creates the dimer core, the dimer is further stabilized by interactions between the twofold related β1 strands, which swap to form an antiparallel β-sheet. Residues from α1 and α5 interact with β1 to further seal the top of the dimer. The dimer interface formed by these interactions is predominantly hydropho- bic and buries a large 1;200 Å2 of subunit surface from solvent. TubR Forms a “Recognition Helix Dimer”: Implications for DNA Bind- ing. Gel filtration studies on TubR confirmed that it is a dimer in solution. However, the finding from the structure that the TubR α4 recognition helices are buried in the dimer core has important implications in terms of its DNA-binding mechanism. Indeed, it suggests that, although TubR contains a structurally canonical HTH, it is not utilized for DNA binding in a manner typical of HTH proteins. A second crystal form (I222) of TubR, which was solved to 2.5-Å resolution, revealed the same TubR dimer. The presence of the identical dimer in two different crystal forms and its large buried surface area supports that the dimer observed in the crystal structures is physiologically relevant. However, to test this, we mutated residues within the recognition helices that the structure indicates are critical for dimerization and assayed the ability of the mutant proteins to dimerize via gel filtration. Specifically, we mutated Ser-63 and Ala-67 individually to tryp- tophan and arginine. The structure shows that residues occupying positions 63 and 67 must be small and largely hydrophobic to permit the proper packing of the α4 helices in the dimer (Fig. S1A). Hence, the in- troduction of the bulky side chain of tryptophan and, in particu- lar, the large as well as charged side chain of arginine would be predicted to be highly disruptive to dimerization. Gel filtration analyses on purified mutant proteins clearly showed that the ar- ginine mutants exist primarily as monomers in solution (>80%), whereas the tryptophan mutants were able to maintain the di- meric state (Fig. 2A and Fig. S1B). However, all mutant proteins showed reduced or loss of DNA-binding activity as ascertained by fluorescence polarization (FP) studies, which examined TubR protein binding to its centromere site (Fig. 2B and Fig. S1 C and D) (9). The fact that the monomeric mutants were severely impaired in DNA binding was not surprising. However, the finding that the tryptophan mutants, which were largely dimeric, displayed reduced DNA-binding activity suggested that their oli- gomer structures might be altered. To address this issue, the struc- ture of S63W TubR was solved to 2.8-Å resolution, resulting in Rwork∕Rfree values of 20.4%∕26.6% (Table S1). The subunit struc- ture of S63W TubR is essentially identical to that of WT TubR as revealed by their superimposition (rmsd of 1.2 Å) (Fig. 2 C and D). However, this single subunit overlay shows that the S63W TubR dimer, although the same as the WT in general arrange- ment, is forced into a more open oligomer conformation in which one subunit is rotated 20° away from its dimer mate compared to WT. This rotation is required to accommodate the bulky S63W side chains (Fig. 2D). The N-terminal β1–β1′-sheet interaction appears to play a key role in holding the TubR subunits together. In addition, the tight stacking of the twofold related Trp63 indole groups (3.5 Å) provides a compensatory interaction that, com- bined with the β1–β1′ interaction, apparently permits retention of the dimer state, indicating why the TubR tryptophan mutants were able to maintain the dimer state, albeit an altered dimer state relative to WT (Fig. 2 C and D). By contrast, the TubR arginine mutations, which introduced both bulk and charge with- in the predominantly hydrophobic dimer interface, were highly destabilizing for dimerization. In addition, the finding that the S63W TubR mutant forms an altered dimer explains the severe effect on DNA binding because a correct dimer orientation is likely essential for binding to its palindromic DNA sites (9). TubR-DNA Model. Because all but the N-terminal residues of the TubR recognition helices are buried in the dimer interface, TubR must use a different mode of DNA binding than the ArsR or other HTH containing proteins (23). Examination of surface electro- statics of TubR reveals that one face of the protein is electro- negative, whereas the other is strongly electropositive (Fig. 3A). Notably, the positive region is composed of one large and contig- uous basic patch. Basic residues in this region correspond to Arg- 74, Arg-77, and Lys-79, in the wing and Lys-43, which is on α3, the helix preceding α4 in the HTH motif. These residues were mutated singly to alanine to examine their roles in DNA binding (Fig. 3B). FP experiments showed that mutation of the basic wing residues resulted in either a complete (R74A and R77A) or nearly com- plete (K79A) abrogation of DNA binding, indicating that the wings play a major role in TubR DNA binding. Residue Lys-43 is surface exposed and located at the center of the basic region on the TubR dimer (Fig. 3A). The K43A mutant also showed no binding to TubR, supporting the notion that the continuous basic patch of TubR represents its DNA-binding surface. To gain insight into the structural mechanism of DNA binding, a DNA duplex was docked onto the basic patch of the TubR di- Fig. 1. B. thuringiensis pBtoxis TubR structure. (A) One TubR subunit is red and the other cyan. Secondary structural elements and N and C termini are labeled. (B) Superimposition of one subunit of TubR (Red) onto a S. aureus CzrA subunit (Yellow). Regions with different structures are labeled. (C) Same superimposition as B showing the location of the other subunit in the TubR and CzrA dimers after one subunit is overlaid. A–C are in the same orienta- tion to highlight differences. Figs. 1 A–C, 2 C and D, 3 A, C, and E, 4B, and 5B were made by using PyMOL (31). 11764 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1003817107 Ni et al. mer by using the location of the mutations that affected DNA binding as a guide (Fig. 3C). This model revealed that the wings are positioned to interact with successive minor grooves, with either the bases or the phosphate backbone depending on the ability of the DNA to deform. In the model TubR interacts with a minimum of 14 bp of DNA. However, the centromere bound by Fig. 2. The TubR “recognition helix” dimer. (A) Gel fil- tration studies on TubR mutants S63W and S63R show- ing that the S63W mutant remains dimeric whereas S63R is >80% monomer. (B) Fluorescence polarization studies examining the ability of WT TubR, S63R, and S63W TubR mutants to bind iteronic DNA. Fluorescence polarization units (millipolarization) and TubR concen- tration (nM) are along the y and x axes, respectively. The Kd of WT TubR for the centromere DNA is 8  2 nm. (C) Superimposition of WT TubR (Green) onto the TubR S63W mutant structure (Tan). (D) Close-up of the site of the S63W mutation in the expanded TubR S63W dimer showing stacking interactions between the twofold related tryptophans. Fig. 3. TubR-DNA binding. (A) Electrostatic surface potential of the TubR dimer. Blue and red represent electropositive and electronegative regions, respec- tively. (Left) The electronegative side of the TubR dimer, and the side on the right is the electropositive side. Labeled on the left side are the locations of the mutated residues. (B) FP binding isotherms showing the DNA binding of WT TubR and the K43A, R74A, R77A, and K79A mutants. Fluorescence polarization units (millipolarization) and TubR concentration (nM) are along the y and x axes, respectively. (C) TubR-DNA model showing TubR electrostatic potential. (D) Stoichiometry of TubR (subunit) binding showing titration curve of TubR into the 48-mer iteron resulting in a molar ratio of TubR subunit to DNA of eight (or four) dimers. (E) Left: Ribbon diagram of the TubR-DNA model with the recognition helices colored yellow. Right: Ribbon diagram showing a canonical HTH–DNA interaction (the λ repressor-DNA complex) with the recognition helices colored yellow (32). Ni et al. PNAS ∣ June 29, 2010 ∣ vol. 107 ∣ no. 26 ∣ 11765 BIOCHEMISTRY TubR consists of four 12-bp sites with the consensus T(T/A)(T/A) (C/A)(G/A)GTTTA(A/C)(A/C) (9). Thus, we used FP to ascer- tain the binding stoichiometry of TubR for its 48-bp centromere. As shown in Fig. 3D, eight TubR subunits or four TubR dimers bind the 48-mer centromere, consistent with a dimer of TubR binding each palindrome. Thus, either TubR distorts its DNA or the TubR dimers bind with some degree of overlap on their DNA sites perhaps imparting cooperativity, as observed in other centromere-binding protein–DNA interactions (13, 14). In addi- tion to the insertion of the wings, a striking outcome of the mod- eling was the finding that the N termini of the recognition helices, which interact with each other in a parallel, coiled-coil-like man- ner, are in position to insert into a single major groove. Structures of HTH proteins bound to DNA have thus far shown that the recognition helices insert singly into successive major grooves by using residues in the first few turns or the central portion of the recognition helix to contact the DNA bases (Fig. 3E). Thus, in the TubR-DNA model, the HTH–DNA interaction is drama- tically different from any displayed previously by a HTH protein. TubZ Binds TubR-DNA. A characteristic feature exhibited by partition centromere-binding proteins is the ability to bind their partner NTPase (13, 14). To determine if TubR binds TubZ, we utilized a FP assay and found that full length (FL) TubZ bound avidly to the TubR-centromere complex (Fig. 4A). However, unlike other partition systems in which the NTPase must be com- plexed with nucleotide to bind its centromere-binding protein, the interaction of TubR with TubZ did not require the presence of GTP-Mg2þ. Previous studies have shown that the C-terminal regions of tubulin and FtsZ mediate key binding events with their target proteins (24–26). We noted that the terminal region of TubZ, consisting of residues 407–484, is the most divergent region between TubZ proteins and between TubZ and tubulin/FtsZ pro- teins, suggesting that it may be similarly utilized and bind TubR. To test this hypothesis, we constructed various TubZ truncations, TubZ(1-407), TubZ(1-442), TubZ(1-460), and TubZ(1-470), and examined the ability of each protein to bind TubR-DNA. TubZ (1-407) showed no binding to TubR, whereas the remaining trun- cation mutants bound weakly to TubR-DNA (Fig. 4A). Thus, the data indicate that the last 14 amino acids of TubZ are critical for the ability of TubZ to form a tight interaction with TubR but that residues 408–470 also play an important role in this interaction. These data demonstrate that TubR acts as a partition partner for TubZ, linking it to pBtoxis plasmid DNA. Although TubZ has been shown to form polymers in a GTP-dependent manner, the TubZ protein displays limited sequence similarity to tubu- lin/FtsZ, suggesting potential differences in TubZ and tubulin/ FtsZ structures (7–9). To gain insight into TubZ function, we next determined structures of B. thuringiensis pBtoxis TubZ. Structure of TubZ. Crystallization of FLTubZ was not successful, in either its apo form or bound to guanine nucleotides. We noted that FL TubZ degraded over time whereby C-terminal residues were proteolyzed. Therefore, truncated TubZ proteins were utilized in crystallization trials, and crystals were obtained of apo TubZ (1-428) and the structure solved by MAD (Table S2). The model consists of residues 1–79 and 91–404 and has an Rwork∕Rfree of 21.4∕24.9% to 2.3-Å resolution (Fig. 4B). No discernible oligomer- ization of apo TubZ(1-428) was observed in the crystal packing, and gel filtration analyses confirmed that it is monomeric (Fig. S2). The overall TubZ structure can be divided into two main domains: an N domain (residues 25–235) and a C domain (resi- dues 258–377). These domains are connected by a long, core helix, H7. The TubZ N domain has a Rossman fold and consists of six parallel β-strands with topology 3-2-1-4-5-6. The resulting β-sheet is sandwiched by five α-helices, with two helices on one side and three on the other. The C domain consists of four β-strands with the topology 1-4-2-3. The C-domain β-strands are arranged nearly perpendicular to those in the N domain. In addition to these main protein domains, there are two helices: one at the N terminus, H0, and a long helix at the C terminus, H11 (Fig. 4B). Database searches showed that TubZ indeed belongs to the tubulin/FtsZ fa- mily of proteins and is similar to both eukaryotic and prokaryotic members of the family; TubZ can be optimally superimposed with rmsds of 3.4 Å onto both bovine α tubulin and Pseudomonas aeruginosa FtsZ (1, 5). Whereas the two-domain architectures of tubulin, FtsZ, and TubZ are similar in overall structure, the extreme N- and C-terminal regions of these proteins are very di- vergent (Fig. S3 A–D). N-Terminal and C-Terminal Differences in TubZ, FtsZ, and Tubulin: Im- plications for Polymer Formation and Target Protein Binding. Tubulin proteins do not contain significant N-terminal extensions, whereas FtsZ proteins from different organisms show structural variability within their N-terminal regions. For instance, in the Escherichia coli FtsZ structure the N-terminal residues are disor- dered, whereas Methanococcus jannaschii FtsZ has an extra N-terminal helix, H0, which is flexibly attached to the body of the protein and has been captured in multiple orientations (5). Although H0 is not conserved in FtsZ proteins, one M. jannaschii FtsZ structure revealed a semicontinuous polymer in the crystal, thought to closely represent in vivo protofilaments, which utilizes H0 in subunit-subunit contacts (4). This finding suggests that the flexibly attached H0 is stabilized in a specific orientation by protofilament formation, at least in the M. jannaschii protein. The TubZ H0 helix extends in the opposite direction compared to that of the protofilament stabilized FtsZ H0 helix. Moreover, in TubZ, H0 is not flexibly attached to the N domain but is tightly anchored to the C domain through numerous interactions with the core helix and C-domain residues. The large number of inter- actions involving H0, and the fact that it covers what would other- wise be a surface exposed hydrophobic patch, indicate that the TubZ H0 does not undergo conformational changes during protofilament formation and is important for the general fold of TubZ (Fig. S3 A and C). Data suggest that FtsZ and tubulin form protofilaments with similar longitudinal contacts (4). However, the TubZ structure reveals key differences, primarily in its C-domain and C-terminal regions, which suggest that it forms protofilaments distinct from those formed by tubulin/FtsZ. A notable difference is the struc- ture of loop 7 (L7). This loop inserts into the adjacent subunit providing the key catalytic residues required for GTP hydrolysis. In tubulin/FtsZ proteins, L7 has the consensus GXXNXDXAD. In TubZ, the loop is very divergent in conformation compared Fig. 4. TubZ interacts with TubR-DNA and contains a tubulin/FtsZ fold. (A) FP assay measuring binding of FL TubZ, TubZ(1-470), TubZ(1-460), TubZ(1-442), and TubZ(1-407) to TubR-DNA. Below is the control (TubZ titrated into DNA alone). Millipolarization units and TubZ concentration (nM) are along the y and x axes, respectively. (B) TubZ(1-428) structure. The N domain or GTP-bind- ing domain is colored salmon and the C domain purple. The interdomain helix, H7, is red. TubZ also contains an N-terminal helix, H0 (Yellow), and a C-terminal helix, H11 (White). 11766 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1003817107 Ni et al. to that in FtsZ/tubulin and consists of the sequence 256- DNVTYDPSD-266. In addition to the N-terminal region, the extreme C-terminal extensions of tubulin, FtsZ, and TubZ are structurally divergent (Fig. S3B). In FtsZ, the C-terminal region forms a small, two-stranded β-sheet and continues into an ex- tended region that is involved in binding adaptor proteins such as FtsA and ZipA (6, 25). By contrast, the C-terminal regions of tubulin proteins consist of a two-helix bundle followed by an extended region. Like FtsZ, however, these regions interact with numerous target proteins such as the microtubule-associated pro- teins (MAPs) (24). Consistent with this, the C-terminal regions of tubulin have been shown to face the outside of the microtubule. A characteristic feature of the extreme C-terminal extensions of tubulin proteins is their highly acidic nature (3). This acidic region has been shown to be critical for binding to several MAPs that harbor a substantial basic character, such as tau, MAP2, and MAP4 (24, 27). The TubZ C-terminal region is also helical, but it contains a single, long helix. Notably, the TubZ-tubulin overlay shows that the long C-terminal helix of TubZ would dramatically clash with the adjacent subunit in a polymer, providing support for the no- tion that TubZ forms protofilaments different from tubulin/FtsZ (Figs. S3B and S4). Interestingly, and in contrast to tubulin pro- teins, the flexible C-terminal region of TubZ that follows H11 is highly basic, in particular the last 14 residues. We have shown that these residues play a central role in TubR binding (Fig. 4A). TubR uses its electropositive face for DNA binding, leaving exposed its opposite face for TubZ interaction. Notably, this exposed face is strongly electronegative and hence would complement the basic C-terminal tail of TubZ (Fig. 3A). Tubulin/FtsZ protofilaments combine to form higher-order structures. In tubulin, the protofilaments interact in a parallel manner to form microtubules. Central to microtubule formation are lateral contacts between protofilaments from the so-called M loop, between H10 and S9. In tubulin, this loop is composed of 13 residues (1–3). The corresponding loop is much shorter in FtsZ proteins, consistent with the fact that FtsZ does not form tubulin microtubule-like structures (5, 28–30). In TubZ, the M loop is even shorter than in FtsZ, spanning only four residues. In fact, the TubZ/tubulin overlay shows that the side of the molecule con- taining the M loop is the most divergent between these proteins. These combined findings suggest that TubZ not only forms pro- tofilaments with distinct longitudinal contacts compared to FtsZ and tubulin, but it also does not form tubulin-like microtubule structures. TubZ Interactions with Guanine Nucleotides. Consistent with TubZ being a member of the tubulin/FtsZ family, our isothermal titra- tion calorimetry (ITC) studies showed that TubZ binds guanine nucleotides with high affinity; Kds for GTP-γ-S and GDP were ∼0.69 and 26 μM, respectively (Fig. 5A). We next determined the structure of the TubZ-GTP-γ-S complex by soaking GTP-γ-S into preformed TubZ(1-428) crystals. The TubZ-GTP-γ-S struc- ture contains TubZ residues 1–79 and 91–404, one GTP-γ-S, and has Rwork∕Rfree ¼ 21.8%∕25.5% (Fig. 5B and Table S2). The structure shows that TubZ binds GTP-γ-S in the same GTP binding pocket as tubulin/FtsZ (1–5). Comparison of the apo and GTP-γ-S bound TubZ structures indicated that, like FtsZ, guanine nucleotide binding does not lead to significant conforma- tional changes (5). The phosphate binding pocket is formed by two of the most highly conserved regions between TubZ and tubulin/FtsZ called loops 1 and 4 (L1 and L4) (1–3). L1 contacts the GTP-γ-S α- and β-phosphate groups via the Gln-32 and Lys-33 amide nitrogens. The L1 region of FtsZ and tubulin contain the sequences GQ(A/G)G and GQCG, respectively, whereas in TubZ the motif is 31-GQKG-34. However, the alanine/glycine and cysteine residues in FtsZ and tubulin do not contact the bound nucleotide; the TubZ Lys-33 side chain makes stacking interac- tions with the guanine base (Fig. 5B). L4 represents the so-called signature motif [GGGTG(T/S)G], which serves as an identifier of tubulin/FtsZ family members. Like FtsZ and tubulin, the L4 region of TubZ-GTP-γ-S makes phosphate interactions via its glycine amide nitrogens. Whereas L1 and L4 residues of the N domain mediate phosphate contacts, the GTP-γ-S guanine moiety is specified from residues in the core helix, H5, and C domain. In this regard, an important motif is loop 6 (L6). In FtsZ and tubulin, L6 has the consensus (F/Y)XXX(N/D) and the conserved (F/Y) residue functions in guanine base stacking. This region in TubZ, 236-WKXXXN-241, is in an altered conformation compared to FtsZ and tubulin structures. Despite the presence of the trypto- phan, which might be expected to interact with the guanine, the side chain of Lys-237 instead stacks with the guanine ring. Hence, in the TubZ-GTP-γ-S structure, the guanine base does not interact with aromatic residues as in tubulin/FtsZ but is sand- wiched between the aliphatic portions of two lysine side chains, Lys-33 and Lys-237. Finally, two asparagine residues, Asn-213 and Asn-241, from L6 effectively read the guanine N2/N3 and N1/O6 atoms, respectively, providing high specificity in TubZ’s in- teraction with guanine nucleotides. pBtoxis DNA Segregation: TubR Plasmid Partition Model. Our data show that TubR binds to the flexible, C-terminal, basic region of TubZ. The flexibility and location of the TubZ C-terminal extension suggest that it is not required for polymerization and thus may be exposed on the surface of TubZ filaments. Indeed, negative stain EM images show that TubZ(1-407) forms polymers in a GTP-dependent manner similar to the FL protein (Fig. S5). Recent data suggesting that TubZ filaments are stabilized by Fig. 5. TubZ-guanine nucleotide interactions. (A) ITC binding isotherms showing TubZ-GDP (Left) and TubZ-GTP-γ-S interaction (Right). (B) Left: Overall structure of the TubZ-GTP-γ-S complex. β-strands are colored magen- ta and helices cyan, and the GTP-γ-S molecule is shown as cpk. Right: Close-up view of the GTP binding pocket with the initial Fo-Fc electron density map (Blue Mesh), contoured at 4.5σ, and calculated before the GTP-γ-S was included in refinement. Ni et al. PNAS ∣ June 29, 2010 ∣ vol. 107 ∣ no. 26 ∣ 11767 BIOCHEMISTRY the presence of a GTP cap and undergo treadmilling are consis- tent with the notion that TubZ displays tubulin-like polymer dy- namics (12). Thus, on the basis of the combined data, we suggest a model for TubR/TubZ mediated pBtoxis plasmid segregation shown in Fig. 6. In this model, multiple TubR dimers first bind to the iteronic DNA on the pBtoxis plasmid leading to the crea- tion of a high local concentration of TubR, which can recruit a TubZ polymer, likely by interactions between the acidic TubR dimer face and the basic C-terminal TubZ region. Importantly, this interaction serves to attach the pBtoxis plasmid to the TubZ polymer, which undergoes treadmilling, adding subunits at the þ end and losing subunits at the −end. The bound TubR-pBtoxis can be handed off from the −end to the molecules in the growing þ end, leading to the transport of the pBtoxis plasmid to the cell pole. Interestingly, it has been shown that once TubZ polymers reach and interact with the cell pole, they bend around the curved pole and continue growing in the other direction (7). The force of the interaction with the membrane likely causes the release of TubR-pBtoxis, the net result being transport of pBtoxis to the cell pole. Of course, this model is simplified and many questions re- main. For example, how directionality is achieved and how the replicated plasmids are driven to opposite cell poles is not clear. However, given the large size of the pBtoxis plasmid (8), it may be that only one TubR-pBtoxis “tram” can be bound at once by the rapidly treadmilling TubZ polymer and that, once one such a tram is unloaded after reaching the cell pole, another engages when the now reversed polymer treadmills toward the opposite cell pole. Materials and Methods Summary Detailed methods are provided in SI Materials and Methods. Briefly, the tubR and tubZ genes were codon optimized (for E. coli expression), subcloned into pET15b, expressed, and pur- ified. WT TubR crystals were grown with NaCl and phosphate. TubR S63W was crystallized with PEG and ethylene glycol and TubZ with sodium formate. Detailed assay conditions for FP, ITC, electron microscopy, and gel filtration are provided in SI Materials and Methods. ACKNOWLEDGMENTS. This work was supported by the Burroughs Wellcome Career Development Award 992863 and National Institutes of Health Grant GM074815 (to M.A.S.). 1. Nogales E, Wolf SG, Downing KH (1998) Structure of the αβ tubulin dimer by electron crystallography. Nature 391:199–203. 2. Nogales E, Whittaker M, Milligan RA, Downing KH (1999) High-resolution model of the microtubule. Cell 96:79–88. 3. Nogales E (2000) Structural insights into microtubule function. Annu Rev Biochem 69:277–302. 4. Oliva MA, Cordell SC, Löwe J (2004) Structural insights into FtsZ protofilament formation. Nat Struct Mol Biol 11:1243–1250. 5. Oliva MA, Trambaiolo D, Löwe J (2007) Structural insights into the conformational variability of FtsZ. J Mol Biol 373:1229–1242. 6. Margolin W (2005) FtsZ and the division of prokaryotic cells and organelles. Nat Rev Mol Cell Biol 6:862–871. 7. Larsen RA, et al. (2007) Treadmilling of a prokaryotic tubulin-like protein, TubZ, required for plasmid stability in Bacillus thuringiensis. Genes Dev 21:1340–1352. 8. Tang M, Bideshi DK, Park H-W, Federici BA (2006) Minireplicon from pBtoxis of Bacillus thuringiensis subsp. israelensis. App Environ Microbiol 72:6948–6954. 9. Tang M, Bideshi DK, Park H-W, Federici BA (2007) Iteron-binding ORF157 and FtsZ-like ORF156 proteins encoded by pBtoxis play a role in its replication in Bacillus thuringien- sis subsp. israelensis. J Bacteriol 189:8053–8058. 10. Anand SP, Akhtar P, Tinsley E, Watkins SC, Khan SA (2008) GTP-dependent polymer- ization of the tubulin-like RepX replication protein encoded by the pXO1 plasmid of Bacillus anthracis. Mol Microbiol 67:881–890. 11. Berry C, et al. (2002) Complete sequence and organization of pBtoxis, the toxin- coding plasmid of Bacillus thuringiensis subsp israelensis. Appl Environ Microbiol 68:5082–5095. 12. Chen Y, Erickson HP (2008) In vitro assembly studies of FtsZ/tubulin-like proteins (TubZ) from Bacillus plasmids: Evidence for a capping mechanism. J Biol Chem 283:8102–8109. 13. Hayes F, Barillà D (2006) The bacterial segrosome: A dynamic nucleoprotein machine for DNA trafficking and segregation. Nat Rev Microbiol 4:133–143. 14. Schumacher MA (2008) Structural biology of plasmid partition: Uncovering the molecular mechanisms of DNA segregation. Biochem J 412:1–18. 15. Gerdes K, Møller-Jensen J, Bugge Jensen R (2000) Plasmid and chromosome partitioning: Surprises from phylogeny. Mol Microbiol 37:455–466. 16. Møller-Jensen J, et al. (2003) Bacterial mitosis: ParM of plasmid R1 moves plasmid DNA by an actin-like insertional polymerization mechanism. Mol Cell 12:1477–1487. 17. Popp D, et al. (2008) Molecular structure of the ParM polymer and the mechanism leading to its nucleotide-driven dynamic instability. EMBO J 27:570–579. 18. Salje J, Löwe J (2008) Bacterial actin: Architecture of the ParMRC DNA partitioning complex. EMBO J 27:2230–2238. 19. Dunham TD, Xu W, Funnell BE, Schumacher MA (2009) Structural basis for ADP- mediated transcriptional regulation by P1 and P7 ParA. EMBO J 28:1792–1802. 20. Gajiwala KS, Burley SK (2000) Winged helix proteins. Curr Opin Struct Biol 10:110–116. 21. Eicken C, et al. (2003) A metal-ligand-mediated intersubunit allosteric switch in related SmtB/ArsR zinc sensor proteins. J Mol Biol 333:683–695. 22. Pennella M, Giedroc DP (2005) Structural determinants of metal selectivity in prokaryotic metal-responsive transcriptional regulators. Biometals 18:413–428. 23. Arunkumar AI, Campanello GC, Giedroc DP (2009) Solution structure of a paradigm ArsR family zinc sensor in the DNA-bound state. Proc Natl Acad Sci USA 106:18177–18182. 24. Downing KH (2000) Structural basis for the interaction of tubulin with proteins and drugs that affect microtubule dynamics. Annu Rev Cell Dev Biol 16:89–111. 25. Adams DW, Errington J (2009) Bacterial cell division: Assembly, maintenance and disassembly of the Z ring. Nat Rev Microbiol 7:642–653. 26. Errington J, Daniel RA, Scheffers DJ (2003) Cytokinesis in bacteria. Microbiol Mol Biol Rev 67:52–65. 27. Chau MF, et al. (1998) The microtubule-associated protein tau cross-links to two distinct sites on each alpha and beta tubulin monomer via separate domains. Biochemistry 37:17692–17703. 28. Bi EF, Lutkenhaus J (1991) FtsZ ring structure associated with division in Escherichia coli. Nature 354:161–164. 29. Erickson HP, Taylor D, Taylor KA, Bramhill D (1996) Bacterial cell division protein FtsZ assembles into protofilament sheets and minirings, structural homologs of tubulin polymers. Proc Natl Acad Sci USA 93:519–523. 30. Osawa M, Anderson DE, Erickson HP (2008) Reconstitution of contractile FtsZ rings in liposomes. Science 320:792–794. 31. Delano WL (2002) The PyMOL Molecular Graphics System (DeLano Scientific, San Carlos, CA). 32. Pabo CO, Lewis M (1982) The operator-binding domain of λ repressor: Structure and DNA recognition. Nature 298:443–447. Fig. 6. pBtoxis DNA partition model. In the first step, TubR, which is bound to its centromere on one of the replicated pBtoxis plasmids, contacts the TubZ C-terminal region (indicated by lines pointing from the TubZ “circles”) in a treadmilling TubZ polymer. TubZ subunits are lost from the −end and are added to the þ end. TubR is pulled along the growing polymer by its TubR-TubZ interaction until it reaches the cell pole and is knocked off when it comes into contact with the membrane at the cell pole. TubZ reverses direction and may pick up the other TubR-pBtoxis complex and deliver it similarly to the opposite cell pole. 11768 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1003817107 Ni et al.
3M8F
Protein structure of type III plasmid segregation TubR mutant
Plasmid protein TubR uses a distinct mode of HTH- DNA binding and recruits the prokaryotic tubulin homolog TubZ to effect DNA partition Lisheng Ni, Weijun Xu, Muthiah Kumaraswami, and Maria A. Schumacher1 Department of Biochemistry and Molecular Biology, University of Texas, M. D. Anderson Cancer Center, Unit 1000, Houston, TX 77030 Edited by Robert T. Sauer, Massachusetts Institute of Technology, Cambridge, MA, and approved May 11, 2010 (received for review March 22, 2010) The segregation of plasmid DNA typically requires three elements: a DNA centromere site, an NTPase, and a centromere-binding protein. Because of their simplicity, plasmid partition systems represent tractable models to study the molecular basis of DNA segregation. Unlike eukaryotes, which utilize the GTPase tubulin to segregate DNA, the most common plasmid-encoded NTPases contain Walker-box and actin-like folds. Recently, a plasmid stabi- lity cassette on Bacillus thuringiensis pBtoxis encoding a putative FtsZ/tubulin-like NTPase called TubZ and DNA-binding protein called TubR has been described. How these proteins collaborate to impart plasmid stability, however, is unknown. Here we show that the TubR structure consists of an intertwined dimer with a winged helix-turn-helix (HTH) motif. Strikingly, however, the TubR recognition helices mediate dimerization, making canonical HTH– DNA interactions impossible. Mutagenesis data indicate that a basic patch, encompassing the two wing regions and the N termini of the recognition helices, mediates DNA binding, which indicates an unusual HTH–DNA interaction mode in which the N termini of the recognition helices insert into a single DNA groove and the wings into adjacent DNA grooves. The TubZ structure shows that it is as similar structurally to eukaryotic tubulin as it is to bacterial FtsZ. TubZ forms polymers with guanine nucleotide-binding characteristics and polymer dynamics similar to tubulin. Finally, we show that the exposed TubZ C-terminal region interacts with TubR-DNA, linking the TubR-bound pBtoxis to TubZ polymerization. The combined data suggest a mechanism for TubZ-polymer pow- ered plasmid movement. T he cytoskeletons of eukaryotic cells are constructed of three primary elements: actin, tubulin, and intermediate filaments. Although it had long been presumed that the proteins forming these elements were absent in prokaryotes, it is now known that prokaryotes contain structural homologs to all three components. These prokaryotic proteins appear to carry out distinct functions compared to their eukaryotic counterparts; however, their roles are similar enough to indicate a likely common ancestor. The best known prokaryotic homolog of tubulin is FtsZ. Both FtsZ and tubulin form long filamentous structures by head to tail associa- tion in a manner regulated by GTP, which binds between adjacent subunits (1–4). However, unlike tubulin, FtsZ does not function in DNA segregation but rather cell division. Specifically, it forms a cytokinetic ring called the Z ring at midcell, which mediates septation (5, 6). Recently, however, prokaryotic proteins encoded on large plasmids harbored in bacilli showing 15–20% sequence similarity to both FtsZ and tubulin have been identified and dubbed TubZ (7–12). Studies showed that the Bacillus thuringien- sis TubZ protein from the pBtoxis plasmid is essential for plasmid DNA segregation. DNA segregation of most low copy number plasmids is carried out by specific partition (par) systems. These systems require only three elements: a centromere DNA site, a centromere-binding protein, and a partition NTPase (13, 14). Partition systems have been classified into two main types on the basis of the kind of NTPase present (15). Type I systems contain NTPases with deviant Walker A-type ATPase folds, whereas type II systems uti- lize actin-like NTPases. Interestingly, both types of NTPases form polymers in NTP-dependent manners that are implicated to play a role in plasmid DNA separation (16–19). The recent discovery of TubZ NTPases has led to the designation of “type III” par sys- tems (13, 14). The best studied of these systems is that found on the pBtoxis plasmid in B. thuringiensis. This plasmid stability sys- tem is represented by an operon encoding two proteins: ORF156 (TubZ) and ORF157 (TubR) (7–9, 11). TubR is a 11.6 kDa DNA- binding protein that shows no sequence homology to any known protein. Studies showed that TubR binds a 48-bp centromere con- taining four repeat sites in the pBtoxis plasmid and also autore- gulates tubRZ transcription (8, 9). TubZ is a 54.4 kDa protein that can assemble into filaments in a GTP-dependent manner (12). Both proteins were found to be required for plasmid stability (9). However, how the TubR and TubZ proteins work together to effect pBtoxis plasmid segregation is not known. To gain insight into the molecular mechanism utilized by these proteins in DNA segregation, we carried out structural and biochemical studies on the pBtoxis TubR and TubZ proteins. The TubR structure reveals that it employs a helix-turn-helix (HTH) motif in a previously un- described manner to bind DNA. TubZ contains a tubulin/FtsZ fold but has structural distinctions from these proteins indicating that it forms distinct protofilaments. TubR binds the flexible C-terminal region of TubZ, thus attaching the TubZ filament to the pBtoxis plasmid, providing a mechanism for plasmid move- ment and, ultimately, segregation. Results and Discussion Overall Structure of pBtoxis TubR. The crystal structure of the 107- residue pBtoxis TubR protein was solved to 2.0-Å resolution by selenomethionine multiple wavelength anomalous diffraction (MAD) methods (Table S1). The structure contains two TubR molecules in the crystallographic asymmetric unit and consists of residues 6–102 of one subunit and 4–100 of the second subunit, and has Rwork∕Rfree ¼ 23.8%∕27.0%. The TubR structure forms a highly intertwined dimer with dimensions 30 × 30 × 60 Å3 (Fig. 1A). Each TubR subunit has the topology β1-α1-α2-α3-α4-β2- β3-α5, which is similar to winged HTH motifs found in a number of DNA-binding proteins in both prokaryotes and eukaryotes (20). In TubR, α3-α4 forms the HTH motif and the loop between Author contributions: M.A.S. designed research; L.N., W.X., M.K., and M.A.S. performed research; M.A.S. analyzed data; and M.A.S. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Freely available online through the PNAS open access option. Data deposition: The atomic coordinates and structure factor amplitudes for the WT TubR (C2), WT TubR (I222), TubR(S63W), TubZ, and TubZ-GTP-γ-S structures have been deposited with the Protein Data Bank, www.pdb.org (PDB ID codes 3M8E, 3M9A, 3M8F, 3M8K, and 3M89). 1To whom correspondence may be addressed. E-mail: maschuma@mdanderson.org. This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1003817107/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1003817107 PNAS ∣June 29, 2010 ∣vol. 107 ∣no. 26 ∣11763–11768 BIOCHEMISTRY β2 and β3, the wing. Indeed, each TubR subunit shows the stron- gest structural similarity to members of the ArsR family of prokaryotic winged helix transcription regulators, in particular the Staphylococcus aureus CzrA protein (21, 22). Superposition of one subunit of TubR onto that of CzrA results in a rmsd of 2.7 Å. This similarity includes the core regions of the winged HTH, because the loops and N-terminal regions of the proteins are structurally distinct. For example, TubR contains a β-strand in its N-terminal region compared to a long helix in the CzrA struc- ture (Fig. 1B). This structural similarity initially suggested that TubR may be a member of the ArsR family of proteins. However, the arrangement of the TubR dimer was found to be strikingly different from the dimer organization exhibited by the ArsR proteins (Fig. 1C). ArsR family members are involved in metal-regulated tran- scription processes whereby they act as repressors in their apo forms and are induced off their DNA sites upon metal binding (22). The specific dimer structures of the ArsR proteins are cri- tical for creation of their metal-binding motifs. Not only does TubR form a very different dimer from the ArsR proteins, it also does not harbor any of their metal-binding signatures nor does it contain any other characterized metal-binding motif. Consistent with this, we find that the addition of metals has no effect on TubR DNA binding. Dimerization of the ArsR proteins is imparted by residues from the N-terminal regions, α1 and α5, which, impor- tantly, leaves its recognition helices exposed for DNA interaction (22). By contrast, the TubR dimer is formed primarily by contacts between its twofold related “recognition helices” α4 and α4·. This interaction results in a near complete burial of these helices, leav- ing only the N-terminal residues exposed to solvent. Whereas the α4 and α4· interaction creates the dimer core, the dimer is further stabilized by interactions between the twofold related β1 strands, which swap to form an antiparallel β-sheet. Residues from α1 and α5 interact with β1 to further seal the top of the dimer. The dimer interface formed by these interactions is predominantly hydropho- bic and buries a large 1;200 Å2 of subunit surface from solvent. TubR Forms a “Recognition Helix Dimer”: Implications for DNA Bind- ing. Gel filtration studies on TubR confirmed that it is a dimer in solution. However, the finding from the structure that the TubR α4 recognition helices are buried in the dimer core has important implications in terms of its DNA-binding mechanism. Indeed, it suggests that, although TubR contains a structurally canonical HTH, it is not utilized for DNA binding in a manner typical of HTH proteins. A second crystal form (I222) of TubR, which was solved to 2.5-Å resolution, revealed the same TubR dimer. The presence of the identical dimer in two different crystal forms and its large buried surface area supports that the dimer observed in the crystal structures is physiologically relevant. However, to test this, we mutated residues within the recognition helices that the structure indicates are critical for dimerization and assayed the ability of the mutant proteins to dimerize via gel filtration. Specifically, we mutated Ser-63 and Ala-67 individually to tryp- tophan and arginine. The structure shows that residues occupying positions 63 and 67 must be small and largely hydrophobic to permit the proper packing of the α4 helices in the dimer (Fig. S1A). Hence, the in- troduction of the bulky side chain of tryptophan and, in particu- lar, the large as well as charged side chain of arginine would be predicted to be highly disruptive to dimerization. Gel filtration analyses on purified mutant proteins clearly showed that the ar- ginine mutants exist primarily as monomers in solution (>80%), whereas the tryptophan mutants were able to maintain the di- meric state (Fig. 2A and Fig. S1B). However, all mutant proteins showed reduced or loss of DNA-binding activity as ascertained by fluorescence polarization (FP) studies, which examined TubR protein binding to its centromere site (Fig. 2B and Fig. S1 C and D) (9). The fact that the monomeric mutants were severely impaired in DNA binding was not surprising. However, the finding that the tryptophan mutants, which were largely dimeric, displayed reduced DNA-binding activity suggested that their oli- gomer structures might be altered. To address this issue, the struc- ture of S63W TubR was solved to 2.8-Å resolution, resulting in Rwork∕Rfree values of 20.4%∕26.6% (Table S1). The subunit struc- ture of S63W TubR is essentially identical to that of WT TubR as revealed by their superimposition (rmsd of 1.2 Å) (Fig. 2 C and D). However, this single subunit overlay shows that the S63W TubR dimer, although the same as the WT in general arrange- ment, is forced into a more open oligomer conformation in which one subunit is rotated 20° away from its dimer mate compared to WT. This rotation is required to accommodate the bulky S63W side chains (Fig. 2D). The N-terminal β1–β1′-sheet interaction appears to play a key role in holding the TubR subunits together. In addition, the tight stacking of the twofold related Trp63 indole groups (3.5 Å) provides a compensatory interaction that, com- bined with the β1–β1′ interaction, apparently permits retention of the dimer state, indicating why the TubR tryptophan mutants were able to maintain the dimer state, albeit an altered dimer state relative to WT (Fig. 2 C and D). By contrast, the TubR arginine mutations, which introduced both bulk and charge with- in the predominantly hydrophobic dimer interface, were highly destabilizing for dimerization. In addition, the finding that the S63W TubR mutant forms an altered dimer explains the severe effect on DNA binding because a correct dimer orientation is likely essential for binding to its palindromic DNA sites (9). TubR-DNA Model. Because all but the N-terminal residues of the TubR recognition helices are buried in the dimer interface, TubR must use a different mode of DNA binding than the ArsR or other HTH containing proteins (23). Examination of surface electro- statics of TubR reveals that one face of the protein is electro- negative, whereas the other is strongly electropositive (Fig. 3A). Notably, the positive region is composed of one large and contig- uous basic patch. Basic residues in this region correspond to Arg- 74, Arg-77, and Lys-79, in the wing and Lys-43, which is on α3, the helix preceding α4 in the HTH motif. These residues were mutated singly to alanine to examine their roles in DNA binding (Fig. 3B). FP experiments showed that mutation of the basic wing residues resulted in either a complete (R74A and R77A) or nearly com- plete (K79A) abrogation of DNA binding, indicating that the wings play a major role in TubR DNA binding. Residue Lys-43 is surface exposed and located at the center of the basic region on the TubR dimer (Fig. 3A). The K43A mutant also showed no binding to TubR, supporting the notion that the continuous basic patch of TubR represents its DNA-binding surface. To gain insight into the structural mechanism of DNA binding, a DNA duplex was docked onto the basic patch of the TubR di- Fig. 1. B. thuringiensis pBtoxis TubR structure. (A) One TubR subunit is red and the other cyan. Secondary structural elements and N and C termini are labeled. (B) Superimposition of one subunit of TubR (Red) onto a S. aureus CzrA subunit (Yellow). Regions with different structures are labeled. (C) Same superimposition as B showing the location of the other subunit in the TubR and CzrA dimers after one subunit is overlaid. A–C are in the same orienta- tion to highlight differences. Figs. 1 A–C, 2 C and D, 3 A, C, and E, 4B, and 5B were made by using PyMOL (31). 11764 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1003817107 Ni et al. mer by using the location of the mutations that affected DNA binding as a guide (Fig. 3C). This model revealed that the wings are positioned to interact with successive minor grooves, with either the bases or the phosphate backbone depending on the ability of the DNA to deform. In the model TubR interacts with a minimum of 14 bp of DNA. However, the centromere bound by Fig. 2. The TubR “recognition helix” dimer. (A) Gel fil- tration studies on TubR mutants S63W and S63R show- ing that the S63W mutant remains dimeric whereas S63R is >80% monomer. (B) Fluorescence polarization studies examining the ability of WT TubR, S63R, and S63W TubR mutants to bind iteronic DNA. Fluorescence polarization units (millipolarization) and TubR concen- tration (nM) are along the y and x axes, respectively. The Kd of WT TubR for the centromere DNA is 8  2 nm. (C) Superimposition of WT TubR (Green) onto the TubR S63W mutant structure (Tan). (D) Close-up of the site of the S63W mutation in the expanded TubR S63W dimer showing stacking interactions between the twofold related tryptophans. Fig. 3. TubR-DNA binding. (A) Electrostatic surface potential of the TubR dimer. Blue and red represent electropositive and electronegative regions, respec- tively. (Left) The electronegative side of the TubR dimer, and the side on the right is the electropositive side. Labeled on the left side are the locations of the mutated residues. (B) FP binding isotherms showing the DNA binding of WT TubR and the K43A, R74A, R77A, and K79A mutants. Fluorescence polarization units (millipolarization) and TubR concentration (nM) are along the y and x axes, respectively. (C) TubR-DNA model showing TubR electrostatic potential. (D) Stoichiometry of TubR (subunit) binding showing titration curve of TubR into the 48-mer iteron resulting in a molar ratio of TubR subunit to DNA of eight (or four) dimers. (E) Left: Ribbon diagram of the TubR-DNA model with the recognition helices colored yellow. Right: Ribbon diagram showing a canonical HTH–DNA interaction (the λ repressor-DNA complex) with the recognition helices colored yellow (32). Ni et al. PNAS ∣ June 29, 2010 ∣ vol. 107 ∣ no. 26 ∣ 11765 BIOCHEMISTRY TubR consists of four 12-bp sites with the consensus T(T/A)(T/A) (C/A)(G/A)GTTTA(A/C)(A/C) (9). Thus, we used FP to ascer- tain the binding stoichiometry of TubR for its 48-bp centromere. As shown in Fig. 3D, eight TubR subunits or four TubR dimers bind the 48-mer centromere, consistent with a dimer of TubR binding each palindrome. Thus, either TubR distorts its DNA or the TubR dimers bind with some degree of overlap on their DNA sites perhaps imparting cooperativity, as observed in other centromere-binding protein–DNA interactions (13, 14). In addi- tion to the insertion of the wings, a striking outcome of the mod- eling was the finding that the N termini of the recognition helices, which interact with each other in a parallel, coiled-coil-like man- ner, are in position to insert into a single major groove. Structures of HTH proteins bound to DNA have thus far shown that the recognition helices insert singly into successive major grooves by using residues in the first few turns or the central portion of the recognition helix to contact the DNA bases (Fig. 3E). Thus, in the TubR-DNA model, the HTH–DNA interaction is drama- tically different from any displayed previously by a HTH protein. TubZ Binds TubR-DNA. A characteristic feature exhibited by partition centromere-binding proteins is the ability to bind their partner NTPase (13, 14). To determine if TubR binds TubZ, we utilized a FP assay and found that full length (FL) TubZ bound avidly to the TubR-centromere complex (Fig. 4A). However, unlike other partition systems in which the NTPase must be com- plexed with nucleotide to bind its centromere-binding protein, the interaction of TubR with TubZ did not require the presence of GTP-Mg2þ. Previous studies have shown that the C-terminal regions of tubulin and FtsZ mediate key binding events with their target proteins (24–26). We noted that the terminal region of TubZ, consisting of residues 407–484, is the most divergent region between TubZ proteins and between TubZ and tubulin/FtsZ pro- teins, suggesting that it may be similarly utilized and bind TubR. To test this hypothesis, we constructed various TubZ truncations, TubZ(1-407), TubZ(1-442), TubZ(1-460), and TubZ(1-470), and examined the ability of each protein to bind TubR-DNA. TubZ (1-407) showed no binding to TubR, whereas the remaining trun- cation mutants bound weakly to TubR-DNA (Fig. 4A). Thus, the data indicate that the last 14 amino acids of TubZ are critical for the ability of TubZ to form a tight interaction with TubR but that residues 408–470 also play an important role in this interaction. These data demonstrate that TubR acts as a partition partner for TubZ, linking it to pBtoxis plasmid DNA. Although TubZ has been shown to form polymers in a GTP-dependent manner, the TubZ protein displays limited sequence similarity to tubu- lin/FtsZ, suggesting potential differences in TubZ and tubulin/ FtsZ structures (7–9). To gain insight into TubZ function, we next determined structures of B. thuringiensis pBtoxis TubZ. Structure of TubZ. Crystallization of FLTubZ was not successful, in either its apo form or bound to guanine nucleotides. We noted that FL TubZ degraded over time whereby C-terminal residues were proteolyzed. Therefore, truncated TubZ proteins were utilized in crystallization trials, and crystals were obtained of apo TubZ (1-428) and the structure solved by MAD (Table S2). The model consists of residues 1–79 and 91–404 and has an Rwork∕Rfree of 21.4∕24.9% to 2.3-Å resolution (Fig. 4B). No discernible oligomer- ization of apo TubZ(1-428) was observed in the crystal packing, and gel filtration analyses confirmed that it is monomeric (Fig. S2). The overall TubZ structure can be divided into two main domains: an N domain (residues 25–235) and a C domain (resi- dues 258–377). These domains are connected by a long, core helix, H7. The TubZ N domain has a Rossman fold and consists of six parallel β-strands with topology 3-2-1-4-5-6. The resulting β-sheet is sandwiched by five α-helices, with two helices on one side and three on the other. The C domain consists of four β-strands with the topology 1-4-2-3. The C-domain β-strands are arranged nearly perpendicular to those in the N domain. In addition to these main protein domains, there are two helices: one at the N terminus, H0, and a long helix at the C terminus, H11 (Fig. 4B). Database searches showed that TubZ indeed belongs to the tubulin/FtsZ fa- mily of proteins and is similar to both eukaryotic and prokaryotic members of the family; TubZ can be optimally superimposed with rmsds of 3.4 Å onto both bovine α tubulin and Pseudomonas aeruginosa FtsZ (1, 5). Whereas the two-domain architectures of tubulin, FtsZ, and TubZ are similar in overall structure, the extreme N- and C-terminal regions of these proteins are very di- vergent (Fig. S3 A–D). N-Terminal and C-Terminal Differences in TubZ, FtsZ, and Tubulin: Im- plications for Polymer Formation and Target Protein Binding. Tubulin proteins do not contain significant N-terminal extensions, whereas FtsZ proteins from different organisms show structural variability within their N-terminal regions. For instance, in the Escherichia coli FtsZ structure the N-terminal residues are disor- dered, whereas Methanococcus jannaschii FtsZ has an extra N-terminal helix, H0, which is flexibly attached to the body of the protein and has been captured in multiple orientations (5). Although H0 is not conserved in FtsZ proteins, one M. jannaschii FtsZ structure revealed a semicontinuous polymer in the crystal, thought to closely represent in vivo protofilaments, which utilizes H0 in subunit-subunit contacts (4). This finding suggests that the flexibly attached H0 is stabilized in a specific orientation by protofilament formation, at least in the M. jannaschii protein. The TubZ H0 helix extends in the opposite direction compared to that of the protofilament stabilized FtsZ H0 helix. Moreover, in TubZ, H0 is not flexibly attached to the N domain but is tightly anchored to the C domain through numerous interactions with the core helix and C-domain residues. The large number of inter- actions involving H0, and the fact that it covers what would other- wise be a surface exposed hydrophobic patch, indicate that the TubZ H0 does not undergo conformational changes during protofilament formation and is important for the general fold of TubZ (Fig. S3 A and C). Data suggest that FtsZ and tubulin form protofilaments with similar longitudinal contacts (4). However, the TubZ structure reveals key differences, primarily in its C-domain and C-terminal regions, which suggest that it forms protofilaments distinct from those formed by tubulin/FtsZ. A notable difference is the struc- ture of loop 7 (L7). This loop inserts into the adjacent subunit providing the key catalytic residues required for GTP hydrolysis. In tubulin/FtsZ proteins, L7 has the consensus GXXNXDXAD. In TubZ, the loop is very divergent in conformation compared Fig. 4. TubZ interacts with TubR-DNA and contains a tubulin/FtsZ fold. (A) FP assay measuring binding of FL TubZ, TubZ(1-470), TubZ(1-460), TubZ(1-442), and TubZ(1-407) to TubR-DNA. Below is the control (TubZ titrated into DNA alone). Millipolarization units and TubZ concentration (nM) are along the y and x axes, respectively. (B) TubZ(1-428) structure. The N domain or GTP-bind- ing domain is colored salmon and the C domain purple. The interdomain helix, H7, is red. TubZ also contains an N-terminal helix, H0 (Yellow), and a C-terminal helix, H11 (White). 11766 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1003817107 Ni et al. to that in FtsZ/tubulin and consists of the sequence 256- DNVTYDPSD-266. In addition to the N-terminal region, the extreme C-terminal extensions of tubulin, FtsZ, and TubZ are structurally divergent (Fig. S3B). In FtsZ, the C-terminal region forms a small, two-stranded β-sheet and continues into an ex- tended region that is involved in binding adaptor proteins such as FtsA and ZipA (6, 25). By contrast, the C-terminal regions of tubulin proteins consist of a two-helix bundle followed by an extended region. Like FtsZ, however, these regions interact with numerous target proteins such as the microtubule-associated pro- teins (MAPs) (24). Consistent with this, the C-terminal regions of tubulin have been shown to face the outside of the microtubule. A characteristic feature of the extreme C-terminal extensions of tubulin proteins is their highly acidic nature (3). This acidic region has been shown to be critical for binding to several MAPs that harbor a substantial basic character, such as tau, MAP2, and MAP4 (24, 27). The TubZ C-terminal region is also helical, but it contains a single, long helix. Notably, the TubZ-tubulin overlay shows that the long C-terminal helix of TubZ would dramatically clash with the adjacent subunit in a polymer, providing support for the no- tion that TubZ forms protofilaments different from tubulin/FtsZ (Figs. S3B and S4). Interestingly, and in contrast to tubulin pro- teins, the flexible C-terminal region of TubZ that follows H11 is highly basic, in particular the last 14 residues. We have shown that these residues play a central role in TubR binding (Fig. 4A). TubR uses its electropositive face for DNA binding, leaving exposed its opposite face for TubZ interaction. Notably, this exposed face is strongly electronegative and hence would complement the basic C-terminal tail of TubZ (Fig. 3A). Tubulin/FtsZ protofilaments combine to form higher-order structures. In tubulin, the protofilaments interact in a parallel manner to form microtubules. Central to microtubule formation are lateral contacts between protofilaments from the so-called M loop, between H10 and S9. In tubulin, this loop is composed of 13 residues (1–3). The corresponding loop is much shorter in FtsZ proteins, consistent with the fact that FtsZ does not form tubulin microtubule-like structures (5, 28–30). In TubZ, the M loop is even shorter than in FtsZ, spanning only four residues. In fact, the TubZ/tubulin overlay shows that the side of the molecule con- taining the M loop is the most divergent between these proteins. These combined findings suggest that TubZ not only forms pro- tofilaments with distinct longitudinal contacts compared to FtsZ and tubulin, but it also does not form tubulin-like microtubule structures. TubZ Interactions with Guanine Nucleotides. Consistent with TubZ being a member of the tubulin/FtsZ family, our isothermal titra- tion calorimetry (ITC) studies showed that TubZ binds guanine nucleotides with high affinity; Kds for GTP-γ-S and GDP were ∼0.69 and 26 μM, respectively (Fig. 5A). We next determined the structure of the TubZ-GTP-γ-S complex by soaking GTP-γ-S into preformed TubZ(1-428) crystals. The TubZ-GTP-γ-S struc- ture contains TubZ residues 1–79 and 91–404, one GTP-γ-S, and has Rwork∕Rfree ¼ 21.8%∕25.5% (Fig. 5B and Table S2). The structure shows that TubZ binds GTP-γ-S in the same GTP binding pocket as tubulin/FtsZ (1–5). Comparison of the apo and GTP-γ-S bound TubZ structures indicated that, like FtsZ, guanine nucleotide binding does not lead to significant conforma- tional changes (5). The phosphate binding pocket is formed by two of the most highly conserved regions between TubZ and tubulin/FtsZ called loops 1 and 4 (L1 and L4) (1–3). L1 contacts the GTP-γ-S α- and β-phosphate groups via the Gln-32 and Lys-33 amide nitrogens. The L1 region of FtsZ and tubulin contain the sequences GQ(A/G)G and GQCG, respectively, whereas in TubZ the motif is 31-GQKG-34. However, the alanine/glycine and cysteine residues in FtsZ and tubulin do not contact the bound nucleotide; the TubZ Lys-33 side chain makes stacking interac- tions with the guanine base (Fig. 5B). L4 represents the so-called signature motif [GGGTG(T/S)G], which serves as an identifier of tubulin/FtsZ family members. Like FtsZ and tubulin, the L4 region of TubZ-GTP-γ-S makes phosphate interactions via its glycine amide nitrogens. Whereas L1 and L4 residues of the N domain mediate phosphate contacts, the GTP-γ-S guanine moiety is specified from residues in the core helix, H5, and C domain. In this regard, an important motif is loop 6 (L6). In FtsZ and tubulin, L6 has the consensus (F/Y)XXX(N/D) and the conserved (F/Y) residue functions in guanine base stacking. This region in TubZ, 236-WKXXXN-241, is in an altered conformation compared to FtsZ and tubulin structures. Despite the presence of the trypto- phan, which might be expected to interact with the guanine, the side chain of Lys-237 instead stacks with the guanine ring. Hence, in the TubZ-GTP-γ-S structure, the guanine base does not interact with aromatic residues as in tubulin/FtsZ but is sand- wiched between the aliphatic portions of two lysine side chains, Lys-33 and Lys-237. Finally, two asparagine residues, Asn-213 and Asn-241, from L6 effectively read the guanine N2/N3 and N1/O6 atoms, respectively, providing high specificity in TubZ’s in- teraction with guanine nucleotides. pBtoxis DNA Segregation: TubR Plasmid Partition Model. Our data show that TubR binds to the flexible, C-terminal, basic region of TubZ. The flexibility and location of the TubZ C-terminal extension suggest that it is not required for polymerization and thus may be exposed on the surface of TubZ filaments. Indeed, negative stain EM images show that TubZ(1-407) forms polymers in a GTP-dependent manner similar to the FL protein (Fig. S5). Recent data suggesting that TubZ filaments are stabilized by Fig. 5. TubZ-guanine nucleotide interactions. (A) ITC binding isotherms showing TubZ-GDP (Left) and TubZ-GTP-γ-S interaction (Right). (B) Left: Overall structure of the TubZ-GTP-γ-S complex. β-strands are colored magen- ta and helices cyan, and the GTP-γ-S molecule is shown as cpk. Right: Close-up view of the GTP binding pocket with the initial Fo-Fc electron density map (Blue Mesh), contoured at 4.5σ, and calculated before the GTP-γ-S was included in refinement. Ni et al. PNAS ∣ June 29, 2010 ∣ vol. 107 ∣ no. 26 ∣ 11767 BIOCHEMISTRY the presence of a GTP cap and undergo treadmilling are consis- tent with the notion that TubZ displays tubulin-like polymer dy- namics (12). Thus, on the basis of the combined data, we suggest a model for TubR/TubZ mediated pBtoxis plasmid segregation shown in Fig. 6. In this model, multiple TubR dimers first bind to the iteronic DNA on the pBtoxis plasmid leading to the crea- tion of a high local concentration of TubR, which can recruit a TubZ polymer, likely by interactions between the acidic TubR dimer face and the basic C-terminal TubZ region. Importantly, this interaction serves to attach the pBtoxis plasmid to the TubZ polymer, which undergoes treadmilling, adding subunits at the þ end and losing subunits at the −end. The bound TubR-pBtoxis can be handed off from the −end to the molecules in the growing þ end, leading to the transport of the pBtoxis plasmid to the cell pole. Interestingly, it has been shown that once TubZ polymers reach and interact with the cell pole, they bend around the curved pole and continue growing in the other direction (7). The force of the interaction with the membrane likely causes the release of TubR-pBtoxis, the net result being transport of pBtoxis to the cell pole. Of course, this model is simplified and many questions re- main. For example, how directionality is achieved and how the replicated plasmids are driven to opposite cell poles is not clear. However, given the large size of the pBtoxis plasmid (8), it may be that only one TubR-pBtoxis “tram” can be bound at once by the rapidly treadmilling TubZ polymer and that, once one such a tram is unloaded after reaching the cell pole, another engages when the now reversed polymer treadmills toward the opposite cell pole. Materials and Methods Summary Detailed methods are provided in SI Materials and Methods. Briefly, the tubR and tubZ genes were codon optimized (for E. coli expression), subcloned into pET15b, expressed, and pur- ified. WT TubR crystals were grown with NaCl and phosphate. TubR S63W was crystallized with PEG and ethylene glycol and TubZ with sodium formate. Detailed assay conditions for FP, ITC, electron microscopy, and gel filtration are provided in SI Materials and Methods. ACKNOWLEDGMENTS. This work was supported by the Burroughs Wellcome Career Development Award 992863 and National Institutes of Health Grant GM074815 (to M.A.S.). 1. Nogales E, Wolf SG, Downing KH (1998) Structure of the αβ tubulin dimer by electron crystallography. Nature 391:199–203. 2. Nogales E, Whittaker M, Milligan RA, Downing KH (1999) High-resolution model of the microtubule. Cell 96:79–88. 3. Nogales E (2000) Structural insights into microtubule function. Annu Rev Biochem 69:277–302. 4. Oliva MA, Cordell SC, Löwe J (2004) Structural insights into FtsZ protofilament formation. Nat Struct Mol Biol 11:1243–1250. 5. Oliva MA, Trambaiolo D, Löwe J (2007) Structural insights into the conformational variability of FtsZ. J Mol Biol 373:1229–1242. 6. Margolin W (2005) FtsZ and the division of prokaryotic cells and organelles. Nat Rev Mol Cell Biol 6:862–871. 7. Larsen RA, et al. (2007) Treadmilling of a prokaryotic tubulin-like protein, TubZ, required for plasmid stability in Bacillus thuringiensis. Genes Dev 21:1340–1352. 8. Tang M, Bideshi DK, Park H-W, Federici BA (2006) Minireplicon from pBtoxis of Bacillus thuringiensis subsp. israelensis. App Environ Microbiol 72:6948–6954. 9. Tang M, Bideshi DK, Park H-W, Federici BA (2007) Iteron-binding ORF157 and FtsZ-like ORF156 proteins encoded by pBtoxis play a role in its replication in Bacillus thuringien- sis subsp. israelensis. J Bacteriol 189:8053–8058. 10. Anand SP, Akhtar P, Tinsley E, Watkins SC, Khan SA (2008) GTP-dependent polymer- ization of the tubulin-like RepX replication protein encoded by the pXO1 plasmid of Bacillus anthracis. Mol Microbiol 67:881–890. 11. Berry C, et al. (2002) Complete sequence and organization of pBtoxis, the toxin- coding plasmid of Bacillus thuringiensis subsp israelensis. Appl Environ Microbiol 68:5082–5095. 12. Chen Y, Erickson HP (2008) In vitro assembly studies of FtsZ/tubulin-like proteins (TubZ) from Bacillus plasmids: Evidence for a capping mechanism. J Biol Chem 283:8102–8109. 13. Hayes F, Barillà D (2006) The bacterial segrosome: A dynamic nucleoprotein machine for DNA trafficking and segregation. Nat Rev Microbiol 4:133–143. 14. Schumacher MA (2008) Structural biology of plasmid partition: Uncovering the molecular mechanisms of DNA segregation. Biochem J 412:1–18. 15. Gerdes K, Møller-Jensen J, Bugge Jensen R (2000) Plasmid and chromosome partitioning: Surprises from phylogeny. Mol Microbiol 37:455–466. 16. Møller-Jensen J, et al. (2003) Bacterial mitosis: ParM of plasmid R1 moves plasmid DNA by an actin-like insertional polymerization mechanism. Mol Cell 12:1477–1487. 17. Popp D, et al. (2008) Molecular structure of the ParM polymer and the mechanism leading to its nucleotide-driven dynamic instability. EMBO J 27:570–579. 18. Salje J, Löwe J (2008) Bacterial actin: Architecture of the ParMRC DNA partitioning complex. EMBO J 27:2230–2238. 19. Dunham TD, Xu W, Funnell BE, Schumacher MA (2009) Structural basis for ADP- mediated transcriptional regulation by P1 and P7 ParA. EMBO J 28:1792–1802. 20. Gajiwala KS, Burley SK (2000) Winged helix proteins. Curr Opin Struct Biol 10:110–116. 21. Eicken C, et al. (2003) A metal-ligand-mediated intersubunit allosteric switch in related SmtB/ArsR zinc sensor proteins. J Mol Biol 333:683–695. 22. Pennella M, Giedroc DP (2005) Structural determinants of metal selectivity in prokaryotic metal-responsive transcriptional regulators. Biometals 18:413–428. 23. Arunkumar AI, Campanello GC, Giedroc DP (2009) Solution structure of a paradigm ArsR family zinc sensor in the DNA-bound state. Proc Natl Acad Sci USA 106:18177–18182. 24. Downing KH (2000) Structural basis for the interaction of tubulin with proteins and drugs that affect microtubule dynamics. Annu Rev Cell Dev Biol 16:89–111. 25. Adams DW, Errington J (2009) Bacterial cell division: Assembly, maintenance and disassembly of the Z ring. Nat Rev Microbiol 7:642–653. 26. Errington J, Daniel RA, Scheffers DJ (2003) Cytokinesis in bacteria. Microbiol Mol Biol Rev 67:52–65. 27. Chau MF, et al. (1998) The microtubule-associated protein tau cross-links to two distinct sites on each alpha and beta tubulin monomer via separate domains. Biochemistry 37:17692–17703. 28. Bi EF, Lutkenhaus J (1991) FtsZ ring structure associated with division in Escherichia coli. Nature 354:161–164. 29. Erickson HP, Taylor D, Taylor KA, Bramhill D (1996) Bacterial cell division protein FtsZ assembles into protofilament sheets and minirings, structural homologs of tubulin polymers. Proc Natl Acad Sci USA 93:519–523. 30. Osawa M, Anderson DE, Erickson HP (2008) Reconstitution of contractile FtsZ rings in liposomes. Science 320:792–794. 31. Delano WL (2002) The PyMOL Molecular Graphics System (DeLano Scientific, San Carlos, CA). 32. Pabo CO, Lewis M (1982) The operator-binding domain of λ repressor: Structure and DNA recognition. Nature 298:443–447. Fig. 6. pBtoxis DNA partition model. In the first step, TubR, which is bound to its centromere on one of the replicated pBtoxis plasmids, contacts the TubZ C-terminal region (indicated by lines pointing from the TubZ “circles”) in a treadmilling TubZ polymer. TubZ subunits are lost from the −end and are added to the þ end. TubR is pulled along the growing polymer by its TubR-TubZ interaction until it reaches the cell pole and is knocked off when it comes into contact with the membrane at the cell pole. TubZ reverses direction and may pick up the other TubR-pBtoxis complex and deliver it similarly to the opposite cell pole. 11768 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1003817107 Ni et al.
3M8K
Protein structure of type III plasmid segregation TubZ
Plasmid protein TubR uses a distinct mode of HTH- DNA binding and recruits the prokaryotic tubulin homolog TubZ to effect DNA partition Lisheng Ni, Weijun Xu, Muthiah Kumaraswami, and Maria A. Schumacher1 Department of Biochemistry and Molecular Biology, University of Texas, M. D. Anderson Cancer Center, Unit 1000, Houston, TX 77030 Edited by Robert T. Sauer, Massachusetts Institute of Technology, Cambridge, MA, and approved May 11, 2010 (received for review March 22, 2010) The segregation of plasmid DNA typically requires three elements: a DNA centromere site, an NTPase, and a centromere-binding protein. Because of their simplicity, plasmid partition systems represent tractable models to study the molecular basis of DNA segregation. Unlike eukaryotes, which utilize the GTPase tubulin to segregate DNA, the most common plasmid-encoded NTPases contain Walker-box and actin-like folds. Recently, a plasmid stabi- lity cassette on Bacillus thuringiensis pBtoxis encoding a putative FtsZ/tubulin-like NTPase called TubZ and DNA-binding protein called TubR has been described. How these proteins collaborate to impart plasmid stability, however, is unknown. Here we show that the TubR structure consists of an intertwined dimer with a winged helix-turn-helix (HTH) motif. Strikingly, however, the TubR recognition helices mediate dimerization, making canonical HTH– DNA interactions impossible. Mutagenesis data indicate that a basic patch, encompassing the two wing regions and the N termini of the recognition helices, mediates DNA binding, which indicates an unusual HTH–DNA interaction mode in which the N termini of the recognition helices insert into a single DNA groove and the wings into adjacent DNA grooves. The TubZ structure shows that it is as similar structurally to eukaryotic tubulin as it is to bacterial FtsZ. TubZ forms polymers with guanine nucleotide-binding characteristics and polymer dynamics similar to tubulin. Finally, we show that the exposed TubZ C-terminal region interacts with TubR-DNA, linking the TubR-bound pBtoxis to TubZ polymerization. The combined data suggest a mechanism for TubZ-polymer pow- ered plasmid movement. T he cytoskeletons of eukaryotic cells are constructed of three primary elements: actin, tubulin, and intermediate filaments. Although it had long been presumed that the proteins forming these elements were absent in prokaryotes, it is now known that prokaryotes contain structural homologs to all three components. These prokaryotic proteins appear to carry out distinct functions compared to their eukaryotic counterparts; however, their roles are similar enough to indicate a likely common ancestor. The best known prokaryotic homolog of tubulin is FtsZ. Both FtsZ and tubulin form long filamentous structures by head to tail associa- tion in a manner regulated by GTP, which binds between adjacent subunits (1–4). However, unlike tubulin, FtsZ does not function in DNA segregation but rather cell division. Specifically, it forms a cytokinetic ring called the Z ring at midcell, which mediates septation (5, 6). Recently, however, prokaryotic proteins encoded on large plasmids harbored in bacilli showing 15–20% sequence similarity to both FtsZ and tubulin have been identified and dubbed TubZ (7–12). Studies showed that the Bacillus thuringien- sis TubZ protein from the pBtoxis plasmid is essential for plasmid DNA segregation. DNA segregation of most low copy number plasmids is carried out by specific partition (par) systems. These systems require only three elements: a centromere DNA site, a centromere-binding protein, and a partition NTPase (13, 14). Partition systems have been classified into two main types on the basis of the kind of NTPase present (15). Type I systems contain NTPases with deviant Walker A-type ATPase folds, whereas type II systems uti- lize actin-like NTPases. Interestingly, both types of NTPases form polymers in NTP-dependent manners that are implicated to play a role in plasmid DNA separation (16–19). The recent discovery of TubZ NTPases has led to the designation of “type III” par sys- tems (13, 14). The best studied of these systems is that found on the pBtoxis plasmid in B. thuringiensis. This plasmid stability sys- tem is represented by an operon encoding two proteins: ORF156 (TubZ) and ORF157 (TubR) (7–9, 11). TubR is a 11.6 kDa DNA- binding protein that shows no sequence homology to any known protein. Studies showed that TubR binds a 48-bp centromere con- taining four repeat sites in the pBtoxis plasmid and also autore- gulates tubRZ transcription (8, 9). TubZ is a 54.4 kDa protein that can assemble into filaments in a GTP-dependent manner (12). Both proteins were found to be required for plasmid stability (9). However, how the TubR and TubZ proteins work together to effect pBtoxis plasmid segregation is not known. To gain insight into the molecular mechanism utilized by these proteins in DNA segregation, we carried out structural and biochemical studies on the pBtoxis TubR and TubZ proteins. The TubR structure reveals that it employs a helix-turn-helix (HTH) motif in a previously un- described manner to bind DNA. TubZ contains a tubulin/FtsZ fold but has structural distinctions from these proteins indicating that it forms distinct protofilaments. TubR binds the flexible C-terminal region of TubZ, thus attaching the TubZ filament to the pBtoxis plasmid, providing a mechanism for plasmid move- ment and, ultimately, segregation. Results and Discussion Overall Structure of pBtoxis TubR. The crystal structure of the 107- residue pBtoxis TubR protein was solved to 2.0-Å resolution by selenomethionine multiple wavelength anomalous diffraction (MAD) methods (Table S1). The structure contains two TubR molecules in the crystallographic asymmetric unit and consists of residues 6–102 of one subunit and 4–100 of the second subunit, and has Rwork∕Rfree ¼ 23.8%∕27.0%. The TubR structure forms a highly intertwined dimer with dimensions 30 × 30 × 60 Å3 (Fig. 1A). Each TubR subunit has the topology β1-α1-α2-α3-α4-β2- β3-α5, which is similar to winged HTH motifs found in a number of DNA-binding proteins in both prokaryotes and eukaryotes (20). In TubR, α3-α4 forms the HTH motif and the loop between Author contributions: M.A.S. designed research; L.N., W.X., M.K., and M.A.S. performed research; M.A.S. analyzed data; and M.A.S. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Freely available online through the PNAS open access option. Data deposition: The atomic coordinates and structure factor amplitudes for the WT TubR (C2), WT TubR (I222), TubR(S63W), TubZ, and TubZ-GTP-γ-S structures have been deposited with the Protein Data Bank, www.pdb.org (PDB ID codes 3M8E, 3M9A, 3M8F, 3M8K, and 3M89). 1To whom correspondence may be addressed. E-mail: maschuma@mdanderson.org. This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1003817107/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1003817107 PNAS ∣June 29, 2010 ∣vol. 107 ∣no. 26 ∣11763–11768 BIOCHEMISTRY β2 and β3, the wing. Indeed, each TubR subunit shows the stron- gest structural similarity to members of the ArsR family of prokaryotic winged helix transcription regulators, in particular the Staphylococcus aureus CzrA protein (21, 22). Superposition of one subunit of TubR onto that of CzrA results in a rmsd of 2.7 Å. This similarity includes the core regions of the winged HTH, because the loops and N-terminal regions of the proteins are structurally distinct. For example, TubR contains a β-strand in its N-terminal region compared to a long helix in the CzrA struc- ture (Fig. 1B). This structural similarity initially suggested that TubR may be a member of the ArsR family of proteins. However, the arrangement of the TubR dimer was found to be strikingly different from the dimer organization exhibited by the ArsR proteins (Fig. 1C). ArsR family members are involved in metal-regulated tran- scription processes whereby they act as repressors in their apo forms and are induced off their DNA sites upon metal binding (22). The specific dimer structures of the ArsR proteins are cri- tical for creation of their metal-binding motifs. Not only does TubR form a very different dimer from the ArsR proteins, it also does not harbor any of their metal-binding signatures nor does it contain any other characterized metal-binding motif. Consistent with this, we find that the addition of metals has no effect on TubR DNA binding. Dimerization of the ArsR proteins is imparted by residues from the N-terminal regions, α1 and α5, which, impor- tantly, leaves its recognition helices exposed for DNA interaction (22). By contrast, the TubR dimer is formed primarily by contacts between its twofold related “recognition helices” α4 and α4·. This interaction results in a near complete burial of these helices, leav- ing only the N-terminal residues exposed to solvent. Whereas the α4 and α4· interaction creates the dimer core, the dimer is further stabilized by interactions between the twofold related β1 strands, which swap to form an antiparallel β-sheet. Residues from α1 and α5 interact with β1 to further seal the top of the dimer. The dimer interface formed by these interactions is predominantly hydropho- bic and buries a large 1;200 Å2 of subunit surface from solvent. TubR Forms a “Recognition Helix Dimer”: Implications for DNA Bind- ing. Gel filtration studies on TubR confirmed that it is a dimer in solution. However, the finding from the structure that the TubR α4 recognition helices are buried in the dimer core has important implications in terms of its DNA-binding mechanism. Indeed, it suggests that, although TubR contains a structurally canonical HTH, it is not utilized for DNA binding in a manner typical of HTH proteins. A second crystal form (I222) of TubR, which was solved to 2.5-Å resolution, revealed the same TubR dimer. The presence of the identical dimer in two different crystal forms and its large buried surface area supports that the dimer observed in the crystal structures is physiologically relevant. However, to test this, we mutated residues within the recognition helices that the structure indicates are critical for dimerization and assayed the ability of the mutant proteins to dimerize via gel filtration. Specifically, we mutated Ser-63 and Ala-67 individually to tryp- tophan and arginine. The structure shows that residues occupying positions 63 and 67 must be small and largely hydrophobic to permit the proper packing of the α4 helices in the dimer (Fig. S1A). Hence, the in- troduction of the bulky side chain of tryptophan and, in particu- lar, the large as well as charged side chain of arginine would be predicted to be highly disruptive to dimerization. Gel filtration analyses on purified mutant proteins clearly showed that the ar- ginine mutants exist primarily as monomers in solution (>80%), whereas the tryptophan mutants were able to maintain the di- meric state (Fig. 2A and Fig. S1B). However, all mutant proteins showed reduced or loss of DNA-binding activity as ascertained by fluorescence polarization (FP) studies, which examined TubR protein binding to its centromere site (Fig. 2B and Fig. S1 C and D) (9). The fact that the monomeric mutants were severely impaired in DNA binding was not surprising. However, the finding that the tryptophan mutants, which were largely dimeric, displayed reduced DNA-binding activity suggested that their oli- gomer structures might be altered. To address this issue, the struc- ture of S63W TubR was solved to 2.8-Å resolution, resulting in Rwork∕Rfree values of 20.4%∕26.6% (Table S1). The subunit struc- ture of S63W TubR is essentially identical to that of WT TubR as revealed by their superimposition (rmsd of 1.2 Å) (Fig. 2 C and D). However, this single subunit overlay shows that the S63W TubR dimer, although the same as the WT in general arrange- ment, is forced into a more open oligomer conformation in which one subunit is rotated 20° away from its dimer mate compared to WT. This rotation is required to accommodate the bulky S63W side chains (Fig. 2D). The N-terminal β1–β1′-sheet interaction appears to play a key role in holding the TubR subunits together. In addition, the tight stacking of the twofold related Trp63 indole groups (3.5 Å) provides a compensatory interaction that, com- bined with the β1–β1′ interaction, apparently permits retention of the dimer state, indicating why the TubR tryptophan mutants were able to maintain the dimer state, albeit an altered dimer state relative to WT (Fig. 2 C and D). By contrast, the TubR arginine mutations, which introduced both bulk and charge with- in the predominantly hydrophobic dimer interface, were highly destabilizing for dimerization. In addition, the finding that the S63W TubR mutant forms an altered dimer explains the severe effect on DNA binding because a correct dimer orientation is likely essential for binding to its palindromic DNA sites (9). TubR-DNA Model. Because all but the N-terminal residues of the TubR recognition helices are buried in the dimer interface, TubR must use a different mode of DNA binding than the ArsR or other HTH containing proteins (23). Examination of surface electro- statics of TubR reveals that one face of the protein is electro- negative, whereas the other is strongly electropositive (Fig. 3A). Notably, the positive region is composed of one large and contig- uous basic patch. Basic residues in this region correspond to Arg- 74, Arg-77, and Lys-79, in the wing and Lys-43, which is on α3, the helix preceding α4 in the HTH motif. These residues were mutated singly to alanine to examine their roles in DNA binding (Fig. 3B). FP experiments showed that mutation of the basic wing residues resulted in either a complete (R74A and R77A) or nearly com- plete (K79A) abrogation of DNA binding, indicating that the wings play a major role in TubR DNA binding. Residue Lys-43 is surface exposed and located at the center of the basic region on the TubR dimer (Fig. 3A). The K43A mutant also showed no binding to TubR, supporting the notion that the continuous basic patch of TubR represents its DNA-binding surface. To gain insight into the structural mechanism of DNA binding, a DNA duplex was docked onto the basic patch of the TubR di- Fig. 1. B. thuringiensis pBtoxis TubR structure. (A) One TubR subunit is red and the other cyan. Secondary structural elements and N and C termini are labeled. (B) Superimposition of one subunit of TubR (Red) onto a S. aureus CzrA subunit (Yellow). Regions with different structures are labeled. (C) Same superimposition as B showing the location of the other subunit in the TubR and CzrA dimers after one subunit is overlaid. A–C are in the same orienta- tion to highlight differences. Figs. 1 A–C, 2 C and D, 3 A, C, and E, 4B, and 5B were made by using PyMOL (31). 11764 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1003817107 Ni et al. mer by using the location of the mutations that affected DNA binding as a guide (Fig. 3C). This model revealed that the wings are positioned to interact with successive minor grooves, with either the bases or the phosphate backbone depending on the ability of the DNA to deform. In the model TubR interacts with a minimum of 14 bp of DNA. However, the centromere bound by Fig. 2. The TubR “recognition helix” dimer. (A) Gel fil- tration studies on TubR mutants S63W and S63R show- ing that the S63W mutant remains dimeric whereas S63R is >80% monomer. (B) Fluorescence polarization studies examining the ability of WT TubR, S63R, and S63W TubR mutants to bind iteronic DNA. Fluorescence polarization units (millipolarization) and TubR concen- tration (nM) are along the y and x axes, respectively. The Kd of WT TubR for the centromere DNA is 8  2 nm. (C) Superimposition of WT TubR (Green) onto the TubR S63W mutant structure (Tan). (D) Close-up of the site of the S63W mutation in the expanded TubR S63W dimer showing stacking interactions between the twofold related tryptophans. Fig. 3. TubR-DNA binding. (A) Electrostatic surface potential of the TubR dimer. Blue and red represent electropositive and electronegative regions, respec- tively. (Left) The electronegative side of the TubR dimer, and the side on the right is the electropositive side. Labeled on the left side are the locations of the mutated residues. (B) FP binding isotherms showing the DNA binding of WT TubR and the K43A, R74A, R77A, and K79A mutants. Fluorescence polarization units (millipolarization) and TubR concentration (nM) are along the y and x axes, respectively. (C) TubR-DNA model showing TubR electrostatic potential. (D) Stoichiometry of TubR (subunit) binding showing titration curve of TubR into the 48-mer iteron resulting in a molar ratio of TubR subunit to DNA of eight (or four) dimers. (E) Left: Ribbon diagram of the TubR-DNA model with the recognition helices colored yellow. Right: Ribbon diagram showing a canonical HTH–DNA interaction (the λ repressor-DNA complex) with the recognition helices colored yellow (32). Ni et al. PNAS ∣ June 29, 2010 ∣ vol. 107 ∣ no. 26 ∣ 11765 BIOCHEMISTRY TubR consists of four 12-bp sites with the consensus T(T/A)(T/A) (C/A)(G/A)GTTTA(A/C)(A/C) (9). Thus, we used FP to ascer- tain the binding stoichiometry of TubR for its 48-bp centromere. As shown in Fig. 3D, eight TubR subunits or four TubR dimers bind the 48-mer centromere, consistent with a dimer of TubR binding each palindrome. Thus, either TubR distorts its DNA or the TubR dimers bind with some degree of overlap on their DNA sites perhaps imparting cooperativity, as observed in other centromere-binding protein–DNA interactions (13, 14). In addi- tion to the insertion of the wings, a striking outcome of the mod- eling was the finding that the N termini of the recognition helices, which interact with each other in a parallel, coiled-coil-like man- ner, are in position to insert into a single major groove. Structures of HTH proteins bound to DNA have thus far shown that the recognition helices insert singly into successive major grooves by using residues in the first few turns or the central portion of the recognition helix to contact the DNA bases (Fig. 3E). Thus, in the TubR-DNA model, the HTH–DNA interaction is drama- tically different from any displayed previously by a HTH protein. TubZ Binds TubR-DNA. A characteristic feature exhibited by partition centromere-binding proteins is the ability to bind their partner NTPase (13, 14). To determine if TubR binds TubZ, we utilized a FP assay and found that full length (FL) TubZ bound avidly to the TubR-centromere complex (Fig. 4A). However, unlike other partition systems in which the NTPase must be com- plexed with nucleotide to bind its centromere-binding protein, the interaction of TubR with TubZ did not require the presence of GTP-Mg2þ. Previous studies have shown that the C-terminal regions of tubulin and FtsZ mediate key binding events with their target proteins (24–26). We noted that the terminal region of TubZ, consisting of residues 407–484, is the most divergent region between TubZ proteins and between TubZ and tubulin/FtsZ pro- teins, suggesting that it may be similarly utilized and bind TubR. To test this hypothesis, we constructed various TubZ truncations, TubZ(1-407), TubZ(1-442), TubZ(1-460), and TubZ(1-470), and examined the ability of each protein to bind TubR-DNA. TubZ (1-407) showed no binding to TubR, whereas the remaining trun- cation mutants bound weakly to TubR-DNA (Fig. 4A). Thus, the data indicate that the last 14 amino acids of TubZ are critical for the ability of TubZ to form a tight interaction with TubR but that residues 408–470 also play an important role in this interaction. These data demonstrate that TubR acts as a partition partner for TubZ, linking it to pBtoxis plasmid DNA. Although TubZ has been shown to form polymers in a GTP-dependent manner, the TubZ protein displays limited sequence similarity to tubu- lin/FtsZ, suggesting potential differences in TubZ and tubulin/ FtsZ structures (7–9). To gain insight into TubZ function, we next determined structures of B. thuringiensis pBtoxis TubZ. Structure of TubZ. Crystallization of FLTubZ was not successful, in either its apo form or bound to guanine nucleotides. We noted that FL TubZ degraded over time whereby C-terminal residues were proteolyzed. Therefore, truncated TubZ proteins were utilized in crystallization trials, and crystals were obtained of apo TubZ (1-428) and the structure solved by MAD (Table S2). The model consists of residues 1–79 and 91–404 and has an Rwork∕Rfree of 21.4∕24.9% to 2.3-Å resolution (Fig. 4B). No discernible oligomer- ization of apo TubZ(1-428) was observed in the crystal packing, and gel filtration analyses confirmed that it is monomeric (Fig. S2). The overall TubZ structure can be divided into two main domains: an N domain (residues 25–235) and a C domain (resi- dues 258–377). These domains are connected by a long, core helix, H7. The TubZ N domain has a Rossman fold and consists of six parallel β-strands with topology 3-2-1-4-5-6. The resulting β-sheet is sandwiched by five α-helices, with two helices on one side and three on the other. The C domain consists of four β-strands with the topology 1-4-2-3. The C-domain β-strands are arranged nearly perpendicular to those in the N domain. In addition to these main protein domains, there are two helices: one at the N terminus, H0, and a long helix at the C terminus, H11 (Fig. 4B). Database searches showed that TubZ indeed belongs to the tubulin/FtsZ fa- mily of proteins and is similar to both eukaryotic and prokaryotic members of the family; TubZ can be optimally superimposed with rmsds of 3.4 Å onto both bovine α tubulin and Pseudomonas aeruginosa FtsZ (1, 5). Whereas the two-domain architectures of tubulin, FtsZ, and TubZ are similar in overall structure, the extreme N- and C-terminal regions of these proteins are very di- vergent (Fig. S3 A–D). N-Terminal and C-Terminal Differences in TubZ, FtsZ, and Tubulin: Im- plications for Polymer Formation and Target Protein Binding. Tubulin proteins do not contain significant N-terminal extensions, whereas FtsZ proteins from different organisms show structural variability within their N-terminal regions. For instance, in the Escherichia coli FtsZ structure the N-terminal residues are disor- dered, whereas Methanococcus jannaschii FtsZ has an extra N-terminal helix, H0, which is flexibly attached to the body of the protein and has been captured in multiple orientations (5). Although H0 is not conserved in FtsZ proteins, one M. jannaschii FtsZ structure revealed a semicontinuous polymer in the crystal, thought to closely represent in vivo protofilaments, which utilizes H0 in subunit-subunit contacts (4). This finding suggests that the flexibly attached H0 is stabilized in a specific orientation by protofilament formation, at least in the M. jannaschii protein. The TubZ H0 helix extends in the opposite direction compared to that of the protofilament stabilized FtsZ H0 helix. Moreover, in TubZ, H0 is not flexibly attached to the N domain but is tightly anchored to the C domain through numerous interactions with the core helix and C-domain residues. The large number of inter- actions involving H0, and the fact that it covers what would other- wise be a surface exposed hydrophobic patch, indicate that the TubZ H0 does not undergo conformational changes during protofilament formation and is important for the general fold of TubZ (Fig. S3 A and C). Data suggest that FtsZ and tubulin form protofilaments with similar longitudinal contacts (4). However, the TubZ structure reveals key differences, primarily in its C-domain and C-terminal regions, which suggest that it forms protofilaments distinct from those formed by tubulin/FtsZ. A notable difference is the struc- ture of loop 7 (L7). This loop inserts into the adjacent subunit providing the key catalytic residues required for GTP hydrolysis. In tubulin/FtsZ proteins, L7 has the consensus GXXNXDXAD. In TubZ, the loop is very divergent in conformation compared Fig. 4. TubZ interacts with TubR-DNA and contains a tubulin/FtsZ fold. (A) FP assay measuring binding of FL TubZ, TubZ(1-470), TubZ(1-460), TubZ(1-442), and TubZ(1-407) to TubR-DNA. Below is the control (TubZ titrated into DNA alone). Millipolarization units and TubZ concentration (nM) are along the y and x axes, respectively. (B) TubZ(1-428) structure. The N domain or GTP-bind- ing domain is colored salmon and the C domain purple. The interdomain helix, H7, is red. TubZ also contains an N-terminal helix, H0 (Yellow), and a C-terminal helix, H11 (White). 11766 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1003817107 Ni et al. to that in FtsZ/tubulin and consists of the sequence 256- DNVTYDPSD-266. In addition to the N-terminal region, the extreme C-terminal extensions of tubulin, FtsZ, and TubZ are structurally divergent (Fig. S3B). In FtsZ, the C-terminal region forms a small, two-stranded β-sheet and continues into an ex- tended region that is involved in binding adaptor proteins such as FtsA and ZipA (6, 25). By contrast, the C-terminal regions of tubulin proteins consist of a two-helix bundle followed by an extended region. Like FtsZ, however, these regions interact with numerous target proteins such as the microtubule-associated pro- teins (MAPs) (24). Consistent with this, the C-terminal regions of tubulin have been shown to face the outside of the microtubule. A characteristic feature of the extreme C-terminal extensions of tubulin proteins is their highly acidic nature (3). This acidic region has been shown to be critical for binding to several MAPs that harbor a substantial basic character, such as tau, MAP2, and MAP4 (24, 27). The TubZ C-terminal region is also helical, but it contains a single, long helix. Notably, the TubZ-tubulin overlay shows that the long C-terminal helix of TubZ would dramatically clash with the adjacent subunit in a polymer, providing support for the no- tion that TubZ forms protofilaments different from tubulin/FtsZ (Figs. S3B and S4). Interestingly, and in contrast to tubulin pro- teins, the flexible C-terminal region of TubZ that follows H11 is highly basic, in particular the last 14 residues. We have shown that these residues play a central role in TubR binding (Fig. 4A). TubR uses its electropositive face for DNA binding, leaving exposed its opposite face for TubZ interaction. Notably, this exposed face is strongly electronegative and hence would complement the basic C-terminal tail of TubZ (Fig. 3A). Tubulin/FtsZ protofilaments combine to form higher-order structures. In tubulin, the protofilaments interact in a parallel manner to form microtubules. Central to microtubule formation are lateral contacts between protofilaments from the so-called M loop, between H10 and S9. In tubulin, this loop is composed of 13 residues (1–3). The corresponding loop is much shorter in FtsZ proteins, consistent with the fact that FtsZ does not form tubulin microtubule-like structures (5, 28–30). In TubZ, the M loop is even shorter than in FtsZ, spanning only four residues. In fact, the TubZ/tubulin overlay shows that the side of the molecule con- taining the M loop is the most divergent between these proteins. These combined findings suggest that TubZ not only forms pro- tofilaments with distinct longitudinal contacts compared to FtsZ and tubulin, but it also does not form tubulin-like microtubule structures. TubZ Interactions with Guanine Nucleotides. Consistent with TubZ being a member of the tubulin/FtsZ family, our isothermal titra- tion calorimetry (ITC) studies showed that TubZ binds guanine nucleotides with high affinity; Kds for GTP-γ-S and GDP were ∼0.69 and 26 μM, respectively (Fig. 5A). We next determined the structure of the TubZ-GTP-γ-S complex by soaking GTP-γ-S into preformed TubZ(1-428) crystals. The TubZ-GTP-γ-S struc- ture contains TubZ residues 1–79 and 91–404, one GTP-γ-S, and has Rwork∕Rfree ¼ 21.8%∕25.5% (Fig. 5B and Table S2). The structure shows that TubZ binds GTP-γ-S in the same GTP binding pocket as tubulin/FtsZ (1–5). Comparison of the apo and GTP-γ-S bound TubZ structures indicated that, like FtsZ, guanine nucleotide binding does not lead to significant conforma- tional changes (5). The phosphate binding pocket is formed by two of the most highly conserved regions between TubZ and tubulin/FtsZ called loops 1 and 4 (L1 and L4) (1–3). L1 contacts the GTP-γ-S α- and β-phosphate groups via the Gln-32 and Lys-33 amide nitrogens. The L1 region of FtsZ and tubulin contain the sequences GQ(A/G)G and GQCG, respectively, whereas in TubZ the motif is 31-GQKG-34. However, the alanine/glycine and cysteine residues in FtsZ and tubulin do not contact the bound nucleotide; the TubZ Lys-33 side chain makes stacking interac- tions with the guanine base (Fig. 5B). L4 represents the so-called signature motif [GGGTG(T/S)G], which serves as an identifier of tubulin/FtsZ family members. Like FtsZ and tubulin, the L4 region of TubZ-GTP-γ-S makes phosphate interactions via its glycine amide nitrogens. Whereas L1 and L4 residues of the N domain mediate phosphate contacts, the GTP-γ-S guanine moiety is specified from residues in the core helix, H5, and C domain. In this regard, an important motif is loop 6 (L6). In FtsZ and tubulin, L6 has the consensus (F/Y)XXX(N/D) and the conserved (F/Y) residue functions in guanine base stacking. This region in TubZ, 236-WKXXXN-241, is in an altered conformation compared to FtsZ and tubulin structures. Despite the presence of the trypto- phan, which might be expected to interact with the guanine, the side chain of Lys-237 instead stacks with the guanine ring. Hence, in the TubZ-GTP-γ-S structure, the guanine base does not interact with aromatic residues as in tubulin/FtsZ but is sand- wiched between the aliphatic portions of two lysine side chains, Lys-33 and Lys-237. Finally, two asparagine residues, Asn-213 and Asn-241, from L6 effectively read the guanine N2/N3 and N1/O6 atoms, respectively, providing high specificity in TubZ’s in- teraction with guanine nucleotides. pBtoxis DNA Segregation: TubR Plasmid Partition Model. Our data show that TubR binds to the flexible, C-terminal, basic region of TubZ. The flexibility and location of the TubZ C-terminal extension suggest that it is not required for polymerization and thus may be exposed on the surface of TubZ filaments. Indeed, negative stain EM images show that TubZ(1-407) forms polymers in a GTP-dependent manner similar to the FL protein (Fig. S5). Recent data suggesting that TubZ filaments are stabilized by Fig. 5. TubZ-guanine nucleotide interactions. (A) ITC binding isotherms showing TubZ-GDP (Left) and TubZ-GTP-γ-S interaction (Right). (B) Left: Overall structure of the TubZ-GTP-γ-S complex. β-strands are colored magen- ta and helices cyan, and the GTP-γ-S molecule is shown as cpk. Right: Close-up view of the GTP binding pocket with the initial Fo-Fc electron density map (Blue Mesh), contoured at 4.5σ, and calculated before the GTP-γ-S was included in refinement. Ni et al. PNAS ∣ June 29, 2010 ∣ vol. 107 ∣ no. 26 ∣ 11767 BIOCHEMISTRY the presence of a GTP cap and undergo treadmilling are consis- tent with the notion that TubZ displays tubulin-like polymer dy- namics (12). Thus, on the basis of the combined data, we suggest a model for TubR/TubZ mediated pBtoxis plasmid segregation shown in Fig. 6. In this model, multiple TubR dimers first bind to the iteronic DNA on the pBtoxis plasmid leading to the crea- tion of a high local concentration of TubR, which can recruit a TubZ polymer, likely by interactions between the acidic TubR dimer face and the basic C-terminal TubZ region. Importantly, this interaction serves to attach the pBtoxis plasmid to the TubZ polymer, which undergoes treadmilling, adding subunits at the þ end and losing subunits at the −end. The bound TubR-pBtoxis can be handed off from the −end to the molecules in the growing þ end, leading to the transport of the pBtoxis plasmid to the cell pole. Interestingly, it has been shown that once TubZ polymers reach and interact with the cell pole, they bend around the curved pole and continue growing in the other direction (7). The force of the interaction with the membrane likely causes the release of TubR-pBtoxis, the net result being transport of pBtoxis to the cell pole. Of course, this model is simplified and many questions re- main. For example, how directionality is achieved and how the replicated plasmids are driven to opposite cell poles is not clear. However, given the large size of the pBtoxis plasmid (8), it may be that only one TubR-pBtoxis “tram” can be bound at once by the rapidly treadmilling TubZ polymer and that, once one such a tram is unloaded after reaching the cell pole, another engages when the now reversed polymer treadmills toward the opposite cell pole. Materials and Methods Summary Detailed methods are provided in SI Materials and Methods. Briefly, the tubR and tubZ genes were codon optimized (for E. coli expression), subcloned into pET15b, expressed, and pur- ified. WT TubR crystals were grown with NaCl and phosphate. TubR S63W was crystallized with PEG and ethylene glycol and TubZ with sodium formate. Detailed assay conditions for FP, ITC, electron microscopy, and gel filtration are provided in SI Materials and Methods. ACKNOWLEDGMENTS. This work was supported by the Burroughs Wellcome Career Development Award 992863 and National Institutes of Health Grant GM074815 (to M.A.S.). 1. Nogales E, Wolf SG, Downing KH (1998) Structure of the αβ tubulin dimer by electron crystallography. Nature 391:199–203. 2. Nogales E, Whittaker M, Milligan RA, Downing KH (1999) High-resolution model of the microtubule. Cell 96:79–88. 3. Nogales E (2000) Structural insights into microtubule function. Annu Rev Biochem 69:277–302. 4. Oliva MA, Cordell SC, Löwe J (2004) Structural insights into FtsZ protofilament formation. Nat Struct Mol Biol 11:1243–1250. 5. Oliva MA, Trambaiolo D, Löwe J (2007) Structural insights into the conformational variability of FtsZ. J Mol Biol 373:1229–1242. 6. Margolin W (2005) FtsZ and the division of prokaryotic cells and organelles. Nat Rev Mol Cell Biol 6:862–871. 7. Larsen RA, et al. (2007) Treadmilling of a prokaryotic tubulin-like protein, TubZ, required for plasmid stability in Bacillus thuringiensis. Genes Dev 21:1340–1352. 8. Tang M, Bideshi DK, Park H-W, Federici BA (2006) Minireplicon from pBtoxis of Bacillus thuringiensis subsp. israelensis. App Environ Microbiol 72:6948–6954. 9. Tang M, Bideshi DK, Park H-W, Federici BA (2007) Iteron-binding ORF157 and FtsZ-like ORF156 proteins encoded by pBtoxis play a role in its replication in Bacillus thuringien- sis subsp. israelensis. J Bacteriol 189:8053–8058. 10. Anand SP, Akhtar P, Tinsley E, Watkins SC, Khan SA (2008) GTP-dependent polymer- ization of the tubulin-like RepX replication protein encoded by the pXO1 plasmid of Bacillus anthracis. Mol Microbiol 67:881–890. 11. Berry C, et al. (2002) Complete sequence and organization of pBtoxis, the toxin- coding plasmid of Bacillus thuringiensis subsp israelensis. Appl Environ Microbiol 68:5082–5095. 12. Chen Y, Erickson HP (2008) In vitro assembly studies of FtsZ/tubulin-like proteins (TubZ) from Bacillus plasmids: Evidence for a capping mechanism. J Biol Chem 283:8102–8109. 13. Hayes F, Barillà D (2006) The bacterial segrosome: A dynamic nucleoprotein machine for DNA trafficking and segregation. Nat Rev Microbiol 4:133–143. 14. Schumacher MA (2008) Structural biology of plasmid partition: Uncovering the molecular mechanisms of DNA segregation. Biochem J 412:1–18. 15. Gerdes K, Møller-Jensen J, Bugge Jensen R (2000) Plasmid and chromosome partitioning: Surprises from phylogeny. Mol Microbiol 37:455–466. 16. Møller-Jensen J, et al. (2003) Bacterial mitosis: ParM of plasmid R1 moves plasmid DNA by an actin-like insertional polymerization mechanism. Mol Cell 12:1477–1487. 17. Popp D, et al. (2008) Molecular structure of the ParM polymer and the mechanism leading to its nucleotide-driven dynamic instability. EMBO J 27:570–579. 18. Salje J, Löwe J (2008) Bacterial actin: Architecture of the ParMRC DNA partitioning complex. EMBO J 27:2230–2238. 19. Dunham TD, Xu W, Funnell BE, Schumacher MA (2009) Structural basis for ADP- mediated transcriptional regulation by P1 and P7 ParA. EMBO J 28:1792–1802. 20. Gajiwala KS, Burley SK (2000) Winged helix proteins. Curr Opin Struct Biol 10:110–116. 21. Eicken C, et al. (2003) A metal-ligand-mediated intersubunit allosteric switch in related SmtB/ArsR zinc sensor proteins. J Mol Biol 333:683–695. 22. Pennella M, Giedroc DP (2005) Structural determinants of metal selectivity in prokaryotic metal-responsive transcriptional regulators. Biometals 18:413–428. 23. Arunkumar AI, Campanello GC, Giedroc DP (2009) Solution structure of a paradigm ArsR family zinc sensor in the DNA-bound state. Proc Natl Acad Sci USA 106:18177–18182. 24. Downing KH (2000) Structural basis for the interaction of tubulin with proteins and drugs that affect microtubule dynamics. Annu Rev Cell Dev Biol 16:89–111. 25. Adams DW, Errington J (2009) Bacterial cell division: Assembly, maintenance and disassembly of the Z ring. Nat Rev Microbiol 7:642–653. 26. Errington J, Daniel RA, Scheffers DJ (2003) Cytokinesis in bacteria. Microbiol Mol Biol Rev 67:52–65. 27. Chau MF, et al. (1998) The microtubule-associated protein tau cross-links to two distinct sites on each alpha and beta tubulin monomer via separate domains. Biochemistry 37:17692–17703. 28. Bi EF, Lutkenhaus J (1991) FtsZ ring structure associated with division in Escherichia coli. Nature 354:161–164. 29. Erickson HP, Taylor D, Taylor KA, Bramhill D (1996) Bacterial cell division protein FtsZ assembles into protofilament sheets and minirings, structural homologs of tubulin polymers. Proc Natl Acad Sci USA 93:519–523. 30. Osawa M, Anderson DE, Erickson HP (2008) Reconstitution of contractile FtsZ rings in liposomes. Science 320:792–794. 31. Delano WL (2002) The PyMOL Molecular Graphics System (DeLano Scientific, San Carlos, CA). 32. Pabo CO, Lewis M (1982) The operator-binding domain of λ repressor: Structure and DNA recognition. Nature 298:443–447. Fig. 6. pBtoxis DNA partition model. In the first step, TubR, which is bound to its centromere on one of the replicated pBtoxis plasmids, contacts the TubZ C-terminal region (indicated by lines pointing from the TubZ “circles”) in a treadmilling TubZ polymer. TubZ subunits are lost from the −end and are added to the þ end. TubR is pulled along the growing polymer by its TubR-TubZ interaction until it reaches the cell pole and is knocked off when it comes into contact with the membrane at the cell pole. TubZ reverses direction and may pick up the other TubR-pBtoxis complex and deliver it similarly to the opposite cell pole. 11768 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1003817107 Ni et al.
3M8L
Crystal Structure Analysis of the Feline Calicivirus Capsid Protein
JOURNAL OF VIROLOGY, June 2010, p. 5550–5564 Vol. 84, No. 11 0022-538X/10/$12.00 doi:10.1128/JVI.02371-09 Copyright © 2010, American Society for Microbiology. All Rights Reserved. Conformational Changes in the Capsid of a Calicivirus upon Interaction with Its Functional Receptor Robert J. Ossiboff,1† Yi Zhou,2† Patrick J. Lightfoot,1 B. V. Venkataram Prasad,2 and John S. L. Parker1* Baker Institute for Animal Health, College of Veterinary Medicine, Cornell University, Ithaca, New York 14853,1 and Verna and Marrs McClean Department of Biochemistry and Molecular Biology, Baylor College of Medicine, Houston, Texas 77030 2 Received 10 November 2009/Accepted 19 March 2010 Nonenveloped viral capsids are metastable structures that undergo conformational changes during virus entry that lead to interactions of the capsid or capsid fragments with the cell membrane. For members of the Caliciviridae, neither the nature of these structural changes in the capsid nor the factor(s) responsible for inducing these changes is known. Feline junctional adhesion molecule A (fJAM-A) mediates the attachment and infectious viral entry of feline calicivirus (FCV). Here, we show that the infectivity of some FCV isolates is neutralized following incubation with the soluble receptor at 37°C. We used this property to select mutants resistant to preincubation with the soluble receptor. We isolated and sequenced 24 soluble receptor-resistant (srr) mutants and characterized the growth properties and receptor-binding activities of eight mutants. The location of the mutations within the capsid structure of FCV was mapped using a new 3.6-Å structure of native FCV. The srr mutations mapped to the surface of the P2 domain were buried at the protruding domain dimer interface or were present in inaccessible regions of the capsid protein. Coupled with data showing that both the parental FCV and the srr mutants underwent increases in hydrophobicity upon incubation with the soluble receptor at 37°C, these findings indicate that FCV likely undergoes conformational change upon interaction with its receptor. Changes in FCV capsid conformation following its interaction with fJAM-A may be important for subsequent interactions of the capsid with cellular membranes, membrane penetration, and genome delivery. The interactions between viruses and receptors on the sur- face of host cells strongly influence viral pathogenesis and regulate morbidity and mortality in the host. Virus-receptor interactions determine the types of cells that can be infected, the pathway of entry into the cell, and the efficiency of pro- ductive infection. Interactions between nonenveloped virus capsids and their receptor(s) trigger one or more steps re- quired for infectious entry. These steps can include interaction with other receptors, exposure to low pH or endosomal pro- teases, or other factors. Ultimately, one or more of these in- teractions induce changes in capsid conformation that result in the exposure of hydrophobic regions or release of a lipid- seeking factor that can interact with and disrupt the limiting cellular membrane to allow the capsid and/or the genome to be delivered to the interior of the cell (reviewed in reference 60). The Caliciviridae are small nonenveloped viruses containing a positive-sense RNA genome (7 to 8 kb). Several important disease-causing members of the Caliciviridae, including human noroviruses and rabbit hemorrhagic disease virus, cannot be propagated in tissue culture systems (19, 56). This has slowed progress on studies of the mechanisms of cellular entry of these viruses. In contrast, feline caliciviruses (FCVs) propagate readily in tissue culture, and two cell surface receptor mole- cules, feline junctional adhesion molecule A (fJAM-A) and 2,6 sialic acid, are known (29, 55). The FCV receptor, fJAM-A, is a type I transmembrane glycoprotein (molecular size of 36 to 41 kDa) member of the immunoglobulin superfamily (IgSF); it consists of an amino- terminal signal peptide, an extracellular domain (composed of two Ig-like domains—a membrane-distal D1 and a membrane- proximal D2), a transmembrane domain, and a short cytoplas- mic domain that contains a type II PDZ domain-binding motif (11, 30). We have previously shown that the D1 domain of the fJAM-A ectodomain is necessary and sufficient for FCV bind- ing and that preincubation of FCV with soluble fJAM-A (sfJAM-A) results in virus neutralization (35). Additional roles that fJAM-A might play in FCV entry, however, have not been investigated. Caliciviruses are composed of 180 copies of a single capsid protein. Atomic resolution structures of recombinant virus-like particles of Norwalk virus (genus Norovirus) and native San Miguel sea lion virus (SMSV) virions (genus Vesivirus) indicate that the virion consists of 90 dimers of the capsid protein arranged in T3 icosahedral symmetry (5, 41). Each capsid monomer contains three structural domains—an N-terminal arm (NTA), the shell (S), and a protruding domain (P) that is further subdivided into P1 and P2 subdomains. The distal sub- domain, P2, is structurally conserved between Norwalk virus and SMSV, but there is little sequence conservation. In the primary sequence of the FCV capsid, there are two hypervari- able regions that contain neutralizing epitopes (18, 34, 58). The corresponding hypervariable regions (HVRs) of the SMSV capsid structure map to surface-exposed loops. Surface resi- dues at the dimeric interface between two capsid monomers * Corresponding author. Mailing address: Baker Institute for Ani- mal Health, College of Veterinary Medicine, Cornell University, Hun- gerford Hill Road, Ithaca, NY 14853. Phone: (607) 256-5626. Fax: (607) 256-5608. E-mail: jsp7@cornell.edu. † These authors contributed equally to this publication.  Published ahead of print on 31 March 2010. 5550 are conserved within individual calicivirus genera, and it has been suggested that this interface is involved in receptor bind- ing (5). A cryo-electron microscopy (cryo-EM) reconstruction of the FCV vaccine strain F9 complexed with the ectodomain of fJAM-A (modeled on the crystal structures of SMSV and human JAM-A, respectively) shows that fJAM-A engages the top of the P2 domain and that binding causes a rotation in the P dimer (1). However, the relatively low resolution and the lack of atomic resolution structures of FCV and fJAM-A pre- vented precise identification of residues on the viral capsid that contact fJAM-A. A classical approach for identifying virus residues that di- rectly bind or modulate the binding of a receptor is to select for mutant viruses resistant to neutralization with soluble recep- tors (6, 23, 46). Soluble receptor-resistant (srr) mutants of poliovirus revealed that both surface-exposed and internal res- idues regulate receptor attachment and conformational changes in the capsid (6, 42). Here, we report 24 srr mutants and the location of their capsid mutations on a 3.6-Å structure of FCV. In addition, we describe the growth kinetics and re- ceptor-binding properties of a subpanel of eight srr mutants and examine changes in capsid hydrophobicity concurrent with the interaction of FCV capsids with sfJAM-A. MATERIALS AND METHODS Cells and viruses. Crandell-Reese feline kidney (CRFK; ATCC CCL-94) cells were grown in Eagle’s minimal essential medium (EMEM; CellGro) supple- mented with 5% fetal bovine serum (FBS; HyClone), 100 U ml1 penicillin, 100 g ml1 streptomycin, 0.25 g ml1 amphotericin B, 1 mM sodium pyruvate, and nonessential amino acids (CellGro). Suspension Chinese hamster ovary (CHO-S) cells (Invitrogen) were grown in CHO serum-free medium (Gibco). The F9 vaccine strain (VR-782) of FCV was obtained from the ATCC. The viral isolates FCV-5, Deuce, Kaos, FCV-127, FCV-131, and FCV-796 were previously characterized (36). FCV-Urbana viral stock was generated from a full-length infectious clone as previously described (50, 51). Briefly, CRFK cells were infected with vaccinia virus expressing T7 RNA polymerase (MVA/T7) and then transfected using FuGene 6 (Roche) 1 h later with the FCV-Urbana infec- tious clone (pQ14). Samples were lysed at 18 h postinfection (h pi) by freeze- thawing, and viral plaques were selected. The viral capsid gene was sequenced to verify that FCV-Urbana was present. Third-passage viral stocks were prepared from twice-plaque-purified viruses amplified in CRFK cells. Plaque assays were performed as previously described (36). Antibodies and reagents. A rabbit antiserum against the purified fJAM-A ectodomain was previously described (35). Virus was detected with either an anti-FCV mouse monoclonal antibody (MAb) (S1-9; Custom Monoclonal Anti- bodies International) or rabbit anti-FCV-5 antiserum. The fJAM-A extracellular domain (amino acid residues 26 to 232) was purified as a glutathione S-trans- ferase (GST)-tagged fusion as previously described (35). Recombinant protein was purified by glutathione affinity chromatography. When GST-free fJAM-A was desired, N-terminal GST and His tags were cleaved from the recombinant protein by using a human rhinovirus (HRV) 3C protease (Novagen) according to the manufacturer’s directions. Neutralization of FCV by fJAM-A and selection of soluble receptor-resistant (srr) mutants of FCV-5. To investigate the neutralization of FCV by fJAM-A, 1  105 PFU of virus in Dulbecco’s modified Eagle’s medium (DMEM; CellGro) plus 0.1% bovine serum albumin (BSA; Calbiochem) were incubated in either the presence or absence of sfJAM-A (concentrations of receptor, temper- atures of incubation, and lengths of incubation varied by experiment; please refer to Results for experimental details). The infectious titer of the samples was determined by plaque assay. srr mutants of FCV-5 were selected by plaque titrating virus neutralized as described above on CRFK cells in six-well plates and overlaid with EMEM containing 5% FBS and 1% Bacto agar. Following incu- bation at 37°C in humidified 5% CO2 for 48 h, individual plaques were selected and amplified without selection on CRFK cells in 12-well plates (Corning). Once wells showed 100% cytopathic effect (CPE), a small aliquot from each well was again incubated with GST–fJAM-A and used to infect CRFK cells in 12-well plates. In wells that developed CPE, a sample of the lysate was subjected to an additional round of GST–fJAM-A neutralization, and samples were amplified on CRFK cells in a 96-well plate (Corning). Twenty-four wells showing CPE were chosen at random and twice plaque-purified. To verify that the selected srr mutants were resistant to neutralization by soluble receptor, a subpanel of eight srr mutants was incubated with a soluble receptor, and the infectious titer of the samples was determined by plaque assay. Sequencing the major (VP1) and minor (VP2) capsid proteins of the srr mutants. Recovery of viral RNA, reverse transcription-PCR (RT-PCR), and sequencing of the major capsid protein of the srr mutants were performed as described previously (36). Briefly, total RNA was extracted from infected cell lysates (RNeasy minikit; Qiagen), and first-strand cDNA synthesis of the capsid region of the genome was performed using a degenerate primer and AccuScript high-fidelity reverse transcriptase (Stratagene). The major capsid open reading frame (ORF) was then amplified from the first-strand cDNA template (primers available upon request). The resulting 2.1-kb PCR products, encompassing the entirety of the capsid ORF, were purified and sequenced directly. First-strand cDNA synthesis of the region of the genome containing VP2 was performed using RNA extracted from infected cell lysates as described above and using the 3 RACE system for amplification of cDNA ends (Invitrogen). The minor capsid ORF was amplified from the first-strand cDNA template, and the resulting PCR products were purified and sequenced directly. The VP2 sequence of FCV-5 was submitted to GenBank (accession no. HM001263). Purification of FCV. Roller bottles (Corning) containing confluent CRFK monolayers were infected with FCV (multiplicity of infection [MOI]  5) and incubated for 8 h at 37°C. At 8 h pi, cells were removed from the plastic surface by physical agitation, and the cells were collected by centrifugation of the me- dium at 500  g. The cell pellet was resuspended in 250 mM NaCl and 85 mM Tris base (pH 7.5) and frozen at 80°C. The pellet was thawed and briefly sonicated to lyse the cells, and trichlorotrifluoroethane was used to extract the virus from the cell lysate. To clarify the phases, the trichlorotrifluoroethane- lysate mixture was centrifuged for 10 min at 5,850  g. The aqueous phase was removed and the trichlorotrifluoroethane extraction step repeated. Following the second centrifugation, the aqueous phase was placed on a CsCl gradient (1.30 to 1.45 g/ml) in an ultracentrifuge tube (Beckman) and centrifuged at 97,000  g for 16 h. Typically, two bands were visible—a lower band containing infectious particles and an upper band containing noninfectious particles. Occasionally, a third minor band would be visible between the upper and lower bands; infectivity of this band when present was variable. Each band was collected separately and dialyzed exhaustively against virion buffer (150 mM NaCl, 10 mM Tris base, 15 mM MgCl2 [pH 7.2]) for 48 to 72 h at 4°C. Purified virus was stored at 4°C. Crystallization. “Full” infectious particles of FCV-5 were used for crystalliza- tion. Crystallization trials were carried out using the hanging drop vapor diffusion method at room temperature. Crystal screens 1 and 2 from Hampton Research were used to explore various conditions. Typically, for each condition, 2 l of the virus solution, at a concentration of 3 mg/ml, was mixed with 2 l of the well solution and equilibrated with 0.5 ml of the well solution. Initial screening produced crystals under a few conditions, including the crystal screen 1 condition 3 (0.4 M ammonium phosphate) and condition 13 (30% polyethylene glycol [PEG] 400, 0.1 M Tris HCl [pH 8.5], 0.2 M sodium citrate). After optimization of crystallization conditions and synchrotron on-site screening, crystals of the best diffraction quality were obtained using a well solution containing 0.45 M ammonium phosphate with additive screen reagent C10 (1.0 M glycine). Cryo- protection of crystals was carried out by soaking the crystals in the mother liquor with increasing concentrations of PEG 400 through six steps, as follows: 5%, 10%, 15%, 20%, 22.5%, and 25%. The equilibration time at each solution was at least 10 min. The crystals were then flash frozen in liquid nitrogen and shipped to synchrotron sites for data collection. Data collection and analysis. The diffraction data from single crystals were collected under cryo-EM conditions at an APS 19ID station with monochromatic X-rays (wavelength  0.9795 Å) and a detector to crystal distance of 480 mm on an ADSC 3-by-3 charge-coupled-device (CCD) detector, using an oscillation angle of 0.3° and an exposure time of 3 s. Indexing, integration, scaling, postre- finement, and reduction of the data were carried out using the HKL-2000 (37) and d*TREK (40) packages. Analysis of the diffraction data clearly indicated that the FCV-5 crystal belonged to the orthorhombic space group I222 (Table 1). The higher-resolution reflections were generally weak and appeared to be sensitive to radiation damage. A similar trend was observed with several other crystals that were used for data collection. The best crystal, which diffracted to 3.6 Å, was used for data collection. The data between 30 and 3.6 Å from 300 frames were scaled with an Rmerge factor of about 22%. The overall completeness of the data to 3.6-Å resolution is about 90.1%. To determine the precise orientation of the virus particle in the unit cell, self-rotation functions (45) were calculated using the program GLRF (59). These calculations showed all the peaks corresponding VOL. 84, 2010 CALICIVIRUS CONFORMATIONAL CHANGE 5551 to 5-fold (  72°), 3-fold (  120°), and 2-fold (  180°) axes, which are expected from a particle with icosahedral symmetry. In the I222 space group, consistent with the unit cell dimensions and the particle radius, which is esti- mated to be 200 Å, the particle position is uniquely defined by the intersection of the crystallographic 2-fold axes. In such a setting, an icosahedral particle can assume one of the two possible orientations. This ambiguity in orientation was clearly resolved by the self-rotation function calculations by use of the X-ray diffraction data. Each crystallographic asymmetric unit is composed of one-fourth of the virion with a 15-fold noncrystallographic redundancy due to one set of icosahedral 5- and 3-fold symmetry. Crystal structure determination and refinement. A properly positioned and oriented SMSV-4 capsid was placed in an FCV-5 unit cell. One initial model at a resolution of 10 Å was calculated from the 15 copies of SMSV-4 (Protein Data Bank [PDB] ID 2GH8) capsid protein related first by icosahedral 5-fold symmetry and then by icosahedral 3-fold symmetry. Phase refinement and extension were carried out by iterative cycles of real-space electron density averaging, solvent flattening, and back transformation with the RAVE and CCP4 (3, 25) program packages. The phases were extended to the final 3.6-Å resolution with each step being less than one reciprocal space point. An initial mask was constructed from the cryo-EM recon- struction with the program MAMA, and the mask was edited at each step to avoid truncation of the density. The averaged 3.6-Å density map is of good quality and readily interpretable (see Fig. 2A). After initial model building using COOT (12) and the final refinement using CNS (3), the structure has an Rfree (2) of 0.39, calculated using 10% of the data that were not included in the refinement, and a final R factor of 0.37. The atomic coordinates and associated structure factors have been deposited into the PDB (www.pdb.org) (PDB ID 3M8L). Virus-binding assay by flow cytometry. FCV binding to CHO-S cells transiently expressing fJAM-A was detected as previously described (35). Briefly, CHO-S cells (106 cells) were transiently transfected with a plasmid encoding fJAM-A. At 24 h posttransfection, the cells were washed and incubated with FCV (MOI  5) in phosphate-buffered saline (PBS) (137 mM NaCl, 3 mM KCl, 8 mM Na2HPO4 [pH 7.5]) plus 1% BSA (PBSA) on ice for 30 min. After being washed to remove unbound virions, the cells were incubated with fJAM-A rabbit antiserum and anti- FCV mouse MAb in PBSA on ice for 30 min. The cells were then washed in PBSA, fixed in 2% paraformaldehyde in PBS, and incubated with Alexa 647-conjugated goat anti-rabbit and Alexa 488-conjugated goat anti-mouse IgG for 30 min at room temperature. The cells were analyzed using a FACSCalibur (Becton Dickinson). Cells expressing fJAM-A on their surface were gated and analyzed for virus binding (10,000 cells were gated for each sample). Plaque reduction of srr mutants by fJAM-A antiserum. Confluent CRFK monolayers in 60-mm dishes (Corning) were incubated in PBSA with or without fJAM-A rabbit antiserum for 30 min at room temperature. Virus (100 PFU per dish) was adsorbed for 1 h at room temperature. The monolayers were then overlaid with EMEM containing 5% FBS and 1% Bacto agar. After incubation at 37°C for 48 h in humidified 5% CO2, the overlay was removed, and the cells were fixed with 10% buffered formalin and stained with 1% (wt/vol) crystal violet in 20% ethanol. The plaques were counted, and percent reduction from controls was calculated. Single- and multiple-cycle growth kinetics of srr mutants. CRFK cell mono- layers were inoculated with srr mutants at MOIs of 5 and 0.01 for single- and multiple-cycle growth curves, respectively. Virus was adsorbed to cells for 1 h at room temperature in DMEM plus 0.1% BSA; EMEM plus 5% FBS supple- mented as described above was then added, and the cells were incubated at 37°C in 5% CO2. At various times postinfection, samples were collected and stored at 80°C for later titration. Prior to plaque assay, all samples were frozen and thawed three times. Virus yield at each time point was calculated by subtracting the log10(PFU ml1) at T  0 from the log10(PFU ml1) measured at the time point. ELISA binding of FCV to fJAM-A. Purified fJAM-A ectodomain (1 g ml1) in carbonate/bicarbonate buffer (Sigma) was bound to 96-well enzyme-linked immunosorbent assay (ELISA) plates (100 ng per well) at 4°C overnight. The wells were blocked with PBS containing 0.05% Tween 20 (PBS-T) plus 0.5% BSA for 1 h at room temperature and washed three times with PBS-T. Twofold dilutions of purified virus (starting concentration, 200 g ml1) were prepared in PBS-T plus BSA; 100 l of each dilution of the virus was bound to plates for 3 h on ice. After the plates were washed, bound virus was detected with rabbit anti-FCV-5, followed by horseradish peroxidase (HRP)-conjugated goat anti- rabbit IgG. Antibodies diluted in PBS-T plus BSA were incubated with the plates for 90 min on ice. Following washing, substrate 2,2-azino-bis(3-ethylbenzthia- zoline-6-sulfonic acid) (Sigma) in a citric acid-sodium phosphate buffer was added to each well and incubated for 60 min at room temperature. Absorbance was measured at 595 nm on a Microplate biokinetics reader (BioTek Instru- ments). The mean standard error (SE) of three plates (two replicates per plate) is shown. ELISA competitive binding of FCV in the presence of sfJAM-A. Purified fJAM-A ectodomain (1 g ml1) in carbonate/bicarbonate buffer was bound to 96-well ELISA plates (100 ng per well) at 4°C overnight. The wells were blocked with PBS-T plus BSA for 1 h at room temperature and washed three times with PBS-T. Twofold dilutions of fJAM-A ectodomain were prepared in PBS-T plus BSA with 750 ng purified virus (starting fJAM-A concentration, 5.3 M); 100 l of each dilution of the fJAM-A–virus mixture was bound to plates for 3 h on ice. Plates were washed, and virus binding was detected as described above for the assay to determine virus binding to fJAM in the absence of soluble receptor. Absorbance readings for each virus were normalized, and the mean SE of three plates (two replicates per plate) is shown. Detection of bis-ANS binding to FCV by fluorescence spectroscopy. Purified virus (60 g ml1, a 0.5 M concentration of the FCV VP1 dimer) or 1.5 M purified receptor (fJAM-A or hJAM-A ectodomain) diluted in virion buffer was incubated in a 400-l fluorometer cell (Varian) at 4° or 37°C for 10 min to allow temperature equilibration. To detect bis-(8-anilinonaphthalene-1-sulfonate) (bis-ANS) binding to native virus and receptor, bis-ANS was added to the sample to a final concentration of 3 M. Bis-ANS was excited at 395 nm (slit width, 5 nm) and emission collected at 495 nm (slit width, 10 nm) at 2-s intervals for 10 min on a Cary Eclipse spectrofluorometer (Varian). To investigate the binding of bis-ANS to preincubated virus and receptor, virus and receptor were mixed and incubated for 10 min at 4°or 37°C. Bis-ANS was added (3 M) and sample readings were performed as detailed above. All spectrofluorometer readings were taken with the photomultiplier tube set at 700 V. Statistical analyses. The Analyze-it (Analyze-it Software) statistical analysis add-in for Microsoft Excel was used to perform analysis of variance (ANOVA) where necessary. Graphs were prepared using Kaleidagraph (Synergy Software) and Adobe Illustrator CS3 (Adobe). Protein images were prepared using PyMOL (DeLano Scientific) (8). RESULTS FCV-5 is neutralized by preincubation with the sfJAM-A ectodomain at 37°C. Preincubation of nonenveloped viruses, such as rhinoviruses and polioviruses, with their functional receptors neutralizes virus infectivity (17, 23). In a previous study, we showed that preincubation on ice of FCV-5 with a purified soluble GST-fused form of the ectodomain of fJAM-A modestly reduced infectivity (35); FCV-5 is a field isolate re- sponsible for highly virulent systemic disease in cats. During these experiments, we noted that when the virus-receptor pre- incubation was carried out at 37°C rather than 4°C, substan- tially more infectivity was lost. To quantify the extent to which FCV-5 was neutralized by preincubation with soluble receptor TABLE 1. Data collection and refinement statistics Parameter Data Data collection Wavelength (Å) .................................................0.97925 Space group........................................................I222 Cell dimensions a, b, c (Å) .......................................................427.08, 450.73, 467.59 , , (°)........................................................90.0, 90.0, 90.0 Resolution (Å)...................................................30–3.6 (3.73–3.60)a Rsym or Rmerge (%) ............................................22 (55) Completeness (%) .............................................90.1 (87.0) Total no. of reflections .....................................1,233,681 No. of unique reflections..................................462,277 Refinement Resolution (Å)...................................................30–3.6 No. of reflections ...............................................462,277 Rwork/Rfree ...........................................................0.37/0.36 a Values in parentheses are for the highest-resolution shell. 5552 OSSIBOFF ET AL. J. VIROL. at 37°C and to assess the influence of receptor concentration and the form of sfJAM-A ectodomain on neutralization, we preincubated 105 PFU of FCV-5 with increasing concentra- tions of purified GST–fJAM-A or with a soluble form of fJAM-A in which the GST had been removed (sfJAM-A) for 30 min at 37°C. We found a clear relationship between the concentration of soluble receptor and the extent of virus neu- tralization with concentrations of GST–fJAM-A of 27.5 M (Fig. 1A) and of sfJAM-A of 2.8 M (Fig. 1B), causing an 250-fold loss of viral titer. Cleavage of GST from the N terminus of the fJAM-A ectodomain enhanced the capacity of the protein to neutralize FCV-5 10-fold. As GST was fused to the D1 portion of the ectodomain, which we have shown is necessary for binding to FCV, it seems likely that GST steri- cally hindered the FCV-ectodomain interaction. As we had previously found only a modest reduction in FCV-5 infectivity when virus and GST–fJAM-A were preincu- bated at 4°C, we assessed the loss of infectivity following pre- incubation of virus with sfJAM-A (4.3 M) at 4°C for 3 h, followed by an additional incubation at 4, 16, 20, or 37°C for 30 min. We found a temperature dependence with the loss of infectivity, with a 4-fold loss of infectivity at 4°C compared to 250-fold loss of infectivity at 37°C (Table 2). We conclude that soluble forms of the fJAM-A ectodomain can substantially neutralize FCV infectivity when preincubated at 37°C and that neutralization is temperature and concentration dependent and saturable. Selection of soluble receptor-resistant (srr) mutants of FCV. Because some residual infectivity always remained after pre- incubation of FCV-5 with sfJAM-A at 37°C, we hypothesized that resistant mutant viruses were being selected. As we ex- pected that analysis of these mutants would shed light on possible capsid residues important in binding to fJAM-A or receptor-induced conformational changes in the capsid associ- ated with neutralization, we selected 24 mutants that were resistant to preincubation with GST–fJAM-A (55 M) at 37°C for 30 min. All 24 mutants contained one or two point muta- tions within the major capsid protein (Table 3), giving a total of 20 unique mutations. Three of these mutations were present in the leader sequence of the capsid, which is cleaved from the capsid precursor protein during capsid assembly by a virus- encoded proteinase (4, 33, 54); however, the mutants that contained mutations in the leader sequence also had an addi- tional mutation in another region of the capsid protein. Several residue changes were found in more than one mutant. These changes were at positions G329 (three mutants), T438 (two mutants), K480 (three mutants), V516 (five mutants), and K572 (two mutants). Only two mutants had mutations within the minor capsid protein, VP2 (Table 3), and both of these mutants also contained mutations in VP1. Overview of the crystal structure. To identify the location of srr mutations on the capsid, we determined the crystal struc- ture of FCV-5 at 3.6-Å resolution by molecular replacement and real-space averaging (Fig. 2). The structure was refined to a crystallographic R factor of 0.37 (Rfree  0.39) at 3.6-Å resolution. As in other T3 viruses, the icosahedral asymmet- rical unit consisted of three independent but chemically iden- tical subunits (Fig. 2C). These subunits are traditionally des- ignated A, B, and C. The present model of the FCV-5 icosahedral asymmetrical unit consists of residues 129 to 662 of the A and C subunits and residues 133 to 662 of the B subunit. Polypeptide fold. The structure of each subunit comprises an NTA, an S domain, and a P domain (Fig. 2C), similar to that FIG. 1. Neutralization of FCV-5 by sfJAM-A. FCV-5 (1  105 PFU) was incubated with various concentrations of soluble GST– fJAM-A (A) or cleaved fJAM-A at 37°C for 30 min (B). The virus samples were then plaque titrated on CRFK cell monolayers. Each data point represents the average of two replicates from a single representative experiment. TABLE 2. Temperature dependence of FCV-5 neutralization Incubation temp (°C)a Avg log titer (PFU ml1) after incubation withb: Change in log titer SD Buffer 4.4 M sfJAM-A 4 4.8 4.2 0.6 0.2 16 5.2 3.8 1.4 0.3 24 4.7 2.6 2.1 0.2 37 4.4 2.0 2.4 0.1 a All samples were incubated at 4°C for 3 h and then for 30 min at the temperature indicated. b Average log titer (as determined by plaque assay on CRFK cells) of at least three replicates. TABLE 3. Capsid mutations in srr mutants VP1 mutation(s)a VP2 mutationa No. of mutants A168S F486L 1 T202A T447I 1 G329D 1 G329S 1 D15N G329S 1 E401K I15V 1 D434N I661T 1 T438I N13D 1 T438A 1 N443S 2 I94T S465N 1 K480R 1 K480Q 1 K480E N483S 1 V516I 2 V516A 3 P116T I535V 1 K572E 1 K572Q 1 I94T Y575H 1 a Mutations are identified by the wild-type amino acid position in ORF 2 (VP1; including leader sequence) or ORF 3 (VP2) of FCV-5. VOL. 84, 2010 CALICIVIRUS CONFORMATIONAL CHANGE 5553 seen in the recombinant Norwalk virus capsid (rNV) and SMSV-4 (Fig. 3C) (5, 41). Residues 170 to 329 fold into a canonical eight-stranded -barrel structure similar to that found in rNV, SMSV-4, and many other viral capsid proteins. These eight strands, conventionally denoted as B to I, form two four-stranded sheets, BIDG and CHEF. Two  helices are situated between strands C and D and strands E and F, respectively. Structural comparison of the S domains of FCV-5 and SMSV-4 revealed that the most significant vari- ations were located in the loop region connecting strands E and F (Fig. 3C). The P domain of FCV-5 is comprised of residues 330 to 662 and is linked to the S domain by a flexible hinge. The P domain is subdivided into P1 and P2 subdomains. The P1 subdomain, consisting of two segments of polypeptide (residues 330 to 381 and 551 to 662), folds into a structure similar to that seen in SMSV-4. The P2 subdomain, the outermost portion of the capsid protein, is a large insertion into the P1 subdomain. The P2 subdomains of SMSV-4 and FCV-5 form a compact barrel structure consisting of six strands, sequentially labeled as A to F (Fig. 3A and B). However, the six strands in FCV-5 vary in length and relative orientations, compared to their counter- parts in SMSV-4. These six strands are connected by loops of various lengths. Structure superimposition showed that loop CD in FCV-5 is significantly shorter than that of SMSV-4 (Fig. 3C). The P2 subdomain consists of residues 382 to 550 and contains seven fewer residues than that of SMSV-4. Struc- tural comparison showed that those “inserted” residues are mainly distributed in the loop connecting strands C and D. Structural comparison between different subunits. Pairwise superposition calculations indicated that the structure of the P domain of FCV-5 adopts similar structures in all three sub- units. The S domain, however, shows larger variations between the different subunits to facilitate the assembly of a T3 cap- sid. In addition, the A and C subunits have four more disor- dered N-terminal residues than the B subunit. Superposition of the whole subunit structure revealed that the A subunit forms a more compacted structure, as evidenced by the closer association of the S and P domains. The interac- tions between the S and P domains are less prominent in the B subunit, resulting in a more relaxed structure. The C subunit FIG. 2. X-ray structure of FCV (FCV-5) capsid. (A) Two sample regions in the calculated electron density map with modeled amino acid residues 197 to 207 (S domain) and 586 to 596 (P domain), respectively. (B) X-ray structure of FCV-5 viewed along the icosahedral 2-fold axis. Location of a set of A/B and C/C dimers and icosahedral 5-fold and 3-fold axes are shown. (C) Ribbon representation of the B subunit structure. The NTA (green), S domain (blue), and P1 (red) and P2 (yellow) subdomains are indicated. 5554 OSSIBOFF ET AL. J. VIROL. has the most open conformation, with the fewest interactions between the S and P domains. These differences contribute to the bent and flat conformations of the A/B and C/C dimers, respectively (Fig. 3D). Mapping of neutralizing epitopes and conserved surface regions on the P2 subdomain of FCV-5. Hypervariable regions (C and E regions) of the FCV linear capsid sequence were previously identified, and neutralizing epitopes were mapped FIG. 3. Ribbon representation of the P2 subdomain in FCV-5 (A) and SMSV-4 (B). The -strands are labeled from A to F in each case. (C) The superposition of the P domain of FCV-5 (red) and SMSV-4 (blue) shows that the six -strands in the P2 subdomains are similarly spatially arranged into a compact barrel, despite the low sequence similarity. Arrows indicate the loops connecting -strands C and D, and E and F, of the FCV-5 P2 subdomain. (D) The FCV-5 A/B dimer (top) as viewed along the line joining the icosahedral 3- and 5-fold axes and the C/C dimer (bottom) as viewed along the line joining the adjacent icosahedral 3-fold axes demonstrate the bent and flat conformations, respectively, assumed by the subunits. In both cases, the dimeric 2-fold axis is vertical. (E to G) Positions of antigenic sites and conserved residues on the P2 subdomain of the C/C dimer of FCV-5. (E) Neutralizing epitopes mapped on P2 (gray). Antigenic sites (Ags1 to -4; colored pink) represent B-cell linear epitopes identified by Radford et al. (43). MAb escape mutations identified by Tohya et al. are colored red (58). (F) Degree of conservation of surface-exposed residues on the P2 subdomain as predicted by Consurf 3.0. The solid yellow line indicates dimer interface; the dashed yellow line indicates the axis of conservation. (G) Locations of N-terminal and C-terminal HVRs, region C and central conserved portion of region E mapped on surface of P2 subdomain. VOL. 84, 2010 CALICIVIRUS CONFORMATIONAL CHANGE 5555 predominantly to residues within the E region (Fig. 3G) (43, 48, 57, 58). Using the FCV-5 structure, we mapped the loca- tions of four linear B-cell epitopes identified in FCV-F9 (43). All of these sites lay on the margins of the P2 subdomain, with one site, Ags4, at the dimer interface (Fig. 3E). Similarly, antibody escape mutations found in mutants of FCV-F4 are also located on the margins of the P2 subdomain within hy- pervariable regions (Fig. 3E) (58). To more accurately assess the conservation of individual surface-exposed residues on the P2 subdomain, we used Consurf 3.0, which maps the rate of evolution as determined by an empirical Bayesian estimation among homologous proteins onto the three-dimensional (3-D) structure (26). As input, we used an alignment of 47 different FCV capsid sequences to derive an evolutionary conservation score that was used to color code individual residues on the surface of the FCV-5 capsid structure (Fig. 3F). This analysis revealed a patchwork of conserved surface residues along the dimeric interface of the P domain that also extended laterally at a 45° angle to the dimer interface on the surface of each monomer. Highly variable residues on the margins of the P2 subdomain surrounded the more central conserved residues. Mapping srr mutations on the FCV-5 structure. Sequencing of the capsid protein of the 24 srr mutants revealed 17 unique amino acid changes from the parental FCV-5 mature capsid protein sequence. We mapped the locations of these changed residues on the structure of FCV-5 (Fig. 4). Three mutations were found in the NTA and S domains, and 14 mutations were present in the P domain (Fig. 4A). Of the 24 mutants, 21 FIG. 4. Positions of mutations found in srr mutants in the structure of FCV-5. (A) The residues mutated in the srr mutants are illustrated on a schematic of the FCV capsid protein. The locations of the different structural domains in the primary capsid sequence are indicated. (B) Locations of the mutations in the FCV-5 atomic resolution structure. 5556 OSSIBOFF ET AL. J. VIROL. (87.5%) contained mutations that mapped to the P domain of the capsid, and 18 of those mutations were to residues within the P2 subdomain. Of the mutations found in the NTA and S domains, one change was to a residue (A168S) within the NTA, but this mutation was accompanied by a second amino acid change in the P2 domain (F486L). Similarly, a mutant with a change in the shell domain (T202A) also had a change in the P2 domain (T447I). Three srr mutants contained mutations at residue G329 in the shell domain on a loop that connects the shell with the protrusion domain. Residue G329 is located at the point of the flexible hinge, and mutations at G329 were present alone or in addition to a change in the leader sequence of the capsid (D15N) (Fig. 4B and Table 3). Mutations to residue G329 were the only changes identified in the S domain that occurred independently of an additional mutation within the P domain. Three of the 21 mutated residues identified in the P domain were present within the P1 subdomain (K572E/Q, Y575H, and I661T). Residues K572 and Y575 were buried at the dimer interface of the capsid structure (Fig. 4B). Residue I661 was located at the base of the P1 subdomain, in close proximity to the S domain (Fig. 4B); the mutant containing the I661T change also had a second change within the P2 domain (D434N). The other 11 P domain mutations were in the P2 subdomain. Eight of these mutated residues were within sur- face-exposed loops of P2: E401K, D434N, T438A, N443S, T447I, S465N, K480R/Q/E, V516I/A, and I535V (Fig. 4B). Three of the eight surface-exposed mutations were present in more than one srr mutant: N443 (two mutants), K480 (three mutants), and V516 (five mutants). Two mutants had muta- tions to residues at the P2 dimer interface; one mutant pos- sessed a change in the surface-exposed residue N483, while another mutant contained a change to the buried residue F486. Lastly, two mutants had mutations to residue T438 that is buried in the present structure and maps to a loop near the surface of the P2 subdomain. In summary, the location of mutations present in srr isolates mapped to three general re- gions of the structure of FCV-5: (i) surface-exposed residues on the P2 domain; (ii) residues that are buried or not easily accessible from the surface of the capsid; and (iii) the hinge region of the capsid. srr mutants are resistant to neutralization by sfJAM-A but retain the capacity to bind fJAM-A and require cellular fJAM-A for infection. From the original 24 srr mutants, we selected eight mutant viruses to further characterize. We se- lected these mutants based on the location of their mutated residues in an effort to have representative mutations from all regions of VP1. Additionally, we selected mutants based on the frequency that particular residues were mutated, with overrep- resented mutations to residues included. All of the eight mu- tant viruses in the subpanel were resistant to neutralization with sfJAM-A under the conditions in which they were se- lected (Fig. 5). To verify that the eight selected srr mutants still bound to fJAM-A expressed on the cell surface, we used flow cytometry to measure the capacity of each of the mutants to bind to CHO-S cells expressing fJAM-A on their surface (Fig. 6A and B). We used CHO-S cells transiently expressing fJAM-A rather than a stable fJAM-A-expressing CHO-K1 cell line that we previously described (35), because of technical difficulties using adherent CHO-K1 cells in flow cytometry assays. We found that all eight mutants bound to CHO-S cells expressing fJAM-A. One mutant, the K572E mutant, however, bound to 50% fewer cells than did the parental FCV-5. In addition, all eight mutants still required fJAM-A as a func- tional receptor, as pretreatment of CRFK cells with a rabbit antiserum against fJAM-A completely blocked infection (data not shown). We conclude from these findings that the subpanel of eight srr mutants still binds and requires cellular fJAM-A for cell infection. At low MOI, the growth of some srr mutants differs from that of parental FCV-5. To evaluate the growth properties of the srr mutants, we generated single- and multiple-cycle growth curves (Fig. 7A and B, respectively) for each of the eight selected srr mutants. The growth kinetics and final yields fol- lowing a single replicative cycle (MOI  5) for the mutant viruses were similar to those of the parental FCV-5 (Fig. 7A and data not shown). However, we found that during multiple replicative cycles (MOI  0.01) six of the mutants (G329D, V516I, K572E, T438I, N443S, and E401K mutants) had signif- icantly decreased infectivity at 4 h pi compared to parental FCV-5 as determined by ANOVA (P  0.005) (Fig. 7B and data not shown). Two of the eight mutants (K480Q and A168S/ F486L mutants) were similar to FCV-5, producing new infec- tious virions by 4 h pi. By 8 h pi, all of the viruses had attained similar increases in titer, and the final yields were similar. We conclude that six of the panel of eight srr mutants have slower growth kinetics at early times postinfection than the parental FCV-5 in cells infected at a low multiplicity. Binding of srr mutants to fJAM-A. In order to assess the kinetics of binding of the srr mutants to fJAM-A, we purified three mutants (G329D, V516I, and K572E mutants) and com- pared their in vitro binding to plate-bound sfJAM-A by ELISA to that of parental FCV-5 (Fig. 8A). All the viruses showed saturable concentration-dependent binding to fJAM-A. The binding kinetics of FCV-5 and the G329D mutant were similar, and saturation of binding to fJAM-A occurred at similar con- FIG. 5. Neutralization of srr mutants by sfJAM-A. A subpanel of eight srr mutants and FCV-5 (1  105 PFU) was incubated in the presence or absence of soluble GST–fJAM-A (55 M) at 37°C for 30 min. The infectivity of each sample was assayed by plaque titration. The change in log titer was calculated by subtracting the titer of samples incubated with receptor from that of samples incubated with- out receptor. The mean change in log10 titer standard deviation (SD) of three replicates of a representative experiment is shown. VOL. 84, 2010 CALICIVIRUS CONFORMATIONAL CHANGE 5557 centrations of virus. The binding of V516I and K572E mutants to plate-bound fJAM-A saturated at slightly higher virus con- centrations than did those of FCV-5 and the G329D mutant. One possible explanation for this difference is differential de- tection by the polyclonal rabbit FCV-5 antisera. We conclude that G329D, V516I, and K572E srr mutants and FCV-5 all bind immobilized recombinant fJAM-A with relatively similar ki- netics under in vitro conditions. Relative affinities of srr mutants for fJAM-A. To determine if the V516I, K572E, and G329D srr mutants differed from FCV-5 in their affinity for fJAM-A, we performed a competi- tion ELISA (Fig. 8B). Different concentrations of sfJAM-A were used to compete for binding of virus to plate-bound fJAM-A at 4°C. We found that binding of FCV to the plate- bound receptor was strongly inhibited at the highest concen- trations of sfJAM-A. The binding of FCV to plate-bound fJAM-A correlated inversely to the concentration of soluble receptor. A 50% decrease in binding of all the tested viruses to plate-bound fJAM-A occurred at concentrations of sfJAM-A between 0.33 and 0.66 M. All the viruses exhibited similar competition curves; however, binding of the V516I mutant to plate-bound fJAM-A was 50% inhibited at a slightly higher concentration of sfJAM-A, as can be seen by the rightward FIG. 6. Binding of srr mutants to CHO cells expressing fJAM-A. (A) Nonpermissive CHO-S cells were transfected with a DNA con- struct encoding fJAM-A. At 24 h posttransfection, FCV (MOI  5) was adsorbed to the cells on ice for 30 min. After being washed with cold PBS to remove unbound virus, the cells were fixed and bound virus and cell surface fJAM-A were detected with mouse anti-FCV MAb and rabbit anti-fJAM-A antibodies, followed by Alexa 488-con- jugated goat anti-mouse IgG and Alexa 647-conjugated goat anti- rabbit IgG. Virus binding and receptor expression were analyzed by flow cytometry. Flow cytometry dot plots for four representative sam- ples are shown. (B) Virus binding was assayed by determining the percentage of fJAM-A-positive cells that bound virus. The means of the results for three independent experiments (two samples of 1  104 cells per experiment) SE are shown. FIG. 7. Growth of srr mutants. CRFK monolayers were infected with FCV-5 and the indicated srr mutants under single-cycle (MOI  5) (A) or multiple-cycle (MOI  0.01) (B) conditions. For clarity, the data from only four mutants have been included in each graph. The change in virus titer was determined by plaque assay. The mean log10 titer (log10 titer at each time point  log10 titer at T  0) for each time point is shown. The error bars shown at 4 and 8 h pi for the multiple- cycle growth represent the standard deviation of four replicates; for other data points, error bars have been omitted from the figure for clarity. The asterisks indicate significant differences between viral iso- lates as determined by ANOVA at these time points. 5558 OSSIBOFF ET AL. J. VIROL. shift in the competition binding curve (Fig. 8B). Statistical analysis (ANOVA) at three concentrations of soluble receptor (0.04, 0.33, and 2.7 M) showed significantly more viral bind- ing of the V516I mutant than that of FCV-5 for all but the lowest concentration of receptor (P  0.05). We found that preincubation of virus with low concentrations of sfJAM-A significantly enhanced binding to plate-bound fJAM-A com- pared to virus alone. As fJAM-A can form dimers, and FCV binding does not seem to require or occlude the fJAM-A dimerization interface (1, 35), this enhanced binding may be mediated by fJAM-fJAM interactions between immobilized and virus-bound soluble receptor. An alternative explanation is that the binding of sfJAM-A to FCV capsids enhances the binding to the immobilized receptor allosterically. We con- clude that the G329D, V516I, and K572E srr mutants demon- strate similar affinities for fJAM-A. Bis-ANS binding indicates a change in FCV surface hydro- phobicity following preincubation with sfJAM-A. We hypoth- esized that the interaction of soluble receptor with FCV caused a change in conformation of the viral capsid. We, therefore, used binding of bis-ANS as a probe to investigate changes in hydrophobicity that occurred upon interaction of FCV with fJAM-A in solution. Bis-ANS is a fluorophore that binds to exposed hydrophobic regions of protein molecules (44). The sequestration of bis-ANS in hydrophobic pockets of proteins is accompanied by a large increase in the fluorescence quantum yield of the probe. Thus, bis-ANS can be used to probe for structural changes in viral proteins that are accompanied by changes in surface hydrophobicity (10, 14, 15). We first prein- cubated virus or receptor alone in a cuvette in the spectroflu- orometer at 37°C for 10 min and then added bis-ANS and collected fluorescence intensity readings over a period of 10 min at 37°C (Fig. 9A to C). To quantify differences in bis-ANS fluorescence, we waited until the fluorescence emission read- ings had stabilized and then averaged the fluorescence inten- sity measurements over 1 min. Fluorescence intensity in- creased during the first 5 to 10 min after addition of bis-ANS to any of the protein samples, but it stabilized after 10 min. This increase in fluorescence is likely caused by bis-ANS bind- ing to surface-exposed hydrophobic pockets on the proteins (Fig. 9A to C). To determine if the interaction of FCV with fJAM-A trig- gered a conformational change that would change the binding of bis-ANS to the proteins, we preincubated FCV with either hJAM-A (hJAM-A does not bind FCV [35; Fig. 9A and D]) or fJAM-A (Fig. 9B and E) at 37°C for 10 min and then added bis-ANS and monitored changes in fluorescence intensity over 10 min. We found that the average stabilized bis-ANS fluores- cence intensity of samples containing FCV-5 plus hJAM-A was equal to the sum of the intensities associated with incubation of bis-ANS with FCV-5 and hJAM-A individually (Fig. 9D). These findings were expected and indicate that the overall binding of bis-ANS to either protein was unchanged when they were mixed. In contrast, the average stabilized fluorescence intensity of samples containing FCV-5 plus fJAM-A was greater than the sum of the bis-ANS fluorescence intensities of the two proteins when incubated individually (Fig. 9E). This finding suggests that an interaction between virus and fJAM-A changed the conformation of one or both proteins, resulting in the exposure of more hydrophobic patches on the protein and thus an increase in the amount of bis-ANS binding and result- ing fluorescence. As we found that FCV-5 neutralization by sfJAM-A was temperature dependent (Table 2), we investigated if changes in bis-ANS binding to the fJAM-A–FCV-5 mixture were affected by temperature (compare Fig. 9E and F). We found that, unlike our findings at 37°C (Fig. 9E), the fluorescence intensity of samples of FCV-5 and fJAM-A preincubated together at 4°C was less than the sum of the bis-ANS intensities of virus and receptor incubated at 4°C with the fluorescent probe alone (Fig. 9F). Our interpretation of these findings is that binding of FIG. 8. Binding of srr mutants to immobilized fJAM-A in the ab- sence (A) or presence (B) of increasing concentrations of sfJAM-A. ELISA plates were coated with 100 ng of sfJAM-A ectodomain. Var- ious concentrations of purified FCV-5 and G329D, V516I, and K572E srr mutants were incubated with the immobilized protein for 3 h on ice. To investigate the effect of sfJAM-A on the binding of virus to immo- bilized receptor, 750 ng of each virus and various concentrations of soluble receptor were incubated together with plate-bound fJAM-A for 3 h on ice. Bound FCV was detected with rabbit anti-FCV serum, followed by HRP-conjugated goat anti-rabbit IgG. Colorimetric HRP substrate was added, and the amount of bound FCV-5 was quantified by absorbance at 595 nm. The means of three and four plates, respec- tively (two replicates per plate), SE are shown. ANOVA was per- formed on three concentrations of soluble receptor (0.04, 0.33, and 2.7 M) to determine statistical differences; significant differences are indicated by asterisks. VOL. 84, 2010 CALICIVIRUS CONFORMATIONAL CHANGE 5559 FIG. 9. Changes in bis-ANS binding upon mixing of FCV with sfJAM-A. At time zero, bis-ANS was added to the indicated purified virions, soluble JAM-A (human [A] or feline [B and C]), or virions preincubated with JAM-A. Fluorescence was recorded every 2 s continuously for 10 min at excitation and emission wavelengths of 395 and 495 nm, respectively. (D to G) Stabilized fluorescence intensities measured during the last minute for each sample were averaged. Broken lines indicate the additive average fluorescence intensity measured for both native virus and receptor. The means SE (n  3) are shown. 5560 OSSIBOFF ET AL. J. VIROL. fJAM-A to FCV-5 at 4°C blocks hydrophobic patches previ- ously available for bis-ANS binding on the virus and/or fJAM-A. In contrast, the increase in binding of bis-ANS to the fJAM-A–FCV mixture at 37°C indicates a change in confor- mation of the viral capsid and/or the receptor. The binding of bis-ANS at 37°C to V516A, K572E, and G329D srr mutants and to field isolates Kaos and FCV-131 was also investigated. Following incubation of virus and receptor at 37°C, these samples showed an increase in fluorescence inten- sities similar to that seen for FCV-5 (Fig. 9C and G). In contrast, incubation of the vaccine strain F9 with fJAM-A led to a fluorescence intensity maximum that was less than what would be expected from the sum of the intensities associated with the receptor and virus alone; this pattern was similar to that of FCV-5 incubated with receptor at 4°C. From these findings, we conclude that interaction of fJAM-A with FCV at 37°C leads to changes in the binding of bis-ANS that are indicative of changes in hydrophobicity of virus and/or recep- tor. For FCV-5, these changes in bis-ANS binding are depen- dent upon temperature, as virus and receptor incubated at 4°C did not display the same increase in fluorescence intensity. Not all FCV isolates are neutralized by sfJAM-A. The FCV-5 isolate that we used to select for srr mutants was re- covered from a cat suffering from severe systemic disease. As this form of FCV disease is uncommon, we hypothesized that not all isolates of FCV would be neutralized by incubation with soluble receptor at 37°C. To test this hypothesis, we incubated 105 PFU of six additional FCV isolates with 45 M GST– fJAM-A at 37°C; these virus isolates included the vaccine strain F9, two additional isolates recovered from cats with virulent systemic disease, the FCV-Urbana isolate, and three other field isolates recovered from cats with mild- to moderate- severity FCV disease. Several additional FCV isolates were neutralized by preincubation with GST–fJAM-A at 37°C; how- ever, the vaccine strain F9, FCV-Urbana, and the two isolates recovered from cats with milder disease were not neutralized under these conditions (Fig. 10). We conclude that there are natural differences among FCV isolates in their responses to incubation with sfJAM-A at 37°C. DISCUSSION Several nonenveloped viruses are neutralized by preincuba- tion with their functional receptors (23, 31, 49). However, the extent to which infectivity is lost at physiologic temperatures depends upon the mechanism of receptor-mediated neutral- ization. For polioviruses and some HRVs, preincubation with their receptors induces irreversible inactivating conformational changes in the virus (6, 17). We found that some but not all isolates of FCV were neutralized by preincubation with a sol- uble form of fJAM-A. Although it is possible that neutraliza- tion was due to steric hindrance or occupation of receptor- binding sites, the prominent temperature dependence of neutralization strongly suggests that receptor-induced confor- mational changes are involved. Receptor-induced conforma- tional changes in both HRVs and poliovirus are temperature dependent (16, 20). Polioviruses undergo an irreversible neu- tralizing transition from a 160S particle to a 135S particle when they are incubated with soluble receptor at 37°C, but this transition does not occur at temperatures below 33°C (16, 61). In contrast, although FCV-5 receptor-mediated neutralization was clearly temperature dependent, no obvious temperature threshold was observed. Rather, a gradual diminishment of the neutralizing effect as the temperature decreased from 37°C to 4°C was found (Table 2). We propose that some FCV isolates undergo a temperature-dependent, fJAM-A-mediated neutral- ization following incubation at 37°C; however, as not all FCV isolates undergo neutralization under these conditions, an ad- ditional cellular factor or environmental condition is likely required by some FCV isolates for neutralization. Analysis of soluble receptor-resistant (srr) mutants in polio- virus and murine coronavirus identified virus residues that either directly or indirectly facilitated receptor binding (6, 23, 46). Of the 24 srr mutants of FCV-5, mutations to 20 unique VP1 residues were identified; six mutations mapped to surface- exposed residues on the P2 domain of the capsid that poten- tially could directly interact with the receptor. Several mutated residues were buried or inaccessible from the surface of the structure and could possibly indirectly affect binding. Similarly, poliovirus srr mutants contained mutations to both internal and surface-exposed residues. The mutated residues for polio- virus mapped to a hydrophobic pocket in the structure. It is believed that changes to residues in this hydrophobic pocket regulate receptor attachment and conformational changes within the virus particle (6, 42). One to two copies of the VP2 protein are found in FCV virions, and VP2 is essential for the synthesis and maturation of infectious FCV virions (52, 53). It was, therefore, important to ensure that resistance to neutralization of the srr mutants was not due to changes to VP2. Sequencing of the ORF en- coding the minor capsid protein of FCV revealed that only two mutants of the 24-mutant panel contained coding mutations to VP2. As both mutants possessing VP2 mutations also had VP1 mutations, and the remainder of the srr mutants contained only FIG. 10. Ability of sfJAM-A to neutralize select FCV isolates. A panel of six FCV field isolates (FCV-5, Kaos, Deuce, FCV-127, FCV- 131, and FCV-796) and two tissue culture-adapted strains (F9 and Urbana) (1  105 PFU) was incubated in the presence or absence of soluble GST–fJAM-A (45 M) at 37°C for 30 min. The remaining infectivity in each sample was assayed by plaque titration. The change in log titer was calculated by subtracting the titer of samples incubated with receptor from the titer of samples incubated without receptor. The mean change in log10 titer SD (n  4) is shown. VOL. 84, 2010 CALICIVIRUS CONFORMATIONAL CHANGE 5561 VP1 mutations, changes to VP2 cannot explain the resistance of the mutants to receptor-mediated neutralization. Our analysis of the FCV srr mutants indicates that residues within critical regions of the capsid are necessary for structural flexibility or receptor interactions. In particular, changes to residue G329 (mutated in three srr mutants) suggest that movement between the S and P domains of the capsid protein occurs upon receptor interaction and is involved in receptor- mediated neutralization. The glycine residue at position 329 is 100% conserved in the FCV capsid sequence and is located in a short loop between the S and P domains of the capsid pro- tein. A similar glycine residue is found in this hinge region of the capsid in all caliciviruses. Cryo-EM reconstructions of mu- rine norovirus in complex with neutralizing Fab fragments show that this hinge region is extended, moving the P domain radially away from the surface of the shell domain (24). In addition, a cryo-EM structure of FCV-F9 complexed with the fJAM-A ectodomain showed that the P domain of the capsid moved with respect to the S domain when bound to fJAM-A, indicating that movement in this region occurs (1). Our find- ings further support the hypothesis that flexibility in this hinge region is important to permit conformational changes of the structural domains of the calicivirus capsid in relation to each other and indicate that this region actively moves during re- ceptor interactions. Moreover, our findings indicate that changes at residue G329 substantially alter the susceptibility of FCV to neutralization by sfJAM-A. Several srr mutants had mutated residues at the P domain dimer interface. These mutations could alter receptor binding indirectly, or these residues could become accessible for re- ceptor binding following a conformational change of the cap- sid. The altered residue in the K572E srr mutant is in P1 and is unlikely to be accessible from the surface of the capsid. This mutant had decreased binding to fJAM-A on cells. However, the K572E mutant had binding kinetics similar to that of pa- rental FCV-5 in in vitro ELISA binding assays. A possible explanation for this difference is that unlike fJAM-A purified from bacteria, JAM-A on cells is N-glycosylated (32). Poliovirus srr mutants had reduced binding affinity for their receptor (6). It was, therefore, unexpected that only one mu- tant (the K572E mutant) had decreased cell surface fJAM-A binding and that none of the FCV srr mutants we examined in detail had a substantive change in binding to fJAM-A in vitro. In fact, the V516I mutant showed higher affinity for immobi- lized fJAM-A (Fig. 8B). These findings indicate that resistance of the srr mutants to receptor-induced neutralization cannot be explained by changes in their binding affinity for fJAM-A. Single-step growth kinetics of poliovirus srr mutants were similar, but all mutants had lower titers than did wild-type P1/Mahoney poliovirus at early time points postinfection. This delay in replication was attributed to an assembly defect, as all the srr mutants bound as efficiently as wild-type poliovirus to cells at room temperature (6). In contrast, we found that the FCV srr mutants and parental FCV-5 had similar single-step growth kinetics and titers throughout the replicative cycle. We found a difference during multiple-cycle growth experiments, where six of the eight srr mutants investigated had significantly decreased titers in comparison to parental FCV-5 at 4 h pi. One explanation for these findings is decreased efficiency of viral entry (receptor binding, cellular trafficking, and/or un- coating) of the mutants with a consequent lag in the onset of viral replication. As small differences in the efficiency of entry may be masked when a high MOI is utilized, experiments using low input titers of virus can be more sensitive. The K572E mutant demonstrated decreased binding to cell surface-ex- pressed fJAM-A and also had delayed multiple-cycle growth kinetics. Therefore, the decreased early growth kinetics of this mutant could be due to decreased receptor binding. However, G329D and V516I srr mutants bound cell-expressed and plate- bound fJAM-A to the same extent as did parental FCV-5 or had increased affinity for plate-bound fJAM-A and also had delayed early growth kinetics. We previously showed that low- er-virulence-field or tissue culture-adapted FCV isolates had delayed multiple cycle growth kinetics compared to more-vir- ulent isolates (including FCV-5) (36). Although the virulence of these srr mutants is currently unknown, it is possible that by selecting for srr mutants, we may have also selected for viruses with altered virulence. Following attachment, nonenveloped viruses must undergo conformational alterations that mediate membrane penetra- tion and deliver their genetic material (60). The temperature dependence of the sfJAM-A neutralization of FCV and the identification of srr mutant residues buried in the capsid struc- ture suggested that fJAM-A induces a conformational change in the FCV capsid. Our finding that mixtures of FCV-5 and sfJAM-A incubated at 37°C had bis-ANS fluorescence inten- sities greater than the sum of the fluorescence intensities as- sociated with virus and sfJAM-A individually provided further support for this concept. Bis-ANS binding has previously been used to detect conformational changes associated with changes in hydrophobicity in the envelope glycoproteins of Newcastle disease virus and HIV-1 upon incubation with gangliosides and CD4, respectively (14, 22). Additionally, our finding that pre- incubation of FCV-5 and fJAM-A at 4°C resulted in lower bis-ANS fluorescence than the sum of the intensities of the two components is consistent with the idea that binding of fJAM-A to FCV-5 at 4°C occludes potential bis-ANS binding sites and that exposure of new hydrophobic patches as a consequence of conformational change does not occur or is decreased at 4°C (Fig. 9E and F). An alternative explanation is that the in- creases in hydrophobicity we observed were due to changes in the receptor, not the virus. However, as there is substantial evidence for viral protein conformational changes following interaction with cellular receptors, and such conformational changes are likely a requirement of all nonenveloped viruses, this explanation is less likely. For those viruses that when mixed with sfJAM-A showed an increase in total bis-ANS fluorescence intensity, the intensity was 11 to 30% greater for the mixtures of FCV and sfJAM-A than the sum of the bis- ANS fluorescence intensities of each component incubated separately. Although this is an in vitro assay, we believe that taken together with our other data, these changes are biolog- ically significant. Additional investigation will be required to determine the nature of the conformational changes and their functional significance. Upon incubation with soluble receptor, the srr mutants and parental FCV-5 demonstrated similar increases in hydropho- bicity. These findings raise the possibility that FCV neutraliza- tion by soluble receptor does not involve a conformational change in the capsid. However, a more likely possibility is that 5562 OSSIBOFF ET AL. J. VIROL. more than one intermediate capsid conformer forms upon interaction of FCV with fJAM-A. Following interaction of picornaviruses with their functional receptors, three interme- diate conformers of the capsid are formed—the 135S altered particle and two forms of the 80S empty particle (7, 9, 13, 27, 28). Therefore, our working model is that mutants resistant to neutralization undergo an initial conformational change upon interacting with fJAM-A that either is reversible or results in a stable, infectious intermediate. A cryo-EM reconstruction of the FCV-F9 strain complexed with fJAM-A at 18 Å revealed that fJAM-A interacted with the top of the P2 subdomain of the F9 capsid, causing an 13° counterclockwise rotation of the P dimers about their 2-fold axes of symmetry (1). Interestingly, we found that in contrast to the other FCV isolates we investigated, the F9–sfJAM-A com- plex bound less bis-ANS than did the F9 virus alone (Fig. 9G). Given the structural evidence for a conformational change in the F9 capsid, one possible explanation is that fJAM-A binding induces a conformational change that decreases capsid surface hydrophobicity. In the face of our observations of increased bis-ANS binding to other FCV isolates when preincubated with fJAM-A, it is possible that the capsid conformational changes reported by Bhella et al. (1) for the vaccine isolate F9 differ from changes that occur in the capsids of other FCV field isolates. We found that some FCV isolates, including low-passage field isolates and tissue culture-adapted lab strains, were not susceptible to fJAM-A neutralization (Fig. 10). A natural iso- late that was resistant to receptor neutralization, FCV-131, also exhibited increased hydrophobicity following incubation with soluble receptor. These findings suggest that similar to the srr mutants, field isolates undergo a conformational alteration following interaction with fJAM-A but do not lose infectivity. Other isolates, however, may undergo more drastic, global conformational alterations that result in either a complete loss of or significant decrease in infectivity. Interestingly, we found that sensitivity of different FCV isolates to neutralization cor- related with their virulence in cats. FCV is a common pathogen of cats that typically causes mild or unapparent disease, with fatal disease being unusual. However, highly virulent isolates of FCV cause systemic signs of disease and result in mortality rates as high as 50% (21, 38, 39, 47). In our panel, the most virulent isolates were susceptible to fJAM-A neutralization, while low-virulence isolates were resistant. It is possible that the in vitro assay of receptor neutralization is highlighting an important determinant of in vivo virulence. ACKNOWLEDGMENTS We thank Christian Nelson, Meg Crapster-Pregont, Sarah Caddy, and Rachel Mays for excellent technical assistance. We thank Terry Dermody and Kristen Guglielmi for the generous gift of reagents and Stanislav Sosnovtsev for technical advice. We acknowledge the SBC- CAT 19ID beamline at the Advanced Photon Source (supported by the U.S. Department of Energy, Basic Energy Sciences, Office of Science, under contract no.W-31-109-Eng-38) and its staff for their help during data collection. This work was supported by grants from The Cornell Feline Health Center and the Winn Foundation to J.S.L.P., National In- stitutes of Health (PO1 AI057788), and R. Welch Foundation (Q- 1279) to B.V.V.P. R.J.O. is the recipient of a scholarship from Cornell University. REFERENCES 1. Bhella, D., D. Gatherer, Y. Chaudhry, R. Pink, and I. G. Goodfellow. 2008. Structural insights into calicivirus attachment and uncoating. J. Virol. 82: 8051–8058. 2. Bru¨nger, A. T. 1992. Free R value: a novel statistical quantity for assessing the accuracy of crystal structures. Nature 355:472–475. 3. Bru¨nger, A. T., P. D. Adams, G. M. Clore, W. L. DeLano, P. Gros, R. W. Grosse-Kunstleve, J. S. Jiang, J. Kuszewski, M. Nilges, N. S. Pannu, R. J. Read, L. M. Rice, T. Simonson, and G. L. Warren. 1998. Crystallography & NMR system: a new software suite for macromolecular structure determi- nation. Acta Crystallogr. D Biol. Crystallogr. 54:905–921. 4. Carter, M. J., I. D. Milton, P. C. Turner, J. Meanger, M. Bennett, and R. M. Gaskell. 1992. Identification and sequence determination of the capsid pro- tein gene of feline calicivirus. Arch. Virol. 122:223–235. 5. Chen, R., J. D. Neill, M. K. Estes, and B. V. Prasad. 2006. X-ray structure of a native calicivirus: structural insights into antigenic diversity and host spec- ificity. Proc. Natl. Acad. Sci. U. S. A. 103:8048–8053. 6. Colston, E., and V. R. Racaniello. 1994. Soluble receptor-resistant poliovirus mutants identify surface and internal capsid residues that control interaction with the cell receptor. EMBO J. 13:5855–5862. 7. Crowell, R. L., and L. Philipson. 1971. Specific alterations of coxsackievirus B3 eluted from HeLa cells. J. Virol. 8:509–515. 8. Delano, W. L. 2002. The PyMOL molecular graphics system. DeLano Sci- entific, Palo Alto, CA. 9. De Sena, J., and B. Mandel. 1977. Studies on the in vitro uncoating of poliovirus. II. Characteristics of the membrane-modified particle. Virology 78:554–566. 10. de Sousa, P. C., Jr., R. Tuma, P. E. Prevelige, Jr., J. L. Silva, and D. Foguel. 1999. Cavity defects in the procapsid of bacteriophage P22 and the mecha- nism of capsid maturation. J. Mol. Biol. 287:527–538. 11. Ebnet, K., C. U. Schulz, M. K. Meyer Zu Brickwedde, G. G. Pendl, and D. Vestweber. 2000. Junctional adhesion molecule interacts with the PDZ do- main-containing proteins AF-6 and ZO-1. J. Biol. Chem. 275:27979–27988. 12. Emsley, P., and K. Cowtan. 2004. Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 60:2126–2132. 13. Fenwick, M. L., and P. D. Cooper. 1962. Early interactions between polio- virus and ERK cells: some observations on the nature and significance of the rejected particles. Virology 18:212–223. 14. Ferreira, L., E. Villar, and I. Munoz-Barroso. 2004. Conformational changes of Newcastle disease virus envelope glycoproteins triggered by gangliosides. Eur. J. Biochem. 271:581–588. 15. Gaspar, L. P., J. E. Johnson, J. L. Silva, and A. T. Da Poian. 1997. Partially folded states of the capsid protein of cowpea severe mosaic virus in the disassembly pathway. J. Mol. Biol. 273:456–466. 16. Go´mez Yafal, A., G. Kaplan, V. R. Racaniello, and J. M. Hogle. 1993. Characterization of poliovirus conformational alteration mediated by soluble cell receptors. Virology 197:501–505. 17. Greve, J. M., C. P. Forte, C. W. Marlor, A. M. Meyer, H. Hoover-Litty, D. Wunderlich, and A. McClelland. 1991. Mechanisms of receptor-mediated rhinovirus neutralization defined by two soluble forms of ICAM-1. J. Virol. 65:6015–6023. 18. Guiver, M., E. Littler, E. O. Caul, and A. J. Fox. 1992. The cloning, sequenc- ing and expression of a major antigenic region from the feline calicivirus capsid protein. J. Gen. Virol. 73(Pt. 9):2429–2433. 19. Guix, S., M. Asanaka, K. Katayama, S. E. Crawford, F. H. Neill, R. L. Atmar, and M. K. Estes. 2007. Norwalk virus RNA is infectious in mammalian cells. J. Virol. 81:12238–12248. 20. Hoover-Litty, H., and J. M. Greve. 1993. Formation of rhinovirus-soluble ICAM-1 complexes and conformational changes in the virion. J. Virol. 67: 390–397. 21. Hurley, K. E., P. A. Pesavento, N. C. Pedersen, A. M. Poland, E. Wilson, and J. E. Foley. 2004. An outbreak of virulent systemic feline calicivirus disease. J. Am. Vet. Med. Assoc. 224:241–249. 22. Jones, P. L., T. Korte, and R. Blumenthal. 1998. Conformational changes in cell surface HIV-1 envelope glycoproteins are triggered by cooperation be- tween cell surface CD4 and co-receptors. J. Biol. Chem. 273:404–409. 23. Kaplan, G., D. Peters, and V. R. Racaniello. 1990. Poliovirus mutants resis- tant to neutralization with soluble cell receptors. Science 250:1596–1599. 24. Katpally, U., C. E. Wobus, K. Dryden, H. W. Virgin IV, and T. J. Smith. 2008. Structure of antibody-neutralized murine norovirus and unexpected differ- ences from viruslike particles. J. Virol. 82:2079–2088. 25. Kleywegt, G. J., and T. A. Jones. 1999. Software for handling macromolec- ular envelopes. Acta Crystallogr. D Biol. Crystallogr. 55:941–944. 26. Landau, M., I. Mayrose, Y. Rosenberg, F. Glaser, E. Martz, T. Pupko, and N. Ben-Tal. 2005. ConSurf 2005: the projection of evolutionary conservation scores of residues on protein structures. Nucleic Acids Res. 33:W299–W302. 27. Levy, H. C., M. Bostina, D. J. Filman, and J. M. Hogle. 2010. Catching a virus in the act of RNA release: a novel poliovirus uncoating intermediate char- acterized by cryo-electron microscopy. J. Virol. 84:4426–4441. 28. Lonberg-Holm, K., and B. D. Korant. 1972. Early interaction of rhinoviruses with host cells. J. Virol. 9:29–40. VOL. 84, 2010 CALICIVIRUS CONFORMATIONAL CHANGE 5563 29. Makino, A., M. Shimojima, T. Miyazawa, K. Kato, Y. Tohya, and H. Akashi. 2006. Junctional adhesion molecule 1 is a functional receptor for feline calicivirus. J. Virol. 80:4482–4490. 30. Mandell, K. J., and C. A. Parkos. 2005. The JAM family of proteins. Adv. Drug Deliv. Rev. 57:857–867. 31. Marlin, S. D., D. E. Staunton, T. A. Springer, C. Stratowa, W. Sommergru- ber, and V. J. Merluzzi. 1990. A soluble form of intercellular adhesion molecule-1 inhibits rhinovirus infection. Nature 344:70–72. 32. Naik, U. P., Y. H. Ehrlich, and E. Kornecki. 1995. Mechanisms of platelet activation by a stimulatory antibody: cross-linking of a novel platelet receptor for monoclonal antibody F11 with the Fc gamma RII receptor. Biochem. J. 310(Pt. 1):155–162. 33. Neill, J. D., I. M. Reardon, and R. L. Heinrikson. 1991. Nucleotide sequence and expression of the capsid protein gene of feline calicivirus. J. Virol. 65:5440–5447. 34. Neill, J. D., S. V. Sosnovtsev, and K. Y. Green. 2000. Recovery and altered neutralization specificities of chimeric viruses containing capsid protein do- main exchanges from antigenically distinct strains of feline calicivirus. J. Vi- rol. 74:1079–1084. 35. Ossiboff, R. J., and J. S. Parker. 2007. Identification of regions and residues in feline junctional adhesion molecule required for feline calicivirus binding and infection. J. Virol. 81:13608–13621. 36. Ossiboff, R. J., A. Sheh, J. Shotton, P. A. Pesavento, and J. S. Parker. 2007. Feline caliciviruses (FCVs) isolated from cats with virulent systemic disease possess in vitro phenotypes distinct from those of other FCV isolates. J. Gen. Virol. 88:506–517. 37. Otwinowski, Z., and W. Minor. 1997. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 276:307–326. 38. Pedersen, N. C., J. B. Elliott, A. Glasgow, A. Poland, and K. Keel. 2000. An isolated epizootic of hemorrhagic-like fever in cats caused by a novel and highly virulent strain of feline calicivirus. Vet. Microbiol. 73:281–300. 39. Pesavento, P. A., N. J. MacLachlan, L. Dillard-Telm, C. K. Grant, and K. F. Hurley. 2004. Pathologic, immunohistochemical, and electron microscopic findings in naturally occurring virulent systemic feline calicivirus infection in cats. Vet. Pathol. 41:257–263. 40. Pflugrath, J. W. 1999. The finer things in X-ray diffraction data collection. Acta Crystallogr. D Biol. Crystallogr. 55:1718–1725. 41. Prasad, B. V., M. E. Hardy, T. Dokland, J. Bella, M. G. Rossmann, and M. K. Estes. 1999. X-ray crystallographic structure of the Norwalk virus capsid. Science 286:287–290. 42. Racaniello, V. R. 1996. Early events in poliovirus infection: virus-receptor interactions. Proc. Natl. Acad. Sci. U. S. A. 93:11378–11381. 43. Radford, A. D., K. Willoughby, S. Dawson, C. McCracken, and R. M. Gaskell. 1999. The capsid gene of feline calicivirus contains linear B-cell epitopes in both variable and conserved regions. J. Virol. 73:8496–8502. 44. Rosen, C. G., and G. Weber. 1969. Dimer formation from 1-amino-8-naph- thalenesulfonate catalyzed by bovine serum albumin. A new fluorescent molecule with exceptional binding properties. Biochemistry 8:3915–3920. 45. Rossmann, M. G., and D. M. Blow. 1962. The detection of sub-units within the crystallographic asymmetric unit. Acta Crystallogr. 15:24–31. 46. Saeki, K., N. Ohtsuka, and F. Taguchi. 1997. Identification of spike protein residues of murine coronavirus responsible for receptor-binding activity by use of soluble receptor-resistant mutants. J. Virol. 71:9024–9031. 47. Schorr-Evans, E. M., A. Poland, W. E. Johnson, and N. C. Pedersen. 2003. An epizootic of highly virulent feline calicivirus disease in a hospital setting in New England. J. Feline Med. Surg. 5:217–226. 48. Seal, B. S., and J. D. Neill. 1995. Capsid protein gene sequence of feline calicivirus isolates 255 and LLK: further evidence for capsid protein config- uration among feline caliciviruses. Virus Genes 9:183–187. 49. Silberstein, E., G. Dveksler, and G. G. Kaplan. 2001. Neutralization of hepatitis A virus (HAV) by an immunoadhesin containing the cysteine-rich region of HAV cellular receptor-1. J. Virol. 75:717–725. 50. Sosnovtsev, S., and K. Y. Green. 1995. RNA transcripts derived from a cloned full-length copy of the feline calicivirus genome do not require VpG for infectivity. Virology 210:383–390. 51. Sosnovtsev, S., S. Sosnovtseva, and K. Y. Green. 1997. Proceedings of the First International Symposium on Caliciviruses, Reading, United Kingdom. 52. Sosnovtsev, S. V., G. Belliot, K. O. Chang, O. Onwudiwe, and K. Y. Green. 2005. Feline calicivirus VP2 is essential for the production of infectious virions. J. Virol. 79:4012–4024. 53. Sosnovtsev, S. V., and K. Y. Green. 2000. Identification and genomic map- ping of the ORF3 and VPg proteins in feline calicivirus virions. Virology 277:193–203. 54. Sosnovtsev, S. V., S. A. Sosnovtseva, and K. Y. Green. 1998. Cleavage of the feline calicivirus capsid precursor is mediated by a virus-encoded proteinase. J. Virol. 72:3051–3059. 55. Stuart, A. D., and T. D. Brown. 2007. Alpha2,6-linked sialic acid acts as a receptor for feline calicivirus. J. Gen. Virol. 88:177–186. 56. Thiel, H. J., and M. Konig. 1999. Caliciviruses: an overview. Vet. Microbiol. 69:55–62. 57. Tohya, Y., K. Masuoka, E. Takahashi, and T. Mikami. 1991. Neutralizing epitopes of feline calicivirus. Arch. Virol. 117:173–181. 58. Tohya, Y., N. Yokoyama, K. Maeda, Y. Kawaguchi, and T. Mikami. 1997. Mapping of antigenic sites involved in neutralization on the capsid protein of feline calicivirus. J. Gen. Virol. 78(Pt. 2):303–305. 59. Tong, L., and M. G. Rossmann. 1997. Rotation function calculations with GLRF. Methods Enzymol. 276:594–611. 60. Tsai, B. 2007. Penetration of nonenveloped viruses into the cytoplasm. Annu. Rev. Cell Dev. Biol. 23:23–43. 61. Tsang, S. K., B. M. McDermott, V. R. Racaniello, and J. M. Hogle. 2001. Kinetic analysis of the effect of poliovirus receptor on viral uncoating: the receptor as a catalyst. J. Virol. 75:4984–4989. 5564 OSSIBOFF ET AL. J. VIROL.