Upload batch_0

batch_0/PMC2527531.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527531",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527531\nAUTHORS: Ran Taube, Quan Zhu, Chen Xu, Felipe Diaz-Griffero, Jianhua Sui, Erick Kamau, Markryan Dwyer, Daniel Aird, Wayne A. Marasco\n\nABSTRACT:\nBackgroundIsolation of human antibodies using current display technologies can be limited by constraints on protein expression, folding and post-translational modifications. Here we describe a discovery platform that utilizes self-inactivating (SIN) lentiviral vectors for the surface display of high-affinity single-chain variable region (scFv) antibody fragments on human cells and lentivirus particles.Methodology/Principal FindingsBivalent scFvFc human antibodies were fused in frame with different transmembrane (TM) anchoring moieties to allow efficient high-level expression on human cells and the optimal TM was identified. The addition of an eight amino acid HIV-1 gp41 envelope incorporation motif further increased scFvFc expression on human cells and incorporation into lentiviral particles. Both antibody-displaying human cells and virus particles bound antigen specifically. Sulfation of CDR tyrosine residues, a property recently shown to broaden antibody binding affinity and antigen recognition was also demonstrated. High level scFvFc expression and stable integration was achieved in human cells following transduction with IRES containing bicistronic SIN lentivectors encoding ZsGreen when scFvFc fusion proteins were expressed from the first cassette. Up to 106-fold enrichment of antibody expressing cells was achieved with one round of antigen coupled magnetic bead pre-selection followed by FACS sorting. Finally, the scFvFc displaying human cells could be used directly in functional biological screens with remarkable sensitivity.Conclusions/SignificanceThis antibody display platform will complement existing technologies by virtue of providing properties unique to lentiviruses and antibody expression in human cells, which, in turn, may aid the discovery of novel therapeutic human mAbs.\n\nBODY:\nIntroductionMonoclonal antibodies (mAbs) have been used with increasing frequency to treat a wide spectrum of human diseases, including heart disease, infections and immune disorders [1]–[5]. The mAb based immunotherapies are now standard of care in an increasing number of human cancers including Erb2+ breast cancer, Non-Hodgkin's Lymphoma, colon cancer and others [1], [6], [7].Since 2001, human mAbs developed through recombinant DNA techniques have constituted the largest number entering clinical study [1]. This shift, toward de novo human mAb isolation and their clinical use, is in part due to new antibody display and other library screening techniques, which are now being exploited to isolate human antibodies with high affinity and specificity. The microbial surface display technologies for screening antibody libraries include phage, yeast and bacteria. Phage-display is widely used due to its simplicity, versatility and ability to be adapted to many specific conditions, including selection on whole cells and tissues [8]. Yeast and bacteria display platforms have several advantages over the phage system including use of flow-cytometry and sorting techniques to enable finer affinity discrimination of selected antibodies [9], [10]. Among the non-microbial systems is ribosomal display that has the capacity to screen libraries of greater size as well as facilitating diversity and efficient antibody maturation in vitro.Although isolation of human antibodies from the above mentioned systems has been successful, there can be unexpected problems with subsequent therapeutic mAb development due to constraints of protein expression, correct folding and post-translational modifications. This has been particularly true for antibodies isolated by phage-display technology. There has been great interest in screening antibodies directly from mammalian cells due to their ability to provide proper posttranslational modification, as well as the existence of the natural chaperones that assist in antibody folding. Animal cells have been used for the direct screening of hypermutating antibodies [11], [12] and during antibody selection from a retroviral-antibody display library [13]. Transient antibody expression on the surface of human 293T has also been recently reported as a system to perform in vitro affinity maturation of human antibodies [14]. Furthermore, sulfation of tyrosine residues in the CDR residues of human antibodies can markedly affect antigen recognition [15], [16] and contribute bidirectionally to the binding activity of antibodies [17]. These latter findings suggest that antibody selection and expression on the surface of human cells may not only identify a population of antibodies that would be difficult or even impossible to detect in other microbial or cell-free display systems, which lack the ability to sulfonate CDR tyrosines, but may also be able to select against antibodies that may otherwise loose activity upon transferring to mammalian expression systems.In this report, we show that bivalent functional human scFvFc fusion proteins can be efficiently expressed on surface of lentiviral transduced human cells, as well as incorporated onto the surface of lentiviral particles. The displayed scFvFc antibodies can undergo post-translational CDR tyrosine sulfation. Combined magnetic bead and FACS selections on transduced human cells have provided, proof-in-principle, that 106-fold enrichments of specific antibodies can be achieved in a single, rapid selection step. In addition, scFvFc displaying human cells could be used directly in functional biological screens with remarkable sensitivity.ResultsOptimization of scFv surface expression in mammalian cellsPS11 scFv, an antibody targeting the Tat-recognition motif (TRM) of cyclin T1 [18], was chosen as a model for optimizing functional expression of scFv on the surface of mammalian cells. To gain bivalency and increase the sensitivity of detecting antigens bound to surface antibody, the PS11 scFv was expressed as an scFvFc fusion protein [18]–[20]. For anchoring to the cell membrane, PS11 scFvFc protein was fused, in frame, to a transmembrane (TM) moiety. TM domains of HIV-1 gp41, CD8 and CD28 were tested for maximal surface expression of the scFvFc. As shown in Figure 1, all anchoring moieties consist of a short extracellular region, an entire TM domain and a cytoplasmic tail. Eight residues of the most membrane-proximal HIV-1 gp41 cytoplasmic tail, previously shown to provide a putative “envelope (env) incorporation motif” [21], were also tested for their ability to promote efficient pseudotyping of the scFvFc fusion proteins onto HIV virions or subsequently the cell surface, as a direct fusion to the gp41 TM or attached to the carboxy terminus of the CD8 or CD28 cytoplasmic tails.10.1371/journal.pone.0003181.g001Figure 1Diagram of constructs used in the study.ScFv antibodies were inserted between the leader peptide (LP) and the Fc region of a human IgG1 molecule. The Fc domain was linked in-frame to a short segment of extracellular domain of HIV-1 gp41 (blue), CD8 (green) or CD28 (purple), followed by their respective transmembrane domains (TM; horizontal stripes) and cytoplasmic domains (vertical stripes). In the case of HIV-gp41, the last 19 residues of the extracellular region (solid blue) are followed by a TM spanner (22 residues; blue horizontal stripes) and a cytoplasmic tail (blue vertical stripes). Either the full length 151 residues of the cytoplasmic domain or a truncated region that includes only the first eight residues of the cytoplasmic tail were used. Numbering is according to p160 of HIV-1 HXB2. For CD8, the most membrane-proximal 12 residues of the extracellular domain (solid green) and 11 residues of the cytoplasmic domain (green vertical stripes) flank 21 residues of the TM region (horizontal green stripes). For CD28, an extracellular region consisting of 40 residues (solid purple) and a cytoplasmic region of 13 residues (purple vertical stripes) flank the 27-residue TM domain (horizontal purple stripes). To facilitate scFvFc-TM incorporation into virions, an eight-residue “env incorporation motif”, which encodes the membrane proximal part of the gp41 cytoplasmic tail (NRVRQGYS; single blue line-amino acids 706–713), was attached to the carboxy-terminal ends of the cytoplasmic domains of CD8 and CD28. A nine amino acid C9 tag (red box) is positioned at C-terminus of all Fc domains to facilitate detection/quantitation of scFvFc expression on the cell surface. The gene cassette was cloned into pCDNA3.1 or the modified pHAGE lentiviral vector between Sfi-I and Pac-I sites. A CMV promoter controls expression of the scFvFc-TM transgenes.Surface expression of the PS11-scFvFc-TM proteins was initially analyzed by FACS analysis of transiently transfected 293T cells, stained with APC-conjugated anti-human-Fc antibody (Figure 2). While transfection efficiency with each of the constructs was relatively equal, as monitored through the expression level of a co-transfected GFP plasmid (data not shown), depending on the transmembrane moiety, differences in cell-surface expression of PS11-scFvFc were observed in both the percentage of cells that were positive for scFvFc expression (Panel a), and more pronounced by their respective MFI values (Panel b). The data indicate that PS11-scFvFc antibodies anchored by the TM of CD8 (lanes 6 and 7) or CD28 (lanes 8 and 9) were highly expressed on the surface of mammalian cells, compared to PS11-scFvFc fused to HIV-gp41 TM (either a long or short cytoplasmic tails; lanes 4 and 5) that were poorly surface expressed, and their MFI values were low.10.1371/journal.pone.0003181.g002Figure 2Optimization of scFvFc cell-surface expression using different transmembrane domains.293T cells were transfected with the pcDNA 3.1 based constructs encoding PS11-scFvFc antibodies of different configurations as described in Figure 1 and labeled under each lane in Panels a and b. pcDNA3.1-CMV-GFP was co-transfected as an internal control for transfection efficiency. At 48 hours post transfection, cells were harvested and analyzed for GFP and scFv-Fc expression by FACS analysis. Panels a and b, represent results from FACS analysis of the percentage of cells that are positive for APC-anti-human Fc staining (a) and their respective MFI values (b). Error bars represent the standard deviation of the average of three experiments. Panel c. Cellular localization of the PS11-scFvFc-TM analyzed by confocal immunomicroscopy. 293T cells were transfected with either ZsGreen expression vector alone, or with a bicistronic vector expressing both the PS11 scFvFc-TM fusion proteins and ZsGreen. At 48 hours post transfection, cells were stained with a rhodamine-conjugated anti-human Fc for the detection of scFvFc expression as visualized by a confocal microscope. Image a, cells transfected with ZsGreen only vector; Images b\n and \nc, cells transfected with vectors expressing either PS11-scFvFc-gp41 (665–856)-IRES ZsGreen or PS11-scFvFc-CD28-gp41 (706–713)-IRES-ZsGreen, respectively. Absence of the ZsGreen fluorescence in some of the APC+ cells is likely the result of low level expression of ZsGreen from the second cassette of the bi-cistronic message. Panel d. PS11-scFv-CD28-gp41 is present as a dimer in transfected cells. 293T cells expressing pCDNA3.1-PS11-scFvFc-CD28-gp41 fusion protein were metabolically labeled with [35S]-cysteine and [35S]-methionine mixture. Cell lysates were immunoprecipitated with protein A sepharose beads, resuspended with 2× SDS non-reducing (lane 1) or reducing buffer (lane 2), and subjected to SDS-PAGE and autoradiogram.To directly visualize the distribution and localization of scFvFc-TM expression, cells transfected with a bicistronic IRES-ZsGreen expression vector encoding PS11-scFvFc-gp41 (665–856), or PS11-scFvFc-CD28-gp41 (706–713) were labeled with a rhodamine conjugated anti-human Fc antibody for immunofluorescence analysis. As shown in Figure 2c, cells expressing PS11-scFvFc-gp41 (665–856) demonstrated punctate staining with large aggregates and exhibited an overall low level of cell-surface expression (image b). In contrast, PS11-scFvFc-CD28-gp41 (706–713) proteins were evenly distributed on the cell surface and also had a reticular staining pattern, consistent with efficient ER folding and expression (image c). As a control, rhodamine conjugated anti-human Fc staining was not detected on cells transfected with ZsGreen encoding vector alone (image a). These results are consistent with the FACS data shown in Figure 2a and 2b. Low expression and possible aggregation of PS11-scFvFc-gp41 (665–856) may be a result of poor folding, as the natural Fc moiety forms dimers, while gp41 forms trimers through its TM. Finally, radio-immunoprecipitation and SDS-PAGE analysis confirmed that the membrane bound PS11-scFv-CD28-gp41 (706–713) protein was dimeric (Figure 2d).Specific antigen binding by mammalian cell-surface expressed scFvFc-TM antibodiesTo confirm that the cell surface-anchored PS11-scFvFc remains functional for binding to its antigen, 293T cells transfected with different PS11-scFvFc-TM fusion proteins were stained with a biotinylated-TRM peptide and analyzed by FACS, using APC-conjugated streptavidin. As seen in Figure 3, all PS11-scFvFc-TM proteins were functional for binding the biotinylated TRM peptide. The binding was specific, since an irrelevant X48-scFvFc, recognizing a biotinylated CXCR4 peptide [17], showed only a relatively low level of APC-streptavidin staining (lane 4). Overall, the PS11-scFvFc-CD8-gp41 and PS11-scFvFc-CD28-gp41 proteins were the most competent in binding the peptide and exhibited higher levels of peptide binding compared to the corresponding constructs without the envelope incorporation motifs (Figure 3b, compare lanes 8 and 10 to lanes 7 and 9, respectively). This result suggests that this motif may stabilize surface antibody expression, by promoting its proper folding and membrane association. The lower surface expression of PS11-scFvFc linked to the gp41 TM region (Figure 2a and b) lead to a dramatically lower binding capacity to the biotinylated TRM peptide, as compared with the PS11-scFvFc fused to the CD8 or CD28 TM regions, evidenced by both a lower percentage of antigen binding cells and MFI values (Figure 3a and b; compare lanes 5 and 6 with lanes 7 and 9).10.1371/journal.pone.0003181.g003Figure 3Cell surface expressed scFvFc proteins bind their cognate antigens.293T cells were transfected with the same constructs as described in Figure 2 and labeled under each lane in Panels a & b. Two additional constructs encoding antibodies against CXCR4, X20- and X48-scFvFc-CD28-gp41, were also transfected. pcDNA3.1-CMV-GFP was again co-transfected as an internal control for transfection efficiency. At 48 hours post transfection, cells were harvested and stained for biotinylated-TRM and streptavidin-APC, followed by FACS analysis. GFP expression was also analyzed to ensure equal transfection efficiencies. Panel a and b, depict the percentage of positive cells that express a functional PS11 scFvFc as determined by staining with streptavidin-APC (Panel a) and their respective MFI values (Panel b). Error bars represent the standard deviation of the average of three experiments. P values<0.05 above the designated bars, represent statistically significant difference in MFI values. Panel c. Post-translational sulfation occurs in selected surface displayed scFvFc antibodies. 293T cells expressing cell surface X48 or X20-scFcFc-CD28-gp41 fusion proteins (lanes 2 and 3, respectively) were labeled with [35S]-cysteine and [35S]-methionine mixture (upper panel; Cys/Met) or with [35S]-sulfate (lower panel; SO4) with or without 100 mM sodium chlorate treatment. Cell lysates were immunoprecipitated with protein A sepharose beads, washed and analyzed by SDS-PAGE and autoradiography. pcDNA3.1 backbone empty vector was also used as negative control (lane 1).Tyrosine sulfation of the mammalian cell-surface expressed scFvFc-CD28-gp41 antibodiesWe have demonstrated that in self-reactive human anti-CXCR4 antibodies, tyrosine sulfation occurs in novel areas of the V-region genes and contributes bidirectionally to antibody binding activity [17]. To determine if tyrosine sulfation could also occur on surface displayed scFvFc, the self-reactive human anti-CXCR4 antibodies X20- and X48-scFv were analyzed as scFvFc-CD28-gp41 (706–713) fusion proteins. Radioimmunoprecipitation studies confirmed that sulfation indeed occurred in the surface displayed X20-scFvFc-CD28-gp41 but not with X48-scFvFc-CD28-gp41 (Figure 3c, compare lower lanes 3 and 2, respectively). Treatment of transfected cells with sodium-chlorate, a sulfation inhibitor [15], [22], decreased expression but more significantly abolished sulfation of scFvFc proteins (Figure 3c lower and upper panels, respectively). This is consistent with the results for each corresponding soluble scFv, where sulfation was mapped to tyrosine in VH CDR2 and VL FW3 regions of X20 and required for maximal binding and antigen recognition activity [17].Incorporation of functional scFvFc-TM proteins into lentivirus particlesThe IgG leader-scFvFc-CD28/CD28-gp41 (706–713) coding sequences were next cloned into the first cassette of a bicistronic self-inactivating (SIN) lentivector containing an IRES-ZsGreen reporter gene. Viruses encoding PS11-scFvFc, X48-scFvFc, as well as two SARS-CoV specific antibodies 80R-scFvFc (19, 23) and 11A-scFvFc (Sui et al, submitted) that recognize Tor2 or GD03 Spike protein, respectively, were produced through co-transfecting cells with HIV packaging plasmid and a VSV-G envelope DNA plasmid providing surface binding and fusogenic activity for viral entry. These viruses were analyzed for incorporation of scFvFc into the viral envelope and their capacity to bind specific antigens; and were further used to establish a mammalian cell display of surface-bound antibodies through transduction.Incorporation of scFvFc-TM was first examined by western blot analysis of equal amounts of viral particles, as determined by p24 levels (Figure 4a, lower panel). Using an anti-human Fc antibody, both PS11-scFvFc-CD28-gp41 (706–713) and PS11-scFvFc-CD28 were detected in the purified viral particles (Figure 4a, upper panel, lanes 2 and 3, respectively), while a control CMV-GFP lentivirus showed no reactivity (upper lane 1). Most importantly, the gp41 (706–713) env incorporation motif-encoding viruses exhibited higher Fc expression (compare upper lanes 2 and 3), confirming a more efficient incorporation of the PS11-scFvFc-CD28-gp41 into lentiviral particles, which is in agreement with the results obtained in Figure 3b.10.1371/journal.pone.0003181.g004Figure 4Functional scFvFc are incorporated into lentivirus particles.\nPanel a. An HIV-1 gp41 incorporation motif enhances scFvFc incorporation into viral particles-equal loads of lentiviruses encoding ZsGreen (lane 1), PS11-scFvFc-CD28-gp41-IRES ZsGreen (lane 2) or PS11-scFvFc-CD28-IRES ZsGreen (lane 3) were subjected to SDS-PAGE analysis, followed by western blotting using either HRP-conjugated anti-human Fc (upper panel) or anti-HIV-1 p24 (lower panel) antibodies. Panel b. Immunostaining of viruses. Viral particles were attached to Hela cells for 2 hour at 4°C. Cells were then fixed and stained with anti-HIV p24 antibody followed by Cy2-conjugated anti-mouse IgG or a rhodamine-conjugated anti-human Fc for surface Fc staining. Following these procedures, cells were washed and analyzed by confocal microscope. Shown separately are viruses stained for detection of p24 (image a) and scFvFc (image b) along with a merged image (image c). Panel c. Antigen specific capture of lentiviruses displaying corresponding scFvFc antibodies. Equal amounts of lentivirus particles expressing on their surface either the PS11-scFvFc or the 11A-scFvFc were loaded on a 96-well plate that was coated with the following specific antigens: TRM-peptide (PS11 specific), GD03-Fc (11A specific) or BSA. Following incubation to allow capture of the viruses to the antigens, wells were washed extensively and viral particles were eluted and quantitated by RT assay. Presented are normalized RT counts, where RT counts of particles bound to their antigens were divided by the RT counts of virus bound to the BSA control. Values are the average of duplicated samples and data are representative of two separate experiments.Upon binding and fixation of purified viral particles to Hela cells, an immunostaining protocol was performed to confirm that viral particles incorporated the scFvFc fusion proteins. Confocal microscopic images shown in Figure 4b indicated that viruses expressing PS11-scFvFc-CD28-gp41 (706–713) on their surface were stained with both anti-HIV-1 p24 antibody (image a) and anti-human Fc antibody (image b). Merging of the two staining profiles confirmed co-localization of the core with scFvFc (image c). Control CMV-GFP viral particles, with no scFvFc molecules on their surface, stained positively with the anti-p24 antibody only (data not shown).To further verify that scFvFc-CD28-gp41 (706–713) proteins are displayed on the surface of lentivirus and remain functional, a virion capture experiment was performed. Equal amounts of viral particles [based on reverse transcription (RT) value], expressing on their surface either PS11- or 11A-scFvFcs, were incubated in a 96-well plate where wells were coated with either biotinylated TRM peptide (PS11-scFvFc specific) or GD03-Fc protein (11A-scFvFc specific) antigen. BSA served as a negative control. Following binding and extensive washing, the amount of captured viral particles in each well was determined by RT assay. As shown in Figure 4c, each virus bound to its own target with a very high selectivity/specificity. Hence, recombinant lentiviral particles could be efficiently pseudotyped with functionally intact scFvFc-CD28-gp41 fusion proteins.Characterization of scFvsFc expressed on the surface of lentivirus transduced cellsTo establish conditions for mammalian cell display of scFvFc antibodies, 293T cells were transduced with PS11-scFvFc-CD28-gp41-IRES-ZsGreen encoding lentiviruses. As shown in Figure 5a, transduced cells efficiently expressed both PS11-scFvFc, as detected by APC-anti-human Fc and ZsGreen. The difference in transgene expression could be a result of either higher sensitivity of APC-anti-human Fc staining and/or less efficient CAP-independent IRES driven expression of ZsGreen. Importantly, cell-surface expression of scFvFc was detectable at very low multiplicity of infection. Quantification analysis revealed that at MOI of 1, there was about 5000–8000 of PS11-scFvFc-CD28-gp41 surface expressed molecules per transduced human cell (see Material and Methods for details on quantification methods).10.1371/journal.pone.0003181.g005Figure 5Expression of scFvFc on the surface of lentivirus transduced cells.\nPanel a 293T cells were transduced with increasing dilutions (different MOIs as indicated) of lentivirus encoding the PS11-scFvFc-CD28-gp41-IRES-ZsGreen. Transduced cells were harvested, stained for Fc-surface expression, and analyzed by FACS. Expression of ZsGreen was measured to monitor levels of transduction. The graphs depict the percentage of transduced cells that express ZsGreen (blue diamonds) and the percentage values of transduced cells that express PS11-scFvFc as monitored by staining with APC-conjugated anti-human Fc IgG (pink squares). Panels b and c, cell -surface expressed scFvFc proteins are functional. 293T cells were transduced with a lentivirus encoding PS11-scFv-Fc-CD28-gp41-IRES-ZsGreen (Panel b) or 11A-scFvFc-CD28-gp41-IRES-ZsGreen (Panel c). Cells were incubated with biotinylated GD03-Fc, a specific antigen for 11A-scFvFc, and stained for streptavidin-APC as described in Methods, and then analyzed by FACS. Note that two clusters of cells in Panel c represent high (R2 gated) and low (R3 gated) levels of 11A-scFvFc on their surface as measured by APC staining. These could reflect variations in cell-surface expression levels resulting from multiple integration events of the scFvFc cassette following transduction, or the difference in transgene integration site, i.e., its proximity to active transcriptional units. R2 = 1390 and R3 = 220 are MFI values of 11A scFvFc expressing cells, where percentage of positive cells in each gate is 31% and 49% respectively. Panel d. a summary of specific antigen binding by the scFvFc displayed on the lentivirus transduced cells. The table shows the percentage of transduced cells expressing X48-scFvFc-CD28-gp41 or PS11 scFvFc-CD28-gp41 that bind to their cognate or irrelevant biotinylated antigens as visualized by APC staining and their corresponding MFI values.Antigen binding specificity of surface expressed scFvFcs were confirmed by incubating cells transduced with CD28-gp41-IRES-ZsGreen lentiviruses encoding either 11A-, PS11- or X48-scFvFc with biotinylated antigens, followed by APC-streptavidin staining. As shown in Figure 5, upon incubation with its specific biotinylated GD03-Fc protein antigen, 11A-scFvFc expressing cells could be easily detected (Panel c), while the control PS11-scFvFc expressing cells exhibited only background levels of APC-streptavidin staining (Panel b). Similarly, following incubation with a biotinylated N-terminal CXCR4 peptide, 37% of cells transduced with the X48-scFvFc-CD28-gp41-IRES-ZsGreen lentivirus stained positive with streptavidin-APC (Figure 5d). In contrast, only background staining was detected when the same transduced cells were incubated with biotinylated GD03-Fc protein. Comparable results were seen with PS11-scFvFc, which specifically binds TRM peptide but not an irrelevant GD03-Fc protein.Selection and enrichment of rare scFvFc antibodies displayed on the surface of lentivirus transduced human cellsTo determine if transduced cells expressing scFvFc fusion proteins on their surface could serve as a platform for isolating new scFvs, 11A-scFvFc- scFvFc-CD28-gp41 and PS11-scFvFc-CD28-gp41 cells were mixed at decreasing concentrations of the former and the sensitivity of the isolation and enrichment process was evaluated. Our initial results indicated that, at one-week post-viral transduction, a single round of selection by direct FACS sorting of high antigen binding/ZsGreen expressing cells, resulted in a three log enrichment of antigen specific 11A-scFvFc surface displayed cells, from a background cell population. However, isolation of 11A-scFvFc expressing cells could not be reliably achieved at a mixing ratio below 1∶1000 (data not shown).We hence modified the selection procedure in order to improve sensitivity for specific antibody detection. Lentivirus transduced cells were pre-sorted for ZsGreen expression soon after transduction. Upon further propagation, 11A- and PS11-scFvFc cells were mixed at different ratios; incubated with a fixed concentration of biotin-GD03-Fc protein and streptavidin-APC; and an enrichment step, using MACS-anti-APC microbeads (Miltenyi), was performed prior to FACS analysis and sorting of streptavidin-APC positive cells. FACS analysis showed that the enrichment procedure was highly efficient at cell mixing ratio of 1∶106, reaching at least 45-fold (compare Figure 6a R2 gates of left and middle panels; note 0.1% cells within R2 gate are non-distinguishable from a background value). Following magnetic bead enrichment, high (R2 gate) and low (R3 gate) APC-stained and ZsGreen expressing cells were again sorted and the two cell populations (about 500–1000 cells) were propagated for one week. A portion of cells was further propagated to reach a sufficient number for re-staining with biotinylated GD03-Fc protein and APC-conjugated streptavidin, while scFv genes from the remaining cells were rescued by PCR amplification of genomic DNA for rapid recloning and scFv DNA sequence analysis. As shown in Figure 6a (right panel), upon propagation and re-staining, the majority of originally high APC staining cells were positive for GD03-Fc binding (81% within the R2 gate), while cells isolated from R3-gate expressed low levels of ZsGreen and human Fc staining, but did not bind to biotin-GDO3-Fc (data not shown). DNA analysis confirmed that 51/56 of the clones generated from the R2-gated cells were positive for the 11A scFvFc gene (Figure 6b). In contrast, 40/45 clones from the R3-gated cells encoded the PS11 scFvFc gene and only 2/45 clones expressed the 11A scFvFc. Overall, magnetic beads enrichment combined with FACS sorting of high antigen binding/ZsGreen expressing cells resulted in a 106 fold-enrichment of antigen binding cells in a tandem two-step round of selection.10.1371/journal.pone.0003181.g006Figure 6Selection of rare scFvFc expressing cells by a two-step magnetic bead and FACS sorting procedure.293T cells were transduced with either PS11-scFvFc-CD28-gp41-IRES-ZsGreen or 11A-scFvFc-CD28-gp41-IRES-ZsGreen encoding lentiviruses. These two transduced cell populations were mixed at different 11A- to PS11-scFvFc ratios with a total cell number of ∼109. Mixed cells were incubated with the biotinylated GD03-Fc protein antigen and APC-streptavidin. APC-positive cells were subjected to an enrichment using anti-APC magnetic micro-beads followed by FACS sorting. Cells from either R2 gate (high APC expressing cells) or R3 gated (low APC expressing cells) were isolated, propagated, re-stained for biotin-GD03-Fc binding and also analyzed for their scFv gene content by PCR rescue and DNA sequencing. Panel a, the FACS dot-blot profiles of APC- and ZsGreen positive cells within the 11A∶PS11 scFvFc expressing cell population (initial mixing ratio at 1∶106), before magnetic beads enrichment (left panel), following magnetic bead enrichment (middle panel), and the R2 gated cells after sorting and expansion (right panel). Panel b, DNA sequencing results of individual clones recovered from the R2 or R3 gated pools of cells sorted from different 11A∶PS11 scFvFc expressing cell ratios (1∶103–1∶106). Note that the irrelevant sequencing data are most likely originated from cloning background.Neutralization of infection mediated by cell-surface displayed scFvFcIt was next determined if transduced human cells expressing the unique surface anchored scFvFc could be used directly in a biological screen. The anti-SARS-CoV 80R antibody was chosen as a model system for these studies [19], [23]. Luciferase expressing lentivirus pseudotyped with the cognate Tor2 Spike protein was absorbed with increasing numbers of 80R-scFvFc or irrelevant PS11-scFvFc expressing cells, prior to their single round infection of permissive cells expressing the SARS-CoV receptor, ACE-2. Non-specific virus absorption by scFvFc expressing 293T cells was also controlled with VSV-G pseudotyped viral particles. As shown in Figure 7, close to 100% neutralization of TOR2 pseudotyped virus infection was achieved following incubation with 80R-scFvFc surface expressing cells as compared with 40% inhibition seen by incubation with the PS11 scFvFc expressing cells. In contrast, neither PS11 nor 80R scFvFcs could neutralize the infection of permissive cells by VSV-G pseudotyped particles. Thus, very few numbers of transduced human cells expressing a unique scFvFc can be used directly in biological screens with exquisite sensitivity.10.1371/journal.pone.0003181.g007Figure 7Neutralization of SARS-CoV TOR2 spike protein pseudotyped lentiviral infection of ACE2 expressing cells mediated by cell-surface displayed anti-TOR2 spike 80R-scFvFC antibodies.Single round, TOR2 spike protein pseudotyped luciferase expressing lentiviral particles were incubated with increasing concentrations of 293T cells expressing on their surface the 80R scFvFc (blue diamond) or the control PS11 scFvFc (pink circle). As a control for non-specific reporter virus absorption, a VSV-G pseudotyped luciferase reporter lentivirus was also incubated with 80R scFvFc (red triangle) or PS11 scFvFc expressing cells (brown square). Following incubation, the supernatant containing remaining lentivirus was used to infect permissive cells that express the ACE2 receptor for SARS CoV. At 48 hours post transduction, cells were harvested, luciferase activity was measured and relative inhibition of reporter virus infection was calculated. Asterisks in the designated points represent a statistical analysis that was performed to verify significant differences in % of inhibition between viral absorptions with the 80R- or PS11-scFvFc at a specific cell number point (P<0.05).DiscussionThis study demonstrates that human cells, the natural host of human antibodies, can serve as a scaffold for antibody surface expression, screening and isolation using lentivirus display. Several thousand functional, bivalent scFvFc fusion proteins were stably expressed on the surface of human cells. The scFvFc antibodies when fused to the CD28/CD8 transmembrane moieties were evenly distributed on the cell surface. A number of technical features of this lentiviral display system were explored in this study and deserve further comment.First, transduction of human cells with a self-inactivating, bicistronic lentiviral vector encoding scFvFc-CD28-gp41 and ZsGreen proteins, combined with magnetic beads enrichment and FACS sorting, resulted in a 106-fold enrichment of specific antibody expressing cells in one single tandem, two-step procedure, to levels comparable or superior to those achieved by other microbial display systems [24], [25]. Together with optimized scFv PCR rescue and re-expression, this system should allow the development of rapid, iterative antibody enrichment procedures. Indeed, in the scFv gene PCR rescue experiment described in Figure 6b, only one week of cell propagation was used and shorter times are clearly possible. It should be noted that two cell populations were commonly detected as low (R3 gate) and high (R2 gate) ZsGreen expression and GD03-Fc antigen binding. This occurred despite optimized amounts of biotinylated GD03-Fc protein that were used to limit APC-streptavidin cross-reactive staining of irrelevant PS11 scFvFc transduced cells (Figure 5b). Sequencing analysis confirmed that the majority of the R3-gated cells expressed the irrelevant PS11 scFvFc, while the vast majority of the R2-gated cells encoded 11A scFvFc (Figure 6b). Thus, careful consideration of the background threshold for each antigen, and sorting for only high antigen binding and ZsGreen expressing cells, are useful guidelines to maximize the recovery of antigen binding positive cells. In addition, the positive signal to noise MFI ratio could potentially be increased through adjusting levels of biotin conjugation to an antigen probe or through an extra biotin/avidin amplification step. Initial MOI for transductions should also be controlled to ensure both library diversity and antibody expression.Second, although not explored in detail in this study, human B-cells are also efficiently transduced and display high levels of scFvFc-CD28-gp41 antibody on their surface (data not shown). The endogenous biochemical pathways that are responsible for hypermutation of antibodies are constitutively expressed in human B-cells [11], [12] and/or can be further manipulated in an inducible manner through retroviral gene transfer [26], [27]. Thus, affinity maturation by somatic hypermutation of displayed antibodies may be possible and would have the advantage of occurring in stable transduced cell lines, allowing easy subcloning and recovering of the scFv gene.Third, the lentiviral display system should complement existing antibody display technologies by virtue of providing properties unique to antibody expression in human cells. For example, surface display of human antibodies modified by CDR tyrosine-sulfation has not been reported for other antibody display systems. Based on the fact that V-region tyrosine sulfation did occur on individual surface displayed scFvs (Figure 3c), expression of scFvFc proteins on human cells should allow isolation of antibodies exhibiting unique properties of functional tyrosine sulfation that could otherwise be missed through expression by ribosomal display or other microbial display systems, where post-translational sulfation of tyrosine residues by Golgi associated tyrosine-O-sulfonyl transferases does not occur [28]. In addition, the antibodies isolated via human cell screening should express efficiently in mammalian cell systems, without the unpredictable problems that are frequently seen in expression of antibodies selected by phage display.A fourth distinct advantage of this system is that the scFvFc displaying human cells could be used in direct biological screens with remarkable sensitivity. Studies described in Figure 7 showed that virus neutralization activity could be detected with as few as 320 scFvFc expressing cells. Although not yet tested, it also seems likely that transduced cells expressing bivalent surface displayed antibodies may mediate cross-linking of antigen molecules expressed on the surface of human target cells following co-incubation, which could lead to positive or negative modulation of signaling and biological responses of the target cells. This may provide an early and direct screen to interrogate the desired biological activity of the antibodies without the need to initiate costly soluble antibody production and purification procedures until the lead antibodies are identified.Finally, an important feature of this system is that the functional scFvFc antibodies were successfully pseudotyped and expressed on lentiviral surface (Figure 4). Thus, the scFvFc pseudotyped lentiviral particles could also serve as a highly specific targeted gene delivery vehicle, particularly when fusion functions are provided in trans, as has been recently reported [29], [30].In summary, relative ease in generating high titer lentiviral stocks [14], [31]–[33] combined with the high permissiveness of 293T cells to lentiviral transduction provides a platform which, we believe, could be easily scaled up to host a large diverse human scFvFc library. A human scFvFc-display master cell bank would serve as a rich source for screening and isolation of high affinity human scFv. By fully exploiting this lentivirus antibody display system, the isolation of new human antibodies with unique structural and biochemical properties complementing existing display systems should be possible. Transfer of large and diverse human scFv libraries from phage to the lentivirus mediated scFvFc cell surface display platform and panning against common target antigens using these alternative screening systems are ongoing. These comparative studies, as have been similarly performed for yeast and phage [34], will help define the value of lentivirus display in the discovery of novel therapeutic human mAbs.Materials and MethodsConstruction of mammalian cell surface displayAll scFv antibodies used in this study were originally derived from the Mehta I/II non-immune human scFv-phage libraries [18]. ScFv were cloned into a pcDNA 3.1-based expression vector as an Sfi-I/Not-I 856 bp insert, and fused in frame with a human Fc-region (hinge-CH2-CH3) that had been amplified by PCR and cloned into pCDNA 3.1 as a Not-I/Xba-I fragment. Transmembrane (TM) anchoring moieties were amplified by PCR, using the appropriate primers and templates (sequences available upon request) and were cloned, in-frame, as Xba-I/Pac-I digested PCR fragments, into the pCDNA 3.1-PS11-scFvFc expression vector. A C9 sequence [34] was inserted N-terminus of all TM domains. Plasmids DNA were sequenced and verified for cell-expression.FACS analysis of functional expression of scFvFc on the surface of transiently transfected cells293T cells were seeded a day before transfection on a 6 well plate. At the day of transfection, the 95% confluent cells were co-transfected with 4 µg of each of the plasmids DNA (see Figure 1) and 0.1 µg of pCDNA3.1-CMV-GFP, using lipofectamine 2000 (Invitrogen). APC-conjugated anti-human Fc antibody (Jackson ImmunoResearch) was used to determine cell-surface expression level of scFvFc. 5×105 293T cells were harvested using 5 mM EDTA at 48 hour post transfection, washed with PBS, and incubated on ice for 1 hour with a PBS staining solution containing 1 µl of APC-conjugated anti-human Fc antibody (Jackson ImmunoResearch) and 2% BSA per sample. Following incubation, cells were washed 3 times with PBS+2% BSA and analyzed by FACS. Mock- transfected cells incubated with the APC-conjugated anti-human Fc antibody were used as controls. Transfection efficiency of each sample was verified through GFP expression and analyzed concurrently by FACS.To analyze if cell-surface expressed PS11-scFvFc proteins remain functional for specific antigen binding and whether the different TM domains have an effect on scFvFc function, 293T cells were transiently transfected and harvested as described above followed by incubation with biotinylated Tat Recognition Motif (TRM) peptide (“Macromolecular Resources”, CO) at 10 µM final concentration/sample on ice for 30 minutes and washed 3× with PBS+2% BSA. Cells were further incubated on ice, in the dark, with a staining PBS solution containing 2% BSA and 1 µl of streptavidin-APC (Jackson ImmunoResearch). Finally, cells were washed 3 times with PBS+2% BSA and analyzed for APC staining by FACS. GFP expression of transfected cells was also analyzed to standardize transfection efficiency.Confocal microscopy analysis of scFvFc cellular localization in transiently transfected 293T cells293T cells were transfected with DNA encoding for a PS11-scFvFc-CD28-gp41-IRES-ZsGreen, PS11-scFvFc-gp41-IRES-ZsGreen or CMV-ZsGreen plasmids. The latter served as a negative control for cells that do not express scFvFc on their surface. At 48 hours post transfection, samples were fixed with 3.9% paraformaldehyde (SIGMA) for 30 minutes and washed once with PBS. Subsequently, cells were incubated with PBS/0.1 M glycine (SIGMA) for 10 minutes, followed by washing once with PBS and permeabilization with PBS/0.05% saponin (SIGMA) for additional 30 minutes. Upon further washing, cells were blocked with PBS supplemented with 2% BSA and PBS/0.05% saponin for 30 minutes and incubated in the dark with an anti-human Fc-rhodamine antibody (Jackson ImmunoResearch) for 1 hour. Finally, cells were washed and mounted for flurescence microscopy by using ProLong antifade kit (Invitrogen). Images were acquired by using BIO-Rad Radiance 2000 laser scanning confocal microscope using a Nikon 60× camera.Metabolically radioisotope labeling and immunoprecipitation of scFvFc-CD28-gp41 fusion proteinsSurface displayed scFvFc post-translational sulfation analysis was performed as described earlier [17]. Briefly, 293T cells were transiently transfected with pCDNA3.1-scFvFc-CD28-gp41 expression plasmids using lipofectamine 2000 (Invitrogen). Eighteen hours later, cells were washed twice with PBS. To determine protein sulfation, one set of cells was incubated with sulfate-free media (Sigma), supplemented with 500 µCi of [35S]-sulfate (PerkinElmer), with or without 100 mM sodium chlorate. For analyzing protein expression, another set of cells was incubated in parallel with L-Methionine and L-cysteine free DMEM medium (GIBCO), supplemented with 200 µCi of [35S]-labeled cysteine-methionine mixture (PerkinElmer), with or without 100 mM sodium chlorate. Cells were collected 24 hours later and lysed with solubilization buffer containing 100 mM (NH4)2SO4, 20 mM Tris (pH 7.5), 20% glycerol, and 1% 3-[(-cholamidopropyl) dimethylammonio]-2-hydroxyl-1-propanesulfonic acid (CHAPSO, Anatrace, Maumee, Ohio) in the presence of 1× complete protease inhibitor mixture (Roche Molecular Biochemicals, Indianapolis, Ind.). Cell lysates were incubated overnight at 4°C with protein A sepharose beads (GE Healthcare) and washed three times with PBS containing 0.1%Tween 20. Finally, proteins were eluted from the beads by 2XSDS buffer separated by 12% SDS-PAGE, and visualized by autoradiogram.Generation of scFvFc-TM/VSV-G pseudotyped lentiviruses and 293T cells stably expressing scFvFc-TM through lentiviral transductionA self-inactivating pHAGE lentivector (a gift from R. Mulligan) was modified to accommodate subcloning of Sfi-I/Pac-I scFvFc-TM DNA fragments. pHAGE lentivector has a CMV promoter that drives the expression of the transgene. It also expresses the Zoanthus Green Fluorescent protein (ZsGreen) gene via an IRES sequence, which can be used to monitor and normalize transduction efficiencies.For the production of VSV-G pseudotyped lentiviral particles, total of 5×106 293T cells were seeded on 100 mm diameter plates, and co-transfected the next day with 10 µg of the pHAGE lentivector and packaging plasmids (10 µg HIV-1 Gag-Pol, 1 µg pCMV-Rev1b, and 2.5 µg pCMV-VSV-G), using the Ca-Phosphate method. At 48 and 72 hours post transfection, supernatant was collected and cleared by centrifugation (2100 rpm; 15 minutes) and sterile filtered through a 0.45 µm filter. Viral supernatant was concentrated by ultracentrifugation (21,000 rpm; 2 hours) through a 20% sucrose cushion, aliquoted and stored at −80°C. The titer of the pseudotyped virus particles was evaluated either by RT assay or by transduction of HeLa cells with increasing dilutions of the lentivirus stock and measurement of ZsGreen marked cells or APC-anti-human Fc staining by FACS analysis. Importantly, incorporation of the scFvFc-CD28/CD28-gp41 fusion protein did not lower the titer of viral particles (data not shown).293T cells were transduced, in the presence of 8 µg/ml polybrene, with above generated recombinant lentiviruses at different MOI as indicated for 4 hours. Functional scFvFc expression on the transduced cell surface was analyzed by FACS, following a similar protocol as described above, first at 48 hours post-trandsuction then again after longer periods of propagation to confirm stable scFvFc expression. Biotinylated peptide or protein antigens and their concentration used for specific antibody binding analysis are described in the figure legends.Incorporation of functional scFvFc-TM proteins into lentivirus particlesViral capture assayA 96-well plate was coated with different antigens overnight at 4°C, using a coating buffer-NaCO3/HCO3 (pH 9.6). The following antigens were used: commercial biotinylated TRM peptide 50 µM, biotinylated GD03 (S1-RBD)-Fc 20 µg/ml, or 20 µg/ml BSA. Duplicated wells were blocked with 1% BSA at 4°C for 1.5 hour prior to incubation with equal amounts of concentrated lentiviral particles (based on reverse transcription activity) expressing on their surface either the 11A or PS11 scFvFc antibodies for additional 1.5 hour at 4°C. Wells were then washed 5 times with 200 µl PBS and reverse transcription analysis was performed on eluted particles.Western blot analysisEqual amounts (based on p24 antigen levels) of concentrated lentiviruses were lysed using RIPA lysis buffer (50 mM Tris-HCl pH 7.4; 150 mM NaCl; 1 mM PMSF; 1 mM EDTA; 1% NP-40; 1% sodium deoxycholate; 0.1% SDS) supplemented with protease inhibitors (Roche, Indianapolis, Ind.) and resolved by SDS-PAGE under reducing conditions. Upon transfer, nitrocellulose membranes were blocked with 5% skim milk and proteins were probed with a mouse anti HIV-1-p24-HRP antibody (Immuno-Diagnostics), or a goat anti-human Fc-HRP antibody (Pierce) followed by detection using an enhanced chemiluminescence kit (Amersham).Immunofluorescence of pseudotyped virusesHela cells grown overnight on 12 mm cover-slips were incubated at 4°C for 30 minutes with media containing 10 mM HEPES, pH-8.0, followed by incubation for additional 2 hours with VSV-G pseudotyped viral particles expressing surface PS11-scFvFc-CD28-gp41-ZsGreen, or control ZsGreen virus. Samples were fixed with 3.9% paraformaldehyde (SIGMA) for 30 minutes. Cells were washed once more with PBS and incubated with PBS/0.1 M glycine (SIGMA) for 10 minutes, followed by another wash in PBS and permeabilization with PBS 0.05% saponin (SIGMA) for 30 minutes. Samples were then blocked with PBS supplemented with 2% BSA and 0.05% saponin for 30 minutes and incubated with either a mouse anti-p24 antibody (AG3.0; NIH AIDS Research and Reference Reagent Program) or with an anti-human Fc-rhodamine antibody (Jackson ImmunoResearch) for 1 hour. Cells were then washed and incubated with Cy2-labeled anti-mouse IgG antibody (Jackson ImmunoResearch) for an additional hour. Controls include virus-bound cells incubated with the secondary antibody alone. Finally, samples were mounted for fluorescence microscopy by using ProLong antifade kit (Invitrogen). Images were acquired by using BIO-Rad Radiance 2000 laser scanning confocal microscope using a Nikon 60×.ScFvFc quantification on the surface of transfected or transduced cellsTo quantify the number of scFv-Fc molecules on the surface of transiently transfected or lentivirus transduced cells, the “Quantum Simply cellular anti-mouse IgG kit” was used (Bangs Laboratories, Inc.). Briefly, cells expressing on their surface the scFvFc with a C9 tag were incubated with 10 µg of an APC conjugated mouse anti-C9 monoclonal antibody 1D4 (Invitrogen-Molecular Probes) for 1 hour on ice in PBS supplemented with 2% BSA. Calibration beads with known binding capacity of mouse IgG molecules on their surface were treated with the same conditions as the cells. Upon washing, both cells and calibration beads were analyzed for APC staining intensity by FACS. Number of scFv-Fc molecules on cell surface was determined by plotting a calibration curve, using the QuickCal quantitative software from Bangs labs.Isolation and enrichment of rare 11A scFvFc expressing cellsTransduction, ZsGreen sorting, and antigen staining of scFvFc expressing cellsAs a model for the isolation of rare antibody from a population of scFvFc expressing cells, 293T cells were transduced with lentiviruses encoding either 11A scFvFc or PS11 scFvFc at MOI of one. The transduced cells were propagated for one week to ensure stable expression and then sorted based on their ZsGreen expression. Upon further propagation, ZsGreen sorted cells were counted and mixed at the depicted ratio of 11A/PS11 scFvFc. For the detection and isolation of 11A scFvFc transduced cells, total mixed cells were blocked with 2% BSA for 30 minutes followed by staining with biotinylated GD03-Fc at an optimized concentration of 2.65 µg/ml (GD03-Fc protein was biotinylated using the EZ-link NHS-biotin, PIERCE) for 30 minutes on ice. Cells were then washed 3 times with PBS and stained with streptavidin-APC in PBS/2% BSA.Magnetic beads enrichment and FACS sorting of APC-positive 11A scFvFc expressing cellsFollowing staining with APC-Streptavidin, cells were washed and re-suspended in 500 µl PBS buffer containing 0.5% BSA and 2 mM EDTA. Cells were then labeled with MACS-anti-APC magnetic beads (20 µl/107 cells, Miltenyi) for 10 minutes at room temperature. Single-cell suspension was loaded on pre-washed MS magnetic columns that were placed in a magnetic field. Upon removing unbound cells through 3 times of washing, APC labeled cells were recovered by removing the columns from the magnetic field and plunging the cells using 1 ml of the above buffer. Cells isolated by anti-APC magnetic beads were analyzed by FACS. Two populations of APC positive cells, R2 and R3, were sorted. Upon propagation for one week, APC positive R2 and R3 sorted cells were divided into two portions. One part was subjected to PCR isolation of scFvFc fragments. Amplified scFvFc fragments were cloned into the TOPO TA cloning vector (Invitrogen) followed by DNA sequencing. Rest of the cells were further propagated for 1–2 weeks to achieve adequate numbers for a repeated staining of cell-surface scFvFc by biotinilated GD03-Fc and APC-Streptavidin.Inhibition of SARS-CoV spike protein pseudotyped lentiviral infection by cell-surface displayed 80R scFvFcSARS-CoV spike protein (TOR2 strain) pseudotyped lentiviruses, expressing a luciferase reporter gene, were incubated at 4°C for 30 minutes with increasing concentrations of 293T cells expressing on their surface the anti-TOR2 spike 80R-scFvFc or with cells expressing PS11-scFvFc on their surface as a control. Both 80R- and PS11-scFvFc surface-expressing cells were also incubated with VSV-G pseudotyped viral particles for the analysis of non-specific viral absorption. Following absorption, viral supernatant was used to infect 293T cells expressing the ACE2 receptor as described [23]. Neutralization of infection was determined by measuring the luciferase activity in the target cells following transduction, using EG&G Berthold Microplate Luminometer.\n\nREFERENCES:\n1. ReichertJMRosensweigCJFadenLBDewitzMC\n2005\nMonoclonal antibody successes in the clinic.\nNat Biotechnol\n23\n1073\n1078\n16151394\n2. AdamsGPWeinerLM\n2005\nMonoclonal antibody therapy of cancer.\nNat Biotechnol\n23\n1147\n1157\n16151408\n3. WeinerLMCarterP\n2005\nTunable antibodies.\nNat Biotechnol\n23\n556\n557\n15877072\n4. SouriauCHudsonPJ\n2003\nRecombinant antibodies for cancer diagnosis and therapy.\nExpert Opin Biol Ther\n3\n305\n318\n12662144\n5. GuraT\n2002\nTherapeutic antibodies: magic bullets hit the target.\nNature\n417\n584\n586\n12050630\n6. ReichertJPavoluA\n2004\nMonoclonal antibodies market.\nNat Rev Drug Discov\n3\n383\n384\n15152631\n7. WaldmannTA\n2003\nImmunotherapy: past, present and future.\nNat Med\n9\n269\n277\n12612576\n8. SmithGP\n1985\nFilamentous fusion phage: novel expression vectors that display cloned antigens on the virion surface.\nScience\n228\n1315\n1317\n4001944\n9. DaughertyPSIversonBLGeorgiouG\n2000\nFlow cytometric screening of cell-based libraries.\nJ Immunol Methods\n243\n211\n227\n10986416\n10. VanAntwerpJJWittrupKD\n2000\nFine affinity discrimination by yeast surface display and flow cytometry.\nBiotechnol Prog\n16\n31\n37\n10662486\n11. CumbersSJWilliamsGTDaviesSLGrenfellRLTakedaS\n2002\nGeneration and iterative affinity maturation of antibodies in vitro using hypermutating B-cell lines.\nNat Biotechnol\n20\n1129\n1134\n12379869\n12. SeoHMasuokaMMurofushiHTakedaSShibataT\n2005\nRapid generation of specific antibodies by enhanced homologous recombination.\nNat Biotechnol\n23\n731\n735\n15924134\n13. UrbanJHSchneiderRMCompteMFingerCCichutekK\n2005\nSelection of functional human antibodies from retroviral display libraries.\nNucleic Acids Res\n33\ne35\n15731328\n14. HoMNagataSPastanI\n2006\nIsolation of anti-CD22 Fv with high affinity by Fv display on human cells.\nProc Natl Acad Sci U S A\n103\n9637\n9642\n16763048\n15. ChoeHLiWWrightPLVasilievaNVenturiM\n2003\nTyrosine sulfation of human antibodies contributes to recognition of the CCR5 binding region of HIV-1 gp120.\nCell\n114\n161\n170\n12887918\n16. HuangCCVenturiMMajeedSMooreMJPhogatS\n2004\nStructural basis of tyrosine sulfation and VH-gene usage in antibodies that recognize the HIV type 1 coreceptor-binding site on gp120.\nProc Natl Acad Sci U S A\n101\n2706\n2711\n14981267\n17. XuCSuiJTaoHZhuQMarascoWA\n2007\nHuman anti-CXCR4 antibodies undergo VH replacement, exhibit functional V-region sulfation, and define CXCR4 antigenic heterogeneity.\nJ Immunol\n15;179(4)\n2408\n18\n18. BaiJSuiJZhuRYTallaricoASGennariF\n2003\nInhibition of Tat-mediated transactivation and HIV-1 replication by human anti-hCyclinT1 intrabodies.\nJ Biol Chem\n278\n1433\n1442\n12401780\n19. SuiJLiWMurakamiATaminAMatthewsLJ\n2004\nPotent neutralization of severe acute respiratory syndrome (SARS) coronavirus by a human mAb to S1 protein that blocks receptor association.\nProc Natl Acad Sci U S A\n101\n2536\n2541\n14983044\n20. YuanQASimmonsHHRobinsonMKRussevaMMarascoWA\n2006\nDevelopment of engineered antibodies specific for the Mullerian inhibiting substance type II receptor: a promising candidate for targeted therapy of ovarian cancer.\nMol Cancer Ther\n5 (8)\n2096\n2105\n16928831\n21. MooreMJDorfmanTLiWWongSKLiY\n2004\nRetroviruses pseudotyped with the severe acute respiratory syndrome coronavirus spike protein efficiently infect cells expressing angiotensin-converting enzyme 2.\nJ Virol\n78 (19)\n10628\n10635\n15367630\n22. MintzKPFisherLWGrzesikWJHascallVCMiduraRJ\n1994\nChlorate-induced inhibition of tyrosine sulfation on bone sialoprotein synthesized by a rat osteoblast-like cell line (UMR 106-01 BSP).\nJ Biol Chem\n269\n4845\n4852\n8106456\n23. SuiJLiWRobertsAMatthewsLJMurakamiA\n2005\nEvaluation of human monoclonal antibody 80R for immunoprophylaxis of severe acute respiratory syndrome by an animal study, epitope mapping, and analysis of spike variants.\nJ Virol\n79\n5900\n5906\n15857975\n24. HoogenboomHR\n2005\nSelecting and screening recombinant antibody libraries.\nNat Biotechnol\n23\n1105\n1116\n16151404\n25. FeldhausMJSiegelRWOpreskoLKColemanJRFeldhausJM\n2003\nFlow-cytometric isolation of human antibodies from a nonimmune Saccharomyces cerevisiae surface display library.\nNat Biotechnol\n21 (2)\n163\n170\n12536217\n26. OguetaSBYaoFMarascoWA\n2001\nDesign and in vitro characterization of a single regulatory module for efficient control of gene expression in both plasmid DNA and a self-inactivating lentiviral vector.\nMol Med\n7\n569\n579\n11591893\n27. ZhuQRicardoRRZhangLOguetaSBAgrawalRS\n2004\nDevelopment of constitutive and inducible self-inactivating lentiviral vectors and their application in cardiovascular gene transfer.\nGene Therapy and Molecular Biology\n8\n91\n102\n28. MooreKL\n2003\nThe biology and enzymology of protein tyrosine O-sulfation.\nJ Biol Chem\n278\n24243\n24246\n12730193\n29. LinAHKasaharaNWuWStripeckeREmpigCL\n2001\nReceptor-specific targeting mediated by the coexpression of a targeted murine leukemia virus envelope protein and a binding-defective influenza hemagglutinin protein.\nHum Gene Ther\n12\n323\n332\n11242525\n30. YangLBaileyLBaltimoreDWangP\n2006\nTargeting lentiviral vectors to specific cell types in vivo.\nProc Natl Acad Sci U S A\n103\n11479\n11484\n16864770\n31. IkedaYTakeuchiYMartinFCossetFLMitrophanousK\n2003\nContinuous high-titer HIV-1 vector production.\nNat Biotechnol\n21\n569\n572\n12679787\n32. Sena-EstevesMTebbetsJCSteffensSCrombleholmeTFlakeAW\n2003\nOptimized large-scale production of high titer lentivirus vector pseudotypes.\nJ Virol Methods\n122\n131\n139\n33. KlagesNZuffereyRTronoD\n2000\nA stable system for the high-titer production of multiply attenuated lentiviral vectors.\nMol Ther\n2\n170\n176\n10947945\n34. BrowleyDRLabrijinAFZwickMBBurtonDR\n2007\nAntigen selection from an HIV-1 immune antibody library displayed on yeast yields many novel antibodies compared to selection from the same library displayed on phage.\nProtein Eng Des Sel\n20\n81\n90\n17242026"
+}

batch_0/PMC2527559.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527559",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527559\nAUTHORS: Guillaume Bompard, Gabriel Rabeharivelo, Nathalie Morin\n\nABSTRACT:\nBackgroundCytokinesis is the final step of cell division taking place at the end of mitosis during which the cytoplasmic content and replicated chromosomes of a cell are equally partitioned between the two daughter cells. This process is achieved by the formation and the ingression of an actomyosin contractile ring under the control of equatorial microtubules. The mechanisms of contractile ring formation are not fully understood but involve recruitment of preexisting actin filaments and de novo actin polymerisation.ResultsIn this study, we evaluated the role of the actin nucleation factor, Arp2/3 complex, during cytokinesis. We found that the Arp2/3 complex is recruited late to the cleavage furrow suggesting a potential involvement of Arp2/3 complex during this process. Furthermore, wiskostatin a potent inhibitor of N-WASP activity towards the Arp2/3 complex blocked cytokinesis without affecting mitosis. Nonetheless, this inhibition could not be reproduced using alternative approaches targeting the N-WASP/Arp2/3 complex pathway.ConclusionWe conclude that the wiskostatin induced defective cytokinesis does not occur through the inhibition of the N-WASP/Arp2/3 pathway. Wiskostatin is likely to either directly target other proteins required for cytokinesis progression or alternately wiskostatin bound to N-WASP could affect the activity of other factors involved in cytokinesis.\n\nBODY:\nBackgroundCytokinesis is essential to cell growth and has to be tightly regulated to insure equal distribution of the cytoplasm content and newly replicated chromosomes within two daughter cells at the end of mitosis. Defective chromosome segregation or daughter cell separation can alter chromosome number, which is commonly found in tumour cells and potentially involved in cancer development [1]. For these reasons it is essential to understand the signalling pathways that coordinate mitotic exit and cytokinesis.In metazoans, the central spindle forms during anaphase between the two sets of chromosomes. Simultaneously, the cleavage furrow composed by actin and myosin II filaments is established at the cell equator under the control of peripheral microtubules (central spindle and/or peripheral astral microtubules) [2]. Thus, constriction of the actomyosin ring can only start once the two sets of chromosomes are segregated, insuring genome integrity. The two daughter cells stay connected by a cytoplasmic bridge with the midbody in the centre for several hours until the bridge is finally severed during abscission [3].The small GTPase RhoA has been shown to play a major role in cleavage furrow establishment. Indeed, the GTPase exchange factor (GEF) ECT2 is specifically recruited at the cell equator under the control of the centralspindlin complex (MKLP1/MgcRacGAP) also required for central spindle formation [3-7]. Once activated, RhoA induces, through its effector ROCK, the phosphorylation of myosin regulatory light chain (MRLC), which is necessary for contractile ring formation and ingression [3].In mammals, actin filaments present in the contractile ring derive from preexisting F-actin flux and de novo actin polymerisation [8]. In S. pombe, cleavage furrow establishment and activity depend on de novo actin polymerisation driven by the two main actin nucleation factors: the Arp2/3 complex and a formin (Cdc12) [9,10]. Arp2/3 complex is activated by Myo1 and WSP1 at the equator and is involved in the maturation and/or the maintenance of the cleavage furrow [11,12]. In mammals the mDia formins, orthologs of Cdc12 and RhoA effectors, are required for cytokinesis [13,14]. In agreement with a conserved mechanism for cleavage furrow formation and activity, the Arp2/3 complex and its activator WASP/N-WASP, ortholog of WSP1, were found to potentially play a role during cytokinesis [15-19].In this study the functional role of the Arp2/3 complex during cytokinesis in mammalian cells was evaluated. We demonstrated that Arp2/3 complex is recruited late to the cleavage furrow. Consequences of the inhibition of WASP/N-WASP activity towards the Arp2/3 complex during cytokinesis were then studied using the newly identified inhibitor: wiskostatin [20]. We show that wiskostatin strongly inhibits cytokinesis inducing cell binucleation without affecting the cell cycle. However, these results are unlikely related to the Arp2/3 complex function since the wiskostatin induced phenotype could not be validated using N-WASP and/or Arp2/3 RNA interference approach. This novel effect of wiskostatin on cytokinesis is also not related to another known unexpected effect of this compound i.e. ATP depletion [21].This study that was originally designed to study N-WASP/Arp2/3 complex pathway during cytokinesis, unexpectedly highlighted a novel effect of wiskostatin. For this reason, wiskostatin will have to be used with caution in future experiments.ResultsLocalisation of Arp2/3 complex during mitosisIn mammalian cells, the localisation of endogenous Arp2/3 complex during mitosis is currently unknown. In order to study the function of the N-WASP/Arp2/3 complex pathway during cytokinesis, we thus analyzed the subcellular localisation of Arp3 and p34 (ARPC2) subunits of the Arp2/3 complex by indirect immunofluorescence. In interphase HeLa cells using monoclonal anti-Arp3 antibody, the Arp2/3 complex was found to be diffuse and associated with vesicle-like structures in the cytoplasm but also accumulated in cortical F-actin-rich structures, which are likely lamellipodia (data not shown). To better visualize cytoskeleton bound Arp2/3 complex in mitotic cells, the cytosolic unbound fraction was eliminated by permeabilising the cells prior to fixation (Fig. 1). Under these conditions, Arp3 presented a diffuse vesicular-like localisation and was not particularly enriched into the nascent cleavage furrow during anaphase (Fig. 1, a, arrowhead). However, during late telophase, once the cleavage furrow is fully contracted, Arp3 was strongly enriched in the close vicinity of the contractile ring, as visualized by the F-actin staining (Fig. 1, h, arrowhead). Furthermore, the vesicular-like localisation of Arp3 was organised around the minus-end and along microtubules (Fig. 1, e, arrowhead). In cytokinesis, when the cytoplasmic bridge between the two daughter cells extended, Arp3 colocalised with F-actin in the cortex facing the cytoplasmic bridge (Fig. 1, l, arrowhead) and within vesicle-like structures (Fig. 1, i and 1j, arrows). Furthermore, the Arp2/3 complex also localised within the cytoplasmic bridge but not in the midbody (see Additional file 1). This subcellular localisation of Arp2/3 complex was further confirmed by using a polyclonal antibody directed against p34 subunit (see Additional file 1). Thus, the localization of Arp2/3 complex is in agreement with a potential functional role during cytokinesis.Figure 1Arp3 localisation during mitosis and cytokinesis. HeLa cells were enriched in mitosis after thymidine and RO3306 blocks (see Methods for details). Two hours after release from RO3306 block cells were permeabilised and fixed. Cells were treated for indirect Alexa 555 localisation of Arp3 (a, e, and i) and Alexa 350 localisation of β-tubulin (c, g and k) with specific antibodies. F-actin was visualised with FITC-coupled phalloidin (b, f, and j). Merged images are presented (d, h and l). Images are representative of the different stages of mitosis: anaphase (a-d), late telophase (e-h) and cytokinesis (i-l). Bars, 10 μm.Wiskostatin inhibits cytokinesis but not mitosisThe mammalian ortholog of WSP1, WASP/N-WASP, is regulated by autoinhibition essentially relieved by binding of activated small GTPase Cdc42 and/or phosphatidyl inositol 4,5-biphosphate (PtdIns[4,5]P2) [10,22]. GTP loaded Cdc42 binds to the GBD (GTPase binding domain) of WASP/N-WASP, which is preceded by a basic region where PtdIns[4,5]P2 interacts [22]. Wiskostatin is a recently identified WASP/N-WASP chemical inhibitor designed to bind to the GBD [20]. Wiskostatin not only prevents Cdc42 binding, but also stabilises the autoinhibited state of this actin nucleation promoting factor [20]. This led us to use wiskostatin to investigate the role of the Arp2/3 complex activated by WASP/N-WASP during cytokinesis.HeLa cells stably expressing GFP-tagged histone H2B (HeLa GFP-H2B) were synchronised in prometaphase with nocodazole and were released for various times after shake-off in presence of vehicle (DMSO) or increasing concentration of wiskostatin (from 1 – 10 μM). Cell DNA content was determined by flow cytometry at 0, 120 and 300 minutes after nocodazole release. In control conditions, at T = 0, cells had a 4N content confirming mitosis synchronisation mediated by nocodazole. 120 minutes later half of the cells had exit mitosis and displayed a 2N content. Finally, 300 min after release, control cells were mostly in G1 phase (diploid cells, 2N) (Fig. 2A). In contrast, under wiskostatin treatment, the number of tetraploid cells (4N) was abnormally elevated at 120 min after release suggesting a mitotic delay both at 5 and 10 μM doses (Fig. 2A). 300 minutes after nocodazole release under 5 μM wiskostatin, half of the cells had a 2N content while under 10 μM wiskostatin treatment all cells were tetraploid indicating that they did not proceed through mitosis (Fig. 2A). Thus wiskostatin inhibits mitotic progression in a dose dependent manner.Figure 2Wiskostatin inhibits cytokinesis but not mitosis. A. HeLa cells were synchronised in prometaphase with nocodazole and treated with vehicle or indicated concentrations of wiskostatin. Cells were harvested at indicated time points and DNA content determined by FACS analysis. A similar scale was used for each histogram. This result is representative of at least 3 independent experiments. B. HeLa cells synchronised as previously described were treated with vehicle (-) or 10 μM wiskostatin (+) and harvested at indicated time points. Cyclin B1, phosphorylation of Ser 10 of histone H3 and cofilin levels were determined by immunoblot analysis with specific antibodies. This result is representative of at least 5 independent experiments. C. HeLa cells stably expressing GFP-tagged histone H2B were synchronised as previously described, treated with vehicle (Control) or 10 μM wiskostatin (Wiskostatin 10 μM) and fixed after 300 minutes. F-actin was stained and nuclei per cell determined. Percentage of binucleated cells was then calculated. 536 cells treated with vehicle and 569 cells treated with wiskostatin were count from three independent experiments. Error bars represent standard error of the mean (SEM). D. Assynchronous HeLa cells stably expressing GFP-tagged histone H2B were treated with vehicle (Control) or 5 μM wiskostastin (Wiskostatin 5 μM) for 24 hours and were then fixed. Percentages of binucleated cells were determined as previously indicated. 629 cells treated with vehicle and 640 cells treated with wiskostatin were count from three independent experiments. Error bars represent SEM. Representative images from time lapse movies of HeLa cells stably expressing GFP-tagged histone H2B synchronised as previously described, treated with vehicle (E) or 10 μM wiskostatin (F). Images represent merged between DIC and fluorescent (Histone H2B) images. Time indicated as hours: minutes.The increased number of tetraploid cells under wiskostatin could either result from an inhibition of mitosis or cytokinesis. To answer this question, we first analysed the status of cyclin B1, which is degraded at the metaphase/anaphase transition. Mitotic HeLa cells were released in presence of vehicle alone or wiskostatin (10 μM), and cyclin B1 levels were determined by immunoblot analysis (Fig. 2B). No difference in the kinetics of cyclin B1 degradation was observed between control and wiskostatin treated cells. This result was further confirmed by studying another mitotic marker, the phosphorylation of histone H3 on serine 10 (Ser 10) which correlates with chromatin condensation [23]. As expected, histone H3 phosphorylation was not affected by wiskostatin treatment (Fig. 2B). Equal sample loading was confirmed by cofilin immunoblot analysis (Fig. 2B). These results clearly demonstrate that wiskostatin does not inhibit mitotic progression.Taken altogether, our results indicate an effect of wiskostatin on cytokinesis. Cytokinesis failure is generally associated with cell binucleation. To investigate this hypothesis, nocodazole synchronised HeLa GFP H2B cells were grown in presence of vehicle alone or wiskostatin (10 μM), fixed after 300 minutes and binucleated cells were quantified. In control condition, around 7% of cells were found to be binucleated (7.26% SEM ± 2.11, n = 536). In contrast, 85% of wiskostatin treated cells contained two nuclei (84.79% SEM ± 3.44, n = 569) (Fig. 2C). To rule out any cooperative effects of nocodazole and wiskostatin, asynchronous HeLa GFP-H2B cells were grown in presence of 5 μM wiskostatin for 24 hours and binucleated cells were counted as previously described. Under these conditions more than 26% of cells were found to be binucleated (26.24%, SEM ± 0.92, n = 640) compared to 0.3% (0.3%, SEM ± 0.3, n = 629) of cells treated with vehicle alone (Fig. 2D). Similar results were obtained with NIH3T3 and XL2 cells (data not shown). This result clearly demonstrates that wiskostatin affects mitotic exit.In order to analyse wiskostatin induced cytokinesis defects in more details, time-lapse microscopy was used. In vehicle treated HeLa GFP-H2B cells cytokinesis started around 45 minutes after release from nocodazole block (Fig. 2E, see Additional file 2). After complete cleavage furrow contraction, daughter cells were linked to each other by a cytoplasmic bridge, which extended over time with the midbody in its centre (Fig. 2E, arrowheads). Finally, after several hours, abscission occurred. Under wiskostatin treatment, cytokinesis started with the same kinetics as control cells (Fig. 2F, see Additional file 3) but prior and during cleavage furrow contraction blebs were visible at cell poles (Fig. 2F, arrowheads). Finally, once the cleavage furrow fully contracted, the cytoplasmic bridge did not extend, the contractile ring relaxed and daughter cells finally fused-back together (Fig. 2F, arrows).We then studied the organisation of the cytoskeleton in more details. Mitotic HeLa cells stably expressing GFP-tagged β-Tubulin were fixed two hours after nocodazole release in presence of vehicle or 10 μM wiskostatin, and were stained for F-actin and DNA. During cytokinesis, the microtubule bundles of the central spindle, which are not associated with kinetochores, are compacted by the contraction of the cleavage furrow. In control cells, these compacted bundles were clearly visible between the two daughter cells (Fig. 3A, b, arrowhead). Similarly in wiskostatin treated cells, compacted microtubule bundles were also present between newly assembled nuclei but cells were binucleated (Fig. 3A, f, arrowhead). Cleavage furrow relaxation induced by wiskostatin was not associated with obvious F-actin reorganisation and some cleavage furrow remnant could be observed surrounding the central spindle as evidenced by F-actin staining (Fig. 3A, e, arrowhead). Relocalisation of ECT2 from the central spindle in anaphase to the midbody during cytokinesis was also not altered by wiskostatin treatment (data not shown) confirming that wiskostatin did not affect cleavage furrow formation and initial contraction.Figure 3Wiskostatin inhibits late cytokinesis. A. Wiskostatin does not affect microtubule organisation. HeLa cells stably expressing GFP-tagged β-tubulin were synchronised in prometaphase with nocodazole and released after a shake-off in presence of vehicle (Cont) or 10 μM wiskostatin (Wisko). 120 minutes after release cells were fixed and F-actin (a and e) and DNA (c and g) stained respectively with Alexa 555-coupled phalloidin and DAPI. β-tubulin (b and f) was directly visualised. Merged images (d and h) are presented. B. Wiskostatin alters Arp3 localisation. HeLa cells were synchronised in G2/M by thymidine and RO3306 double block. 120 minutes after release cells were fixed and stained as described in Figure 1. Images a and e represent staining of Arp3, b and f staining of F-actin and c and g staining of β-tubulin. Merged images (d and h) are presented. Bars, 10 μm.We next studied the localisation of Arp3 during cytokinesis under wiskostatin treatment. As previously showed, Arp3 was enriched within the cleavage furrow once fully contracted in vehicle treated cells at the end of telophase (Fig. 3B, a, arrowhead). Under wiskostatin treatment, Arp3 localisation within the contractile ring was reduced (Fig. 3B, e, arrowhead). In addition, wiskostatin induced the clustering of Arp3 within vesicle-like structures throughout the cell (Fig. 3B, e, arrow).Altogether, our results showed that wiskostatin inhibited late cell cycle events required for abscission to occur. Wiskostatin induced cleavage furrow regression was associated with a reduced localisation of Arp2/3 complex within the contractile ring.Wiskostatin effects onto cytokinesis are not reproduced by alternative inhibition of the N-WASP/Arp2/3 pathwayIn order to study whether the mislocalisation of Arp3, induced by wiskostatin, was responsible for defective cytokinesis, we used alternative approaches to inhibit N-WASP and Arp2/3 complex activities. We first knocked-down N-WASP and Arp3 protein levels using an siRNA approach. N-WASP and Arp3 expressions were strongly reduced in cells treated for 72 hours with specific siRNA (more than 80% inhibition) as demonstrated by immunoblot analysis (Fig. 4A). As previously shown, Arp3 knockdown is associated with down-regulation of other Arp2/3 subunits [24] as revealed by the strong inhibition of p34 expression in Arp3 null cells (Fig. 4A). The DNA content of siRNA treated cells was then evaluated by flow cytometry in order to study the consequences of N-WASP and/or Arp3 depletion on cytokinesis. No major differences between control cells and cells transfected with N-WASP and/or Arp3 siRNAs were observed (Fig. 4B). In contrast, tetraploid and also octaploid cells accumulated when transfected with siRNA inhibiting ECT2 protein expression (Fig. 4A, more than 90% inhibition), a major effector of cytokinesis (Fig. 4B, arrows). The phenotypes resulting from siRNA transfection were also evaluated by indirect immunofluorescence analysis. Cellular morphology, F-actin organisation and Arp3 staining were not affected in control and N-WASP knockdown cells (see Additional file 4). In contrast, Arp3 staining was greatly reduced in cells transfected with Arp3 siRNA for 72 hours. The morphology of these cells was altered as they appeared rounder and smaller, a phenotype likely due to the disorganization of the F-actin network (see Additional file 4). During cytokinesis, Arp3 recruitment to the contractile ring was only reduced in Arp3 null cells but not in control and N-WASP null cells (see Additional file 4). We obtained similar results using other siRNAs targeting N-WASP and Arp3 sequences (data not shown). Our data suggest that Arp2/3 complex localisation within the cleavage furrow does not rely on N-WASP and is not mandatory for cytokinesis to occur.Figure 4Arp3 and/or N-WASP knockdowns do not affect cytokinesis. A. HeLa cells were transfected with water (Ø) or specific siRNA directed against N-WASP or Arp3 or N-WASP + Arp3 or ECT2. 72 hours after transfection cells were harvested and lysed. N-WASP, Arp3, ECT2, p34 and β-tubulin protein levels were determined by immunoblot analysis using specific antibodies. B. The DNA content of the same HeLa cells was determined by flow cytometry analysis. Arrows indicate abnormal DNA content in ECT2 knockdown cells. A similar scale was used for each histogram. These results are representative of at least three independent experiments. C. Representative DIC images from time-lapse movies of HeLa cells transfected with water (Ø), siRNA directed against N-WASP (N-WASP), against Arp3 (Arp3), against Arp3 and N-WASP (Arp3 + N-WASP) or against ECT2. Images started to be acquired 24 hours post-transfection for 40 hours. Coloured stars indicate the nuclei of dividing cells and of their daughter cells. Time indicated as hours: minutes.In order to investigate possible subtle consequences of N-WASP and/or Arp3 losses on mitosis/cytokinesis, siRNA treated cells were studied by time-lapse microscopy. The time-lapse experiment began 24 hours after siRNA transfection when protein expression was significantly reduced (see Additional file 4) and cells were imaged every 30 minutes for 30 hours. No differences in terms of frequency and kinetics of mitosis and cytokinesis were observed between control cells (Fig. 4C and see Additional file 5), N-WASP null-cells (Fig. 4C and see Additional file 6), Arp3 null-cells (Fig. 4C and see Additional file 7), N-WASP and Arp3 null-cells (Fig. 4C and see Additional file 8). In contrast, ECT2 knockdown strongly interfered with cell division inducing binucleation and tetranucleation over time (Fig. 4C and see Additional file 9). Surprisingly, these results suggest that the loss of N-WASP and/or Arp2/3 complex, are not associated with cytokinesis failure as we were expecting based on the effects of wiskostatin. In order to confirm these unexpected results, various N-WASP mutants known to interfere with Arp2/3 complex activity were overexpressed and their effects on cytokinesis evaluated. WASP family proteins are characterized by the presence, at the carboxy-terminus, of three independent domains: the verprolin homology (V) (a.k.a. WASP homology 2/WH2 domain), central (C) and acidic (A) regions. These domains form the so-called VCA module, which is necessary and sufficient to activate Arp2/3-dependent actin polymerisation in vitro. The V and A regions bind actin monomers and Arp2/3 respectively, whilst the C region binds Arp2/3 and induces crucial changes in the tertiary and quaternary structures of the Arp2/3 complex, thereby regulating the ability of the Arp2/3 complex to induce actin polymerisation [10]. Deletion of the A region within the sequence of WASP family proteins compromises binding to the Arp2/3 complex. This mutation acts as a dominant negative protein. On the other hand, overexpression of VCA and CA modules from WASP family proteins strongly induces mislocalised Arp2/3 complex activation and unproductive Arp2/3 binding respectively [25,26]. These constructs titrate endogenous Arp2/3 complex and inhibit F-actin processes depending onto Arp2/3 activity [10]. Vector expressing GFP, GFP-tagged N-WASP, N-WASP ΔA (deletion of the acidic domain), VCA or CA modules from N-WASP were transfected in NIH3T3 fibroblasts. Three days after transfection cells were fixed and binucleated transfected cells counted. Results from a representative experiment are presented in Table 1. As shown, no major differences were observed between GFP, GFP-tagged N-WASP, N-WASP ΔA and CA overexpressing cells, although, overexpression of GFP-tagged VCA slightly, but consistently, increased cell binucleation (Table 1).Table 1Effect of various N-WASP constructs onto cell ploidy.VectorTotal cell numberNumber of binucleated cells% of binucleated cellsGFP419174.06GFP VCA5276111.57GFP CA629304.77GFP N-WASP351195.41GFP N-WASP ΔA13073.38NIH3T3 cells were transfected with indicated constructs. Three days after transfection cells were harvested and a fifth seeded onto glass coverslips. Two hours after seeding cells were fixed. Nuclei number of transfected cells was determined after F-actin and DNA staining respectively with TRITC-coupled phalloidin and DAPI. Total cell number and the number of binucleated cells from a representative experiment are presented as well as the deduced percentage of binucleated cells.Altogether, our data demonstrated that the inhibition of N-WASP/Arp2/3 pathway is not sufficient to mimic wiskostatin effects on cytokinesis. Therefore, it is likely that wiskostatin affect the activity of additional targets involved in cytokinesis.Wiskostatin effect onto cytokinesis is independent of ATP depletionTaken altogether, our results suggest that wiskostatin effects on cytokinesis progression cannot solely be explained by the inhibition of N-WASP. During the course of our study, wiskostatin was shown to induce cellular ATP depletion [21]. We thus evaluated whether ATP depletion may interfere with cytokinesis progression. Mitotic HeLa cells released in presence of vehicle or wiskostatin (10 μM) from a nocodazole block were harvested at various time points and lysed for 10 minutes at 100°C in distilled water. ATP levels were then determined from clarified lysates using a bioluminescent luciferase assay. In vehicle (at 300 min, 91.2%, SEM ± 5.3, n = 3) or wiskostatin (at 300 min, 82%, SEM ± 1.5, n = 3) treated cells no major variation of ATP levels was observed during mitosis/cytokinesis (Fig. 5A). In contrast, ATP levels strongly decreased in cells released in presence of energy poisons such as sodium azide and 2-deoxy-D-glucose (at 300 min, 41.1%, SEM ± 3.8, n = 3, Fig. 5A).Figure 5Wiskostatin does not affect cellular ATP levels. A. Cellular ATP levels in wiskostatin treated cells. HeLa cells were synchronised in prometaphase with nocodazole and released after shake-off in presence of vehicle (control), 10 mM NaN3 + 6 mM deoxyglucose (ΔATP) or 10 μM wiskostatin (wiskostatin). After indicated time points cells were harvested and lysed in boiling water for 10 minutes. Cellular ATP contents were determined on clarified lysates using ATP determination kit (see Methods). Cellular ATP level at time point 0 from each treatment was defined as 100%. Graph represents mean of three independent experiments. Error bars represent SEM. ATP depletion delays mitosis. B. HeLa cells synchronised and treated as previously described were harvested at indicated time points and DNA content analysed by flow cytometry. C. In similarly treated cells cyclin B1 and γ-tubulin levels were determined by immunoblot analysis using specific antibodies.In order to confirm that inhibition of cytokinesis induced by wiskostatin was independent of ATP depletion, DNA content of HeLa cells synchronised and treated as mentioned was determined by flow cytometry. Whereas vehicle or ATP depleted treated cells were mainly diploids after 300 minutes from nocodazole release, wiskostatin treated cells were massively tetraploids and likely binucleated (Fig. 5B). Interestingly, we observed a delay in exit to mitosis in ATP depleted cells (Fig. 5B, see 90 min time point), which was correlated with a delay of cyclin B1 degradation as demonstrated by immunoblot analysis (Fig. 5C). As previously demonstrated, there was no effect of wiskostatin on cyclin B1 degradation. Time-lapse microscopy confirmed that ATP depletion delayed mitotic exit whereas wiskostatin did not (data not shown).Our results showed that ATP depletion does not block mitosis/cytokinesis but simply delays it. Furthermore, we established that inhibition of cytokinesis by wiskostatin is independent of ATP depletion suggesting additional targets for wiskostatin.DiscussionPartitioning of the two daughter cells following cytokinesis is crucial to allow faithful segregation of chromosomes. Disrupting the final stages of cytokinesis can lead to multinucleated cells and genome instability. The involvement of the actin cytoskeleton during cytokinesis is now well established. The cleavage furrow is formed during anaphase by actin and myosin filaments and its complete ingression allows daughter cell separation at the end of mitosis. However, little data regarding actin dynamics during cytokinesis is available. A recent study showed that actin filaments present within the cleavage furrow of mammalian cells come from the flux of preexisting F-actin and de novo in situ actin polymerisation [8]. However the functional importance of de novo actin polymerisation or the identity of the nucleation factor involved remain unanswered questions.In S. pombe, the two main actin nucleation factors, the Arp2/3 complex and a formin, are required for cleavage furrow establishment and activity [9,11]. In mammals, the involvement of formins during cytokinesis is now well established suggesting a conserved mechanism between yeast and mammals for cleavage furrow formation [3,13,14,27]. However, in metazoan the involvement of WASP/N-WASP and the Arp2/3 complex during cytokinesis is still a matter of debate. Very little information regarding the function of WASP/N-WASP during cell growth in general and cytokinesis in particular are available. In nematodes, knockdown of WSP-1 protein, the WASP/N-WASP ortholog, leads to cytokinesis defects [18]. N-WASP knockout in mouse is embryonic lethal (E12) and primary fibroblasts isolated from knockout embryos are unable to grow without being transformed suggesting an involvement of N-WASP in cell growth [28,29]. Data regarding the involvement of the Arp2/3 complex during cytokinesis in metazoans are more controversial. In nematode, the requirement of the Arp2/3 complex is unclear since two studies using RNA interference lead to conflicting results. Two groups described that knocking-down most of the Arp2/3 complex subunits lead to embryonic lethality, however only the group of Skop et al. showed that these depletions were associated with early or late cytokinesis defects [16,30]. In Drosophila, the picture is more readable since Arp3 knockdown in S2 cells led to cell binucleation [15]. Furthermore, the Arp2/3 complex is required for the assembly of two Drosophila embryonic structures highly related to the cleavage furrow, the pseudocleavage furrow in syncytial embryos and the ring canal [31,32]. Finally, in mammals, p21 (ARPC3) subunit of the Arp2/3 complex is an essential gene, as knockdown results in a block in cell growth phenotype [33]. However, a potential relation with a cytokinesis defect was not reported.In summary, the data available in the literature on the functional importance of the Arp2/3 complex for cell growth are clear, however because of the pleïotropic functions of the actin cytoskeleton, specific requirements of Arp2/3 complex activity for efficient cytokinesis are still questionable.In this study, we showed that endogenous Arp3, despite being localised to the cell cortex, did not accumulate in the nascent contractile ring. However, once the cleavage furrow was completely contracted during cytokinesis, Arp2/3 complex colocalised with F-actin within this structure. Arp2/3 complex was also found within the cytoplasmic bridge linking daughter cells at the end of mitosis in agreement with the biochemical association of Arp2/3 subunits with the central spindle [16]. The absence of early specific localisation of Arp3 within the cleavage furrow suggests that the Arp2/3 complex is not involved in its establishment. This is in agreement with the proposed role of the Arp2/3 complex in fission yeast where it is recruited late to/around the contractile ring and seems to be involved in cleavage furrow maintenance by allowing new plasma membrane addition [12]. New membrane addition relies on vesicle recruitment and is essential for cleavage furrow ingression and to seal up daughter cells at time of abscission [34]. The Arp2/3 complex is involved in endocytosis, vesicular trafficking but also in the polarised delivery of certain proteins to the plasma membrane [35,36]. All these processes are known or supposed to be important during cytokinesis [34]. We showed that Arp2/3 complex was enriched in vesicle-like structure localised on microtubules facing the cytoplasmic bridge between daughter cells.We found that inhibition of N-WASP activity towards the Arp2/3 complex using wiskostatin strongly impaired cytokinesis without affecting mitosis in a dose dependent manner. Wiskostatin inhibited final stages of cytokinesis, which correlated with a decrease of the Arp2/3 complex from the contractile ring and the clustering of the Arp2/3 complex within vesicle-like structures. The late inhibition of cytokinesis by wiskostatin suggests that the Arp2/3 complex is required at the end of cytokinesis and/or for abscission in agreement with its late recruitment within the cleavage furrow. However, we were unable to confirm these results by knocking-down N-WASP and/or Arp3 protein expression by RNA interference. Moreover, N-WASP depletion showed that Arp3 localisation within the cleavage furrow is not dependent upon N-WASP. Different possibilities can explain the discrepancy between the results we obtained using RNA interference versus wiskostatin. First, we cannot exclude that remaining N-WASP and/or Arp3 proteins, in siRNA treated cells, are sufficient to trigger cytokinesis bearing in mind that complete depletion of these proteins is associated with cell growth defects and embryonic lethality [29-31,33]. Second, wiskostatin could have additional targets related or not to N-WASP. Wiskostatin binds purified N-WASP in vitro but nothing is known about binding to native N-WASP. Native N-WASP is predominantly associated with WIP (WASP-interacting protein), a protein involved in N-WASP inhibition but also required for N-WASP activation [37-39]. WIP can regulate F-actin organisation independently of its effects on N-WASP [39]. Wiskostatin could prevent the release or the binding of WIP or other unknown proteins. This could have major consequences for the progression of cytokinesis if such proteins play a role during this process. Furthermore, this could explain the discrepancy between wiskostatin and N-WASP knockdown results. Finally, wiskostatin could inhibit additional targets unrelated to N-WASP. Indeed, wiskostatin was recently shown to deplete cellular ATP in an irreversible manner perturbing membrane transport [21]. ATP in general and membrane trafficking in particular, as mentioned above, play crucial roles during mitosis and cytokinesis. However, we showed that wiskostatin, at the highest concentration used in this study (10 μM), did not affect cellular ATP levels during mitosis/cytokinesis. Thus, inhibition of cytokinesis induced by wiskostatin is likely independent to its ATP depleting activity.The existence of additional targets of wiskostatin is supported by the results we obtained using N-WASP mutants. The expression of a dominant negative mutant of N-WASP towards the Arp2/3 complex (N-WASP ΔA) did not block cytokinesis suggesting again that N-WASP inhibition is not sufficient. In contrast, overexpression of N-WASP VCA domain slightly increased nuclei number. The VCA module alters Arp2/3 activity by titrating and delocalising it. VCA overexpression could either block the activation of the Arp2/3 complex by another factor than N-WASP involved in cytokinesis. On the other hand, the increase of F-actin generated by over-activation of Arp2/3 complex triggered by VCA overexpression could affect cytokinesis. This hypothesis is supported by a recent study showing that expression of a constitutively active mutant of WASP towards the Arp2/3 complex increases cell binucleation in a same extend as VCA [19].ConclusionWiskostatin has been designed, as numerous inhibitors, to specifically inhibit N-WASP activity towards the Arp2/3 complex. Our studies reveal the caveat of using such an in vitro approach to identify new compounds. Indeed, in vivo test may not parallel the in vitro activity and should be included in the compound characterisation. We show that wiskostatin inhibits cytokinesis and it is likely that N-WASP is not its sole target during this process. Our study is the second report of a non-expected effect of wiskostatin, which suggests that this compound will have to be used in future with caution. However, it will be of interest to characterise wiskostatin targets involved in cytokinesis. The identification of these proteins could be achieved using a proteomic approach based upon mass spectrometry using immobilised wiskostatin over mitotic cell extracts.MethodsReagents and antibodiesAll chemicals were purchased from Sigma unless otherwise stated. Monoclonal antibodies against Arp3, β-tubulin (TUB 2.1) and γ-tubulin (GTU-88) were from Sigma. Polyclonal antibody against p34 (ARPC2) was from Upstate Biotechnology. Polyclonal antibodies against N-WASP (H-100), cyclin B1 (GNS1) and ECT2 (C-20) were purchased from Santa-Cruz Biotechnology. Polyclonal antibody against cofilin was from Cytoskeleton. Monoclonal antibody against p-histone H3 (Ser 10) was from Euromedex. Polyclonal antibody against β-tubulin was a generous gift from Jose M. Andreu (Centro de Investigaciones Biológicas, Madrid, Spain). Goat Alexa 350, 488, 555, 680-conjugated anti-mouse and anti-rabbit were from Molecular Probes.Expression plasmidsVector encoding EGFP-tagged bovine N-WASP full-length (aa, 1–505) and ΔA (aa, 1–484) mutant as well as N-WASP CA (aa, 455–505) and VCA (aa, 391–505) regions were from Laura M. Machesky (Beatson Institute, Glasgow, UK).Cell culture and synchronisationHeLa and NIH3T3 cells were cultured in Dulbecco's Modified Eagle's medium (DMEM) supplemented with 10% foetal bovine serum (FBS), 20 mM HEPES and antibiotics. For HeLa cells stably expressing GFP-H2B (gift from E. Julien, IGMM, Montpellier, France) media was supplemented with 0.2 mg/ml G418.Cells were synchronised in prometaphase by nocodazole block. Briefly cells were grown for 12 hours in the presence of 100 ng/ml nocodazole. Mitotic cells obtained by shake-off were plated onto dishes or glass coverslips coated with poly-L-lysine in fresh growth medium containing vehicle alone or wiskostatin. Cells were then harvested at indicated time points, treated for fluorescence-activated cell sorting (FACS), lysed for immunoblot analysis or fixed for indirect immunofluorescence study. In some experiments, cells were alternatively synchronised in G2/M using 10 μM Cdk1 inhibitor (RO3306, Calbiochem). Briefly, 8 hours after seeding cells were blocked in G1/S with 2.5 mM thymidine for 24 hours. After several washings, cells were released in complete medium containing 24 μM deoxycytine and 10 μM RO3306 for 15 hours. Finally, G2/M blocked cells were released in complete medium containing or not wiskostatin after several washings and treated for immunofluorescence 2 hours after release.RNA interferencesiRNA were transfected in HeLa cells using Lipofectamine 2000™ reagent (Invitrogen) according to manufacturer instructions. Proteins were targeted with the following sequences: Arp3 with 5'-GCCAAAACCUAUUGAUGUA-3', N-WASP with 5'-GGUGUUGCUUGUCUUGUUA-3' and ECT2 with 5'-CAUUUGAUAUGAAGCGUUA-3'. Cells were harvested 72 hours after transfection for FACS and immunoblot analysis.ImmnunofluorescenceFor immunofluorescence, cells were stained and mounted on glass slides as previously described [40]. In brief, cells were fixed with 4% paraformaldehyde in PEM (0.1 mM PIPES, pH 6.9; 1 mM EGTA; 0.5 mM MgCl2) containing 0.2% Triton X-100 for 10 min at 37°C, blocked with 1% BSA and stained with phalloidin or the appropriate antibodies diluted in PBS containing 1% BSA. When mentioned cells were permeabilised with PEM containing 0.5% Triton X-100 and 1 mM PMSF for 30 seconds prior fixation with 4% paraformaldehyde. Cells were examined with a DMR A microscope PL APO 63× oil immersion objective with appropriate filters (Leica) and images were recorded with a cooled CCD Micromax camera (Princeton Instruments) driven by MetaMorph (Molecular Devices).Time-lapse microscopy was performed using an Axiovert 200 M microscope PlasDIC 32× objective with appropriate filters (Zeiss) and images were recorded with a cooled CCD Micromax camera (Princeton Instruments) driven by MetaMorph (Molecular Devices).Determination of cellular ATP levels1.5 × 105 mitotic HeLa cells/well in 6 well plates were treated for indicated time with vehicle alone, 10 μM wiskostatin or 10 mM NaN3 + 6 mM 2-deoxyglucose (ATP depletion). Cells were harvested and lysed in boiling distilled water for 10 min. Lysates were clarified by centrifugation at 20,000 × g for 10 min at 4°C. Cellular ATP levels were determined from clarified lysates using ATP determination kit (Proteinkinase, Germany) and following manufacturer instructions. Luminescence was measured using a MicroLumat (Berthold Technologies) luminometer.Flow cytometryHarvested cells were washed in PBS, fixed in 70% ethanol and kept at -20°C until analysis. DNA was stained with propidium iodide and analysed with a FACSCalibur (Becton Dickinson) driven by CellQuest™. The DNA content of 10 000 cells was determined for each condition.AbbreviationsWASP: Wiskott-Aldrich syndrome protein; N-WASP: Neural WASP; Arp: actin-related protein; GBD: GTPase binding domain.Authors' contributionsGB conceived the study, designed and carried out the experiments and wrote the manuscript. GR assisted with carrying out the experiments. NM assisted in designing the experiments and in writing the manuscript. All authors read and approved the final manuscript.Supplementary MaterialAdditional file 1p34 (ARPC2) localisation during the exit of mitosis. A. HeLa cells were enriched in mitosis after thymidine and RO3306 blocks (see Methods for details). Two hours after release from RO3306 block cells were permeabilised and fixed as described in Methods. Cells were treated for indirect Alexa 555 localisation of p34 (a and e) and Alexa 350 localisation of β-tubulin (c and g) with specific antibodies. F-actin was visualised with FITC-coupled phalloidin (b and f). Merged images are presented (d and h). Images are representative of the different stages of mitosis: late telophase (a-d) and cytokinesis (e-h). During late telophase p34 is enriched within the contractile ring (d, arrowhead) During cytokinesis, p34 presented a vesicular-like staining at the basis of microtubule bundles (e, arrowhead) and was found within the cytoplasmic bridge between the two daughter cells (e, arrow). Colocalisation with F-actin was also evident at the cortex facing the cytoplasmic bridge (h, arrowhead). B. Arp3 and p34 localisations are undistinguishable HeLa cells prepared as previously described were treated for indirect Alexa 555 localisation of Arp3 (a) and Alexa 488 localisation of p34 (b) with specific antibodies. DNA was stained with DAPI. A merged imaged is presented (d) and highlight the strong colocalisation between the two Arp2/3 complex subunits in particular within the cleavage furrow. Bars, 10 μm.Click here for fileAdditional file 2Mitosis of vehicle treated HeLa GFP-Histone H2B cells. For details see figure 2E legend. Frames were taken every 5 min s and are played at six frames per second. Time indicated as hours: minutes.Click here for fileAdditional file 3Mitosis of wiskostatin (10 μM) treated HeLa GFP-Histone H2B cells. For details see figure 2F legend. Frames were taken every 5 min and are played at six frames per second. Time indicated as hours: minutes.Click here for fileAdditional file 4Study of siRNA efficiency. A. Arp3 knockdown affects cell morphology. HeLa cells were transfected with indicated siRNA for 72 hours, permeabilised and fixed as previously indicated and treated for indirect Alexa 555 localisation of Arp3 (a, e and i) using specific antibody. F-actin was stained with FITC-coupled phalloidin (b, f and j). Merged images are presented (c, g and k). Exposure times for Alexa 555 and FITC channels were determined for control conditions and applied to N-WASP and Arp3 null cells allowing direct comparison of protein levels based on signal intensities. Arp3 localised at the tip of membrane protrusion (a and e, arrowhead) and in vesicle-like structures where it colocalised with F-actin (c and g). Arp3 staining is greatly reduced in Arp3 null-cells (compare i with a and e). Bar, 20 μm. B. Arp3 recruitment to the contractile ring is not affected in N-WASP null-cells. HeLa in cytokinesis transfected with siRNA as previously described were treated for indirect Alexa 555 localisation of Arp3 (a, e and i) and Alexa 350 localisation of β-tubulin (c, g and k) using specific antibodies. F-actin was stained with FITC-coupled phalloidin (b, f and j). Merged images (d, h and l) are presented. Exposure times in Alexa 555 and FITC channels were fixed as previously described. Arp3 recruitment to the contractile ring is identical between control (a, arrowhead) and N-WASP (b, arrowhead) null-cells but is greatly reduced in Arp3 null-cells (i, arrowhead). Bar, 10 μm. C. Time course depletion of N-WASP, Arp3 and ECT2 proteins by RNA interference. HeLa cells were transfected with indicated siRNA and were harvested 24, 48 and 72 hours after transfection. Proteins levels were determined by immunoblot analysis using specific antibodies. The maximum of N-WASP knockdown is achieved after 48 hours transfection, whereas 72 hours are required for Arp3 and only 24 hours for ECT2.Click here for fileAdditional file 5Mitosis of control siRNA treated cells. For details see figure 4C legend. Frames were taken every 30 min are played at four frames per second. Time indicated as day: hours: minutes.Click here for fileAdditional file 6Mitosis of N-WASP siRNA treated cells. For details see figure 4C legend. Frames were taken every 30 min are played at four frames per second. Time indicated as day: hours: minutes.Click here for fileAdditional file 7Mitosis of Arp3 siRNA treated cells. For details see figure 4C legend. Frames were taken every 30 min are played at four frames per second. Time indicated as day: hours: minutes.Click here for fileAdditional file 8Mitosis of N-WASP + Arp3 siRNA treated cells. For details see figure 4C legend. Frames were taken every 30 min are played at four frames per second. Time indicated as day: hours: minutes.Click here for fileAdditional file 9Mitosis of ECT2 siRNA treated cells. For details see figure 4C legend. Frames were taken every 30 min are played at four frames per second. Time indicated as day: hours: minutes.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2527576.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527576",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527576\nAUTHORS: Christopher E Benejam, Steven G Potaczek\n\nABSTRACT:\nIntroductionLisfranc fracture dislocations of the foot are rare injuries. A recent literature search revealed no reported cases of injury to the tarsometatarsal (Lisfranc) joint associated with sledding.Case presentationA 19-year-old male college student presented to the emergency department with a Lisfranc fracture dislocation of the foot as a result of a high-velocity sledding injury. The patient underwent an immediate open reduction and internal fixation.ConclusionLisfranc injuries are often caused by high-velocity, high-energy traumas. Careful examination and thorough testing are required to identify the injury properly. Computed tomography imaging is often recommended to aid in diagnosis. Treatment of severe cases may require immediate open reduction and internal fixation, especially if the risk of compartment syndrome is present, followed by a period of immobilization. Complete recovery may take up to 1 year.\n\nBODY:\nIntroductionAn unusual case of Lisfranc fracture dislocation of the foot resulting from a high-velocity sledding injury is discussed. A recent literature search revealed no reported cases of injury to the tarsometatarsal (Lisfranc) joint associated with sledding.Case presentationA healthy 19-year-old male college student presented to the emergency department with acute pain in the left foot after sustaining a sledding injury. While sledding in the sitting position and with legs extended, the plantar aspect of his left foot struck a tree limb at high speed. The pain was throbbing and did not radiate. Weight bearing was impossible. Previous medical and surgical records were unremarkable.On physical examination, localized swelling and tenderness of the dorsal aspect of the midfoot prevented weight-bearing or movement of the foot and ankle. Circulation and neurological examinations were normal. The skin was intact.Foot radiograph demonstrated a Lisfranc fracture dislocation (Fig. 1). A subsequent CT scan is shown (Fig. 2).Figure 1Radiograph of the left foot. There is lateral displacement of the first, second, and third metatarsals (tarsometatarsal or Lisfranc joint) with associated fracture of the middle cuneiform.Figure 2Computed tomography of the left foot. There is disruption of the tarsometatarsal (Lisfranc) joint with associated soft tissue swelling.This patient underwent an immediate open reduction and internal fixation of the Lisfranc fracture-dislocation. A postoperative radiograph is shown (Fig. 3). He was treated with a non-weight-bearing cast followed by a weight-bearing boot. He was advised to refrain from strenuous physical activity for 6 weeks after removal of the boot, after which time, normal physical activity was resumed. A non-steroidal anti-inflammatory drug was prescribed for pain. The patient had only mild pain with weight-bearing at 6 months and was ambulating without difficulty; he was pain-free at 2 years.Figure 3Radiograph of the left foot. There is anatomic alignment of the tarsometatarsal (Lisfranc) joint with a screw connecting the first metatarsal and the medial cuneiform, and a screw connecting the second metatarsal and the medial cuneiform.DiscussionThe Lisfranc joint derives its name from Jacques Lisfranc (1790–1847), a surgeon in Napoleon's army. Lisfranc performed amputations through the tarsometatarsal (TMT) joint to treat gangrenous injury of the foot [1]. Injuries of the Lisfranc joint are rare, representing less than 0.2% of all orthopedic traumas [2]. However, as many as 20% of Lisfranc joint injuries are missed upon initial examination [3]. The injury should always be suspected following trauma to the foot [4]. Most commonly, Lisfranc joint sprains and fractures are caused by high-velocity traumas, such as motor vehicle and industrial accidents. Injuries can be sustained during many athletic activities. In this case, injury was caused by direct impact of the foot against a tree trunk resulting in acute plantar flexion. In patients with high-energy trauma foot injury, CT imaging is often recommended to aid in diagnosis [5].Mild sprains to the Lisfranc joint, where there is no evidence of diastasis, may be treated by immobilization [6]. Treatment of more severe cases such as dislocations, however, usually includes open reduction and internal fixation of the joint. Cortical screw fixation is preferred to Kirschner wire fixation for these injuries [7]. The joint is secured to reduce without diastasis the lateral border of the medial cuneiform to the second metatarsal [3]. Surgery may be postponed to allow for reduction in tissue edema. However, if a risk of compartment syndrome is present, surgery should be performed immediately. After surgery, the foot is immobilized in a non-weight-bearing cast for 6 to 8 weeks, after which, the foot may be placed in an immobilizing boot with minimal weight bearing. After an additional 6 to 8 weeks, the boot may be removed and full weight-bearing may be established gradually. Complete recovery often takes up to 1 year [3], although long-term disability is possible. Despite appropriate reduction and fixation, patients may develop chronic post-traumatic arthritis [8]. Primary complete arthrodesis as a salvage procedure [9] is recommended only for severe chronic pain.ConclusionLisfranc injuries are often caused by high-velocity traumas. Careful examination and thorough testing are required to identify the injury correctly, as a patient may present symptoms consistent with sprains or other minor injuries. Treatment of severe cases may require open reduction and internal fixation followed by a period of immobilization. Complete recovery may take up to 1 year.ConsentWritten informed consent was obtained from the patient for publication of this case report and the accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsCB wrote the first draft of the manuscript, obtained patient consent, and reviewed the literature. SP proofread the case report and provided revisions. All authors read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2527604.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527604",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527604\nAUTHORS: Igor B Rogozin, Kira S Makarova, Youri I Pavlov, Eugene V Koonin\n\nABSTRACT:\nAbstractA widespread and highly conserved family of apparently inactivated derivatives of archaeal B-family DNA polymerases is described. Phylogenetic analysis shows that the inactivated forms comprise a distinct clade among archaeal B-family polymerases and that, within this clade, Euryarchaea and Crenarchaea are clearly separated from each other and from a small group of bacterial homologs. These findings are compatible with an ancient duplication of the DNA polymerase gene followed by inactivation and parallel loss in some of the lineages although contribution of horizontal gene transfer cannot be ruled out. The inactivated derivative of the archaeal DNA polymerase could form a complex with the active paralog and play a structural role in DNA replication.ReviewersThis article was reviewed by Purificacion Lopez-Garcia and Chris Ponting. For the full reviews, please go to the Reviewers' Reports section.\n\nBODY:\nFindingsDNA polymerases are enzymes that are essential for genome replication and repair in all cellular life forms [1,2]. There are several distinct families of DNA polymerases some of which are unrelated to each other whereas most show varying degrees of relationship [3,4]. All archaea and eukaryotes encode at least one but typically several paralogous B-family DNA polymerases that play a key role in DNA replication [5]. Euryarchaeota typically posses one or two B-family polymerase, in addition to the apparently unrelated D-family polymerase whereas Crenarchaeota have two or three paralogous B-family polymerases but no D-family polymerases [6-8].When examining the relationships between archaeal B-family DNA polymerases, we unexpectedly observed that in many Crenarchaeota and Euryarchaeota, one of the B-family DNA polymerase paralogs, despite the high level of overall sequence conservation, contains disrupted versions of the sequence motifs that are known to be essential for the catalytic functions. The polymerases of the B-family contain two enzymatic domain, namely, the N-terminal 3'-exonuclease domain that performs proofreading and the C-terminal polymerase proper domain [9,10]. In the apparently inactivated archaeal polymerase, the sequence motifs that in the active forms harbor the catalytic amino acids required for each of these activities are partially disrupted (Fig. 1 and Additional File 1). In particular, the crucial, most conserved YGDTD motif that coordinates Mg2+ in B-family DNA polymerases and in which both aspartates and the tyrosine are essential for the polymerase activity [11-13] harbors at least two replacements in the apparently inactivated derivatives although it is notable that the distal aspartate is conserved and so is likely to retain an important function (Fig. 1). In addition, the most conserved DIE motif (Exo I) of the N-terminal 3'-exonuclease proofreading domain does not have a clear counterpart in the corresponding part of the polymerase derivatives described here (Additional File 1), strongly suggesting that the exonuclease activity is inactivated as well.Figure 1Inactivated derivatives of B-family DNA polymerases in archaea: inactivation of the principal catalytic motif and phylogeny. The right part of the figure shows the sequences of the essential Mg2+-binding motif of B-family DNA polymerases for two families of active archaeal polymerase (top and middle) and the inactivated derivative (bottom). The motifs are shown in the form of sequence LOGOs where the height of the amino acid symbols is a function of the frequency of the given amino acid in the given position [20,21]. The left part of the figure shows a phylogenetic tree for archaeal B-family DNA polymerases. Color code: purple, Euryarchaeota (E); green: Crenarchaeota (C); brown: organisms that might represent distinct arcaheal phyla: Nanoarchaeota (N), Korarchaeota (K), and Thaumoarchaeota (T); black, bacteria (B). Each organism is denoted by the full systematic name and the Gene Identifier (GI) number. Multiple sequence alignment of archaeal B-family polymerases (Additional File 1) was constructed using the MUSCLE program [22]. The tree was built using the maximum likelihood method implemented in the MOLPHY program [23] by local rearrangement of an original Fitch tree [24]. The same program was used to compute bootstrap values which are indicated (%) for selected major branches.Phylogenetic analysis of archaeal B-family polymerases showed that the inactivated forms comprised a distinct clade with a 100% bootstrap support (Fig. 1). Moreover, within this clade, the Euryarchaeal and Crenarchaeal forms were clearly separated, also with full bootstrap support, and joined a third subclade that included similarly inactivated homologs from three diverse bacteria (Fig. 1). The presence of the inactivated DNA polymerases in diverse subsets of both Euryarchaeota and Crenarchaeota is compatible with an ancient duplication of the DNA polymerase gene followed by inactivation and parallel loss in some of the lineages. However, a relatively later duplication, inactivation and subsequent spread via horizontal gene transfer also could be a viable evolutionary scenario for the inactivated polymerases.Inactivation of the enzymatic function of a protein and utilization of the inactivated protein for non-enzymatic (structural) roles, so that it retains substantial sequence conservation, is not without precedent. Perhaps, the cases that are most relevant for the present finding are the apparent inactivation of the small subunits of DNA polymerases and the N-terminal 3'-exonuclease domains in eukaryotes. In all Euryarchaeota, the small subunits of PolD contain all sequence signature of the calcineurine-like superfamily of phosphoesterases [14] and have been shown to possess 3'-exonuclease, proofreading activity [15-17]. By contrast, in the eukaryotic orthologs of these proteins, most of the catalytic residues are replaced, so the protein apparently performs a structural role in the polymerase complex [8,14]. Another example is the inactivation of the N-terminal 3'-exonuclease domain in two eukaryotic members of the polB family, pol α and pol ζ. Despite the overall high level of sequence conservation, the catalytic residues in the exonuclease domains of these polymerases are replaced with amino acid residues that cannot function in the proofreading reaction, so structural roles have been proposed for these domains [2,18]. To our knowledge, the PolB derivative described here is the first case of an evolutionarily conserved and, by implication, functionally important but, apparently, inactive DNA polymerase. Conceivably, this protein is involved in archaeal DNA replication, perhaps, as a structural subunit of the DNA polymerase holoenzyme [14,19]. Determination of the function of this unusual protein should yield significant insight into archaeal DNA replication.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsIBR and YIP made the original observation and performed initial sequence comparisons, KSM performed the final sequence analysis and phylogenetic analysis, EVK wrote the manuscript. All authors edited and approved the final version.Reviewers' commentsReviewer's report 1Purificacion Lopez-Garcia, Universite Paris-SudRogozin and co-workers report here an interesting observation. They identify a highly conserved clade of B-family polymerases grouping euryarchaeotal, crenarchaeotal and also some bacterial sequences whose Mg2+-binding motif has several replacements from the canonical YGDTD, notably the tyrosine and one of the aspartates that are essential during polymerization. They propose that these amino acid changes lead to inactivated enzymes that might have retained a purely structural role. This is a justified in silico prediction that, however, remains to be tested. In particular, it would be important to see whether the enzymatic activity actually disappears or whether some kind of activity can still be performed, or performed under particular conditions and with precise partners. One of the aspartates involved in polymerization remains very well conserved in this motif; the same is true for another aspartate upstream in the sequence. Such sequence conservation at long evolutionary distances might indicate more than an exclusive structural function. I hope that this observation will foster the appropriate biochemical work downstream. I would recommend the authors to insist on the hypothetical nature of their prediction, which they do in parts of the text by speaking about 'apparently inactivated' polymerases, through all of it including the title and the figure legend.Authors' response: We appreciate the comments and included additional qualifications in the text; we also emphasize the conservation and probable functional importance of the distal aspartate in the Mg2+-binding motif. However, we also stress the obvious disruption of the catalytic motifs in the exonuclease domain, an observation that suggests concordant inactivation of both domains of these proteins.Reviewer's report 2Chris Ponting, Oxford UniversityRogozin et al. have made the interesting observation that some bacteria, euryarchaea and crenarchaea possess a homologue of B family DNA polymerase that contains substitutions within the YGDTD active site sequence. The evidence that this represents a homologue whose function is likely to be distinct from that of typical family B polymerases is compelling. What function this might be is uncertain, although the authors tentatively suggest an atypical role in DNA replication. This study shows, lest we ever forget, that demonstration of homology need not always lead to a successful prediction of function.Supplementary MaterialAdditional File 1Multiple sequence alignment of active and inactivated forms of archaeal B-family DNA polymerases.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2527686.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527686",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527686\nAUTHORS: Nicola J. Rutherford, Yong-Jie Zhang, Matt Baker, Jennifer M. Gass, NiCole A. Finch, Ya-Fei Xu, Heather Stewart, Brendan J. Kelley, Karen Kuntz, Richard J. P. Crook, Jemeen Sreedharan, Caroline Vance, Eric Sorenson, Carol Lippa, Eileen H. Bigio, Daniel H. Geschwind, David S. Knopman, Hiroshi Mitsumoto, Ronald C. Petersen, Neil R. Cashman, Mike Hutton, Christopher E. Shaw, Kevin B. Boylan, Bradley Boeve, Neill R. Graff-Radford, Zbigniew K. Wszolek, Richard J. Caselli, Dennis W. Dickson, Ian R. Mackenzie, Leonard Petrucelli, Rosa Rademakers\n\nABSTRACT:\nThe TAR DNA-binding protein 43 (TDP-43) has been identified as the major disease protein in amyotrophic lateral sclerosis (ALS) and frontotemporal lobar degeneration with ubiquitin inclusions (FTLD-U), defining a novel class of neurodegenerative conditions: the TDP-43 proteinopathies. The first pathogenic mutations in the gene encoding TDP-43 (TARDBP) were recently reported in familial and sporadic ALS patients, supporting a direct role for TDP-43 in neurodegeneration. In this study, we report the identification and functional analyses of two novel and one known mutation in TARDBP that we identified as a result of extensive mutation analyses in a cohort of 296 patients with variable neurodegenerative diseases associated with TDP-43 histopathology. Three different heterozygous missense mutations in exon 6 of TARDBP (p.M337V, p.N345K, and p.I383V) were identified in the analysis of 92 familial ALS patients (3.3%), while no mutations were detected in 24 patients with sporadic ALS or 180 patients with other TDP-43–positive neurodegenerative diseases. The presence of p.M337V, p.N345K, and p.I383V was excluded in 825 controls and 652 additional sporadic ALS patients. All three mutations affect highly conserved amino acid residues in the C-terminal part of TDP-43 known to be involved in protein-protein interactions. Biochemical analysis of TDP-43 in ALS patient cell lines revealed a substantial increase in caspase cleaved fragments, including the ∼25 kDa fragment, compared to control cell lines. Our findings support TARDBP mutations as a cause of ALS. Based on the specific C-terminal location of the mutations and the accumulation of a smaller C-terminal fragment, we speculate that TARDBP mutations may cause a toxic gain of function through novel protein interactions or intracellular accumulation of TDP-43 fragments leading to apoptosis.\n\nBODY:\nIntroductionTransactive response DNA binding protein with a molecular weight of 43 kDa (TDP-43) is a ubiquitously expressed nuclear protein encoded by the TARDBP gene, located on chromosome 1p36. TDP-43 was identified as the major disease accumulated protein in ubiquitinated neuronal cytoplasmic (NCI) and neuronal intranuclear inclusions (NII), that define a growing class of neurological diseases, collectively referred to as TDP-43 proteinopathies\n[1]–[5]. These diseases include amyotrophic lateral sclerosis (ALS), frontotemporal lobar degeneration (FTLD) with ubiquitin immunoreactive, tau negative inclusions (FTLD-U) and FTLD with motor neuron disease (FTLD-MND). In TDP-43 proteinopathies, TDP-43 is relocated from the nucleus to the cytoplasm and sequestered into inclusions that are mainly composed of hyperphosphorylated and C-terminally truncated TDP-43 fragments [4],[6],[7]. TDP-43 immunoreactive histopathology has also been reported in 20–30% of patients with Alzheimer's disease (AD), 70% of patients with hippocampal sclerosis (HpScl), 33% of patients with Pick's disease and in a subset of patients with Lewy-body related diseases [8]–[12]. TDP-43 is a highly conserved protein, containing 2 RNA recognition motifs and a C-terminal glycine-rich domain, known to promote protein-protein interactions [13].TDP-43 can bind to the common microsatellite region (GU/GT)n in RNA and DNA, with proposed functions in transcriptional regulation [13]. Most recent research has focused on the role of TDP-43 in the regulation of exon 9 alternative splicing in the cystic fibrosis transmembrane conductance regulator gene, however, additional targets have been identified and others likely await identification [14],[15]. TDP-43 has also been implicated in microRNA biogenesis [16].ALS and FTLD-U are etiologically complex disorders with genetic as well as environmental factors contributing to the disease. A positive family history is reported in 5–10% of ALS patients and in up to 50% of FTLD-U patients, often with an autosomal dominant pattern of inheritance [17]–[19]. Mutations in the Cu/Zn superoxide dismutase gene (SOD1) account for ∼10–20% of familial and 1–2% of apparent sporadic ALS patients [20]. However, TDP-43 inclusions were not present in SOD1 mutation carriers, suggesting a distinct disease mechanism in these patients [21]. The genetic basis of FTLD-U is just starting to be understood [19]. Loss-of-function mutations in the gene encoding the secreted growth factor progranulin (PGRN) are a major known cause of familial FTLD-U [22],[23], explaining up to 25% of patients worldwide [24]. Other rare genetic causes of familial FTLD-U include mutations in the valosin containing protein gene (VCP) and the gene encoding the charged multivesicular body protein 2B (CHMP2B), while some families with a combination of FTLD and ALS show genetic linkage to a locus on chromosome 9p [25]–[29].Since rare missense mutations and multiplications have been identified in genes encoding the major constituents of the pathological deposits in several neurodegenerative diseases, we hypothesized that mutations in TARDBP may contribute to the development of TDP-43 proteinopathies. In fact, the first missense mutations in TARDBP were recently discovered in 2 autosomal dominant ALS families and 2 sporadic ALS patients, supporting the central role for TDP-43 in disease pathogenesis [30],[31]. A large population-based study further identified 8 different missense mutations in 3 familial and 6 sporadic ALS patients and showed accumulation of a detergent-insoluble TDP-43 protein product of ∼28 kDa [32]. Here, we report on the extensive mutation screening of TARDBP in a diverse cohort of clinical and pathological confirmed patients with neurodegenerative diseases characterized by TDP-43 pathology, which led to the identification of 3 additional ALS families with TARDBP mutations. We further show accumulation of proteolytic cleaved fragments with a molecular weight of approximately 35 and 25 kDa in lymphoblastoid cell lines derived from TARDBP mutation carriers.Results\nTARDBP Mutation AnalysesWe performed in silico analyses of the TARDBP gene structure by alignment of human spliced expressed sequence tags listed in the UCSC genome browser (http://genome.ucsc.edu/). This led to the identification of a novel 5′ non-coding exon (exon 0) in addition to the known non-coding exon 1 and the 5 coding exons that are included in the TARDBP reference mRNA sequence (NCBI accession number NM_007375). Sequencing analyses of the 5 coding and 2 non-coding exons of TARDBP in our initial cohort of 176 clinical patients and 120 patients with pathologically confirmed TDP-43 pathology revealed 3 heterozygous missense mutations in 3 of the 116 analyzed ALS patients (2.6%), while no mutations were detected in 180 patients affected with FTLD-U, FTLD-MND, AD, HpScl and Lewy-body disease (Table 1, Figure 1). Since all mutation carriers were index patients of autosomal dominant ALS families, the frequency of TARDBP mutations increased to 3.3% in the subpopulation of familial ALS patients (3/92 patients). One silent mutation (p.Ala66) and 18 additional sequence variants in intronic and non-coding regions were further identified, none of which was predicted to affect the TDP-43 protein (Table S1). Genomic TARDBP copy-number analyses in 208 patients including all 116 ALS patients did not reveal deletions or multiplications.10.1371/journal.pgen.1000193.g001Figure 1Missense mutations identified in TARDBP in familial ALS patients.(A) Pedigrees showing family history of ALS for three probands carrying TARDBP mutations. Black symbols represent patients affected with ALS; white symbols represent unaffected individuals. Pedigrees are constructed based on family history data provided by the NINDS Human Genetics Resource Center DNA and Cell Line Repository (http://ccr.coriell.org/ninds). The alive/dead status of individuals is unknown. Arrowheads indicate the probands. The onset age of ALS symptoms and the TARDBP mutation identified are included below each proband. (B) DNA sequence traces observed in a sample from the proband of each family. The observed single base substitution and predicted amino acid change are indicated below each chromatogram. cDNA numbering is according to the largest TARDBP transcript (NM_007375.3) and starting at the translation initiation codon. Protein numbering is relative to the largest TDP-43 isoform (NP_031401.1).10.1371/journal.pgen.1000193.t001Table 1Patients included in the TARDBP sequencing analyses.Patients (N)Patients with positive family history (N)\nClinical Diagnoses\nALS9592FTLD6050FTLD-ALS2111\nPathological Diagnoses\nALS210FTLD-U2925FTLD-MND178AD (TDP-43+)4621LBD (TDP-43+)42HpScl (TDP-43+)31\nTotal\n296210All TARDBP mutations identified in this study are located in exon 6 (Figure 2). In the index patient of family A (ND10588), we identified the known c.1009 A>G mutation, predicted to substitute valine for methionine at codon 337 (p.M337V), and previously reported to segregate with disease in a large British autosomal dominant ALS kindred. In the index patient of family B (ND08308), a novel mutation c.1035 C>A was identified, predicted to change asparagine to a lysine at codon 345 (p.N345K). Finally, in the index patient of family C (ND08470), a novel mutation c.1147 A>G which predicts an isoleucine for a valine substitution at codon 383 (p.I383V) was identified. Sequence analysis of TARDBP exon 6 in 185 healthy control individuals did not identify these or other sequence variants. Using custom made TaqMan genotyping assays, the presence of p.M337V, p.N345K and p.I383V was further excluded in 640 US control individuals. Genotyping 652 sporadic ALS patients for these mutations did not identify additional mutation carriers. Since all 3 mutation carriers were obtained from the National Institute of Neurological Disorders and Stroke (NINDS) Human Genetics Resource Center DNA and Cell Line Repository (Coriell), DNA samples of relatives were unavailable for genetic studies and segregation of the mutations with disease could therefore not be determined.10.1371/journal.pgen.1000193.g002Figure 2Overview of mutations identified to date in TARDBP.Schematic overview of the 7 TARDBP exons showing coding regions in dark blue and non-coding regions in light blue (top). The TDP-43 protein structure with location of the conserved domains is shown with protein numbering according to the largest isoforms NP_031401.1 (middle). Protein sequence alignment shows strong conservation in the C-terminal region of TDP-43 (bottom). Colored boxes indicate the position of known and novel TDP-43 mutations identified in sporadic (orange) and familial (red) ALS patients. TDP-43 mutations identified in this study are underlined. Orange and red lines in TARDBP gene and TDP-43 protein indicate approximate positions of the mutations. RRM = RNA recognition motif.Clinical Characteristics of TARDBP Mutation CarriersAll 3 TARDBP mutation carriers were identified in the clinical patient series and were diagnosed by El Escorial criteria with probable or probable-lab supported ALS. Electromyography (EMG) examination was performed in 2 patients (ND10588 and ND08470) and was supportive of the diagnosis of ALS. A detailed overview of the distribution of upper and lower motor neuron signs in the TARDBP mutation carriers is included in Table S2. Patients ND10588 and ND08308 showed early onset ages of 38 and 39 years, respectively, while patient ND08470 showed symptom onset at 59 years (Figure 1). The initial presenting symptom in patients ND10588 and ND08470 was upper-limb ALS, while ND08308 suffered from lower-limb onset ALS. No signs of dementia or other atypical features of ALS were reported in any of the mutation carriers or their affected relatives. No autopsy of TARDBP mutation carriers was available.Allele Sharing Analyses of TARDBP p.M337VTo investigate whether our US p.M337V mutation carrier and the previously reported p.M337V family from the UK are descendants of a common founder, we did an allele sharing study with 12 short tandem repeat (STR) markers spanning a region of 6.7 Mb flanking TARDBP, including 5 markers within 1.0 Mb of TARDBP (Table 2). We determined the disease haplotype in the UK family and compared this to the genotypes observed in ND10588 to detect allele sharing. Shared alleles were observed for 6 out of 12 STR markers in the region, however, only one marker (Chr1(AC)_11.06) directly flanking TARDBP was shared and the 264 bp allele identified at this marker was common in the population (62.4%). In addition, potentially shared alleles at all other markers in the region were also common (>28%). These results make it unlikely that p.M337V originated from a common founder.10.1371/journal.pgen.1000193.t002Table 2Allele sharing in p.M337V families.MarkerGenomic positiona\nLinked allele (bp) UK familyFrequency of linked allele (%)Patient ND10588b\nD1S26637.18\n201\n8.9189–199D1S26947.26\n238\n50.0\n238–238D1S4509.51\n255\n10.7249–251D1S163510.91\n160\n14.8147–157\nc.1009A>G (p.M337V)\n11.01\nG\n-A–G\nChr1_11.0611.06\n264\n62.4\n264–270Chr1_11.2811.28\n128\n14.0132–132D1S266711.41\n264\n28.6260–264\nD1S274011.84\n90\n57.4\n90–100D1S48911.97\n143\n37.5\n143–143D1S43412.25\n240\n1.8246–248D1S159713.66\n171\n49.5\n171–171D1S22813.86\n121\n26.8119–123aGenomic position relative to the UCSC genome browser on the Human Mar. 2006 Assembly (http://genome.ucsc.edu/).bAlleles that are shared between the UK family and patient ND10588 are in bold.Biochemical Analysis of TARDBP Mutations in Familial ALS PatientsKabashi and colleagues previously reported a substantial increase in a ∼28 kDa fragment in lymphoblastoid cells with TARDBP mutations in the presence of the proteasomal inhibitor, MG-132, but not in lymphoblastoid cells derived from control individuals or ALS patients suggesting an increase aggregation property of these TDP-43 mutants [32]. Based on this result, we performed a similar study and analyzed the 3 patients with TARDBP mutations identified in our study, 2 sporadic ALS cases and 5 control individuals in the presence or absence of MG-132. Consistent with the previous report, a marked increase in the accumulation of detergent insoluble TDP-43 protein fragments were observed in the lymphoblastoid cell lines treated with MG-132 derived from patients with TARDBP mutations but not those derived from control individuals. In our study, we sized the higher and lower TDP-43 C-terminal fragments at approximately 35 and 25 kDa respectively (Figure 3). A similar increase was also found in individuals with sporadic ALS (Figure 3).10.1371/journal.pgen.1000193.g003Figure 3Biochemical analysis of TDP-43 in lymphoblastoid cell lines of TARDBP mutation carriers.Western blot analyses of protein lysates derived from lymphoblastoid cell lines from 3 familial ALS patients carrying different TARDBP mutations (p.M337V, p.N345K and p.I383V), 2 ALS patients (1 and 2) without TARDBP mutations and 5 healthy control individuals (Control 1–5). In lymphoblastoid cell lines derived from TARDBP mutation carriers and sporadic ALS patients an accumulation of 2 smaller C-terminal fragments of TDP-43 protein of approximately 35 and 25 kDa was observed in detergent-insoluble fractions treated with the proteasome inhibitor, MG-132. In lymphoblastoid cell lines derived from control individuals the levels of the 35 kDa fragment were substantially lower, and the 25 kDa fragment was mostly undetectable. Membranes from the soluble fraction were reprobed for beta-actin to monitor protein loading.We previously demonstrated that the proteolytic cleavage of TDP-43 by caspases can generate insoluble C-terminal fragments (35 and 25 kDa) similar to those found in diseased brains. Therefore, we investigated whether proteasome-induced toxicity was associated with proteolytic processing of endogenous TDP-43 in cell culture models. H4 neuroglioma cells were treated with either vehicle (DMSO) or proteasome inhibitor I (PSI) (10 µM) for 24 hours. In the presence of PSI, TDP-43 was cleaved into ∼35 and ∼25 kDa fragments (Figure 4), similar to the 35 and 25 kDa fragments found in the lymphoblastoid cell lines derived from the TARDBP mutation carriers (Figure 3). Similar results were obtained using MG-132 (data not shown). The inhibitory activity and toxicity of PSI also led to a marked increase in cleaved (active) capase-3 levels, which promotes apoptotic cell death and accumulates upon such inhibition. Furthermore, when we co-treated the cells with PSI and the caspase inhibitor, Z-VAD (OMe)-FMK, the generation of proteolytic TDP-43 fragments was inhibited (Figure 4). HSP70 immunoblot analysis was used to verify the inhibition of the proteasomal machinery. As expected, HSP70 levels were increased after PSI treatment and the levels persisted in the presence of caspase inhibitor Z-VAD (OMe)-FMK (Figure 4). Taken together, these data strongly suggest that proteasome inhibition is sufficient to promote proteolytic cleavage and accumulation of TDP-43 through a mechanism that implicates programmed cell death.10.1371/journal.pgen.1000193.g004Figure 4Proteasome inhibition increases the proteolytic cleavage of TDP-43.Western blot analyses of H4 neuroglioma cells treated with the proteasome inhibitor, PSI (10 µM, 24 hours) and a pan-caspase inhibitor, Z-VAD-FMK (100 µM, 24 hours) separately or in combination. Treatment with PSI revealed an increase in proteolytic cleavage of TDP-43 fragments (35 and 25 kDa) and an increase in caspase-3 activity. Treatment with a pan-caspase inhibitor suppressed PSI-induced TDP-43 cleavage and caspase-3 activity. HSP70 levels were increased after PSI treatment and the levels persisted in the presence of a pan-caspase inhibitor. Similar results were obtained in 3 independent experiments.DiscussionThe identification of rare mutations in genes encoding the major protein component of the pathologic brain depositions observed in familial neurodegenerative diseases has played a critical role in our current understanding of the molecular pathways underlying AD (APP), FTLD (MAPT) and Parkinson's disease (SNCA) [33],[34]. In this study, we performed mutation analyses of TARDBP, encoding TDP-43, in a large cohort of patients with neurodegenerative diseases characterized by TDP-43 pathology to determine if rare mutations or multiplications in TARDBP are involved in the genetic etiology of TDP-43 proteinopathies. Patients with a clinical diagnosis of ALS, FTLD or FTLD-ALS, and patients with pathologically confirmed TDP-43-proteinopathy were included in the analyses. In support of our hypothesis, 14 different pathogenic TARDBP missense mutations were reported by other researchers during the course of this study in familial and sporadic ALS patients [30]–[32],[35].We identified 2 novel TARDBP missense mutations (p.N345K and p.I383V) and the known p.M337V mutation in 3 out of 92 familial ALS patients (3.3%), while no mutations were identified in 24 sporadic ALS patients or 180 patients with other neurodegenerative diseases. p.M337V, p.N345K and p.I383V were excluded in 825 US control individuals and in 652 additional sporadic ALS patients. The TARDBP mutation frequency in our familial ALS cohort is comparable to the frequency reported by Kabashi and colleagues [32] (3/80 patients = 3.8%) but considerably higher than the frequency reported by Sreedharan and colleagues (1/154 patients = 0.6%) [31]. This may reflect the difference in study design, as a significant number of our patients were index patients of autosomal dominant ALS families, including all 3 patients carrying TARDBP mutations. Unfortunately, since all mutation carriers were index patients obtained from the NINDS Human Genetics Resource Center DNA and Cell Line Repository, DNA of affected relatives was not available to determine segregation of the mutations with disease. The absence of TARDBP mutations in patients with neurodegenerative diseases other than ALS in our study, confirms the lack of mutations and genetic association of TARDBP in FTLD populations [30], [36]–[38]. However, without extensive TARDBP sequence analyses in additional cohorts of FTLD and AD patients, TARDBP mutations cannot be excluded as a rare cause of these disorders.All TARDBP mutation carriers identified in this study presented with probable ALS according to El Escorial criteria in the absence of atypical clinical signs, in agreement with the previous reports on TARDBP mutation carriers.The p.M337V mutation has previously been reported to segregate with disease in a British autosomal dominant ALS family [31]. We identified p.M337V in an index patient from a US family with a strong family history of ALS. Our mutation carrier showed upper limb-onset ALS at 38 years of age, 6 years younger than the earliest onset age reported in the British p.M337V family. Signs of dementia were not reported in any of the family members, consistent with the previous report. An allele sharing study using 12 STR markers flanking TARDBP did not support a common ancestor for the UK family and our US patient, although our set of analyzed markers would not have detected a very distant common ancestor [39],[40]. In addition, we cannot exclude the rare possibility that marker Chr1_11.28 mutated in patient ND10588 or that the genomic position of this marker is incorrect, which would leave open the possibility of a shared region of maximum 1.3 Mb (D1S1635-D1S434). In anyway, the identification of p.M337V in two genealogically unrelated ALS families adds further strength to the pathogenicity of TARDBP mutations and justifies mutation screening for TARDBP in patients with familial ALS.Similar to 13 of the 14 previously reported TARDBP mutations, both novel missense mutations identified in this study were located in exon 6 encoding the highly conserved C-terminus of TDP-43, known to be involved in protein-protein interactions (Figure 2). p.N345K was identified in a 43 year old male with a 4 year history of ALS and an autosomal dominant family history. The p.I383V mutation was also identified in a familial ALS patient; however the onset age was 59 years, 2 decades later than the other 2 mutations identified in this study. This may reflect the more conservative amino acid substitution (Iso→Val) or its more C-terminal location in the TDP-43 protein compared to the other mutations, which may induce a different disease mechanism. Alternatively, additional genetic and/or environmental factors may determine the disease expression of TARDBP mutations, as suggested by the wide onset age range (48–83 years) observed in the recently published family with the p.A315T mutation in TARDBP\n[30]. Finally, although there is strong evidence supporting that p.N345K and p.I383V are pathogenic, there remains the possibility that these mutations in fact represent rare benign polymorphisms. Definitive confirmation of their pathogenic nature will depend on finding additional ALS patients carrying these mutations.To determine the pathological significance of TARDBP missense mutations on the post-translational processing of TDP-43, we examined human lymphoblastoid cell lines derived from all 3 familial TARDBP mutation carriers identified in this study, 2 ALS patients without TARDBP mutations and 5 control individuals (Figure 3). Patient cell lines revealed a substantial increase in a proteolytic cleaved fragment with a molecular weight of approximately 35 and 25 kDa consistent with caspase cleavage [7]. These data suggest that TARDBP mutations may cause a toxic gain of function through novel protein interactions or intracellular accumulation, particularly of caspase fragments. Kabashi and colleagues previously reported a similar substantial increase in a fragment of approximately 28 kDa in lymphoblastoid cell lines of TARDBP mutation carriers. This fragment accumulated in the presence of a proteasome inhibitor (MG-132), which led the authors to speculate that this TDP-43 product is likely degraded by the ubiquitin-proteasome system (UPS) [32]. While we can't exclude the enhanced aggregation of their mutants in the presence of the inhibitor, our data suggests that proteasome-induced toxicity enhances proteolytic cleavage of TDP-43 into 35 and 25 kDa fragments, resulting in cleavage fragments similar to those observed in ALS patients (Figure 4). Although we can't exclude the possibility that these fragments may be degraded by the UPS, it is likely that the accumulation of these fragments is primarily mediated by caspase cleavage.In conclusion, our findings support that TARDBP mutations are a rare cause of ALS, but so far are not found in other neurodegenerative diseases. Since all reported TARDBP mutations cluster in exon 6 encoding a highly conserved region of the TDP-43 protein, selective mutation analyses of TARDBP exon 6 in familial and sporadic ALS may be warranted.Materials and MethodsStudy PopulationsOur initial study population comprised a total of 296 patients with TDP-43 related neurodegenerative diseases, including 176 clinically diagnosed patients with ALS, FTLD and FTLD-ALS and 120 patients with pathologically confirmed TDP-43 proteinopathy. The average age at onset in the clinical cohort was 57.8±10.7 (range 31–81 years) and the average age at death in the pathological cohort was 74.8±13.8 (range 38–100 years). Among patients with known ethnicities (N = 214), 95% were Caucasian (N = 203), 3% were Hispanic (N = 7) and 2% were others (African/American (N = 2), East-Indian (N = 1) and Caribbean (N = 1)). A summary of the primary diagnoses and family history of the patients is provided in Table 1. The majority of the pathological confirmed patients (N = 87) were derived from the Mayo Clinic Jacksonville Brain Bank and primarily ascertained through The State of Florida Alzheimer's Disease Initiative funded through the Department of Elder Affairs, The Einstein Aging Study, The Udall Center for Excellence in Parkinson's Disease Research, CurePSP/The Society for Progressive Supranuclear Palsy, the Mayo Alzheimer's Disease Patient Registry (ADPR) and the Florida Alzheimer's Disease Research Center (ADRC). Additional clinical and pathological confirmed patients were ascertained through the Mayo Clinic Jacksonville and Rochester ADRC (N = 60), Mayo Clinic Scottsdale Alzheimer's Disease Center (ADC) (N = 4), the Neurological Institute of New York, Columbia University (N = 2), the University of California, Los Angeles (UCLA) ADC (N = 23), the University of British Columbia (N = 58), the Harvard Brain Bank (N = 5), the Sun Health Research Institute (N = 4), the Drexel University College of Medicine (N = 1), the Northwestern Feinberg School of Medicine (N = 13) and the Coriell Institute for Medical Research (N = 39). A list of the specific samples from the Coriell Institute included in the TARDBP mutation screening is provided as Table S3.To determine the frequency of the TARDBP mutations identified in our initial cohort, an additional cohort of 652 sporadic ALS patients was obtained from the University of British Columbia (N = 140), the Neurological Institute of New York, Columbia University (N = 48) and the Coriell Institute for Medical Research (N = 464). All control individuals (N = 825) included in the study were Caucasian and ascertained through the Mayo Clinics in Jacksonville, Florida and Scottsdale, Arizona.Sequencing AnalysisThe 5 coding and 2 non-coding exons of TARDBP were amplified by polymerase chain reaction (PCR) in standard 25 µl reactions using Qiagen PCR products (Table S4). PCR products were purified using the Agencourt Ampure method and sequenced using Big dye terminator V.3.1 products. Sequencing products were purified using the Agencourt CleanSEQ method and analyzed on an ABI 3730 DNA analyzer (Applied Biosystems, Foster City, CA, USA).GenotypingThe presence of TARDBP mutations c.1009A>G (p.M337V), c.1035 C>A (p.N345K) and c.1147 A>G (p.I383V) in sporadic ALS patients and control individuals was determined with custom-designed TaqMan SNP genotyping assays (Applied Biosystems) (Table S5) and analyzed on an ABI7900 genetic analyzer using SDS2.2.2 software.\nTARDBP Copy-Number AnalysesTaqMan gene expression assays to exons 2, 4 and 6 of TARDBP and to exon 5 of PSEN2 (for use as endogenous control) were designed using File Builder 3.1 software (Applied Biosystems) (Table S6) to test for the presence of genomic TARDBP copy-number mutations in 208 patients selected from our population. This approach was used to detect copy-number mutations affecting exons 2, 4 or 6, as well as complete TARDBP and large N- and C-terminal TARDBP deletions and multiplications. Real-time PCR with 25 ng genomic DNA as template was performed on an ABI7900 using the TaqMan method according to standard procedures. All samples were run in triplicate. The FAM-fluorescent signal was analyzed using SDS2.2.2 software, and genomic copy number determined by relative quantification (ΔΔct method).p.M337V Allele Sharing StudiesTo examine whether the US and UK families carrying the p.M337V mutation shared a common founder, we typed 12 STR markers spanning a region of 6.7 Mb flanking TARDBP in 3 patients and 8 unaffected relatives of the previously published UK family, in the US patient ND10588 and in 2 CEPH samples. STR markers were amplified with one fluorescently labeled primer and PCR fragments were analyzed on an automated ABI3100 DNA analyzer. Alleles were scored using the Genemapper software (Applied Biosystems). CEPH allele frequencies were used to estimate the allele frequency of the shared alleles in control individuals (CEPH genotype database; http://www.cephb.fr/cephdb/). The 2 novel markers were PCR amplified using Chr1_11.06-F: FAM-CAGCATCATGTGGTTTGGCAGT, Chr1_11.06-R: CAGCTCGCAGGGAAGATGAAA, Chr1_11.28-F: FAM-TGGCCATCTTAACAGGAACAGC and Chr1_11.28-R:TTCAAGGGCTTTCGAGGTGAA and allele frequencies were estimated in a population of 93 unrelated US control individuals.Cell Culture and TreatmentH4 neuroglioma cells were grown in Opti-Mem plus 10% FBS and 1% pen-strep. Cells were plated in 6-well plates and at 90% confluency treated with 10 µM proteasome inhibitor I (PSI) (EMD Chemicals, Inc. San Diego, CA) or 100 µM pan-caspase inhibitor (Z-VAD-FMK) (EMD Chemicals, Inc. San Diego, CA) separately or in combination. Twenty-four hours after treatment, cells were harvested for subsequent Western blot analysis in the Co-IP buffer (50 mM Tris-HCl, pH 7.4, 1 M NaCl, 1% Triton-X-100, 5 mM EDTA) plus 1% SDS, PMSF, protease and phosphatase inhibitors. A similar experiment was performed using 10 µM MG-132 (Calbiochem, San Diego, CA) instead of PSI.Fractionation ExperimentLymphoblastoid cells from 5 healthy control individuals, 3 familial ALS patients with TARDBP mutations and 2 ALS patients without TARDBP mutations were grown in RPMI1640 plus 10% FBS and 1% pen-strep. Cells were plated in T25 flasks and treated the following day with MG-132 (20 µM, 6 hours). Cell pellets from each cell line were lysed with the 0.2% Triton X-100-PBS with PMSF, protease and phosphatase inhibitors on ice for 10 minutes. After sonication, samples were centrifuged at 10,000 g for 15 minutes at 4°C. The supernatant was saved as the soluble fraction and the pellet was resuspended, sonicated in 2% SDS-PBS-Urea and saved as the insoluble fraction. The soluble and insoluble fractions were subjected to Western blot analysis.Western Blot AnalysisProtein concentrations of cells lysates were measured by a standard BCA assay (Pierce, Rockford, IL). Next, samples were heated in Laemmli's buffer and equal amounts of protein were loaded into 10-well 10% or 4–20% Tris-glycine gels (Novex, San Diego, CA). After transfer, blots were blocked with 5% nonfat dry milk in TBST (TPS plus 0.1% Triton X-100) for 1 hour, and then incubated with rabbit polyclonal TDP-43 antibody (1∶500; ProteinTech Group, Inc, Chicago, IL), rabbit polyclonal caspase-3 antibody (1∶1000; Cell Signaling, Beverly, MA), HSP70 (1∶2000; Stressgen, Ann Arbor, MI) or mouse monoclonal β-actin antibody (1∶5000, Sigma, Saint Louis, MS) overnight at 4°C. Membranes were washed three times each for 10 minutes with TBST and then incubated with anti-mouse or anti-rabbit IgG conjugated to horseradish peroxidase (1∶2000; Jackson ImmunoResearch, West Grove, PA) for 1 hour. Membranes were then washed three times each for 10 minutes, and protein expression was visualized by ECL treatment and exposure to film.Supporting InformationTable S1Sequence variants identified in TARDBP.(0.07 MB DOC)Click here for additional data file.Table S2Distribution of Upper and Lower Motor Neuron signs in TARDBP mutation carriers.(0.03 MB DOC)Click here for additional data file.Table S3Specific samples from the Coriell Institute included in the TARDBP mutation analyses.(0.07 MB DOC)Click here for additional data file.Table S4\nTARDBP PCR and sequencing primers.(0.03 MB DOC)Click here for additional data file.Table S5Primers and probes for TARDBP copy-number analyses.(0.03 MB DOC)Click here for additional data file.Table S6Detailed Information on TARDBP Taqman genotyping assays.(0.03 MB DOC)Click here for additional data file.\n\nREFERENCES:\n1. AraiTHasegawaMAkiyamaHIkedaKNonakaT\n2006\nTDP-43 is a component of ubiquitin-positive tau-negative inclusions in frontotemporal lobar degeneration and amyotrophic lateral sclerosis.\nBiochem Biophys Res Commun\n351\n602\n611\n17084815\n2. DavidsonYKelleyTMackenzieIRPickering-BrownSDu PlessisD\n2007\nUbiquitinated pathological lesions in frontotemporal lobar degeneration contain the TAR DNA-binding protein, TDP-43.\nActa Neuropathol\n113\n521\n533\n17219193\n3. KwongLKUryuKTrojanowskiJQLeeVM\n2008\nTDP-43 proteinopathies: neurodegenerative protein misfolding diseases without amyloidosis.\nNeurosignals\n16\n41\n51\n18097159\n4. NeumannMSampathuDMKwongLKTruaxACMicsenyiMC\n2006\nUbiquitinated TDP-43 in frontotemporal lobar degeneration and amyotrophic lateral sclerosis.\nScience\n314\n130\n133\n17023659\n5. WiderCWszolekZK\n2008\nEtiology and pathophysiology of frontotemporal dementia, Parkinson disease and Alzheimer disease: lessons from genetic studies.\nNeurodegener Dis\n5\n122\n125\n18322368\n6. WintonMJIgazLMWongMMKwongLKTrojanowskiJQ\n2008\nDisturbance of nuclear and cytoplasmic Tar DNA binding protein (TDP-43) induces disease-like redistribution, sequestration and aggregate formation.\nJ Biol Chem\n7. ZhangYJXuYFDickeyCABurattiEBaralleF\n2007\nProgranulin mediates caspase-dependent cleavage of TAR DNA binding protein-43.\nJ Neurosci\n27\n10530\n10534\n17898224\n8. Amador-OrtizCLinWLAhmedZPersonettDDaviesP\n2007\nTDP-43 immunoreactivity in hippocampal sclerosis and Alzheimer's disease.\nAnn Neurol\n61\n435\n445\n17469117\n9. FreemanSHSpires-JonesTHymanBTGrowdonJHFroschMP\n2008\nTAR-DNA binding protein 43 in Pick disease.\nJ Neuropathol Exp Neurol\n67\n62\n67\n18091558\n10. HigashiSIsekiEYamamotoRMinegishiMHinoH\n2007\nConcurrence of TDP-43, tau and alpha-synuclein pathology in brains of Alzheimer's disease and dementia with Lewy bodies.\nBrain Res\n1184C\n284\n294\n11. Nakashima-YasudaHUryuKRobinsonJXieSXHurtigH\n2007\nCo-morbidity of TDP-43 proteinopathy in Lewy body related diseases.\nActa Neuropathol (Berl)\n114\n221\n229\n17653732\n12. ProbstATaylorKITolnayM\n2007\nHippocampal sclerosis dementia: a reappraisal.\nActa Neuropathol\n114\n335\n345\n17639426\n13. BurattiEBaralleFE\n2008\nMultiple roles of TDP-43 in gene expression, splicing regulation, and human disease.\nFront Biosci\n13\n867\n878\n17981595\n14. BurattiEDorkTZuccatoEPaganiFRomanoM\n2001\nNuclear factor TDP-43 and SR proteins promote in vitro and in vivo CFTR exon 9 skipping.\nEmbo J\n20\n1774\n1784\n11285240\n15. MercadoPAAyalaYMRomanoMBurattiEBaralleFE\n2005\nDepletion of TDP 43 overrides the need for exonic and intronic splicing enhancers in the human apoA-II gene.\nNucleic Acids Res\n33\n6000\n6010\n16254078\n16. GregoryRIYanKPAmuthanGChendrimadaTDoratotajB\n2004\nThe Microprocessor complex mediates the genesis of microRNAs.\nNature\n432\n235\n240\n15531877\n17. Piemonte and Valle d'Aosta Register for Amyotrophic Lateral Sclerosis (PARALS)\n2001\nIncidence of ALS in Italy: evidence for a uniform frequency in Western countries.\nNeurology\n56\n239\n244\n11160962\n18. KunstCB\n2004\nComplex genetics of amyotrophic lateral sclerosis.\nAm J Hum Genet\n75\n933\n947\n15478096\n19. RademakersRHuttonM\n2007\nThe genetics of frontotemporal lobar degeneration.\nCurr Neurol Neurosci Rep\n7\n434\n442\n17764635\n20. RosenDRSiddiqueTPattersonDFiglewiczDASappP\n1993\nMutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis.\nNature\n362\n59\n62\n8446170\n21. MackenzieIRBigioEHIncePGGeserFNeumannM\n2007\nPathological TDP-43 distinguishes sporadic amyotrophic lateral sclerosis from amyotrophic lateral sclerosis with SOD1 mutations.\nAnn Neurol\n61\n427\n434\n17469116\n22. BakerMMackenzieIRPickering-BrownSMGassJRademakersR\n2006\nMutations in progranulin cause tau-negative frontotemporal dementia linked to chromosome 17.\nNature\n442\n916\n919\n16862116\n23. CrutsMGijselinckIvan der ZeeJEngelborghsSWilsH\n2006\nNull mutations in progranulin cause ubiquitin-positive frontotemporal dementia linked to chromosome 17q21.\nNature\n442\n920\n924\n16862115\n24. GassJCannonAMackenzieIRBoeveBBakerM\n2006\nMutations in progranulin are a major cause of ubiquitin-positive frontotemporal lobar degeneration.\nHum Mol Genet\n15\n2988\n3001\n16950801\n25. MoritaMAl-ChalabiAAndersenPMHoslerBSappP\n2006\nA locus on chromosome 9p confers susceptibility to ALS and frontotemporal dementia.\nNeurology\n66\n839\n844\n16421333\n26. SkibinskiGParkinsonNJBrownJMChakrabartiLLloydSL\n2005\nMutations in the endosomal ESCRTIII-complex subunit CHMP2B in frontotemporal dementia.\nNat Genet\n37\n806\n808\n16041373\n27. ValdmanisPNDupreNBouchardJPCamuWMeiningerV\n2007\nThree families with amyotrophic lateral sclerosis and frontotemporal dementia with evidence of linkage to chromosome 9p.\nArch Neurol\n64\n240\n245\n17296840\n28. VanceCAl-ChalabiARuddyDSmithBNHuX\n2006\nFamilial amyotrophic lateral sclerosis with frontotemporal dementia is linked to a locus on chromosome 9p13.2–21.3.\nBrain\n129\n868\n876\n16495328\n29. WattsGDWymerJKovachMJMehtaSGMummS\n2004\nInclusion body myopathy associated with Paget disease of bone and frontotemporal dementia is caused by mutant valosin-containing protein.\nNat Genet\n36\n377\n381\n15034582\n30. GitchoMABalohRHChakravertySMayoKNortonJB\n2008\nTDP-43 A315T mutation in familial motor neuron disease.\nAnn Neurol\n31. SreedharanJBlairIPTripathiVBHuXVanceC\n2008\nTDP-43 mutations in familial and sporadic amyotrophic lateral sclerosis.\nScience\n319\n1668\n1672\n18309045\n32. KabashiEValdmanisPNDionPSpiegelmanDMcConkeyBJ\n2008\nTARDBP mutations in individuals with sporadic and familial amyotrophic lateral sclerosis.\nNat Genet\n33. GoateAChartier-HarlinMCMullanMBrownJCrawfordF\n1991\nSegregation of a missense mutation in the amyloid precursor protein gene with familial Alzheimer's disease.\nNature\n349\n704\n706\n1671712\n34. HuttonM\n2001\nMissense and splice site mutations in tau associated with FTDP-17: multiple pathogenic mechanisms.\nNeurology\n56\nS21\n25\n11402146\n35. Van DeerlinVMLeverenzJBBekrisLMBirdTDYuanW\n2008\nTARDBP mutations in amyotrophic lateral sclerosis with TDP-43 neuropathology: a genetic and histopathological analysis.\nLancet Neurol\n36. GijselinckISleegersKEngelborghsSRobberechtWMartinJJ\n2007\nNeuronal inclusion protein TDP-43 has no primary genetic role in FTD and ALS.\nNeurobiol Aging\n37. RollinsonSSnowdenJSNearyDMorrisonKEMannDM\n2007\nTDP-43 gene analysis in frontotemporal lobar degeneration.\nNeurosci Lett\n38. SchumacherAFriedrichPDiehl-SchmidJIbachBPerneczkyR\n2007\nNo association of TDP-43 with sporadic frontotemporal dementia.\nNeurobiol Aging\n39. Al-ChalabiAAndersenPMChiozaBShawCShamPC\n1998\nRecessive amyotrophic lateral sclerosis families with the D90A SOD1 mutation share a common founder: evidence for a linked protective factor.\nHum Mol Genet\n7\n2045\n2050\n9817920\n40. PartonMJBroomWAndersenPMAl-ChalabiANigel LeighP\n2002\nD90A-SOD1 mediated amyotrophic lateral sclerosis: a single founder for all cases with evidence for a Cis-acting disease modifier in the recessive haplotype.\nHum Mutat\n20\n473"
+}

batch_0/PMC2527784.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527784",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527784\nAUTHORS: P Wolter, C Stefan, B Decallonne, H Dumez, M Bex, P Carmeliet, P Schöffski\n\nABSTRACT:\nSunitinib is approved for the treatment of metastatic renal cell carcinoma (RCC) and imatinib-resistant or -intolerant gastrointestinal stromal tumours (GIST). Several studies have identified unexpected rates of thyroid dysfunction with sunitinib treatment. We performed a prospective observational study with the aim of more accurately defining the incidence and severity of hypothyroidism in RCC or GIST patients receiving sunitinib. Thyroid function was assessed at baseline and on days 1 and 28 of each treatment cycle. Thyroid antibodies were assessed at baseline and during follow-up if abnormal thyroid function tests were recorded. Sixteen patients (27%) developed sub- or clinical hypothyroidism and required hormone replacement and 20 patients (34%) showed at least one elevated thyroid-stimulating hormone not requiring therapeutic intervention. Twenty patients (34%) did not develop any biochemical thyroid abnormality. Thus, sunitinib can induce (sub-) clinical hypothyroidism, warranting close monitoring of thyroid function. We propose a new algorithm for managing this side effect in clinical practise.\n\nBODY:\nSunitinib (SUTENT; Pfizer Inc., New York NY, USA) is an oral multitargeted tyrosine kinase inhibitor with both anti-angiogenic and anti-tumor activities mediated by signal blockade of several receptors, including VEGFR-1, -2 and -3; KIT; PDGFR-α and PDGFR-β; FLT-3; CSF-1 and RET (Chow and Eckhardt, 2007; Roskoski Jr 2007). Its efficacy in metastatic renal cell carcinoma (RCC) (Motzer et al, 2006) and in imatinib-resistant or -intolerant gastrointestinal stromal tumours (GIST) (Demetri et al, 2006) led to Food and Drug Administration (FDA) and European Medicines Agency (EMEA) approval for treatment of these cancers (Goodman et al, 2007; Rock et al, 2007). Sunitinib may also be beneficial in other malignancies, including neuroendocrine, breast, colon and non-small-cell lung cancer (Saltz et al, 2007; Burstein et al, 2008; Socinski et al, 2008).Our group has gained clinical experience with sunitinib administration, primarily in patients with RCC and GIST (Schöffski et al, 2006a, 2006b; Wolter et al, 2007), since 2005. Having first identified sporadic cases of thyroid dysfunction during sunitinib treatment, we evaluated this phenomenon in a limited number of patients, both retrospectively (n=14) and prospectively (n=19) (Schöffski et al, 2006b). In both studies, we found an unexpected high frequency of hypothyroidism (57 and 37%, respectively). This uncommon drug-induced side effect was not previously identified in clinical trials, in which sunitinib toxicities included fatigue, nausea, vomiting, hand-foot-skin reaction, rash, hypertension and/or diarrhoea (Demetri et al, 2006; Faivre et al, 2006; Motzer et al, 2006). An association between sunitinib treatment and thyroid dysfunction has been also described by other groups (Desai et al, 2006; Martorella et al, 2006; Shaheen et al, 2006; Mannavola et al, 2007; Rini et al, 2007; Wong et al, 2007). However, the previous studies were limited by factors related to study design including variable definition of hypothyroidism, incomplete panel of thyroid function tests (TFTs) or lack of measurements for all patients and inclusion of patients with underlying thyroid disorders.Here, we present the results of our prospective evaluation of incidence and severity of the new-onset hypothyroidism in a larger number of sunitinib-treated cancer patients (n=59). We include clearly defined exclusion criteria, baseline thyroid-stimulating hormone (TSH) values in all patients and thyroid antibodies in the majority of patients, as well as systematically evaluating TFTs during treatment. Furthermore, we discuss comparatively our study with the currently available data on sunitinib-induced thyroid dysfunction, and propose for the first time an algorithm for routine treatment of this clinical phenomenon.Patients and methodsPatients and treatment scheduleOur single-centre, prospective, observational study included patients receiving sunitinib in the Department of General Medical Oncology at the University Hospital Gasthuisberg (Leuven, Belgium). Between November 2005 and June 2007, a total of 74 patients received sunitinib for metastatic, immunotherapy-resistant RCC or imatinib-refractory or -intolerant GIST. Patients with abnormal TFTs at baseline (before sunitinib initiation), with previous thyroid hormone replacement due to underlying thyroid disease or under sunitinib treatment for less than 4 weeks, were excluded from the study. Patients had no food restriction, and medication other than sunitinib was not standardised. Sunitinib 50 mg day−1 was administered orally for 4 weeks, followed by a 2-week rest period. Doses were adjusted based on haematological and non-haematological adverse events, according to the manufacturer's recommendations. Study end points included the incidence and severity of sunitinib-induced thyroid dysfunction.Evaluation of thyroid functionAll patients had thyroid function assessed at baseline and on days 1 and 28 of each treatment cycle, which was determined by serum TSH, total triiodotyronine (T3) and the free thyroxine index (FTI) calculated from total thyroxine (T4) and the residual T-uptake. Antibodies against thyroid peroxidase (TPOAb), thyroglobulin (TgAb) and the TSH receptor (TR-Ab) were also assessed at baseline in most patients and during follow-up if abnormal TFTs were recorded. Reverse T3 and Tg were also measured in patients with abnormal TFTs. Serum TSH, total T3, T4 and T-uptake were measured by an electrochemiluminescent immunoassay (ECLIA; Roche, Mannheim, Germany); TPOAb and TgAb by a radioligand assay (Brahms, Henningsdorf, Germany); and TR-Ab by a radio receptor assay (Brahms). Serum samples were collected, handled and analysed according to internal standard operating procedures.Our laboratory reference ranges are 0.27–4.20 mIU l−1 for TSH; 5.1–14.1 μg dl−1 for total T4; 80–200 ng dl−1 for total T3; 4.8–12.7 for FTI; and 80–130% for T-uptake. The cutoff level for TPOAb is 100 U ml−1. Antibodies against thyroglobulin >200 U ml−1 is considered positive. Antibodies against thyroglobulin <1 IU l- is considered negative and >1.5 IU l−1 as positive.Definition of hypothyroidism and hyperthyroidismThe biochemical diagnosis of subclinical hypothyroidism and hyperthyroidism was determined in accordance with guidelines of the American Thyroid Association (ATA), the American Association of Clinical Endocrinologists (AACE) and the Endocrine Society (Surks et al, 1990, 2004). Subclinical hypothyroidism is considered as serum TSH above the upper limit of normal (ULN=4.20 mIU l−1 in our laboratory), with FTI within normal limits. Clinical hypothyroidism is defined as low serum FTI together with elevated TSH. Subclinical hyperthyroidism is defined as serum TSH below the lower limit of normal (LLN=0.27 mIU l−1 in our laboratory), with serum T4 and T3 within normal ranges. Overt hyperthyroidism is considered as low TSH and elevated FTI.Patients with overt hypothyroidism and those with at least two consecutive TSH measurements >10 mIU l−1 and symptoms compatible with hypothyroidism (e.g., fatigue, cold intolerance, constipation or weight gain) received thyroid hormone replacement therapy with L-thyroxine. Target TSH concentration for replacement intervention was 0.5–2 mIU l−1. All patients with overt hyperthyroidism and symptoms compatible with hyperthyroidism (e.g., anxiety, weakness, tremor, palpitations, heat intolerance, increased perspiration and weight loss, despite a normal or increased appetite and diarrhoea) were treated.ResultsPatient characteristicsIn total, 59 out of 74 patients were eligible for evaluation. Of the 15 patients who were excluded from the study, 5 patients had abnormal baseline TFTs, 4 patients were already receiving thyroid hormone replacement due to underlying thyroid disease and 6 patients received sunitinib treatment for less than 4 weeks. The percentage of patients with abnormal baseline TFTs (7%) agrees with the expected prevalence of thyroid dysfunction in the general population. Demographics, disease and treatment characteristics of the 59 eligible patients are presented in Table 1. Almost two-thirds of eligible patients had RCC (n=42), with the majority having clear cell histology, with the remainder belonging to the GIST cohort (n=17). The patient median age/range and gender ratio was similar for the two cohorts. Prior treatment included interleukin-2 and/or interferon-α for the RCC group and imatinib for the GIST group. In the total study population (n=59), median TSH at baseline was 1.47 mIU l−1 (range 0.28–3.94), which was within our laboratory reference range. Overall, median treatment with sunitinib was 29 weeks (range 4–82), with a slightly more prolonged median follow-up of 34 weeks (range 4–88). Haematological or non-haematological grade III and IV toxicities were observed in 18 (43%) RCC and 6 (35%) GIST patients.Thyroid function during sunitinib administrationTable 2 provides a summary of TFT results in both RCC and GIST cohorts during sunitinib treatment. We followed ATA, AACE and the Endocrine Society guidelines to establish a biochemical diagnosis of subclinical hypothyroidism and hyperthyroidism (Surks et al, 1990, 2004). Although there is some controversy as to whether the TSH ULN should be reduced, we considered TSH >4.20 mIU l−1 to be elevated. Only 20 patients (34%) had no biochemical thyroid abnormality. Twenty patients (34%) showed at least one elevated TSH not requiring therapeutic intervention and 16 (27%) patients developed sub- or clinical hypothyroidism requiring treatment. In three patients (5%) at least one serum TSH fell below the LLN (LLN=0.27 mIU l−1), with serum T4 and T3 within normal ranges.We observed TSH levels of up to 120 mIU l−1, approximately 30-fold ULN. Thyroid-stimulating hormone elevation can be quite substantial even within one sunitinib cycle, as observed, for instance, in one patient, from a normal 0.19 to a high of 78.63 mIU l−1. In most patients, such high levels of TSH were accompanied by typical features of hypothyroidism, such as progressive fatigue, myalgia, cold intolerance and constipation. We did not encounter myxoedema coma cases, but such cases may exist, as demonstrated by the study of Mannavola et al (2007).Thyroid function abnormalities were detected relatively early during treatment, with median time to abnormal TSH of 4 weeks (range 2–46). Representative courses of TSH levels in three patients treated with sunitinib are presented in Figure 1A–C. Note the characteristic alternate course of TSH concentrations (zigzag shape), with normal levels at baseline, elevated on day 28 of a 4-week treatment cycle, and normalised at the end of the 2-week rest period. In patients experiencing TFT abnormalities, such pattern was observed during most cycles of sunitinib treatment. In the 16 patients developing (sub)clinical hypothyroidism requiring hormone replacement, median TSH at baseline was 1.45 mIU l−1 (range 0.59–3.94), increased to a median of 8.15 mIU l−1 (range 2.61–26.32) by day 28 of the first cycle, and decreased to a median of 2.5 mIU l−1 (range 0.92–13.23) by day 1 of the second cycle. A similar pattern was also found in the subgroup of patients with at least one elevated TSH without requiring treatment; however, the increase in TSH was more moderate. Thus, for these patients, median TSH at baseline was 1.96 mIU l−1 (range 0.78–3.9), rose by day 28 of the first sunitinib cycle to 3.65 mIU l−1 (range 1.15–7.22), and dropped by day 1 of the second cycle to a median of 2.14 mIU l−1 (range 0.52–4.08) (Figure 1D).The median duration of sunitinib treatment in the group developing hypothyroidism was 48 weeks (range 4–81), which is much longer than in the group with no thyroid function abnormality (21 weeks; range 4–82). The percentage of patients with (sub)clinical hypothyroidism requiring treatment was higher in the RCC group (33%) than in the GIST group (12%). Dose reductions for grades III–IV haematological and non-haematological toxicities were more frequent in the hypothyroid group than in the normal thyroid group.Subclinical hypothyroidism was preceded by a short period of low TSH and elevated T3/T4 in 4 out of the 16 patients developing hypothyroidism. Although TPOAb was negative in all four patients, ultrasound revealed thyroid hypovascularity and serum Tg was elevated in 2, suggesting underlying drug-induced thyroiditis.Low titres of TPOAb were found in only 2 (4%) out of 49 total evaluable patients, whereas TgAb and TR-Ab were positive in 2 and 1 patients, respectively.DiscussionOur study prospectively analysed the occurrence and severity of hypothyroidism in cancer patients receiving sunitinib. Our patient population comprises 59 patients, the majority of whom had metastatic cytokine-resistant RCC, and the rest had imatinib-resistant or -intolerant GIST. Although our study confirms that sunitinib may induce biochemical and clinical hypothyroidism, it adds several features compared with previously published studies (Table 3). The definition of hypothyroidism was somewhat variable in previous studies. Other limitations include lack of complete TFTs before or during treatment, and/or measurements available in some but not all patients. When our manuscript was in preparation, Mannavola et al published results from a prospective evaluation, involving a smaller number of patients (n=24) (Mannavola et al, 2007).Following initiation of sunitinib, one-third of patients did not develop any biochemical TFT abnormalities. We observed a transient elevation of TSH in 34% of cases and (sub)clinical primary hypothyroidism requiring treatment in a further 27% of patients. As all patients had normal thyroid function at baseline, the latter percentage applies to the new-onset hypothyroidism. The frequency of (sub)clinical hypothyroidism in sunitinib-treated patients varies substantially among the reported studies (Table 3), but is up to 10-fold higher than the occurrence of hypothyroidism in the general population (Laurberg et al, 2005; Lazarus, 2007).Hypothyroidism is uncommon in the general population, with a prevalence of 4–8.5% (Surks et al, 1990, 2004). Reliable data on hypothyroidism prevalence in cancer patients are not available; however, some preliminary data suggest a slightly higher frequency in some tumour types such as melanoma and breast cancer (Ellerhorst et al, 2003; Jiskra et al, 2004; Shah et al, 2006).In general, hypothyroidism is particularly common in older age groups and women (Laurberg et al, 2005). However, in our study, the median patient age was 61 years (range 42–77) and male : female ratio was 3 : 1, indicating that these patient characteristics do not contribute substantially to the observed incidence in our patients. Similar patient characteristics were also found in the study by Rini et al (2007).On the basis of initial study protocol, we started hormone replacement therapy in patients with persistent TSH (>10 mIU l−1 on day 1 of two consecutive cycles) showing typical symptoms of hypothyroidism; symptoms resolved in the majority of cases but not all. The clinical presentation of patients with hypothyroidism is highly variable and non-specific, and side effects of sunitinib can be very similar to symptoms of hypothyroidism, certainly in patients with advanced cancer. Therefore, we cannot be sure whether fatigue in sunitinib-treated patients can be exclusively explained by primary hypothyroidism. The hypothesis of Garfield et al (2007) that hypothyroidism may be associated with improved outcome in certain types of cancer warrants further investigation, but hormone replacement therapy should not be withheld if clinically indicated (Garfield et al, 2007).Patients with preexisting thyroid function abnormalities may require higher doses of hormone replacement therapy to maintain an euthyroid state while treated with sunitinib. In our centre, TSH levels increased in three patients with a history of well-controlled hypothyroidism before sunitinib treatment, and adjustment of hormone replacement therapy was necessary. These patients were excluded from the prospective evaluation, however, a similar finding was reported by Rini et al (2007). More recently, De Groot et al (2006) reported the case of a GIST patient who was treated simultaneously with sunitinib and L-thyroxine following previous thyroidectomy and 131I-ablation due to follicular thyroid carcinoma (de Groot et al, 2006). The patient developed overt hypothyroidism while receiving sunitinib, and L-thyroxine doses were increased.The most common cause of hypothyroidism in the general population is chronic autoimmune (Hashimoto's) thyroiditis, resulting in antibody-mediated destruction of the thyroid tissue (Laurberg et al, 2005). More than 90% of such patients have elevated serum TPO or Tg antibodies. Desai et al (2006) reported TPOAb in 2 out of 42 patients, for whom values were normal (Desai et al, 2006). Rini et al (2007) measured TgAb (13 abnormal values of 44 available from 66 enrolled patients) and found no correlation between the presence of antibodies and either incidence or severity of other TFT abnormalities (Rini et al, 2007). Accordingly, in our patients, anti-TPO or anti-Tg antibodies were elevated only in a minority of those with elevated TSH. Similarly, in the study of Mannavola et al, TPOAb remained negative during all cycles in all patients who had normal TFTs at baseline (Mannavola et al, 2007). These findings do not support an underlying autoimmune process for developing hypothyroidism in sunitinib-treated patients.Mild thyrotoxicosis may, however, precede hypothyroidism in some but not all patients. Desai et al (2006) identified one or more TSH concentrations <0.5 mIU l−1, thus indicating the development of subclinical hyperthyroidism before hypothyroidism (Desai et al, 2006). In our prospective study, we identified four such cases, suggesting that drug-induced thyroiditis causing leakage of thyroid hormone into the circulation can potentially be an underlying mechanism for sunitinib-induced hypothyroidism.During sunitinib therapy, most of our patients underwent repeated CT scan with intravenous iodinated contrast, every 6–8 weeks. Iodine-containing contrast media can transiently elevate serum TSH, but not free T3 and T4 (Gartner and Weissel, 2004). Although the effect of repeated exposure to iodine on TSH is difficult to estimate in our patients, it should be noted that its level increased already soon after starting sunitinib administration in our RCC cohort.Seriously ill patients often show abnormal TFTs not necessarily associated with thyroid dysfunction, the so-called ‘non-thyroidal illness (NTI)’ or ‘low T3 syndrome’ (Demers, 2004). Therefore, peripheral thyroid hormone levels should be interpreted with caution in these patients. Serum TSH concentrations remain within reference limits for most NTI patients, although occasionally a mild transient TSH decrease can be observed at the beginning of NTI, followed by a rebound to mildly elevated values during recovery. Thus, NTI cannot be excluded as a confounding factor in some patients and may result in a slight overestimation of the percentage of patients developing slight biochemical TFT abnormalities. In the two other groups defined as ‘(sub)clinical hypo-/hyperthyroidism requiring treatment’, thyroid abnormalities had to be persistent and/or associated with clinical symptoms, which is not compatible with NTI.Primary hypothyroidism is not a common complication of therapeutic drugs. Among the drugs known to affect thyroid function are lithium, thioamides, amiodarone and cytokines, such as interferon and interleukin-2. However, the mechanisms are not well understood and may also differ from that of sunitinib-induced hypothyroidism. In our population, most patients with metastatic RCC had received cytokine therapy before starting sunitinib. We cannot exclude that this may have influenced the course of hypothyroidism. It may also explain the difference in thyroid dysfunction between the RCC (33%) and the GIST (12%) groups, as the latter did not have to receive cytokine therapy before sunitinib, and imatinib is not known to induce thyroid dysfunction.According to our study, there is no clear association between the daily dose of sunitinib and developing thyroid dysfunction. Thus, in contrast to the patient group with normal thyroid function during treatment, patients developing hypothyroidism underwent, for the most part, one or two dose reductions for grades III–IV haematological and non-haematological toxicities, suggesting that hypothyroidism in these patients reflects a particular susceptibility to sunitinib rather than dose dependency.The molecular mechanisms of sunitinib-induced hypothyroidism are currently unknown. Autoimmune-mediated hypothyroidism could not be demonstrated as an aetiological factor in this study, which is in accordance with the results of other groups. Sunitinib may have a direct effect on the thyroid, for example, through the inhibition of VEGFR and/or PDGFR. More recent studies in a mouse model have shown that VEGFR inhibition can induce capillary regression in various organs, including thyroid. Moreover, the vasculature of the thyroid had the greatest regression of all organs (Baffert et al, 2006; Kamba et al, 2006). Remarkably, in this animal model, thyroid capillaries regenerated in the absence of VEGFR inhibition. Such a response may explain the rhythmic pattern of TSH seen in patients treated with the 4/2 schedule. It is interesting that in the studies of Baffert et al (2006) and Kamba et al (2006), TSH also increased in mice treated with VEGF inhibition. Sunitinib may also inhibit TPO activity leading to reduced synthesis of the thyroid hormones, as suggested by in vitro studies (Wong et al, 2007). Indirectly, sunitinib may affect the thyroid by interfering with the metabolism of T4/T3 hormones, or with thyroid hormone action at the pituitary level. In patients treated with sorafenib, another tyrosine kinase inhibitor, thyroid dysfunction seems not to be a frequent side effect (Tamaskar et al, 2008; Wolter et al, manuscript in preparation).To address the clinical management of thyroid dysfunction in patients treated with sunitinib, we suggest an algorithm for TFT monitoring and intervention (Figure 2). In our study, patients who developed (sub)clinical hypothyroidism requiring treatment (n=16) showed TSH elevation early, within median 4 weeks, following sunitinib initiation. By contrast, some previously published studies concluded that thyroid dysfunction can occur later during sunitinib treatment (Table 3). However, these studies are mainly retrospective, and consistent TSH measurements on both days 1 and 28 of sunitinib cycles are lacking, which most probably account for this discrepancy (Desai et al, 2006; Martorella et al, 2006). Our observations that TFT abnormalities occur rather early after sunitinib initiation are, however, in better agreement with the prospective study by Mannavola et al (2007) on 24 GIST patients, where TSH measurements were performed on both days 1 and 28 and hypothyroidism developed at median of 3 cycles (range: 1–6). Moreover, we did not observe any patient with normal TFT during the first cycles of sunitinib who developed clinical hypothyroidism later in the course of treatment.In summary, we recommend that all patients treated with sunitinib have TFTs performed on days 1 and 28 of the first four cycles (Figure 2). By applying this intensive initial screening, we consider that we can detect patients at risk of developing sunitinib-induced thyroid dysfunction early following drug initiation. Moreover, on the basis of our study, monitoring of patients with no TFT abnormalities within the four first cycles can be subsequently performed more rarely, every three cycles, unless clinically indicated. In these cases, measurements are best recommended on day 28 rather than on day 1 of a cycle, as the likelihood to detect TFT abnormalities in highest at this point. Furthermore, we suggest starting hormone replacement therapy in patients with persistent TSH >10 mIU ml−1 and either low T4 or normal T4 but with typical symptoms of hypothyroidism. As TSH level declines and could be normalised by the end of the 2-week rest period of a sunitinib cycle, decisions to start hormone replacement should be based on TSH levels on day 1 of a new sunitinib cycle rather than on day 28 of a current cycle, to avoid overtreatment. We have encountered such a situation initially in our practise of treating sunitinib-induced thyroid dysfunction. Although during our study we initiated hormone replacement after TSH >10 mIU l−1 on day 1 of two consecutive cycles, we concluded for our algorithm that it is more prudent to start the therapy already after one cycle rather than waiting for an additional cycle. Although an elevated TSH at day 28 of a cycle may be associated with transient, reversible thyroid dysfunction, an elevated TSH at day 1 of a cycle, may already indicate a more pronounced damage to the thyroid, supporting our suggestion for treatment initiation at this point. Patients with preexisting thyroid function abnormalities may require higher doses of TSH to maintain an euthyroid state while treated with sunitinib. Although our data in measuring TSH after suninitib retrieval is limited, we also recommend continuing measuring TSH in such cases, as partial recovery of the thyroid function may occur. This observation is also supported by the study of Mannavola et al (2007).In conclusion, it is important that clinicians consider the potential side effect of sunitinib on thyroid function. Sunitinib-induced hypothyroidism is easily manageable with hormone replacement therapy, and should not withhold the use of sunitinib when indicated for cancer treatment. Hormone replacement is necessary to reduce symptoms of hypothyroidism, such as fatigue, but also to avoid the potentially life-threatening complications of severe hypothyroidism, such as myxoedema coma.We admit that our recommendations for thyroid management in patients under sunitinib may need further validation in a prospective randomized trial. We believe, however, that our data so far support the proposed algorithm, and will help clinicians in their daily practise until more validated guidelines become available.\n\nREFERENCES:\n1. Baffert F, Le T, Sennino B, Thurston G, Kuo CJ, Hu-Lowe D, McDonald DM (2006) Cellular changes in normal blood capillaries undergoing regression after inhibition of VEGF signaling. Am J Physiol Heart Circ Physiol\n290: H547–H55916172161\n2. Burstein HJ, Elias AD, Rugo HS, Cobleigh MA, Wolff AC, Eisenberg PD, Lehman M, Adams BJ, Bello CL, De Primo SE, Miller KD (2008) Phase II study of sunitinib malate, an oral multitargeted tyrosine kinase inhibitor, in patients with metastatic breast cancer previously treated with an anthracycline and a taxane. J Clin Oncol\n26: 1810–181618347007\n3. Chow LQ, Eckhardt SG (2007) Sunitinib: from rational design to clinical efficacy. J Clin Oncol\n25: 884–89617327610\n4. de Groot JW, Links TP, van der Graaf WT (2006) Tyrosine kinase inhibitors causing hypothyroidism in a patient on levothyroxine. Ann Oncol\n17: 1719–172016731538\n5. Demers LM (2004) Thyroid disease: pathophysiology and diagnosis. Clin Lab Med\n24: 19–2815157555\n6. Demetri GD, van Oosterom AT, Garrett CR, Blackstein ME, Shah MH, Verweij J, McArthur G, Judson IR, Heinrich MC, Morgan JA, Desai J, Fletcher CD, George S, Bello CL, Huang X, Baum CM, Casali PG (2006) Efficacy and safety of sunitinib in patients with advanced gastrointestinal stromal tumour after failure of imatinib: a randomised controlled trial. Lancet\n368: 1329–133817046465\n7. Desai J, Yassa L, Marqusee E, George S, Frates MC, Chen MH, Morgan JA, Dychter SS, Larsen PR, Demetri GD, Alexander EK (2006) Hypothyroidism after sunitinib treatment for patients with gastrointestinal stromal tumors. Ann Intern Med\n145: 660–66417088579\n8. Ellerhorst JA, Cooksley CD, Broemeling L, Johnson MM, Grimm EA (2003) High prevalence of hypothyroidism among patients with cutaneous melanoma. Oncol Rep\n10: 1317–132012883700\n9. Faivre S, Delbaldo C, Vera K, Robert C, Lozahic S, Lassau N, Bello C, Deprimo S, Brega N, Massimini G, Armand JP, Scigalla P, Raymond E (2006) Safety, pharmacokinetic, and antitumor activity of SU11248, a novel oral multitarget tyrosine kinase inhibitor, in patients with cancer. J Clin Oncol\n24: 25–3516314617\n10. Garfield DH, Hercbergs A, Davis PJ (2007) Re: hypothyroidism in patients with metastatic renal cell carcinoma treated with sunitinib. J Natl Cancer Inst\n99: 975–97617565154\n11. Gartner W, Weissel M (2004) Do iodine-containing contrast media induce clinically relevant changes in thyroid function parameters of euthyroid patients within the first week? Thyroid\n14: 521–52415307941\n12. Goodman VL, Rock EP, Dagher R, Ramchandani RP, Abraham S, Gobburu JV, Booth BP, Verbois SL, Morse DE, Liang CY, Chidambaram N, Jiang JX, Tang S, Mahjoob K, Justice R, Pazdur R (2007) Approval summary: sunitinib for the treatment of imatinib refractory or intolerant gastrointestinal stromal tumors and advanced renal cell carcinoma. Clin Cancer Res\n13: 1367–137317332278\n13. Jiskra J, Limanova Z, Barkmanova J, Smutek D, Friedmannova Z (2004) Autoimmune thyroid diseases in women with breast cancer and colorectal cancer. Physiol Res\n53: 693–70215588139\n14. Kamba T, Tam BY, Hashizume H, Haskell A, Sennino B, Mancuso MR, Norberg SM, O′Brien SM, Davis RB, Gowen LC, Anderson KD, Thurston G, Joho S, Springer ML, Kuo CJ, McDonald DM (2006) VEGF-dependent plasticity of fenestrated capillaries in the normal adult microvasculature. Am J Physiol Heart Circ Physiol\n290: H560–H57616172168\n15. Laurberg P, Andersen S, Bulow PI, Carle A (2005) Hypothyroidism in the elderly: pathophysiology, diagnosis and treatment. Drugs Aging\n22: 23–3815663347\n16. Lazarus JH (2007) Aspects of treatment of subclinical hypothyroidism. Thyroid\n17: 313–31617465860\n17. Mannavola D, Coco P, Vannucchi G, Bertuelli R, Carletto M, Casari P, Beck-Peccoz P, Fugazzola L (2007) A novel tyrosine-kinase selective inhibitor, sunitinib, induces transient hypothyroidism by blocking iodine uptake. J Clin Endocrinol Metab\n92: 3531–353417595247\n18. Martorella AJ, Omry G, Hann LE, Motzer RJ, Robbins RJ (2006) Receptor kinase (RTK) inhibitor SU11248 may cause hypothyroidism in a select group of patients with metastatic renal cell carcinoma (RCC). 88th Annual Meeting of the Endocrine Society, Boston, MA, USA\n19. Motzer RJ, Michaelson MD, Redman BG, Hudes GR, Wilding G, Figlin RA, Ginsberg MS, Kim ST, Baum CM, Deprimo SE, Li JZ, Bello CL, Theuer CP, George DJ, Rini BI (2006) Activity of SU11248, a multitargeted inhibitor of vascular endothelial growth factor receptor and platelet-derived growth factor receptor, in patients with metastatic renal cell carcinoma. J Clin Oncol\n24: 16–2416330672\n20. Rini BI, Tamaskar I, Shaheen P, Salas R, Garcia J, Wood L, Reddy S, Dreicer R, Bukowski RM (2007) Hypothyroidism in patients with metastatic renal cell carcinoma treated with sunitinib. J Natl Cancer Inst\n99: 81–8317202116\n21. Rock EP, Goodman V, Jiang JX, Mahjoob K, Verbois SL, Morse D, Dagher R, Justice R, Pazdur R (2007) Food and Drug Administration drug approval summary: Sunitinib malate for the treatment of gastrointestinal stromal tumor and advanced renal cell carcinoma. Oncologist\n12: 107–11317227905\n22. Roskoski Jr R (2007) Sunitinib: a VEGF and PDGF receptor protein kinase and angiogenesis inhibitor. Biochem Biophys Res Commun\n356: 323–32817367763\n23. Saltz LB, Rosen LS, Marshall JL, Belt RJ, Hurwitz HI, Eckhardt SG, Bergsland EK, Halter DG, Lockhart AC, Rocha Lima CM, Huang X, DePrimo SE, Chow-Maneval E, Chao RC, Lenz HJ (2007) Phase II trial of sunitinib in patients with metastatic with metastatic colorectal cancer after failure of standard therapy. J Clin Oncol\n20: 4793–4799\n24. Schöffski P, Dumez H, Clement P, Hoeben A, Prenen H, Wolter P, Joniau S, Roskams T, Van PH (2006a) Emerging role of tyrosine kinase inhibitors in the treatment of advanced renal cell cancer: a review. Ann Oncol\n17: 1185–119616418310\n25. Schöffski P, Wolter P, Himpe A, Dychter SS, Abraham S, Baum C, Peren H, Wildiers H, Bex M, Dumez H (2006b) Sunitinib-related thyroid dysfunction: a single center retrospective and prospective evaluation. J Clin Oncol\n24(18S): 3092\n26. Shah M, Orengo IF, Rosen T (2006) High prevalence of hypothyroidism in male patients with cutaneous melanoma. Dermatol Online J\n12: 1\n27. Shaheen PE, Tamaskar IR, Salas RN, Rini BI, Garcia J, Wood L, Dreicer R, Bukowski RM (2006) Thyroid function tests abnormalities in patients treated with metastatic renal cell carcinoma treated with sunitinib. J Clin Oncol\n24(18S): 4605\n28. Socinski MA, Novello S, Brahmer JR, Rosell R, Sanchez JM, Belani CP, Govindan R, Atkins JN, Gillenwater HH, Pallares C, Tye L, Selaru P, Chao RC, Scagliotti GV (2008) Multicenter, phase II trial of sunitinib in previously treated, advanced non-small-cell lung cancer. J Clin Oncol\n26: 650–65618235126\n29. Surks MI, Chopra IJ, Mariash CN, Nicoloff JT, Solomon DH (1990) American Thyroid Association guidelines for use of laboratory tests in thyroid disorders. JAMA\n263: 1529–15322308185\n30. Surks MI, Ortiz E, Daniels GH, Sawin CT, Col NF, Cobin RH, Franklyn JA, Hershman JM, Burman KD, Denke MA, Gorman C, Cooper RS, Weissman NJ (2004) Subclinical thyroid disease: scientific review and guidelines for diagnosis and management. JAMA\n291: 228–23814722150\n31. Tamaskar I, Bukowski R, Elson P, Ioachimescu AG, Wood L, Dreicer R, Mekhail T, Garcia J, Rini BI (2008) Thyroid dysfunction test abnormalities in patients with metastatic renal cell carcinoma treated with sorafenib. Ann Oncol\n19: 265–26817962201\n32. Wolter P, Dumez H, Schoffski P (2007) Sunitinib and hypothyroidism. N Engl J Med\n356: 1580–158117429091\n33. Wong E, Rosen LS, Mulay M, Vanvugt A, Dinolfo M, Tomoda C, Sugawara M, Hershman JM (2007) Sunitinib induces hypothyroidism in advanced cancer patients and may inhibit thyroid peroxidase activity. Thyroid\n17: 351–35517465866"
+}

batch_0/PMC2527791.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527791",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527791\nAUTHORS: M T Epping, A A M Hart, A M Glas, O Krijgsman, R Bernards\n\nABSTRACT:\nThe tumour antigen PReferentially expressed Antigen of MElanoma (PRAME) is expressed in a variety of malignancies, including breast cancer. We have analysed PRAME gene expression in relation to clinical outcome for 295 primary breast cancer patients. Kaplan–Meier survival curves show a correlation of PRAME expression levels with increased rates of distant metastases and decreased overall patient survival. This correlation existed both for the entire patient group (n=295) and for the subgroup of patients (n=185) who did not receive adjuvant chemotherapy. Multivariable analysis indicated that PRAME is an independent marker of shortened metastasis-free interval in patients who did not receive adjuvant chemotherapy. PRAME expression was associated with tumour grade and negative oestrogen receptor status. We conclude that PRAME expression is a prognostic marker for clinical outcome of breast cancer, independent of traditional clinicopathological markers.\n\nBODY:\nThe expression of preferentially expressed antigen of melanoma (PRAME) has been detected in a variety of cancers including breast cancer, but its expression is low or absent in normal tissues (Ikeda et al, 1997). The protein PRAME was first detected as a tumour antigen in cells isolated from a melanoma, and high PRAME expression has been detected in 88–95% of primary melanomas (Ikeda et al, 1997). PRAME can inhibit retinoic acid (RA) signalling leading to resistance of melanoma cells to proliferation arrest induced by RA (Epping et al, 2005). The function of PRAME in breast cancer and other cancers in which it is expressed is still elusive (Epping and Bernards, 2006).Among the tumour types expressing PRAME are breast cancers, lung cancers, sarcomas, Wilms' tumours, renal carcinomas, medulloblastomas, head-and-neck cancers, lymphomas, and several types of leukaemias (Ikeda et al, 1997; Neumann et al, 1998; Li et al, 2002; Steinbach et al, 2002b; Boon et al, 2003; Radich et al, 2006; Willenbrock et al, 2006). Although many reports have focused on the detection of PRAME mRNA transcripts, there are few studies that have linked gene expression data directly to clinical information. The expression of PRAME is associated with poor prognosis in neuroblastoma: high PRAME expression is associated with more advanced tumour stage, higher ages of patients at diagnosis, and poor clinical outcome (Oberthuer et al, 2004). PRAME expression has been linked to good prognosis in paediatric AML (Steinbach et al, 2002a), but other studies did not find a significant correlation (Paydas et al, 2005).We previously identified a gene expression profile that is associated with the risk of early distant metastases in young breast cancer patients (van ‘t Veer et al, 2002). The tumours of 78 women with sporadic lymph node-negative breast cancer were selected to search for a prognostic signature in their gene expression profiles. We found that 231 genes were significantly associated with disease outcome, one of which was PRAME. Serial computational analyses of the data were conducted to generate a ‘prognosis classifier’ comprising an optimised number of 70 marker genes. PRAME was part of the set of 231 ‘significant prognosis reporter genes’, but not of the optimal set of 70, which together constitute the prognosis classifier (van ‘t Veer et al, 2002). This classifier allowed the categorisation of patients in a ‘good-prognosis’ and a ‘poor-prognosis’ group, as defined by the occurrence of distant metastases within 5 years after initial diagnosis (van ‘t Veer et al, 2002).In a subsequent validation study, the tumours of a series of 295 consecutive women with breast cancer were used to validate the 70-gene prognostic profile (van de Vijver et al, 2002). The genome-wide gene expression profiles of these tumours demonstrated the prognostic power of this profile in predicting the outcome of disease. The poor-prognosis signature was the strongest predictor of the likelihood of distant metastases in all patients, with a more accurate prediction of disease outcome than clinicopathological criteria and the NIH and St Gallen criteria (van de Vijver et al, 2002). The prognosis classifier could be used to select effectively those high-risk patients who would benefit from adjuvant therapy, while reducing the number of patients who receive unnecessary treatment and may suffer from the side effects. Thus, the prognostic profile potentially provides a powerful tool to tailor adjuvant systemic treatment of breast cancer (Buyse et al, 2006). Moreover, the prognostic power of the 70-gene profile indicates that the ability to metastasise to distant sites is an early and inherent genetic property of breast cancer, which argues against the widely accepted idea that metastatic potential is acquired relatively late during multistep tumorigenesis (Bernards and Weinberg, 2002).Because PRAME was not part of the 70-gene prognosis classifier, we have analysed the gene expression data set of the 295 breast cancer patients from our previous study (van de Vijver et al, 2002) for the expression levels of PRAME and found a remarkable association between PRAME expression and poor clinical outcome. We discuss these findings in the context of the recent insights in PRAME function.Materials and MethodsSelection of patientsTumours from a series of 295 consecutive women with breast cancer were selected from the fresh-frozen tissue bank of the Netherlands Cancer Institute according to the following criteria as described previously (van de Vijver et al, 2002). The tumour was primary invasive breast carcinoma less than 5 cm in diameter at pathological examination (pT1 or pT2); the apical axillary lymph nodes were tumour-negative, as determined by a biopsy of the infraclavicular lymph nodes; the age at diagnosis was 52 years or younger; the calendar year of diagnosis was between 1984 and 1995; and there was no previous history of cancer, except nonmelanoma skin cancer. All patients had been treated by modified radical mastectomy or breast-conserving surgery, including dissection of the axillary lymph nodes, followed by radiotherapy if indicated. Among the 295 patients, 151 had lymph node-negative disease (results on pathological examination, pN0) and 144 had lymph node-positive disease (pN+). All patients were assessed at least annually for a period of at least 5 years. Follow-up information was extracted from the medical registry of the Netherlands Cancer Institute. The median follow-up among all 295 patients was 6.7 years (range, 0.05–18.3). There were no missing data. The study was approved by the medical ethics committee of the Netherlands Cancer Institute.Isolation of RNA and microarray expression profilingThe isolation of RNA, labelling of complementary RNA (cRNA), hybridisation of labelled cRNA to 25 000 gene arrays, and assessment of expression ratios were all performed as previously described (Hughes et al, 2001; van ‘t Veer et al, 2002). In brief, tumour material was snap-frozen in liquid nitrogen within 1 h after surgery. Frozen sections were stained with haematoxylin and eosin; only samples that had more than 50 per cent tumour cells were selected. Thirty 30-μm sections were used for the isolation of RNA. Total RNA was isolated with RNAzolB and dissolved in RNase-free water. Then 25 μg of total RNA was treated with DNase with use of the Qiagen RNase-free DNase kit and RNeasy spin columns, and the RNA was then dissolved in RNase-free water to a final concentration of 0.2 μg per microlitre. Complementary RNA was generated by in vitro transcription with the use of T7 RNA polymerase and 5 μg of total RNA and labelled with Cy3 or Cy5 (Cy Dye; Amersham Pharmacia Biotech, Piscataway, NJ, USA). Five micrograms of Cy-labelled cRNA from one breast cancer tumour was mixed with the same amount of reverse-colour Cy-labelled product from a reference pool that consisted of an equal amount of cRNA from each patient. Labelled cRNAs were fragmented to an average size of approximately 50–100 nucleotides by heating the samples to 60°C in the presence of 10 mM zinc chloride and adding a hybridisation buffer containing 1 M sodium chloride, 0.5 per cent sodium sarcosine, 50 mM morpholino-ethane sulphonic acid (pH 6.5), and formamide (final concentration, 30 per cent at 40°C); the final volume was 3 ml. The microarrays included 24 479 biologic oligonucleotides as well as 1281 control probes. The hybridisations were performed in duplicates and with colour reversal. After hybridisation, the slides were washed and scanned with a confocal laser scanner (Agilent Technologies, Palo Alto, CA, USA). Fluorescence intensities on scanned images were quantified, and the values were corrected for the background level and normalised.Statistical methodsThe PRAME expression data were linked to the clinical database on the basis of the Rosetta Bioinformatics patient identification number (rosid). The probe sequence for PRAME was checked using BLAST and was found to code only for PRAME and did not match any other sequence. PRAME expression was quantified as the logarithm of the intensity ratio with respect to a standard pool of breast cancers. A normal probability plot indicated the existence of two subgroups of patients with either relatively high or low PRAME expression. An expectation–maximisation (EM) algorithm was used to estimate the mean values and standard deviations (s.d.) of the log 2 ratio subgroups, which were treated as normal distributions, an assumption that was supported by histograms (Figure 1). The distance of every sample from the mean of both groups measured in s.d. of that group was calculated. The samples were assigned to the group with the smallest distance in s.d. of that group. This method assigned 98 samples (33%) to the high-expression group and 197 samples (67%) to the low-expression group. The cutoff value for PRAME expression levels in log 2 was −1.45, which equals −0.4365 in log 10.Time in years to distant metastasis as first event was measured from the date of diagnosis of the primary tumour. For metastasis, the occurrence of a metastasis before any other failure (locoregional recurrence or contralateral breast cancer or death) was counted as an event, whereas all other patients were censored at the time of another type of failure or end of follow-up. For overall survival, death from any cause was counted as an event, whereas patients still alive at the end of follow-up were censored at that time. The P-values shown for the Kaplan–Meier (KM) curves were calculated using a log rank test. Cox regression analysis was used to analyse the prognostic value of PRAME in addition to that of known clinicopathological prognosticators.ResultsThe expression levels of PRAME mRNA in the primary breast cancer biopsies of 295 patients used in our previous study (van de Vijver et al, 2002) were analysed and matched to the clinical follow-up data. There were 110 patients who received adjuvant systemic chemotherapy, the majority of whom had lymph node-positive disease (van de Vijver et al, 2002). These patients were separated from the whole group, leaving 185 patients who had not received adjuvant chemotherapy. In both cohorts of patients, two subgroups for PRAME expression existed, with relatively low and relatively high PRAME levels. Histograms were made to confirm the existence of two subgroups for PRAME expression in each cohort (Figure 1). Ninety-eight breast cancer samples (33%) were assigned to the high-expression group and 197 samples (67%) to the low-expression group, using a cutoff value of −1.45 in log 2, which equals −0.4365 in log 10 (see Materials and methods).To evaluate the clinical relevance of PRAME expression, Kaplan–Meier plots for overall survival and metastasis-free interval were made. For the group of patients who did not receive adjuvant chemotherapy, there were significant associations between PRAME expression levels and shortened overall survival (P<0.001), and PRAME expression levels and shortened metastasis-free interval (P<0.001) (Figure 2). These data indicate that PRAME expression is a prognostic marker for breast cancer progression.Subsequently, we evaluated whether there was a relation between PRAME expression levels and clinical follow-up in the whole group of 295 patients. There were negative associations between PRAME expression levels and overall survival (P=0.0034) and metastasis-free interval (P=0.0029), that is high PRAME levels were associated with poor outcome (Figure 3). Thus, PRAME mRNA expression levels were inversely correlated with survival in all the patients, irrespective of treatment, and KM plots and P-values showed a more significant separation between high- and low-risk samples for the patient group without adjuvant chemotherapy compared to all 295 patients.To determine whether PRAME expression is predictive for adjuvant chemotherapy response, the subgroup of 110 patients who received chemotherapy was analysed for PRAME expression and clinical outcome. Remarkably, there was no significant difference between the treated patients with high and low PRAME expressions with regard to overall survival (P=0.95) and metastasis-free interval (P=0.91) (Figure 4). These data indicate that the patients with high PRAME expression have had significant benefit from the adjuvant chemotherapy and that in this patient cohort, PRAME expression was predictive for response to adjuvant chemotherapy.Expression of PRAME was compared with clinicopathological characteristics in a multivariable analysis. Independent prognostic factors for metastasis-free interval were PRAME expression (P=0.006), age, lymph node status, mastectomy and tumour grade, and vascularisation for the 185 patients who did not receive chemotherapy (Table 1). In all, for the 295 patients, age, lymph node status, tumour size, grade and vascularisation, and chemotherapy were independent prognostic factors for metastasis-free interval (Table 2). PRAME expression was not significant in this group (P=0.11), probably due to the strong effect of chemotherapy.To search for a possible relation between PRAME expression and clinicopathological tumour characteristics, associations between PRAME expression and clinical variables (age, tumour diameter, number of positive axillary nodes, histologic grade, oestrogen receptor (ER) status, vascular invasion, lymphatic infiltration) were analysed. Only for grade and ER status, evidence for an association with PRAME was found. In the 295-breast cancer patient group, high PRAME expression levels were associated with poorly differentiated tumours and low PRAME expression with well-differentiated tumours (Kruskal–Wallis test: P=0.0005). Furthermore, the expression of PRAME was associated with negative ER status, as ER-negative patients had mostly high PRAME expression, whereas ER-positive patients had mostly low PRAME expression (Mann–Whitney test: P<0.0001). Although significant, this negative association of PRAME with ER was not consistent among all patients, which may be explained by the two patient subgroups with respect to PRAME mRNA expression (Figure 1). The lymph node status of patients was not directly associated with PRAME expression (P=0.678 for the nontreated patients; P=0.691 for all patients).In conclusion, our analyses provide evidence for an association of high PRAME expression levels with poor clinical outcome of premenopausal breast cancer with increased rates of distant metastases and lower rates of overall survival.DiscussionIn the present study, we have evaluated the prognostic value of PRAME mRNA expression in 295 primary breast cancer biopsies. The full-genome gene expression data of these patients were previously used to validate the ‘poor-prognosis profile’ of 70 genes (van de Vijver et al, 2002; van ‘t Veer et al, 2002). Using this large data set, we have demonstrated that PRAME is a prognostic marker for metastasis-free interval and overall survival. These data are reminiscent of a report showing that the expression of PRAME is associated with poor prognosis in neuroblastoma (Oberthuer et al, 2004).Recently, an association between PRAME expression and unfavourable disease outcome was shown in a study involving 103 breast cancer biopsies in which PRAME mRNA was detected in ∼53% of tumour specimens (Doolan et al, 2008). The presence of PRAME expression was associated with shortened disease-free survival and overall survival in all breast cancer cases (Doolan et al, 2008). In the cases in which adjuvant chemotherapy was administrated, an association existed between PRAME expression and shortened relapse-free survival (Doolan et al, 2008). In our study, we provide evidence that PRAME expression is a prognostic marker for metastasis-free interval and overall survival in primary breast cancer. We also demonstrate that the strongest correlation exists for patients who did not receive adjuvant chemotherapy, indicating that PRAME has prognostic power in primary breast cancer. Therefore, the data of the previous study (Doolan et al, 2008) are fully consistent with the conclusion of the present larger study based on 295 patients and a subgroup of 185 patients who did not receive chemotherapy. Our data differ from those of Doolan et al (2008), in that we find that PRAME expression predicts benefit of chemotherapy (Figure 4). However, given this discrepancy, additional tumour series should be evaluated to address the significance of PRAME as a biomarker of chemotherapy response.We have found recently that PRAME is a corepressor of the RA receptor and harbours seven ‘nuclear receptor boxes’, motifs that allow interaction with nuclear receptors (Epping et al, 2005). This finding begs the question whether the function of PRAME in breast cancer progression is also to inhibit the function of specific nuclear receptors. In our study, PRAME expression was found to be higher in the ER-negative breast tumours, suggesting that PRAME does not act on ER. Consistent with this, we did not observe an effect of PRAME expression on ER or progesterone receptor activity (Epping et al, 2005). Similarly, we observed lower RARα levels in tumours having high PRAME expression, suggesting that PRAME may also not act through RARα in breast cancer. Which nuclear receptor (if any) is targeted by PRAME in breast cancer remains elusive at present.\n\nREFERENCES:\n1. Bernards R, Weinberg RA (2002) A progression puzzle. Nature\n418: 82312192390\n2. Boon K, Edwards JB, Siu IM, Olschner D, Eberhart CG, Marra MA, Strausberg RL, Riggins GJ (2003) Comparison of medulloblastoma and normal neural transcriptomes identifies a restricted set of activated genes. Oncogene\n22: 7687–769414576832\n3. Buyse M, Loi S, van't Veer L, Viale G, Delorenzi M, Glas AM, d'Assignies MS, Bergh J, Lidereau R, Ellis P, Harris A, Bogaerts J, Therasse P, Floore A, Amakrane M, Piette F, Rutgers E, Sotiriou C, Cardoso F, Piccart MJ (2006) Validation and clinical utility of a 70-gene prognostic signature for women with node-negative breast cancer. J Natl Cancer Inst\n98: 1183–119216954471\n4. Doolan P, Clynes M, Kennedy S, Mehta JP, Crown J, O'Driscoll L (2008) Prevalence and prognostic and predictive relevance of PRAME in breast cancer. Breast Cancer Res Treat\n109: 359–36517624586\n5. Epping MT, Bernards R (2006) A causal role for the human tumor antigen preferentially expressed antigen of melanoma in cancer. Cancer Res\n66: 10639–1064217108098\n6. Epping MT, Wang L, Edel MJ, Carlee L, Hernandez M, Bernards R (2005) The human tumor antigen PRAME is a dominant repressor of retinoic acid receptor signaling. Cell\n122: 835–84716179254\n7. Hughes TR, Mao M, Jones AR, Burchard J, Marton MJ, Shannon KW, Lefkowitz SM, Ziman M, Schelter JM, Meyer MR, Kobayashi S, Davis C, Dai H, He YD, Stephaniants SB, Cavet G, Walker WL, West A, Coffey E, Shoemaker DD, Stoughton R, Blanchard AP, Friend SH, Linsley PS (2001) Expression profiling using microarrays fabricated by an ink-jet oligonucleotide synthesizer. Nat Biotechnol\n19: 342–34711283592\n8. Ikeda H, Lethe B, Lehmann F, van Baren N, Baurain JF, de Smet C, Chambost H, Vitale M, Moretta A, Boon T, Coulie PG (1997) Characterization of an antigen that is recognized on a melanoma showing partial HLA loss by CTL expressing an NK inhibitory receptor. Immunity\n6: 199–2089047241\n9. Li CM, Guo M, Borczuk A, Powell CA, Wei M, Thaker HM, Friedman R, Klein U, Tycko B (2002) Gene expression in Wilms' tumor mimics the earliest committed stage in the metanephric mesenchymal–epithelial transition. Am J Pathol\n160: 2181–219012057921\n10. Neumann E, Engelsberg A, Decker J, Storkel S, Jaeger E, Huber C, Seliger B (1998) Heterogeneous expression of the tumor-associated antigens RAGE-1, PRAME, and glycoprotein 75 in human renal cell carcinoma: candidates for T-cell-based immunotherapies? Cancer Res\n58: 4090–40959751617\n11. Oberthuer A, Hero B, Spitz R, Berthold F, Fischer M (2004) The tumor-associated antigen PRAME is universally expressed in high-stage neuroblastoma and associated with poor outcome. Clin Cancer Res\n10: 4307–431315240516\n12. Paydas S, Tanriverdi K, Yavuz S, Disel U, Baslamisli F, Burgut R (2005) PRAME mRNA levels in cases with acute leukemia: clinical importance and future prospects. Am J Hematol\n79: 257–26116044453\n13. Radich JP, Dai H, Mao M, Oehler V, Schelter J, Druker B, Sawyers C, Shah N, Stock W, Willman CL, Friend S, Linsley PS (2006) Gene expression changes associated with progression and response in chronic myeloid leukemia. Proc Natl Acad Sci USA\n103: 2794–279916477019\n14. Steinbach D, Hermann J, Viehmann S, Zintl F, Gruhn B (2002a) Clinical implications of PRAME gene expression in childhood acute myeloid leukemia. Cancer Genet Cytogenet\n133: 118–12311943337\n15. Steinbach D, Viehmann S, Zintl F, Gruhn B (2002b) PRAME gene expression in childhood acute lymphoblastic leukemia. Cancer Genet Cytogenet\n138: 89–9112419593\n16. van ‘t Veer LJ, Dai H, van de Vijver MJ, He YD, Hart AA, Mao M, Peterse HL, van der Kooy K, Marton MJ, Witteveen AT, Schreiber GJ, Kerkhoven RM, Roberts C, Linsley PS, Bernards R, Friend SH (2002) Gene expression profiling predicts clinical outcome of breast cancer. Nature\n415: 530–53611823860\n17. van de Vijver MJ, He YD, van't Veer LJ, Dai H, Hart AA, Voskuil DW, Schreiber GJ, Peterse JL, Roberts C, Marton MJ, Parrish M, Atsma D, Witteveen A, Glas A, Delahaye L, van der Velde T, Bartelink H, Rodenhuis S, Rutgers ET, Friend SH, Bernards R (2002) A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med\n347: 1999–200912490681\n18. Willenbrock K, Kuppers R, Renne C, Brune V, Eckerle S, Weidmann E, Brauninger A, Hansmann ML (2006) Common features and differences in the transcriptome of large cell anaplastic lymphoma and classical Hodgkin's lymphoma. Haematologica\n91: 596–60416670065"
+}

batch_0/PMC2527803.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527803",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527803\nAUTHORS: S Baka, R Califano, R Ferraldeschi, L Aschroft, N Thatcher, P Taylor, C Faivre-Finn, F Blackhall, P Lorigan\n\nABSTRACT:\nThis randomised trial compared platinum-based to anthracycline-based chemotherapy in patients with small-cell lung cancer (limited or extensive stage) and ⩽2 adverse prognostic factors. Patients were randomised to receive six cycles of either ACE (doxorubicin 50 mg/m2 i.v., cyclophosphamide 1 g/m2 i.v. and etoposide 120 mg/m2 i.v. on day 1, then etoposide 240 mg/m2 orally for 2 days) or PE (cisplatin 80 mg/m2 and etoposide 120 mg/m2 i.v. on day 1, then etoposide 240 mg/m2 orally for 2 days) given for every 3 weeks. For patients where cisplatin was not suitable, carboplatin (AUC6) was substituted. A total of 280 patients were included (139 ACE, 141 PE). The response rates were 72% for ACE and 77% for PE. One-year survival rates were 34 and 38% (P=0.497), respectively and 2-year survival was the same (12%) for both arms. For LD patients, the median survival was 10.9 months for ACE and 12.6 months for PE (P=0.51); for ED patients median survival was 8.3 months and 7.5 months, respectively. More grades 3 and 4 neutropenia (90 vs 57%, P<0.005) and grades 3 and 4 infections (73 vs 29%, P<0.005) occurred with ACE, resulting in more days of hospitalisation and greater i.v. antibiotic use. ACE was associated with a higher risk of neutropenic sepsis than PE and with a trend towards worse outcome in patients with LD, and should not be studied further in this group of patients.\n\nBODY:\nThe last decades have seen considerable efforts to improve the outcome for patients with small-cell lung cancer (SCLC) but progress has been slow (Govindan et al, 2006). Small-cell lung cancer is highly sensitive to chemotherapy and combination regimens have been the cornerstone of treatment since the 1970s. Characterisation of SCLC into limited (LD) and extensive disease (ED) as proposed originally by the Veterans Administrator Lung Cancer Group and revised by the International Association for the Study of Lung Cancer (IASCL) has been the basis of treatment choice for a number of years (Mountain, 1986; Zelen, 1973). A number of other independent prognostic factors including performance status (PS) and biochemical parameters (eg serum sodium, alkaline phosphatase and serum lactate dehydrogenase) have been identified and prognostic scores using these variables can reliably identify patients with good, intermediate or poor outcome (Buccheri and Ferrigno, 2004; Cerny et al, 1987; Sagman et al, 1991; Thatcher et al, 1995).When the current trial was designed in 1999, trials comparing platinum-based with anthracycline-based chemotherapy in SCLC were ongoing and the importance of concurrent chemoradiotherapy was just beginning to be understood. The ACE combination was still widely used in Europe (Thatcher et al, 2000; Sambrook and Girling, 2001) and was a reference regimen for the European Organisation for Research and Treatment of Cancer (EORTC) Lung Group (Ardizzoni et al, 2002; Giaccone et al, 1993; Postmus et al, 1996). Median survivals of 9–11 months with 1-year survival of 30–40% were reported in trials of both LD and ED patients with good PS (Giaccone et al, 1993; Thatcher et al, 2000; Urban et al, 1999). However, treatment with ACE was associated with significant neutropenia that contributed to infection-related morbidity and mortality (Bunn et al, 1986; Postmus et al, 1996; Thatcher et al, 2000). Cisplatin and etoposide (PE) were widely used in North America, with similar survival rates to those reported for cyclophosphamide, doxorubicin, vincristine (CAV) (Fukuoka et al, 1991; Roth et al, 1992). Carboplatin was shown to be active in the treatment of SCLC, and one randomised study comparing carboplatin and etoposide with PE showed no difference in survival, though the study was not powered for equivalence (Kosmidis et al, 1994).When this clinical trial was designed in 1999, there were no published data from studies comparing ACE with a platinum/etoposide combination in patients with better-prognosis disease.MethodsThe study design was a randomised phase III comparison of ACE with platinum/etoposide chemotherapy as first-line therapy in patients with better-prognosis SCLC.Eligibility criteriaPreviously untreated patients with histologically or cytologically proven SCLC and a maximum of two adverse prognostic factors (extensive stage disease, PS ⩾2, raised LDH, serum sodium <130 μmol l−1, Alk Phos >1.25 ULN) were eligible. Other eligibility criteria included age ⩾18 years, normal blood count, serum bilirubin <35 μmol l−1 and creatinine clearance >50 ml min−1. In patients with impaired renal function, that is, creatinine clearance >30 ml min−1 but <50 ml min−1, and/or patients with significant cardiovascular disease, carboplatin could be substituted for cisplatin in the first or subsequent cycles.A CT brain scan was not routinely performed, but patients with known brain metastases were not eligible.The study had ethical and local approval and was covered by a DDX, later updated to a CTA after the introduction of EU Regulations. Patients gave their written informed consent. The Trial Management Committee consisted of the PI, the Co-Investigator, the lead research sister and the data manager. Adverse events were discussed by the TMC and those defined as serious and unexpected events were reported to the ethics committee.After the publication in 2004 of two studies suggesting a survival benefit for platinum/etoposide in LD patients (Sundstrom et al, 2002; Thatcher et al, 2005), accrual of the final 12 patients required was limited to patients with extensive stage disease.Treatment and monitoringPatients were randomised to receive six cycles of ACE (doxorubicin 50 mg/m2 i.v., cyclophosphamide 1 g/m2 i.v. and etoposide 120 mg/m2 i.v. on day 1, followed by etoposide 240 mg/m2 orally for 2 days) for 3 weeks or six cycles of PE (cisplatin 80 mg/m2 and etoposide 120 mg/m2 i.v. on day 1, followed by etoposide 240 mg/m2 orally for 2 days every 3 weeks). For patients where cisplatin was not suitable, carboplatin was substituted at an AUC of 6, calculated according to the Calvert formula (ie, carboplatin dose=target AUC of 6 (glomerular filtration rate+25 mg), where glomerular filtration rate was based on EDTA or measured creatinine clearance).Chemotherapy was given if the total WBC was ⩾3000 μ/l, neutrophils ⩾1500 μ/l, platelets ⩾100 000 μ/l and creatinine clearance ⩾30 ml min−1, and there was no evidence of severe toxicity. If these conditions were not fulfilled, treatment was delayed and the blood count was repeated at intervals of not more than 1 week; treatment was given at full dose as soon as the above conditions were met. Dose reduction was not recommended. The use of GCSF as secondary prophylaxis was at the discretion of the investigator.Thoracic radiotherapy was given to patients with limited stage disease achieving a complete or partial response to chemotherapy, beginning 3 weeks after the last cycle of chemotherapy (30 Gy in 10 daily fractions). Patients with ED SCLC received thoracic irradiation only if they had thoracic symptoms amenable to palliation with radiotherapy after completion of chemotherapy. Prophylactic cranial irradiation was considered for all LD patients achieving a complete response; suitable patients received 25 Gy in 10 daily fractions after completion of chemotherapy.Tumour stage was assessed with CT scan of thorax and abdomen. Disease measurement was performed within 4 weeks before the start of treatment. During chemotherapy, patients were assessed on days 1 and 15 with physical examination, and weekly with blood count, biochemistry and WHO Performance Status. A chest X-ray (CXR) was carried out after every second cycle of treatment, but assessment of response was made according to the WHO criteria by CT scanning at the end of chemotherapy unless progressive disease was detected in the interim by CXR. Toxicities were graded according to the National Cancer Institute Common Toxicity Grading Criteria version December 1994 (revised).Statistical designThe study design was a randomised phase III comparison of ACE with platinum/etoposide as first-line therapy for patients with SCLC and a maximum of two adverse prognostic factors. The primary end point was 1-year survival. Secondary end points were 2-year survival, median survival, response rate and toxicity. Survival was calculated from the date of randomisation to the date of death from any cause. Time to progression was taken from the date of randomisation to the date of the progression. A 1-year survival of 40% had been reported for ACE in a recent MRC study (Thatcher et al, 2000). The North American experience suggested a 1-year survival rate of approximately 60% for a platinum-based combination, possibly in a more favourable patient group (Evans et al, 1987). Two hundred eighty patents were required to detect a survival difference of 20% (from 40 to 60%) at 1-year, with 90% power and a two-sided significance level of 5%. Patients were randomized on a 1 : 1 basis to one of two treatment arms. The allocation method was stochastic minimisation as implemented in a bespoke computer application at the randomisation centre. The only factor controlled for in the allocation was centre.ResultsPatient's characteristicsBetween April 1999 and February 2005, 280 patients (ACE=139, PE=141) were randomised at two centres in the UK. The two arms were well balanced for age, stage, gender, PS and prognostic score (Table 1). All patients were included in the survival analysis on an intention to treat basis. Two patients were ineligible because of incorrect histological diagnosis (non-SCLC), one for each arm. A further seven patients were not assessable for response, three in the ACE arm (one patient received PE while was waiting for investigations, one died before cycle 1 and one needed radiotherapy following first cycle) and four patients in the PE arm (one died before cycle 1, one stopped the treatment after first cycle because of toxicity and two lost to follow-up after first cycle) (Figure 1).Treatment receivedA total of 584 cycles of ACE and 696 of PE were administered (P=0.001) (Table 2). Fifty-two (37%) of the 139 patients randomised to ACE and 91 (65%) of patients randomised to platinum/etoposide completed all six cycles (P=0.01). The main reasons for early discontinuation were disease progression (ACE 19%, PE 12%) and toxicity (ACE 30%, PE 12%). Thirty-eight per cent of cycles of ACE were delayed or discontinued due to toxicity compared with 30% for PE (P=0.048). Six patients in the PE arm received carboplatin rather than cisplatin on cycle 1, and carboplatin was substituted for cisplatin during the treatment in a further four patients; one patient changed from PE to single-agent carboplatin.In patients with limited stage disease, 58% patients in the ACE arm and 74% of patients in the PE arm received consolidation thoracic radiotherapy (P=0.04) and 23% of ACE patients and 33% of PE patients received PCI (P=0.24). A further 20 (14%) patients in the ACE arm and 12 (8.5%) in the PE arm received palliative radiotherapy. At disease progression, radiotherapy was given to 31 (22%) patients in the ACE arm and to 19 (13%) patients in the PE arm. Twenty-five (18%) and 21 (15%) of patients received second-line chemotherapy in the ACE arm and PE arm, respectively.ToxicityClinically significant toxicity was more common in patients receiving ACE (Table 3). Grades 3 and 4 anaemia occurred in 37 (27%) patients on the ACE arm and 25 (18%) patients on the PE arm (P=0.038). Grades 3 and 4 neutropenia occurred in 123 (90%) of patients receiving ACE and 78 (57%) of patients receiving PE (P<0.005). Neutropenia was associated with a high incidence of grades 3 and 4 infections, 73% of patients receiving ACE and 29% receiving PE arm (P<0.005). Eighty-two percent of patients in the ACE arm required i.v. antibiotics for one or more days during their treatment compared with 37% of PE patients. Hospitalisation for severe neutropenia and infections was less frequent with PE compared with ACE. The total number of days of hospitalisation for those treated with ACE was 1390 compared with 360 for PE (P<0.005). The median number of days of hospitalization was 9 (0–45) for ACE and 0 (0–18) for PE. Non-haematological toxicity was similar in both arms, but patients receiving PE experienced more grades 2 and 3 nausea.Response and survivalThe response rate was higher for PE patients, with a higher number of complete responders (Table 4). There was no significant difference in overall, median, 1-year or 2-year survival between the two treatment arms (Table 5 and Figure 2). For the LD group, the median survival was 10.9 months for ACE and 12.6 months for PE (P=0.58), with an actual 1-year survival rate of 44 and 54%, respectively (P=0.2). For the ED group, median survival was 8.3 months for ACE and 7.5 months for PE, with an actual 1-year survival rate of 17 and 15%, respectively. (P=0.9).DiscussionAlthough PE is now considered the treatment of choice for fitter patients with SCLC, this was not the case when the current trial was being designed. Several prospective studies had compared PE with anthracycline-based therapy, but most failed to demonstrate superiority of PE (Evans et al, 1987; Fukuoka et al, 1991; Roth et al, 1992). Subsequently, an overview of US National Cancer Institute sponsored trials for ED patients conducted between 1972 and 1990 demonstrated that cisplatin-based regimens were associated with an improved median survival (Chute et al, 1999) and a meta-analysis of 19 trials published between 1981 and 1999 showed a small survival benefit for patients receiving cisplatin-based chemotherapy (Pujol et al, 2001). Etoposide-containing regimens, with or without cisplatin, were also shown to be associated with a significant survival benefit (Mascaux et al, 2000).After this study had begun accrual, a trial from Norway comparing PE with cyclophosphamide, epirubicin, etoposide (CEV) reported a survival benefit for PE (Sundstrom et al, 2002). The median survival was 7.8 months for CEV and 10.2 months for PE, with a 2-year survival of 6 and 14%, respectively (P=0.0004). The advantage was confined to patients with LD, with a median survival of 14.5 months for PE and 9.7 months for CEV (P=0.001), and a 2-year survival of 25 and 10% (P=0.0001), respectively. There was a trend in the same direction for patients with ED, but this was not statistically significant. Although the proportions of patients with extensive stage disease and with PS2 were similar in both studies, the choice and doses of drugs are not directly comparable. The comparator arm in the Norwegian study used a lower dose of anthracycline (epirubicin 50 mg/m2\nvs doxorubicin 50 mg/m2) and used vincristine rather than etoposide. The median survival for PE was similar in both studies (10.6 vs 10.2 months). The different survival seen for the two-comparator arms (7.8 months for CEV vs 9.6 months for ACE) may explain in part the statistically significant survival advantage seen for PE in the Sundstrom study. The survival for ACE in this study is marginally lower than that reported for two studies from the UK Medical Research Council, but these studies had fewer ED patients (Thatcher et al, 2000; Thatcher et al, 2005).The median and 2-year survivals for PE in LD patients in our study were lower than those reported by Sundstrom (12.6 vs 14.5 months, 21 vs 25%, respectively). This may in part reflect lower use of radiotherapy, which has an established role in consolidating the response of the primary tumour to chemotherapy and in reducing the risk of brain metastases as a site of recurrence. (Pignon et al, 1992; Warde and Payne, 1992). Cisplatin plays a central role in concurrent treatment, because of its radiosensitising effect, and there is increasing evidence of a survival benefit for patients receiving early, concurrent, cisplatin-based chemoradiotherapy with 5-year survival ranging between 20 and 25% (Fried et al, 2004; Pijls-Johannesma et al, 2005). In this study, thoracic radiotherapy was given after the completion of chemotherapy. The rate or TRT was lower for patients receiving ACE (58%) than those receiving PE (74%). The reason for this is unclear, but is likely to reflect the worse toxicity seen with ACE, and is similar to the 51–54% reported in comparable studies (Thatcher et al, 2000; Ardizzoni et al, 2002). Higher TRT rates have also been reported for other platinum combinations (Lorigan et al, 2005; Thatcher et al, 2005). The TRT rate in the Sundstrom study was higher in both arms (83 and 88%). Prophylactic cranial irradiation has been shown to be associated with a survival advantage in LD and ED patients responding to chemotherapy (Auperin et al, 1999) (Slotman et al, 2007). In this study, 33% of LD patients receiving PE and 23% of those receiving ACE went on to have PCI. The use of PCI was comparable with the Sundstrom study, but lower than that reported in studies using VICE (35–53%) (Lorigan et al, 2005; Thatcher et al, 2005).We observed that haematological toxicity and the risk of infection were significantly higher for ACE than for PE. This difference was clinically relevant, with a higher rate of grades 3 and 4 infection, higher number of days in hospital and higher rate of treatment discontinuation for ACE. The incidence of grades 3 and 4 infection in this study was 73% for ACE, and 29% for PE. Intravenous antibiotics were used in 82% of patients receiving ACE and 37% of patients receiving PE. The infection rates seen in both arms of this study were substantially higher than the 14–15% reported for ACE and 16% for VICE – toxicity data were not reported for the Sundstrom study. (Thatcher et al, 2005; Thatcher et al, 2000) However, the toxic death rate (1%) is lower than the 4–10% reported in other comparable studies (Thatcher et al, 2000; Pujol et al, 2001; Ardizzoni et al, 2002). The likely explanation for the high-recorded infection rate but low toxic death rate in both arms was that patients were seen weekly and clinical teams had a low threshold for responding to neutropenia. Other factors (low use of GCSF, lack of use of prophylactic antibiotics, use of oral etoposide, higher proportion of ED patients) may also have contributed to the proportionally higher infection rate seen in both arms.The study was initially designed to allow inclusion of patients not fit for cisplatin chemotherapy, allowing these to be treated with carboplatin. We were surprised by the low number of patients (six) who commenced treatment with carboplatin. A further four patients changed to carboplatin during treatment. We have included these patients with cisplatin for both efficacy and toxicity analysis.Further advances in the treatment of patients with SCLC are most likely to come with optimum use of concurrent chemotherapy and radiotherapy, and the identification of new drugs and targets. Until then, the combination of cisplatin and etoposide is standard therapy for patients with SCLC and good PS, and further studies of traditional anthracycline-based regimens are not warranted.\n\nREFERENCES:\n1. Ardizzoni A, Tjan-Heijnen VC, Postmus PE, Buchholz E, Biesma B, Karnicka-Mlodkowska H, Dziadziuszko R, Burghouts J, Van Meerbeeck JP, Gans S, Legrand C, Debruyne C, Giaccone G, Manegold C (2002) Standard vs intensified chemotherapy with granulocyte colony-stimulating factor support in small-cell lung cancer: a prospective European Organization for Research and Treatment of Cancer-Lung Cancer Group Phase III Trial-08923. J Clin Oncol\n20: 3947–395512351591\n2. Auperin A, Arriagada R, Pignon JP, Le Pechoux C, Gregor A, Stephens RJ, Kristjansen PE, Johnson BE, Ueoka H, Wagner H, Aisner J (1999) Prophylactic cranial irradiation for patients with small-cell lung cancer in complete remission. Prophylactic Cranial Irradiation Overview Collaborative Group. N Engl J Med\n341: 476–48410441603\n3. Buccheri G, Ferrigno D (2004) Prognostic factors of small cell lung cancer. Hematol Oncol Clin North Am\n18: 445–46015094181\n4. Bunn Jr PA, Greco FA, Einhorn L (1986) Cyclophosphamide, doxorubicin, and etoposide as first-line therapy in the treatment of small-cell lung cancer. Semin Oncol\n13: 45–533020700\n5. Cerny T, Blair V, Anderson H, Bramwell V, Thatcher N (1987) Pretreatment prognostic factors and scoring system in 407 small-cell lung cancer patients. Int J Cancer\n39: 146–1493026969\n6. Chute JP, Chen T, Feigal E, Simon R, Johnson BE (1999) Twenty years of phase III trials for patients with extensive-stage small-cell lung cancer: perceptible progress. J Clin Oncol\n17: 1794–180110561217\n7. Evans WK, Feld R, Murray N, Willan A, Coy P, Osoba D, Shepherd FA, Clark DA, Levitt M, MacDonald A (1987) Superiority of alternating non-cross-resistant chemotherapy in extensive small cell lung cancer. A multicenter, randomized clinical trial by the National Cancer Institute of Canada. Ann Intern Med\n107: 451–4582820289\n8. Fried DB, Morris DE, Poole C, Rosenman JG, Halle JS, Detterbeck FC, Hensing TA, Socinski MA (2004) Systematic review evaluating the timing of thoracic radiation therapy in combined modality therapy for limited-stage small-cell lung cancer. J Clin Oncol\n22: 4837–484515570087\n9. Fukuoka M, Furuse K, Saijo N, Nishiwaki Y, Ikegami H, Tamura T, Shimoyama M, Suemasu K (1991) Randomized trial of cyclophosphamide, doxorubicin, and vincristine vs cisplatin and etoposide vs alternation of these regimens in small-cell lung cancer. J Natl Cancer Inst\n83: 855–8611648142\n10. Giaccone G, Dalesio O, McVie GJ, Kirkpatrick A, Postmus PE, Burghouts JT, Bakker W, Koolen MG, Vendrik CP, Roozendaal KJ (1993) Maintenance chemotherapy in small-cell lung cancer: long-term results of a randomized trial. European Organization for Research and Treatment of Cancer Lung Cancer Cooperative Group. J Clin Oncol\n11: 1230–12408391065\n11. Govindan R, Page N, Morgensztern D, Read W, Tierney R, Vlahiotis A, Spitznagel EL, Piccirillo J (2006) Changing epidemiology of small-cell lung cancer in the United States over the last 30 years: analysis of the surveillance, epidemiologic, and end results database. J Clin Oncol\n24: 4539–454417008692\n12. Kosmidis PA, Samantas E, Fountzilas G, Pavlidis N, Apostolopoulou F, Skarlos D (1994) Cisplatin/etoposide vs carboplatin/etoposide chemotherapy and irradiation in small cell lung cancer: a randomized phase III study. Hellenic Cooperative Oncology Group for Lung Cancer Trials. Semin Oncol\n21: 23–30\n13. Lorigan P, Woll PJ, O′Brien ME, Ashcroft LF, Sampson MR, Thatcher N (2005) Randomized phase III trial of dose-dense chemotherapy supported by whole-blood hematopoietic progenitors in better-prognosis small-cell lung cancer. J Natl Cancer Inst\n97: 666–67415870437\n14. Mascaux C, Paesmans M, Berghmans T, Branle F, Lafitte JJ, Lemaitre F, Meert AP, Vermylen P, Sculier JP (2000) A systematic review of the role of etoposide and cisplatin in the chemotherapy of small cell lung cancer with methodology assessment and meta-analysis. Lung Cancer\n30: 23–3611008007\n15. Mountain CF (1986) A new international staging system for lung cancer. Chest\n89: 225S–233S3514171\n16. Pignon JP, Arriagada R, Ihde DC, Johnson DH, Perry MC, Souhami RL, Brodin O, Joss RA, Kies MS, Lebeau B, Onoshi T, Osterlind K, Tattersall M, Wagner H (1992) A meta-analysis of thoracic radiotherapy for small-cell lung cancer. N Engl J Med\n327: 1618–16241331787\n17. Pijls-Johannesma MC, De Ruysscher D, Lambin P, Rutten I, Vansteenkiste JF (2005) Early vs late chest radiotherapy for limited stage small cell lung cancer. Cochrane Database Syst Rev, doi: 10.1002/14651858.CD004700.pub2\n18. Postmus PE, Scagliotti G, Groen HJ, Gozzelino F, Burghouts JT, Curran D, Sahmoud T, Kirkpatrick A, Giaccone G, Splinter TA (1996) Standard vs alternating non-cross-resistant chemotherapy in extensive small cell lung cancer: an EORTC Phase III trial. Eur J Cancer\n32A: 1498–15038911108\n19. Pujol JL, Daures JP, Riviere A, Quoix E, Westeel V, Quantin X, Breton JL, Lemarie E, Poudenx M, Milleron B, Moro D, Debieuvre D, Le Chevalier T (2001) Etoposide plus cisplatin with or without the combination of 4′-epidoxorubicin plus cyclophosphamide in treatment of extensive small-cell lung cancer: a French Federation of Cancer Institutes multicenter phase III randomized study. J Natl Cancer Inst\n93: 300–30811181777\n20. Roth BJ, Johnson DH, Einhorn LH, Schacter LP, Cherng NC, Cohen HJ, Crawford J, Randolph JA, Goodlow JL, Broun GO (1992) Randomized study of cyclophosphamide, doxorubicin, and vincristine vs etoposide and cisplatin vs alternation of these two regimens in extensive small-cell lung cancer: a phase III trial of the Southeastern Cancer Study Group. J Clin Oncol\n10: 282–2911310103\n21. Sagman U, Feld R, Evans WK, Warr D, Shepherd FA, Payne D, Pringle J, Yeoh J, DeBoer G, Malkin A (1991) The prognostic significance of pretreatment serum lactate dehydrogenase in patients with small-cell lung cancer. J Clin Oncol\n9: 954–9611851821\n22. Sambrook RJ, Girling DJ (2001) A national survey of the chemotherapy regimens used to treat small cell lung cancer (SCLC) in the United Kingdom. Br J Cancer\n84: 1447–145211384091\n23. Slotman B, Faivre-Finn C, Kramer G, Rankin E, Snee M, Hatton M, de Schaetzen G, Collette L, Senan S (2007) A randomized trial of prophylactic cranial irradiation (PCI) vs no PCI in extensive disease small cell lung cancer after a response to chemotherapy. ASCO Annual Meeting Proceedings\n25: 4\n24. Sundstrom S, Bremnes RM, Kaasa S, Aasebo U, Hatlevoll R, Dahle R, Boye N, Wang M, Vigander T, Vilsvik J, Skovlund E, Hannisdal E, Aamdal S (2002) Cisplatin and etoposide regimen is superior to cyclophosphamide, epirubicin, and vincristine regimen in small-cell lung cancer: results from a randomized phase III trial with 5 years' follow-up. J Clin Oncol\n20: 4665–467212488411\n25. Thatcher N, Anderson H, Burt P, Stout R (1995) The Value of Anatomic Staging and Other Prognostic Factors in Small Cell Lung Cancer Management: A View of European Studies. Semin Radiat Oncol\n5: 19–2610717121\n26. Thatcher N, Girling DJ, Hopwood P, Sambrook RJ, Qian W, Stephens RJ (2000) Improving survival without reducing quality of life in small-cell lung cancer patients by increasing the dose-intensity of chemotherapy with granulocyte colony-stimulating factor support: results of a British Medical Research Council Multicenter Randomized Trial. Medical Research Council Lung Cancer Working Party. J Clin Oncol\n18: 395–40410637255\n27. Thatcher N, Qian W, Clark PI, Hopwood P, Sambrook RJ, Owens R, Stephens RJ, Girling DJ (2005) Ifosfamide, carboplatin, and etoposide with midcycle vincristine vs standard chemotherapy in patients with small-cell lung cancer and good performance status: clinical and quality-of-life results of the British Medical Research Council multicenter randomized LU21 trial. J Clin Oncol\n23: 8371–837916293867\n28. Urban T, Chastang C, Lebas FX, Duhamel JP, Adam G, Darse J, Brechot JM, Lebeau B (1999) The addition of cisplatin to cyclophosphamide-doxorubicin-etoposide combination chemotherapy in the treatment of patients with small cell lung carcinoma: A randomized study of 457 patients. ‘Petites Cellules’ Group. Cancer\n86: 2238–224510590363\n29. Warde P, Payne D (1992) Does thoracic irradiation improve survival and local control in limited-stage small-cell carcinoma of the lung? A meta-analysis. J Clin Oncol\n10: 890–8951316951\n30. Zelen M (1973) Keynote address on biostatistics and data retrieval. Cancer Chemother Rep 3\n4: 31–42"
+}

batch_0/PMC2527804.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527804",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527804\nAUTHORS: U Güth, D J Huang, A Schötzau, R Zanetti-Dällenbach, W Holzgreve, J Bitzer, E Wight\n\nABSTRACT:\nPrevious research evaluating the use of adjuvant endocrine therapy among postmenopausal breast cancer patients showed with 15–50% wide ranges of non-adherence rates. We evaluated this issue by analysing an unselected study group comprising of 325 postmenopausal women, diagnosed from 1997 to 2003 with hormonal receptor-positive invasive breast cancer. The different clinical situations that led to the discontinuation of adjuvant endocrine therapy were clearly defined and differentiated: non-adherence was not simply the act of stopping medication, but rather the manifestation of an intentional behaviour of the patient. Of the 287 patients who initiated endocrine therapy, 191 (66.6%) fully completed this treatment. Thirty-one patients (10.8%) showed non-adherence to therapy. Patients who had follow-up with a general practitioner, rather than in an oncologic unit, were more likely to be non-adherent (P=0.0088). Of 25 patients who changed medication due to therapy-related adverse effects, 20 (80%) patients fully completed the therapy after drug change. In adjuvant endocrine therapy, a lowering of the non-adherence rate to 10.8%, the lowest reported in the literature, is realistic when patients are cared for by a specialised oncologic unit focusing on the individual needs of the patients.\n\nBODY:\nSince the late 1990s, adjuvant systemic endocrine therapy was recommended for the vast majority of postmenopausal patients with hormonal receptor (HR)-positive breast carcinomas. The sixth St Gallen International Consensus Panel on the Treatment of Primary Breast Cancer refined these guidelines in 1998: with the exception of low-risk, node-negative patients, the panel recommended a full 5 years of tamoxifen therapy for all elderly women with HR-positive breast cancer (Goldhirsch et al, 1998). More recently, clinical studies demonstrated additional benefit with aromatase inhibitors in postmenopausal women, either as initial management or following a period of tamoxifen use (Coombes et al, 2004; Coates et al, 2007; Forbes et al, 2008).In long-term therapy, however, there are a multitude of factors and clinical situations that prevent and threaten the completion of the targeted 5-year treatment. Patient refusal to initiate the recommended antihormonal medication and non-adherence to therapy (defined as a composite of compliance and persistence: the patient started therapy, but discontinued the planned treatment) are important, but are not the only contributors. Both factors were evaluated in several studies, in clinical trials (Fisher et al, 1989; Coombes et al, 2004; Coates et al, 2007; Forbes et al, 2008) and in clinical practise settings (Waterhouse et al, 1993; Demissie et al, 2001; Silliman et al, 2002; Partridge et al, 2003, 2008; Fink et al, 2004; Grunfeld et al, 2005; Atkins and Fallowfield, 2006; Lash et al, 2006; Barron et al, 2007; Owusu et al, 2008). These studies, however, evaluated only selected populations and failed to describe all the possible situations that determine compliance and adherence to therapy in everyday clinical settings.Our study depicts the entire picture of the adjuvant treatment setting in postmenopausal women. This approach was made possible by considering the entire group of all postmenopausal patients with non-metastatic breast cancer (diagnosed and surgically treated in a 7-year period from 1997 to 2003), whose carcinomas were HR–positive, and therefore were candidates for adjuvant endocrine therapy. We define and evaluate the diverse clinical situations that hinder the realisation of the target of a complete 5-year therapy.Patients and methodsData regarding all postmenopausal non-metastatic breast cancer patients, who received surgical therapy between 1997 and 2003 at the Department of Gynaecology and Obstetrics of the University Hospital Basel (Basel, Switzerland), form the basis of the current analysis. Of these 393 patients, 325 had HR-positive carcinomas (82.7%) and 68 had HR-negative tumours (17.3%). The 325 patients with HR-positive carcinomas were evaluated in this study. We had complete follow-up in 323 of these patients, whereas two patients were lost to follow-up, one after 3 months and the other after 12 months.The following data were collected from the medical records and were available for all patients: age at initial diagnosis, tumour stage according to the American Joint Committee on Cancer (AJCC)/International Union Against Cancer (UICC) TNM Classification, histologic subtype, grading, oestrogen receptor (ER) and progesterone receptor (PR) status, surgery type, and receipt of adjuvant chemotherapy and/or radiation. HER-2/neu status was available for 273 patients (84.0%).The treatment recommendations for all patients were based on the decision of the interdisciplinary tumour board of the University Hospital Basel. Since 1997, all HR-positive patients, with few exceptions, were recommended to have adjuvant endocrine therapy. All patients received a comprehensive consultation at the departmental oncology unit during which the treatment indication and duration, as well as the potential adverse effects, were extensively discussed.The prescribed antihormonal agent given was recorded. To systematically evaluate the different clinical situations during the course of adjuvant endocrine therapy, we created the following subdivisions:\n(A) Patients who did not initiate therapy\nThis subgroup includes the patients to whom endocrine therapy was not recommended, and the patients who refused the recommended therapy and never began treatment.\n(B) Patients who initiated therapy\nThis subgroup includes the patients who completed the therapy after a 5-year course (for the patients diagnosed in 2003, we had a follow-up for at least 4 years). Patients with extended therapy >5 years were also considered as having fully completed therapy. Further subgroups must also be considered as having completed therapy, although treatment was not administered during the targeted 5 years: discontinuation due to death, as well as discontinuation due to breast cancer recurrence (local and/or distant metastases) and serious medical reasons independent from breast cancer disease and therapy-related adverse effects. We distinguished the above-mentioned subgroups from non-adherent patients who discontinued the planned mode of treatment and refused to continue further endocrine therapy.Furthermore, we recorded any change of endocrine agents and the indication for the change: sequential therapy (switching after 2–3 years of adjuvant tamoxifen therapy to an aromatase inhibitor), extended therapy beyond 5 years of adjuvant therapy or change due to adverse effects. Lastly, we recorded the location of treatment and follow-up of the patients (our own oncology unit, external oncology unit or general practitioner).Information concerning type and length of the medication, as well as the reasons for discontinuation, was retrospectively obtained from the medical record. Patients who had no follow-up at our institution were contacted via telephone. Afterwards, contact was made with the treating physician to confirm the patients' statements.The study design and data collection methods were approved by the Institutional Review Board.Statistical analysisTo identify factors associated with treatment refusal and non-adherence, we created two univariate logistic regression models. Each model included the independent variables: year of the initial diagnosis, patient's age, primary surgical therapy, tumour stage, receipt of previous chemotherapy and/or postoperative radiation, and location of follow-up (the latter only in the non-adherence model). Odds ratios (ORs) with corresponding 95% confidence intervals (CIs) were reported; a P-value less than 0.05 was considered significant. Statistical analyses were performed with R Development Core Team Software (version 2.5.0, Vienna, Austria).ResultsThe clinicopathologic, treatment and follow-up characteristics of the 325 patients in the study are summarised in Tables 1, 2 and 3.(A) Patients who did not initiate endocrine therapy (n=38)In 10 cases, despite HR-positive carcinoma, a recommendation for endocrine therapy was not made. The reasons for this included a low-risk constellation (pT1a/b N0, favourable grading) and/or advanced age with considerable comorbidity. Twenty-eight patients (8.9%) refused the recommended endocrine therapy after extensive counselling and never began this treatment. Univariate analysis revealed that older patients, those who were initially diagnosed at the beginning of the study period and those who did not receive adjuvant chemotherapy were more likely to refuse therapy (advanced age: P=0.019; study period: 1997–2000 vs 2001–2003: P=0.0070; no chemotherapy: P<0.0001;Table 4). Additionally, patients who had a favourable disease stage (stage I) tended to refuse the recommended therapy (P=0.0546).(B) Patients who initiated endocrine therapy (n=287)The initial endocrine agents prescribed are listed in Table 2. Forty-five patients were treated within the randomised, phase III, double-blind Breast International Group (BIG) 1-98 four-arm study (letrozole or tamoxifen as monotherapy or sequential therapy) (Coates et al, 2007). The medication was unblinded in 25 of these patients. Even when the study drug had been unblinded, these cases were listed under the category ‘study medication (BIG 1-98)’. Of the 45 BIG 1-98 study patients, 30 (66.7%) had completed 5 years of treatment according to the study guidelines. Of the 14 patients who prematurely discontinued their randomly assigned trial medication for reasons other than disease recurrence, 8 women fully completed the endocrine therapy over a 5-year course outside of the trial (57.3%).In 50 patients, there was a change of the antihormonal agent prescribed within the planned 5-year period. Of these, four patients changed their medication twice and one patient three times. Twenty-five of the 50 patients changed their medication due to adverse effects of the endocrine therapy (three changed the medication twice and one three times). Of these, 20 patients (80%) fully completed the therapy after drug change.The therapy course, reported in Tables 3 and 5, was as follows:Of the 287 patients who initiated endocrine therapy, 191 (66.6%) fully completed the targeted therapy (four patients interrupted therapy for up to 3 months, another for 5 months). Eleven patients (3.8%) discontinued the therapy due to death, whereas 43 patients (15.0%) ceased therapy due to breast cancer recurrence. In nine cases (3.1%), the therapy was discontinued by the physician due to serious medical reasons independent from breast cancer disease and therapy-related adverse effects (advanced age/dementia/need for nursing home care, n=5; incurable malignancy other than breast cancer, n=3; irreversible coma following severe head trauma, n=1).Non-adherence: Thirty-one patients (10.8%) chose independently to end the therapy. The main reasons for this given by the patients are listed in Table 5. In the seven variables used in our univariate analyses, location of follow-up was the only significant predictor, showing that patients who did not have follow-up in an oncologic unit, but rather with a general practitioner, were more likely to be non-adherent to therapy (P=0.0088; Table 4).Location of follow-up: Of the 289 patients who received adjuvant endocrine therapy, 215 (74.4%) were treated in an oncology unit; of these, 194 patients in the internal oncology unit (67.1%) and 21 (7.3%) in an external oncologic unit. Seventy-two patients (24.9) had further follow-up and treatment through a general practitioner. Two patients (0.7%) were lost to follow-up, and the location of the oncologic care was unknown.DiscussionAlthough several studies performed in clinical practise settings (Waterhouse et al, 1993; Demissie et al, 2001; Silliman et al, 2002; Partridge et al, 2003, 2008; Fink et al, 2004; Grunfeld et al, 2005; Atkins and Fallowfield, 2006; Lash et al, 2006; Barron et al, 2007; Owusu et al, 2008) and clinical trials (Fisher et al, 1989; Coombes et al, 2004; Coates et al, 2007; Forbes et al, 2008) have addressed non-adherence to endocrine therapy for breast cancer, the conclusions that can be drawn from the currently available evidence are significantly limited. The reported rates of non-adherent patients in adjuvant endocrine therapy of breast cancer range considerably from 15 to 50%. The great variability of the data can only be adequately interpreted when the basic methods of each study are closely analysed. Within the current literature, there exist major non-uniformities and shortcomings, which make useful insights for a current general population difficult. Some of the critical points include\nAll studies exclusively analysed selected study cohorts.Some studies used insurance claims and public health data (Partridge et al, 2003, 2008; Barron et al, 2007). However, analyses of public databases, which record patient data only to the extent that was necessary for administrative statistical purposes, are not always able to provide information concerning the often very individual clinical situations and reasons that lead to the cessation of medication.In some studies, it must be assumed that also patients who never took the medication, that is patients who received the prescription but refused therapy from beginning, were also considered (Partridge et al, 2003, 2008; Barron et al, 2007).Most clinical practise studies only analysed a predominantly geriatric study group (Silliman et al, 2002; Partridge et al, 2003; Fink et al, 2004; Lash et al, 2006; Barron et al, 2007; Owusu et al, 2008).The data are difficult to compare due to variable observation periods, which included 17 months (Silliman et al, 2002), over 3 (Demissie et al, 2001; Barron et al, 2007; Partridge et al, 2008), 4 (Partridge et al, 2003), and up to 5 (Fink et al, 2004; Lash et al, 2006; Owusu et al, 2008) years.Large part of the published studies are not applicable today, since initial diagnosis and the start of adjuvant treatment took place in the 1990s when the indication for therapy was usually limited to patients with stage II/III disease (Demissie et al, 2001; Silliman et al, 2002; Partridge et al, 2003; Fink et al, 2004; Lash et al, 2006; Owusu et al, 2008). In contrast, endocrine therapy is recommended for nearly all HR-positive breast cancer patients according to the most current guidelines (Goldhirsch et al, 2007). Only two studies have reported results from data obtained in a contemporary study period after 2001 (Barron et al, 2007; Partridge et al, 2008).It should be the goal of studies concerning the use of endocrine therapy to identify the rate of patient non-compliance and non-adherence and to evaluate the characteristics of these patients. A certain proportion of these patients may be potentially motivated to initiate and to maintain therapy by more intensive care and improved counselling. In our view, non-adherence is not simply the act of stopping medication, but rather the manifestation of an intentional behaviour. The reasons for non-adherence, such as distressing adverse effects, inadequate clarification of the benefits of therapy, and fear and mistrust of the agent prescribed, can be elucidated in most cases. An important aspect of non-adherence to treatment is the ability of physician to intervene and change the attitude that led to the discontinuation. To separately analyse the subgroup of non-adherent patients, the clinical situations that result in a discontinuation of medication must be clearly defined. Particularly, patients who refused the recommended therapy or were non-adherent (whose attitude and behaviour may be potentially influenced) and those whose therapy had to be stopped due to breast cancer recurrence or other serious medical reasons (that is, discontinuation of the therapy was unavoidable) must not be analysed in one collective study group.A realisation of this goal requires a careful and detailed clinical follow-up of the study group and a clear discrimination between the terms and reasons behind ‘discontinuation’ and ‘non-adherence’. The currently available studies from clinical practise settings have not been able to accomplish this. In most of these studies, patients who discontinued therapy, regardless of the reason for cessation, were categorised in a single group, and the terms ‘discontinuation’ and ‘non-adherence’ were used synonymously (Waterhouse et al, 1993; Silliman et al, 2002; Partridge et al, 2003, 2008; Fink et al, 2004; Grunfeld et al, 2005; Atkins and Fallowfield, 2006; Lash et al, 2006; Barron et al, 2007; Owusu et al, 2008).Large clinical trials, with their extensive case documentation requirements, also cannot provide exact information regarding adherence to endocrine therapy in the general setting, as the readiness to participate in a trial already creates a certain selection bias. In trials that compared tamoxifen with an aromatase inhibitor, the rates of non-adherence are clearly lower than those reported in clinical practise studies, and within those trials, the variabilities in rates were minor (tamoxifen: 11–13%, aromatase inhibitors: ∼12%) (Coombes et al, 2004; Coates et al, 2007; Forbes et al, 2008). The NSABP B-14 trial, conducted in the 1980s, revealed a non-adherence rate with tamoxifen of 16.8%; remarkably, non-adherence to placebo was also comparatively high (14.9%) (Fisher et al, 1989). Non-adherence data from clinical trials may not necessarily be transferable to general practise, as the withdrawal of study medication does not imply the stop of endocrine therapy. In our analysis, 8 out of 14 patients who stopped the study medication in the BIG 1-98 trial due to other reasons than breast cancer recurrence continued and completed endocrine therapy outside of the trial.The conclusions of all currently available studies can only give more or less reliable information regarding the use of endocrine therapy for postmenopausal breast cancer in clinical practise today. In our study, we avoided many of the above-mentioned methodological problems:\nWe formulated our study group following current guidelines, which indicate endocrine therapy based on menopausal status, and not on a defined age.We utilised a more population-based approach by analysing and following all surgically treated postmenopausal patients in a 7-year period; through this, we were able to minimise selection bias.The data reflect the current situation, as a period was analysed in which the currently valid guidelines of treatment recommendations (Goldhirsch et al, 2007) were active.We provided careful and detailed clinical follow-up with clear differentiation of the situations that led to discontinuation of therapy.We used a clear definition of the term ‘non-adherence’.We did not differentiate between antihormonal agents prescribed in our study based on the finding from clinical trials that adherence to aromatase inhibitors and tamoxifen did not differ significantly (Chlebowski and Geller, 2006).The rate of non-adherence to adjuvant endocrine therapy in our study was with 10.8% considerably lower than that reported in most other clinical practise settings (21–50%) (Partridge et al, 2003, 2008; Fink et al, 2004; Lash et al, 2006; Barron et al, 2007; Owusu et al, 2008); these studies, however, were plagued by the previously mentioned methodological weaknesses. With a non-adherence rate of 15% after a 3-year observation period, Demissie et al (2001) had findings most similar to ours. Not surprisingly, they avoided the major methodical shortcoming of other studies, in that they excluded patients who stopped therapy due to breast cancer recurrence from the group of non-adherent patients. When compared to the non-adherence rates reported in clinical trials (approximately 12%; Coombes et al, 2004; Coates et al, 2007; Forbes et al, 2008), our results appear realistic.Several studies performed in a clinical practise setting demonstrated that adherence to endocrine therapy was influenced by certain factors such as age, adverse effects, patient beliefs about the risks and benefits of tamoxifen use, history of medication and good patient–doctor relationship (Chlebowski and Geller, 2006). Some of these factors may be influenced by the type of care and expertise of the responsible physician. Our study shows that care in an oncologic unit is significantly associated with higher adherence to therapy (non-adherence in the subgroup of patients who had follow-up in an oncologic unit was even as low as 7.4%). The finding that care in an oncologic unit is associated with higher adherence may not necessarily be a reason for optimism. With more and more numbers of cancer patients being treated with oral agents and limitations on specialists' time, a better strategy might be to develop improved collaboration with primary care physicians.As most of the patients of our study were treated at the oncology unit of our department, the favourable non-adherence rates are probably associated with the fact that all practitioners received a targeted education in the techniques of patient-centred communication through the departmental division of psychosomatic medicine. These communication skills were applied in the follow-up of oncologic patients, in particular regarding the dialogue concerning possible therapy-related side effects and the importance of compliance and adherence to therapy (Mallinger et al, 2005; Stewart et al, 2007). Although endocrine therapy for postmenopausal breast cancer is generally well tolerated and the majority of adverse effects are mild to moderate, there are a certain number of patients who experience these as distressing (above all, hot flushes, musculoskeletal complaints/arthralgia and vaginal dryness). These effects should be carefully evaluated and regarded seriously in the follow-up appointments making the patients feel that their complaints are taken earnestly. As stated by other authors, a change in therapy may be an effective strategy to improve patient adherence (Hadji, 2008). In our study, 80% of the patients who required a change of the prescribed agent due to adverse effects could fully complete endocrine therapy in the course of time.The limitations of our retrospective study, however, must be considered. First, our study relies on information obtained by patients' self-report of adherence. It is possible that in some cases, the patients who reported continuing to take medication had indeed stopped taking it and just gave a socially acceptable answer. Furthermore, we defined non-adherence as an intentional behaviour. Atkins and Fallowfield (2006) demonstrated a high percentage of 55% of non-adherent patients with the vast majority of women reporting a non-intentional non-adherence, that is they often just forgot to take their medication. On the other hand, we think that it is unlikely that women would report that they had stopped taking medication when in fact they had not. In this context, the findings of Waterhouse et al (1993) must also be considered; they examined the influence of methodology on the categorisation of adherence to tamoxifen therapy and found that self-reported adherence fairly consistently underestimates non-adherence as determined by more objective measures, such as pill count and microelectronic monitoring. A second caveat may be that our study comes from a single region of a small country with a high socioeconomic status. All inhabitants of Switzerland have universal access to health care and free access to all prescribed and approved drugs. These facts must be considered while interpreting our relatively low non-adherence rate.Our data show that, when compared with other studies, low non-adherence rates can be realistically achieved. In the future, multifactorial approaches should be further analysed and refined with the goal of further improving compliance and adherence, and by doing so, the outcome of breast cancer patients (Chlebowski and Geller, 2006).\n\nREFERENCES:\n1. Atkins L, Fallowfield L (2006) Intentional and non-intentional non-adherence to medication amongst breast cancer patients. Eur J Cancer\n42: 2271–227616644208\n2. Barron TI, Connolly R, Bennett K, Feely J, Kennedy MJ (2007) Early discontinuation of tamoxifen: a lesson for oncologists. Cancer\n109: 832–83917243168\n3. Chlebowski RT, Geller ML (2006) Adherence to endocrine therapy for breast cancer. Oncology\n71: 1–917344666\n4. Coates AS, Keshaviah A, Thurlimann B, Mouridsen H, Mauriac L, Forbes JF, Paridaens R, Castiglione-Gertsch M, Gelber RD, Colleoni M, Lang I, Del Mastro L, Smith I, Chirgwin J, Nogaret JM, Pienkowski T, Wardley A, Jakobsen EH, Price KN, Goldhirsch A (2007) Five years of letrozole compared with tamoxifen as initial adjuvant therapy for postmenopausal women with endocrine-responsive early breast cancer: update of study BIG 1-98. J Clin Oncol\n25: 486–49217200148\n5. Coombes RC, Hall E, Gibson LJ, Paridaens R, Jassem J, Delozier T, Jones SE, Alvarez I, Bertelli G, Ortmann O, Coates AS, Bajetta E, Dodwell D, Coleman RE, Fallowfield LJ, Mickiewicz E, Andersen J, Lonning PE, Cocconi G, Stewart A, Stuart N, Snowdon CF, Carpentieri M, Massimini G, Bliss JM, van de Velde C (2004) A randomized trial of exemestane after two to three years of tamoxifen therapy in postmenopausal women with primary breast cancer. N Engl J Med\n350: 1081–109215014181\n6. Demissie S, Silliman RA, Lash TL (2001) Adjuvant tamoxifen: predictors of use, side effects, and discontinuation in older women. J Clin Oncol\n19: 322–32811208822\n7. Fink AK, Gurwitz J, Rakowski W, Guadagnoli E, Silliman RA (2004) Patient beliefs and tamoxifen discontinuance in older women with estrogen receptor – positive breast cancer. J Clin Oncol\n22: 3309–331515310774\n8. Fisher B, Costantino J, Redmond C, Poisson R, Bowman D, Couture J, Dimitrov NV, Wolmark N, Wickerham DL, Fisher ER et al (1989) A randomized clinical trial evaluating tamoxifen in the treatment of patients with node-negative breast cancer who have estrogen-receptor-positive tumors. N Engl J Med\n320: 479–4842644532\n9. Forbes JF, Cuzick J, Buzdar A, Howell A, Tobias JS, Baum M (2008) Effect of anastrozole and tamoxifen as adjuvant treatment for early-stage breast cancer: 100-month analysis of the ATAC trial. Lancet Oncol\n9: 45–5318083636\n10. Goldhirsch A, Glick JH, Gelber RD, Senn HJ (1998) Meeting highlights: International Consensus Panel on the Treatment of Primary Breast Cancer. J Natl Cancer Inst\n90: 1601–16089811309\n11. Goldhirsch A, Wood WC, Gelber RD, Coates AS, Thurlimann B, Senn HJ (2007) Progress and promise: highlights of the international expert consensus on the primary therapy of early breast cancer 2007. Ann Oncol\n18: 1133–114417675394\n12. Grunfeld EA, Hunter MS, Sikka P, Mittal S (2005) Adherence beliefs among breast cancer patients taking tamoxifen. Patient Educ Couns\n59: 97–10216198223\n13. Hadji P (2008) Menopausal symptoms and adjuvant therapy-associated adverse events. Endocr Relat Cancer\n15: 73–9018310277\n14. Lash TL, Fox MP, Westrup JL, Fink AK, Silliman RA (2006) Adherence to tamoxifen over the five-year course. Breast Cancer Res Treat\n99: 215–22016541307\n15. Mallinger JB, Griggs JJ, Shields CG (2005) Patient-centered care and breast cancer survivors' satisfaction with information. Patient Educ Couns\n57: 342–34915893218\n16. Owusu C, Buist DS, Field TS, Lash TL, Thwin SS, Geiger AM, Quinn VP, Frost F, Prout M, Yood MU, Wei F, Silliman RA (2008) Predictors of tamoxifen discontinuation among older women with estrogen receptor-positive breast cancer. J Clin Oncol\n26: 549–55518071188\n17. Partridge AH, LaFountain A, Mayer E, Taylor BS, Winer E, Asnis-Alibozek A (2008) Adherence to initial adjuvant anastrozole therapy among women with early-stage breast cancer. J Clin Oncol\n26: 556–56218180462\n18. Partridge AH, Wang PS, Winer EP, Avorn J (2003) Nonadherence to adjuvant tamoxifen therapy in women with primary breast cancer. J Clin Oncol\n21: 602–60612586795\n19. Silliman RA, Guadagnoli E, Rakowski W, Landrum MB, Lash TL, Wolf R, Fink A, Ganz PA, Gurwitz J, Borbas C, Mor V (2002) Adjuvant tamoxifen prescription in women 65 years and older with primary breast cancer. J Clin Oncol\n20: 2680–268812039930\n20. Stewart M, Brown JB, Hammerton J, Donner A, Gavin A, Holliday RL, Whelan T, Leslie K, Cohen I, Weston W, Freeman T (2007) Improving communication between doctors and breast cancer patients. Ann Fam Med\n5: 387–39417893379\n21. Waterhouse DM, Calzone KA, Mele C, Brenner DE (1993) Adherence to oral tamoxifen: a comparison of patient self-report, pill counts, and microelectronic monitoring. J Clin Oncol\n11: 1189–11978501505"
+}

batch_0/PMC2527806.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527806",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527806\nAUTHORS: H Brenner, M Hoffmeister, U Haug\n\nABSTRACT:\nWe assessed incidence and mortality of colorectal cancer (CRC) at various ages among women and men in 38 European countries. The ages at which defined levels of incidence and mortality were reached varied between 9 and 17 years between countries. This variation requires consideration in the definition of screening guidelines.\n\nBODY:\nColorectal cancer (CRC) is the third most common cancer and the fourth most common cancer cause of death globally (Parkin et al, 2005). In Europe, more than 400 000 new cases and more than 200 000 deaths occur per year (Ferlay et al, 2007). On account of its typically slow development, there is a large potential to reduce the burden of the disease by early detection and removal of precancerous lesions or early cancer stages. Various screening examinations, including fecal occult blood testing, sigmoidoscopy and colonoscopy have meanwhile been recommended by expert committees (eg Europe Against Colorectal Cancer, 2007), and nationwide screening programmes are currently being implemented, prepared or under consideration in different European countries. Within the European Union, recommendations on cancer screening typically foresee a defined common starting age (eg Advisory Committee on Cancer Prevention, 2000). However, CRC incidence and mortality strongly vary within Europe. The aim of this study was to assess differences in CRC incidence and mortality within Europe, in view of the potential implications regarding variation of age at screening initiation between countries.Materials And MethodsAnalytic strategyOur analyses are based on the following strategy: We looked at median CRC incidence and mortality among men across Europe at ages 50, 55 and 60, and we determined at what ages these levels of incidence and mortality are reached among men and women in each country. The rationale behind this ‘risk advancement approach’ (Brenner et al, 1993) is as follows: CRC incidence and mortality strongly increase with age. The age at which CRC screening becomes meaningful and cost effective depends, among other factors, on CRC incidence and mortality surpassing some minimum threshold in the absence of screening. This threshold, however defined, is reached at various ages in different countries. All of the aforementioned ages have been recommended or are actually used in practise for initiation of CRC screening (Advisory Committee on Cancer Prevention, 2000; Malila et al, 2005; Pox et al, 2007; West et al, 2007).DatabaseEstimates of CRC incidence and mortality rates for age groups 15–44, 45–54, 55–64, and 65+ years were obtained for 38 European countries from the GLOBOCAN 2002 database (Ferlay et al, 2004).Statistical analysesIn our analyses, we made the following approximations: CRC incidence and mortality at age 40 years were approximated as 0.75 × 3=2.25 times the corresponding rates in the 30-year interval 15–44 years. This approximation is based on the observation that roughly 75% of CRC cases diagnosed in the age range 15–44 years occur in the age interval 35–44 years in populations where more detailed age-specific rates are available, and on the assumption that age interval 35–44 years typically comprises roughly one-third of the population in age interval 15–44 years. CRC incidence and mortality at ages 50 and 60 were assumed to equal the corresponding rates in age groups 45–54 and 55–64 years, respectively. Finally, it was assumed that data on CRC incidence and mortality in men and women in age group 65+ years correspond to the respective levels at the mean age of the male and female population aged 65+ years in each country in the year 2000, which was derived from the United Nations Population Database (United Nations Population Division, 2007).Based on the assumptions described above, defined levels of CRC incidence and mortality were available for ages 40, 50, 60, and for an age between 72 and 75 years (men) or between 73 and 77 years (women), respectively, with the latter varying across countries. Incidence and mortality at single years of age between these defined levels were approximated by linear interpolation. Sensitivity analyses varying the aforementioned approximations within plausible ranges led only to very minor variation of the results and are therefore not shown in detail.ResultsCRC incidence strongly increased with age in all countries. Estimates of median incidence (mi) across countries among men aged 50, 55 and 60 years were 37, 73, and 112 per 100 000 persons per year, respectively. The age at which these levels were reached among men and women in the different countries, denoted agemi50, agemi55, and agemi60, respectively, varied strongly. For example, among men, agemi50 varied between 45 years in Hungary and 55 years in Greece (see Table 1). Similarly, agemi55, and agemi60 varied between 51 and 62 years, and 54 and 67 years among men in the same countries. As illustrated in Figure 1, this is explained by the much steeper rise in CRC incidence with age in Hungarian men compared with Greek men. Overall, among men in Western Europe, incidence rates were higher, and the various levels of incidence were reached at younger ages compared with men from other European regions. However, there was also substantial variation within regions.Among women, incidence rates were generally lower compared with men, which implies that specified levels of incidence were reached at higher ages. With few exceptions, between-country variation was similar among women and men, but it was particularly large among older women, reaching a maximum of 17 years (from 60 to 77 years) for agemi60.CRC mortality likewise strongly increased with age in all countries. Estimates of median mortality (mm) across countries among men aged 50, 55 and 60 years were 14, 33, and 52 per 100 000 persons per year, respectively. The age at which these levels were reached among men and women in the different countries, denoted agemm50, agemm55, and agemm60, respectively, again showed strong between-country variation. The lowest and highest levels of these ages (resulting from highest and lowest levels of mortality) were again mostly seen for Hungary and Greece, respectively (see Table 2 and Figure 2). Agemm50, agemm55, and agemm60 varied up to 12 years between countries among both women and men. Again, these ages tended to be higher among women (ranges across countries: 45–57, 54–64, and 60–69 years, respectively) than among men (ranges across countries: 43–54, 51–62, and 53–65 years, respectively). In Eastern countries, mortality was generally higher, and levels of agemm50, agemm55, and agemm60 were generally lower than in countries from other parts of Europe.DiscussionOur comparative analyses of age- and sex-specific CRC incidence and mortality in 38 European countries indicate that differences in incidence and mortality between countries translate to wide age ranges at which comparable levels of risk are reached. Age-specific CRC incidence and mortality represent important parameters regarding potential benefits of screening, which have to be weighed against costs and potential adverse side effects when choosing the age at screening initiation. Our analyses suggest that the balance in favour of screening is likely to be reached at rather different ages in the various European countries.The reasons for this variation are likely to be manifold. Differences in risk factor profiles, such as dietary habits, across European countries, may be most relevant for variation in CRC incidence. As regards mortality, major differences in survival of CRC patients between European countries also play an important role (Verdecchia et al, 2007).Even though CRC incidence and mortality data are important, there are further, country-specific factors to be considered in the choice of the age range at which screening is offered. For example, an important parameter is remaining life expectancy (Ko and Sonnenberg, 2005). Furthermore, non-epidemiological criteria, such as the availability and costs of various screening modalities in different countries have to be taken into account.Risk adapted variation of ages for CRC screening initiation is already well accepted and established for CRC risk factors other than ‘country’, in particular a family history of CRC (Europe Against Colorectal Cancer, 2007). The ‘risk advancement’ (Brenner et al, 1993) given a positive family history is of similar magnitude than the between country differences of up to 10 years or more observed in the present analyses (Brenner et al, 2008). As regards gender differences, risk advancement among men compared to women is roughly 5 years (Brenner et al, 2007). Taken together, these patterns suggest that appropriate differentiation of age at initiation of CRC screening by gender and country might be similarly or even more relevant from a public health point of view than the widely practised differentiation by family history, an important but relatively infrequent CRC risk factor.In the interpretation of our results, the following limitations should be kept in mind. Our analyses are based on data on CRC incidence and mortality by age and sex. Additional factors, such as potential variation in screening effectiveness by age, for which data are lacking, were not considered. In some of the countries included in this analysis, some form of CRC screening has already been practised before 2002, but, because of mostly small regional coverage (Benson et al, 2008) it is unlikely to have had a major effect on CRC incidence and mortality on the national level. Although estimates on CRC incidence and mortality in 2002 obtained from GLOBOCAN 2002 are based on the best available data sources, they are subject to uncertainties due to lack of or incomplete population coverage of cancer registration in many countries.Analyses were presented in detail for selected ‘threshold levels’ of incidence and mortality only. Latter were defined by median levels across countries among men at ages 50, 55 and 60 years, which may not necessarily be the best ‘threshold levels’. Best threshold levels and optimum age for commencing screening will be based on cost-benefit decisions. However, as can be seen from Figures 1 and 2, differences in ages at which thresholds are reached would be rather similar for any other intermediate levels of incidence and mortality within the assessed range.In summary, our analyses do not allow deriving a general recommendation regarding the best age for initiation of CRC screening in each country. Our results do suggest, however, that the optimal age for screening initiation is likely not to be the same for European countries and that variation by up to 10 years or even more across countries might be warranted because of major differences in CRC incidence and mortality.\n\nREFERENCES:\n1. Advisory Committee on Cancer Prevention (2000) Recommendations on screening in the European Union. Eur J Cancer\n36: 1473–147810930794\n2. Benson VS, Patnick J, Davies AK, Nadel MR, Smith RA, Atkin WS, on behalf of the International Colorectal Cancer Screening Network (2008) Colorectal cancer screening: a comparison of 35 initiatives in 17 countries. Int J Cancer\n122: 1357–136718033685\n3. Brenner H, Gefeller O, Greenland S (1993) Risk and rate advancement periods as measures of exposure impact on the occurrence of chronic diseases. Epidemiology\n4: 229–2368512987\n4. Brenner H, Hoffmeister M, Arndt V, Haug U (2007) Gender differences in colorectal cancer: implications for age at initiation of screening. Brit J Cancer\n96: 828–83117311019\n5. Brenner H, Hoffmeister M, Haug U (2008) Family history and age at initiation of colorectal cancer screening. Am J Gastroenterol (in press)\n6. Europe against colorectal cancer (2007) Declaration of Brussels, http://www.future-health-2007.com/fileadmin/user_upload/Brussels_Declaration.pdf.accessed 4 April 2008\n7. Ferlay J, Autier P, Boniol M, Heanue M, Colombet M, Boyle P (2007) Estimates of the cancer incidence and mortality in Europe in 2006. Ann Oncol\n18: 581–59217287242\n8. Ferlay J, Bray F, Pisani P, Parkin DM, GLOBOCAN 2002 (2004) Cancer Incidence, Mortality and Prevalence Worldwide. IARC Cancer Base No. 5 Version 2.0. IARC Press: Lyon, France\n9. Ko CW, Sonnenberg A (2005) Comparing risks and benefits of colorectal cancer screening in elderly patients. Gastroenterology\n129: 1163–117016230070\n10. Malila N, Anttila A, Hakama M (2005) Colorectal cancer screening in Finland: details of the national screening programme implemented in autumn 2004. J Med Screen\n12: 28–3215814016\n11. Parkin DM, Bray F, Ferlay J, Pisani P (2005) Global cancer statistics, 2002. CA Cancer J Clin\n55: 74–10815761078\n12. Pox C, Schmiegel W, Classen M (2007) Current status of screening colonoscopy in Europe and in the United States. Endoscopy\n39: 168–17317327977\n13. United Nations Population Division (2007) World Population Prospects. The 2006 Revision. Population Database http://esa.un.org/unpp/, accessed 19 October 2007\n14. Verdecchia A, Francisci S, Brenner H, Gatta G, Micheli A, Mangone L, Kunkler I, the EUROCARE-4 Working Group (2007) Recent cancer survival in Europe: a 2000–02 period analysis of EUROCARE-4 data. Lancet Oncol\n8: 784–79617714993\n15. West NJ, Poullis AP, Leicester RJ (2007) The NHS Bowel Cancer Screening Programme–a realistic approach with additional benefits. Colorectal Dis e-pub ahead of print 23 October"
+}

batch_0/PMC2527997.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527997",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527997\nAUTHORS: Carlos F. Pereira, Rémi Terranova, Natalie K. Ryan, Joana Santos, Kelly J. Morris, Wei Cui, Matthias Merkenschlager, Amanda G. Fisher\n\nABSTRACT:\nDifferentiated cells can be reprogrammed through the formation of heterokaryons and hybrid cells when fused with embryonic stem (ES) cells. Here, we provide evidence that conversion of human B-lymphocytes towards a multipotent state is initiated much more rapidly than previously thought, occurring in transient heterokaryons before nuclear fusion and cell division. Interestingly, reprogramming of human lymphocytes by mouse ES cells elicits the expression of a human ES-specific gene profile, in which markers of human ES cells are expressed (hSSEA4, hFGF receptors and ligands), but markers that are specific to mouse ES cells are not (e.g., Bmp4 and LIF receptor). Using genetically engineered mouse ES cells, we demonstrate that successful reprogramming of human lymphocytes is independent of Sox2, a factor thought to be required for induced pluripotent stem (iPS) cells. In contrast, there is a distinct requirement for Oct4 in the establishment but not the maintenance of the reprogrammed state. Experimental heterokaryons, therefore, offer a powerful approach to trace the contribution of individual factors to the reprogramming of human somatic cells towards a multipotent state.\n\nBODY:\nIntroductionReprogramming somatic cells to become ES-like is an important goal in cell replacement therapy since it affords the opportunity to generate and use patient-specific ES derived cells as grafts. Epigenetic reprogramming can be achieved in different ways including nuclear transfer [1]–[4] or the forced expression of one or more transcription factors [5],[6]. Retroviral-mediated expression of four transcriptional regulators, Oct4, Sox2, c-Myc and Klf4, was shown to drive mouse fibroblasts into an ES-like (iPS) state, albeit at low frequency [7],[8]. Reprogramming of human fibroblasts has also recently been achieved in a parallel approach using Oct4, Sox2 and either Nanog plus Lin28 [9] or Klf4 plus c-Myc [10].These pioneering studies have illustrated the importance of several factors for iPS, but also suggested that additional ones may be needed for efficient conversion to pluripotency. Reprogramming can also be achieved by cellular fusion, a process that occurs spontaneously in vitro\n[11], in vivo\n[12] and experimentally using specific agents [13]. For example, fusion of differentiated cells with pluripotent ES cells, embryonic carcinoma (EC) or embryonic germ (EG) cells, induces the expression of pluripotency-associated markers in the hybrid cells [14]–[18] and chromatin remodelling at specific sites in the somatic cell genome [14], [15], [18]–[21]. While these data show that reprogramming occurs through the epigenetic resetting of gene expression programs in the differentiated cell, it has been unclear whether nuclear fusion and genome duplication are absolutely required for successful conversion [22]. Here we investigated the requirements for, and the stability of, dominant reprogramming of human B cells by fusion with mouse ES cells. We show that reprogramming is surprisingly rapid and occurs within heterokaryons in which lymphocyte and ES cell nuclei remain spatially discrete. Furthermore, our data show that while Oct4 is critical for successful reprogramming of human lymphocytes to an ES-like state, Sox2 is not required. Thus our data outline an alternative strategy for defining the factors that are required for inducing a pluripotent state in human somatic cells.ResultsReprogramming of Gene Expression Is Initiated in ES Cell Heterokaryons Prior to Nuclear FusionHuman B cells were fused with mouse ES cells using polyethylene glycol (PEG). The nuclear events in fused cells were monitored by fluorescence microscopy and quantitative RT-PCR to analyse gene expression (Figure 1). To facilitate the identification of fused cells, E14tg2a mouse ES cells were pre-labelled with DiD and human B cells with DiI and dual-stained cells were purified by FACS (typically 10–15% of cells, Figure S1A). Human (B cell-derived) and mouse (ES cell-derived) nuclei were distinguished on the basis of DAPI and human-specific Lamin A/C labelling, and the proportion of cells containing two discrete (heterokaryons) or conjoined nuclei (hybrids) was assessed over time (Figure 1B). Up to 2 days following cell fusion 98–99% of dual labelled cells were identified as heterokaryons in which a single human and a single mouse nucleus were evident (illustrated in Figure 1B, central image). The kinetics of nuclear fusion were also confirmed by fluorescence in situ hybridization (FISH) analysis in which probes specific for mouse chromosomes (γ-satellite, red) or human chromosomes (α-satellite, green) were used to detect interspecies chromosome mixing, indicative of hybrid formation (Figure S1B and Text S1).10.1371/journal.pgen.1000170.g001Figure 1Pluripotent reprogramming of human B-lymphocytes by mouse ES cells is initiated in heterokaryons prior to nuclear fusion and cell division.(A) Shows the experimental strategy used to generate interspecies heterokaryons (hB x mES). Human B-lymphocytes (hB) and mouse embryonic stem cells (mES) were respectively labelled with the cell membrane dyes DiI and DiD and fused in the presence of polyethylene glycol (PEG). Fused cells were FACS sorted and cultured under conditions that promote mES self-renewal. (B) Immunofluorescence analysis of the kinetics of heterokaryon (cells in which parental nuclei share the same cytoplasm but remain discrete) and hybrid formation (where both parental genomes occupy the same nucleus) following PEG induced fusion. In the lower panels hB-derived nuclei were distinguished by mouse nuclei on basis of DAPI (blue) and human Lamin A/C staining (green), and actin staining (red) delineates individual cells. Confocal sections showing a hB cell prior to fusion (left, day 0), a heterokaryon [one mouse (with DAPI intense foci) and one human nucleus (hLamin A/C positive)] (middle, day 2) and a hybrid cell (right, day 3) are shown. Scale bar, 10 µm. n = 100. (C) The expression of hES-specific genes (hOct4, hNanog, hCripto, hDnmt3b, hTle1, hRex1) was assessed by quantitative RT-PCR analysis 0 to 8 days after cell fusion. Positive (hES-NCL1, black bars) and negative (hB) controls for this analysis were included. (D) Activation of embryonic genes is accompanied by silencing of lymphocyte-specific genes (hCD19, hCD37, hCD20, hCD45 and hPax5), while a constitutively expressed gene hHprt remained detectable at similar levels at all time points. Data were normalised to hGapdh expression. Error bars indicate the s.d. of 3 independent experiments. (E) Bisulfite genomic sequencing analysis of DNA methylation at the human Oct4 promoter 0, 2, 4 and 8 days after cell fusion demonstrated the rapid de-methylation of Oct4 induced by fusion with mES cells. Human ES cells (hES, cell line H1; lower panel) are shown as controls. The methylation pattern of Igf2/H19 imprinting control region (ICR) remained unaltered throughout the experiment. The position of CpG sites relative to the transcriptional start site (TSS) is indicated. Open circles represent unmethylated cytosines, black closed circles represent methylated cytosines and grey closed circles represent constitutively methylated cytosines. Regions 1, 2 and 3 indicate CpG sites that are part of the same PCR product.The expression of pluripotency-associated genes and lymphocyte-associated genes by human B cell-derived nuclei was assessed by qRT-PCR, using primers that selectively amplify the human transcripts. Expression of human Oct4, Nanog, Cripto, Dnmt3b and Tle1 was detected in cells as early as 1 day after fusion and human Rex1 after 2 days (Figure 1C and Figure S1C). Although the levels were low in heterokaryons (<1% of that detected in human ES cells, cell line NCL1), these increased over time and were undetectable in non-fused (or self-fused, not shown) human B cells or control mouse ES cells. Expression of hTert was detected from day 4 onwards (Figure S1D), while hHprt expression was equivalent at all stages, as anticipated (Figure 1D). Mouse lymphocyte-specific gene transcripts (mCD19, mCD37 and mCD45) were not detected throughout the analysis (not shown), confirming the dominance of ES cells in conversion [15],[18]. Increased expression of human pluripotency-associated genes over this 8-day period was mirrored by a reduction in expression of several human lymphocyte-associated genes within the second (hCD45, hCD37 and hCD19) or third day (hCD20 and hPax5) of heterokaryon formation (Figure 1D). Collectively these data show that upon dominant reprogramming, activation and silencing of tissue-specific gene programs begins ahead of, and therefore does not require, nuclear fusion and cell division. In addition, since these results examine gene expression at the population level, it is possible that gene expression varied between individual heterokaryons and hybrid cells.As the reprogramming of somatic cells has been previously shown to result in altered DNA methylation at specific loci [15],[18],[21],[23], we examined changes in the methylation status of the human Oct4 gene promoter [24] and as a control, the Igf2/H19 imprinting control region (ICR) [25]. As illustrated in Figure 1E, human B cells prior to fusion showed high levels of DNA methylation throughout the hOct4 promoter and across a single Igf2/H19 allele. Following cell fusion, DNA methylation of hOct4 in reprogrammed B cells declined, consistent with a trend towards a hypomethylated state as seen in the human ES cell line H1. Demethylation of the hOct4 promoter was detected prior to nuclear fusion and cell division, a result that is consistent with active chromatin remodelling of the locus prior to expression. No changes in DNA methylation at Igf2/H19 ICR were detected over this period, consistent with its imprinted status [25].Induction of a Human ES-Specific Gene Expression ProfileA comparison of the relative abundance of gene-specific transcripts in reprogrammed human B cells (Figure 2A, right-hand column), showed a strong similarity with the gene expression profiles of several human ES cell lines (NCL1 [26], HI, H7, H9 [27]; Figure 2A, left-hand column). For example, while Oct4 was abundantly expressed in all human and mouse ES cell lines, Nanog and Cripto expression was consistently much lower than Oct4 (100–1000 fold) for each of the mouse ES cell lines analysed (OS25, CCE, E14, ZHBTc4; Figure 2A, middle panel). In human ES cell lines however, Oct4, Nanog and Cripto transcripts were similarly abundant, consistent with that seen in reprogrammed human B cells. Expression of some pluripotency-associated genes, for example Sox2, was variable and often required extended periods of time (>8 days) for detection (not shown). This could reflect the fact that genes such as Sox2 are subject to multiple layers of repressive epigenetic modifications in B cells including DNA and histone methylation [28],[29] and late replication [30], or that they require a higher threshold of activators for overt expression.10.1371/journal.pgen.1000170.g002Figure 2Gene expression in reprogrammed lymphocytes resembles human rather than mouse ES cell lines.(A) Quantitative RT-PCR analysis of the relative levels of gene expression in several human (NCL1, H1, H7 and H9), mouse (OS25, CCE, E14 and ZHBTc4) ES cell lines and in reprogrammed lymphocytes (hB x mES) at 0, 2, 4 and 8 days after cell fusion. Human ES lines (left panel) and hB x mES (right panel) gene expression data were normalised to hGapdh. Mouse ES lines (middle panel) gene expression data were normalised to mGapdh. Error bars indicate the s.d. of 3–4 independent experiments. (B) After cell fusion, genes involved in the maintenance of undifferentiated human ES cells (hFgfr1, hFgfr2 and hFgf2) were activated while genes selectively expressed by mouse ES cells (hBmp4, hLifr and hJak3) were not induced. Data were normalised to hGapdh expression. Error bars indicate the s.d. of 3 independent experiments. (C) Heterokaryons resulting from human B cell and mouse ES cell fusions (hB x mES) were stained for SSEA4 at 0, 2, 4 and 8 days and expression was analysed by flow cytometry. The results showed that 13.5% (day 2), 16.6% (day 4) and 15.8% (day 8) of total heterokaryons expressed SSEA4, as delineated by the rhomboid gates. Mean intensity fluorescence of positive cells is indicated.Similarities between gene expression profiles of human ES cell lines and hB x mES fused cells prompted us to examine additional markers that are expressed solely by either human or mouse ES cells [31]–[34]. These included fibroblast growth factor receptors (Fgfr1 and Fgfr2) and Fgf2 (expressed by human ES cells), Bmp4 and leukaemia inhibitory factor (Lif) receptor (expressed by mouse ES cells) and SSEA4, a surface glycoprotein selectively expressed by human ES cells [27] (Figure S2B). This analysis revealed that reprogrammed cells expressed increasing amounts hFgfr1, hFgfr2 and hFgf2 but did not express hBmp4 or hLifr or upregulate the downstream kinase hJak3 (Figure 2B). Thus, these data show that while dominant conversion is driven by mouse ES cells (that express Bmp4 and Lifr prior to fusion, Figure S2A), reprogrammed heterokaryons and hybrid cells show a remarkably different expression profile resembling human, rather than mouse ES cell lines. Consistent with this, fusion of mouse ES cells and human B cells resulted in SSEA4 expression by 13–16% of the cells (days 2–8 as shown in Figure 2C). Isolation of SSEA4-positive cells confirmed that this subset contained successfully reprogrammed cells that express hOct4, hNanog and hCripto (Figure S2C), while SSEA4-negative cells were not reprogrammed. The observation that only a proportion of heterokaryons are successfully reprogrammed, as judged by hOct4 DNA demethylation and SSEA4 expression, might partly explain why the levels of transcripts encoding pluripotency factors are lower in reprogrammed cultures than established hES cell lines.To ask whether the reprogramming of human B cells by mouse ES cells resets multi-lineage potential, hB x mES cultures were treated with retinoic acid (RA) 6–8 days after cell fusion in order to induce differentiation (Figure 3). Prior to RA treatment most cells in hB x mES colonies showed alkaline phosphatase (AP) activity (Figure 3A), and expressed human AP transcripts (not shown). Hybrid colonies also expressed several pluripotency-associated markers, including hNanog protein (detected using a human Nanog-specific antibody) and the human embryonic-specific antigens SSEA4, TRA-1-60 and TRA-1-81 [27] (Figure S3). Following treatment with RA, AP activity and expression of hOct4, hNanog and hRex1 was reduced (Figure 3C), while morphological heterogeneity within colonies increased. RA treatment induced the expression of genes associated with extra-embryonic (hCdx2, hHand1 and hGata6), endoderm (hSox7, hHnf4 and hCollagenIVαI), mesoderm (hMixl1, hEbf and hMyoD) and ectoderm (hNestin,\nFigure 3B) differentiation in hB x mES, but not in control hB cells (Figure 3C, blue and black lines respectively). Differentiation also resulted in increased DNA methylation of the hOct4 promoter (Figure 3D) to levels similar with that seen in differentiated human cells (Figure 1E). Taken together, these results show that reprogramming of human B cells by mouse ES cells resets gene expression and multi-lineage potential.10.1371/journal.pgen.1000170.g003Figure 3Multi-lineage potential is reset in reprogrammed human lymphocytes.(A) Hybrid colonies resulting from fusion of human B cells (hB) and mouse ES cells (hB x mES) showed alkaline phosphatase activity (pink), that was reduced upon retinoic acid (RA) treatment. (B) RA treatment of hybrid colonies (day 6) generated cells that expressed Nestin (green) detected by immunostaining using an antibody specific for human (and not mouse) Nestin protein. DAPI counterstaining (blue) is shown. Scale bars, 50 µm. (C) Quantitative RT-PCR analysis of gene expression upon RA treatment of hybrid cells (blue line) showed that levels of pluripotency genes (hOct4, hNanog and hRex1) declined while differentiation-associated genes were upregulated [extra-embryonic (hCdx2, hHand1 and hGata6), endoderm (hSox7, hHnf4 and hCollagenIVαI), mesoderm (hMixl1, hEbf and hMyoD) and ectoderm (hNestin)]. Unfused hB cells were included as controls (black line). Data were normalised to hGapdh expression. (D) Bisulfite genomic sequencing analysis of DNA methylation at the human Oct4 promoter 8 days after RA treatment showed the re-methylation of the Oct4 promoter while the Igf2/H19 imprinted control region (ICR) remains unaltered. The position of CpG sites relative to the transcriptional start site (TSS) is indicated. Open circles represent unmethylated cytosines, black closed circles represent methylated cytosines and grey closed circles represent constitutively methylated cytosines.Interspecies Reprogramming of Human B Cells Requires mOct4 but Not mSox2\nOct4 is part of the core regulatory circuitry in ES cells [35] and it is essential for pluripotency and self-renewal [36]. To assess the potential role of mouse-derived Oct4 as a dominant ‘trans’ acting factor within inter-species heterokaryons we generated ES cells expressing Flag-tagged mouse Oct4 protein (Figure 4A) and fused these with human B lymphocytes (Figure 4B). Flag-tagged Oct4 (derived from mouse ES cells) was seen to accumulate within human nuclei 3 to 6 hours after cell fusion (Figure 4B; complete kinetic experiment shown in Figure S4A). In addition, Oct4 protein was present in heterokaryon nuclei (at 3 hours) before transcription of hOct4 was initiated (at 24 hours). Thus, translocation of the ES-derived Oct4 into human lymphocyte nuclei precedes reprogramming.10.1371/journal.pgen.1000170.g004Figure 4Oct4 is required for successful reprogramming.(A) Mouse ES cells expressing a tagged Oct4 protein (Flag-mOct4) were generated by insertion of Flag-tagged mouse Oct4 cDNA in E14tg2a ES cells (parental cell line). Western blotting with anti-Oct4 and anti-Flag antibodies confirmed the presence of Flag-tagged Oct4 protein by transduced cells. Equivalent protein loading is shown with Lamin B detection. (B) Immunofluorescence analysis of cultured heterokaryons 6 hours after cell fusion showed the presence of ES cell-derived Oct4 (Flag-Oct4, green) in a human nucleus (arrowed). Human nuclei were distinguished from mouse nuclei on basis of diffuse versus punctuate DAPI staining (blue), respectively. Actin labelling (red) delineates the cell membrane. Images are confocal sections of heterokaryons containing a single mouse (with DAPI intense foci) and a single human nucleus. Scale bar, 10 µm. (C) In ZHBTc4 ES cells endogenous Oct4 was replaced by an inducible transgene (Oct4βgeo) which can be downregulated by addition of doxycycline (Dox) [36]. Quantitative RT-PCR analysis showed that 6 hours (+6) and 12 hours (+12) after Dox treatment, mOct4 was progressively downregulated, while expression of other pluripotency-associated genes (mNanog, mCripto, mRex1 and mSox2) was largely unaffected. (D) ES cells expressing normal levels of Oct4 (-), partially reduced (Dox+6) or lacking Oct4 expression (Dox+12) were fused to hB-lymphocytes. Successful reprogramming was assessed by quantifying the abundance of human ES-associated transcripts two days after fusion by qRT-PCR. Activation of pluripotency genes in hB-lymphocytes was reduced or impaired when Oct4 was ablated. Data were normalised to Gapdh expression. Error bars indicate the s.d. of 2–3 independent experiments.Conversion of human fibroblasts to ES-like cells has been shown to require the activation of at least four factors including Oct4, Sox2 and either Nanog plus Lin28 [9] or Klf4 plus c-Myc [10]. Recently it was shown that mouse ES cells lacking Sox2, a factor thought to be vital for preventing extra-embryonic differentiation, can remain pluripotent provided with elevated Oct4 levels [37]. To investigate the relative importance of Oct4 and Sox2 in reprogramming, mouse ES cells that are inducible null (Tet-off) for mOct4 (ZHBTc4 [36]) or for mSox2 (2TS22C [37]) were used as fusion partners with human B cells. These inducible null ES cell lines were constructed and characterised previously [36],[37] and display a rapid (within 24 hours) and complete elimination of Oct4 or Sox2 gene/protein expression upon doxycycline (+Dox) treatment. In our hands, pre-treatment of ZHBTc4 cells with Dox for 6 and 12 hours, resulted in a progressive decrease in mOct4 gene expression (Figure 4C), without significantly affecting the expression of other pluripotency-associated genes in these cells or the efficiency which they fuse with human B cells (Figure S4B). Successful reprogramming, as judged by induction of several human genes (Oct4, Nanog, Cripto, Dnmt3b, Sox2, Tle1, Tert and Rex1) was however reduced (+6 hours) or eliminated (+12 hours) by pre-treatment of ZHBTc4 cells with Dox (Figure 4D, a complete kinetic analysis is provided in Figure S4C). Likewise, knocking down mOct4 using short interference RNA (siRNA) in E14tg2a mES cells (Figure S5A and Text S1) also abolished reprogramming activity (Figure S5B). These results confirm that mOct4 expression is critically important for initiating successful reprogramming, in keeping with previous reports [7]–[10],[38]. The extinction of human lymphocyte-specific genes was however not impaired by Oct4 removal (Figure S4C), a result that may support previous findings that the activation and silencing of gene expression programs in heterokaryons are mechanistically distinct processes [13]. Eliminating mSox2 expression in the mouse ES cell (Figure 5A, 2TS22C) had, in contrast, a relatively mild effect on reprogramming efficiency (Figure 5B, compare values at 0, 12 and 24 hours of Dox treatment). Furthermore, reprogramming was fully restored in fusions using 2O1 cells, a Sox2-deficient mES cell line in which mOct4 expression is up-regulated [37] (Figure 5A,B values shown in red and complete kinetics shown in Figure S6). These data show that Oct4, but not Sox2, is critical for the dominant reprogramming activity of mouse ES cells. Interestingly, using 2O1 cells we observed the enhanced induction of hSox2 (Figure 5B, red arrow), a result that suggests that mouse-derived Oct4 levels may be important for initiating hSox2 expression in somatic nuclei.10.1371/journal.pgen.1000170.g005Figure 5Sox2 is dispensable for reprogramming.(A) In 2TS22C ES cells endogenous Sox2 is replaced by an inducible transgene (Sox2Zeo) which can be downregulated by addition of doxycycline (Dox) [37]. Quantitative RT-PCR analysis showed that 12 hours (+12) and 24 hours (+24) after Dox treatment, mSox2 was downregulated while expression of other pluripotency-associated genes (mNanog, mCripto, mRex1 and mOct4) continued to be expressed. 2O1 ES cells are Sox2-deficient mES cells (asterisk) in which mOct4 expression is up-regulated (red bars). (B) ES cells expressing Sox2 (-), Sox2 depleted cells (Dox+12, Dox+24) and 2O1 cells were fused to hB-lymphocytes. Successful reprogramming was assessed by quantifying the abundance of human ES-associated transcripts two days after fusion by qRT-PCR. Activation of pluripotency genes in hB-lymphocytes occurs in the absence of mSox2. An elevated induction of hSox2 using 2O1 cells as a fusion partner is highlighted by an arrow (red). All data were normalised to Gapdh expression and error bars indicate the s.d. of 2–3 independent experiments.ES-Derived mOct4 Is Dispensable for Maintaining the Reprogrammed Status of Somatic CellsTo assess whether gene expression by the reprogrammed cell is stable (self sustaining) or requires the continuous supply of factors provided by the mouse ES cell, we generated hybrid cells between lymphocytes and ES cells in which Oct4 expression could be conditionally withdrawn (ZHBTc4, experimental outline depicted in Figure 6A). In these experiments fusions were performed between mouse lymphocytes carrying a silent, Oct4-driven GFP transgene (GOF18ΔPE) and mouse ZHBTc4 ES cells, to allow successfully reprogrammed hybrid cells to be identified on the basis of GFP re-expression by day 10 (Figure 6A). Hybrid clones contained a rearranged IgH locus, consistent with their derivation from mouse B cells (Figure S7A and Text S1), displayed twice the DNA content of diploid cells (4n, Figure 6B) and were karyotypically stable over the study period (not shown). As anticipated, hybrid cells expressed ZHBTc4-derived Oct4βgeo transcripts and several pluripotency-associated genes, but did not express B cell markers such as CD19, Pax5 and Ly108 (Figure S7B). Two hybrid clones were selected for study (hybrid 4 and 12) and were treated with Dox to selectively ablate expression of ZHBTc4-derived Oct4βgeo (Figure 6C; Figure S7C shows the strategy used to selectively detect Oct4βgeo transgene expression). Withdrawal of ZHBTc4-derived Oct4 did not alter the expression of mNanog and mSox2 in reprogrammed cells (Figure 6C), and did not precipitate differentiation towards trophectoderm or the up-regulation of mCdx2 and mHand1 expression [36] (Figure 6D and Figure S7D); events that are induced by the removal of Oct4 from the parental ZHBTc4 line (Figure 6D right hand panel and Figure 6E). Thus, our data show that reprogramming of lymphocytes by mouse ES cells induces an epigenetically stable (and heritable) resetting of gene expression in the lymphocyte nucleus.10.1371/journal.pgen.1000170.g006Figure 6Reprogramming is self-sustaining and can be maintained in the absence of ES-derived Oct4.(A) To address whether reprogramming is stable or subject to reversion, we ablated Oct4 expression after hybrid formation. ZHBTc4 ES cells [mES with endogenous Oct4 replaced by an inducible transgene (Oct4βgeo) which can be downregulated by addition of doxycycline (Dox)] were fused with mouse B-lymphocytes (mB) carrying a GFP transgene under the control of Oct4 promoter (GOF18ΔPE). Reprogramming of mB results in the re-activation of GFP in hybrid colonies (d10, lower panels). Kinetic analysis of single cells (upper panels) showed that transgene re-activation occurs in heterokaryons (day 2, 2 arrows), and hybrid cells (day3, arrowhead). mB cells are shown as negative controls. Nuclei were visualised with DAPI staining (blue). Scale bars, 10 µm. (B) Hybrid clones (mES x mB, 4n) that re-expressed GFP were isolated and analysed by FACS. mES, mB and mES x mB hybrid cells unstained (left panel) or stained with propidium iodide (right panel) to assess GFP expression and DNA content, respectively. (C) Hybrid clones (4 and 12) were treated with Dox to ablate ES-derived Oct4, and quantitative RT-PCR confirmed downregulation of Oct4βgeo transcript (upper panels). Removal of mES-derived Oct4βgeo in hybrid clones did not affect gene expression of pluripotency-associated transcripts (lower panels; mOct4, mNanog and mSox2) after 96 hours of Dox treatment. (D) No differentiation was observed after Oct4 removal in hybrid cells. mRex1 expression was retained and the extra-embryonic markers mHand1 and mCdx2 were not induced. In ZHBTc4 ES cells (open bars) upon Dox treatment, mHand1 and mCdx2 were induced [36] and are shown for comparison. Data were normalised to mGapdh expression. Error bars indicate the s.d. of 3 independent experiments. (E) Western blotting with anti-Oct4 antibody confirmed that Oct4 protein is rapidly removed after Dox treatment of ZHBTc4 ES cells (lower panel) but remains detectable at all times in hybrid cells (upper panels). Equivalent protein loading is shown with Lamin B detection.DiscussionIn this study we show that reprogramming human lymphocytes can be achieved using mouse ES cells as a cell fusion partner, a process that induces the re-expression of endogenous human genes normally associated with human blastocyst development and human ES cell lines. Successful interspecies reprogramming is initiated in heterokaryons prior to chromosome intermixing, and generates cells that express human FGF signalling pathway components and human ES-specific surface molecules such as SSEA4, TRA-1-60 and TRA1-81. We show that this reprogramming is critically dependent upon Oct4, since Oct4 deletion abolishes the reprogramming capacity of mES cells. Conversion of human B cells into ES-like cells results in the re-modelling of the somatic genome with loss of DNA methylation at the hOct4 locus. Importantly, once reprogramming is initiated by factors produced by the dominant (ES) nucleus, we show that withdrawal of mOct4 does not compromise the phenotype of hybrid cells. This result implies that the reprogrammed state, once initiated, is both self-sustaining and heritable.One surprising aspect of the reprogramming data shown here is the rapidity of gene conversion and DNA demethylation that occurs within heterokaryons. As successful reprogramming is only achieved in a proportion of heterokaryons (<13%), it is likely that partial DNA replication (or repair) is required for lymphocyte conversion. Previous studies have shown that reprogramming in experimental heterokaryons using adult cells from different lineages [13],[39], can be initiated before genome duplication and cell division. Here we show that conversion of unipotent lymphocytes towards multipotency is achieved in transient heterokaryons prior to cell division. Re-activation of human Oct4 and Nanog by human nuclei, has been shown to occur rapidly upon DNA de-methylation and Tpt1 activation induced by Xenopus oocytes [21],[40]. The rapid re-activation of endogenous pluripotency-associated genes seen in inter-species heterokaryons is consistent with transgene re-activation studies that have reported Oct4gfp expression by MEFs [41] or NSCs [22] fused with mouse ES or EC cells. Collectively these results may have an impact for generating human ES-like cells. Proof that mouse ES cells can dominantly reset the multi-lineage potential in human somatic cells, together with evidence that this process begins prior to nuclear fusion, suggests that improved methods for removing mouse chromosomes from heterokaryons [42] may be applicable for generating human stem cell lines. Alternatively, using conditionally targeted mouse ES cells to dissect the roles of individual proteins thought to be critical for multipotent reprogramming, may provide a rationale for using distinct protein cocktails to directly re-set lineage potential.In the experiments presented here we have shown that reprogrammed human cells express a profile of transcripts, signalling molecules and surface antigens that are similar to those seen in human ES cells, and different from mouse ES cells. This suggests that an early human embryonic “program” of gene expression is initiated in human nuclei by trans-acting (mouse) factors. Differences between the expression profiles of mouse ES cells and human reprogrammed nuclei probably reflect discrepancies in cis-acting regions between the mouse and human genomes. In agreement with this idea, a study in which the entire hTert gene was introduced into mice, showed expression of the transgene was similar to endogenous hTert in humans, rather than mouse endogenous mTert\n[43]. It is interesting to speculate that some of the well-publicised differences between human and mouse ES cells may indeed reflect intrinsic species dissimilarities, rather than temporal differences in stem cells isolation [44],[45]. We show that after fusion of human lymphocytes to mouse ES cells (that are Lif and Bmp dependent), human ES-like cells are generated that express FGF signalling components (and are not dependent of Lif/Bmp). Thus, our data suggest that differences between human and mouse ES cells may reflect distinct signalling and transcriptional networks, rather than necessarily when or where they were isolated during embryogenesis.We show here that Sox2, in contrast to Oct4, is not required to convert human lymphocytes into a multi-potent state. This observation contrasts with results obtained previously using iPS strategies to reprogram mouse and human fibroblasts [7]–[10],[38], mouse hepatocytes and stomach cells [46] and mouse B-lymphocytes [47]. Whether this is because of differences relating to the overexpression of transcription factor cocktails used in iPS, or that reprogramming occurs over an extended time period (pluripotency-associated genes such as Oct4, Nanog and Sox2 are reactivated after 2 weeks of transduction [48],[49]), is not known. However, as Sox2 was recently shown to be dispensable for the activation of Oct–Sox enhancers in mouse ES cells [37], it is also possible that additional Sox family members such as Sox4, Sox11 and Sox15, may have redundant functions with Sox2 in reprogramming. Interestingly, by enhancing Oct4 levels in Sox2-deficient ES cells (ES-2O1) we show elevated expression of hSox2 by reprogrammed human B cells. Recent genome-wide studies have shown that Sox2 is a direct target of Oct4 in both human [35] and mouse [50] ES cells, a fact that could explain why hSox2 is efficiently reprogrammed using ES cells that overexpress mOct4. In our hands, overexpression of exogenous Oct4 in lymphocytes did not induce pluripotent conversion (Pereira & Terranova, unpublished results), a finding that argues that additional chromatin remodelling factors, perhaps including those known to interact with Oct4 [51],[52] or associated with the process of DNA demethylation, may be critical for successful reprogramming. Collectively, our data show that interspecies heterokaryons can provide an interesting and complimentary approach to iPS, allowing the factors that are required to directly induce pluripotency to be defined individually and in combination.Materials and MethodsCell CultureEBV-transformed hB clones were maintained in RPMI supplemented with 10% foetal calf serum (FCS), 2 mM L-glutamine and antibiotics (10 µg/ml Penicillin and Streptomycin). The Abelson transformed Oct4-GFP B-cell line was derived from the Oct4-GFP transgenic mice (GOF18ΔPE) [53] bone marrow, cloned and grown in RPMI supplemented with 20% FCS, non-essential amino acids, L-glutamine, 50 µM 2-mercaptoethanol, antibiotics and IL-7 (5 ng/ml; R&D systems, Minneapolis, MN). Mouse ES cells were grown and maintained undifferentiated either on irradiated SNL feeder layers (E14Tg2a, Hprt\n−/− ES cells; CCE and E14) or directly on 0.1% gelatin-coated surfaces (OS25, ZHBTc4 and 2TS22C feeder-free ES cell lines). ES cells were grown in KO-DMEM medium plus 10% FCS, non-essential amino acids, L-glutamine, 2-mercaptoethanol, antibiotics and 1000 U/ml of leukaemia inhibitory factor (ESGRO-LIF). Feeder-free ES cell lines were cultured in GMEM-BHK21 medium plus 10% FCS, non-essential amino acids, sodium pyruvate, sodium bicarbonate, 2 mM L-glutamine, 2-mercaptoethanol, antibiotics and 1000 U/ml of LIF. Doxycycline (1 µg/ml, Sigma) or Retinoic acid (10−6 M, Sigma) were added to the media when indicated. The Flag-mOct4 cell lines were derived by the overexpression of Flag-tagged mouse Oct4 in E14tg2a ES cells. Briefly, mouse Oct4 cDNA was cloned in the pDFLAG-cDNAIII vector (Invitrogen). The cDNA, including two flag sequences at the 5′ end, was excised and sub-cloned into a suitable vector for expression in ES cells (pCBA), with expression driven by the chicken β-actin promoter. The vector was then linearised and transfected by electroporation into mouse ES cells. G418 selection (400 µg/ml; Invitrogen) was applied 48 hrs after and resistant clones were manually picked and screened by Western blot. Human ES cell lines H1, H7 and H9 cells [27] were cultured in medium conditioned by mitotically inactivated MEFs supplemented with 8 ng/ml bFGF (Peprotech, London, UK) on matrigel-coated plates, as previously described [54]. Cells were routinely passaged at a 1∶3 dilution by treatment with 200 U/ml collagenase IV (Invitrogen, Carlsbad, CA) and mechanical dissociation.Experimental HeterokaryonsHeterokaryons were generated by fusing ES cells and B-lymphocytes using 50% polyethylene glycol, pH7.4 (PEG 1500; Roche Diagnostics, Mannheim, Germany). Briefly, ES cells and hB-lymphocytes were respectively labelled with Vibrant 1,1′-dioctadecyl-3, 3, 3′, 3′ tetramethylindodicarbocyanine (DiD) and 1,1′-dioctadecyl-3, 3, 3′, 3′-tetramethylindocarbocyanine perchlorate (DiI) cell labelling solutions (Molecular Probes, Eugene, OR). Cells were resuspended at 1×106 cells/ml in DMEM and labelled with 5 µl/ml of dye at 37°C, 15 min. ES and hB were then mixed in an appropriate ratio (ES∶hB ratio 1∶1; ES∶Oct4-GFPB ratio 1∶5), and were washed twice in PBS. The supernatant was completely removed and 1 ml of PEG (37°C) was added to the pellet of cells over 60 sec and incubated at 37°C for 90 sec with constant stirring. Then, 4 ml of serum-free medium (DMEM) were carefully added over a period of 3 min, followed by 10 ml of DMEM and incubation at 37°C for 3 min. After centrifugation (1350 rpm, 5 min), the pellet was allowed to swell in complete medium for 3 min. Cell mixtures were then resuspended and cultured under conditions promoting the maintenance of undifferentiated mouse ES cells at 0.5×106 cells/cm2. To eliminate unfused hB cells, Ouabain (10−5 M; Sigma) was added to the medium 4 hours after cell fusion. When OS25, ZHBTc4 and 2TS22C cell lines were used, proliferating ES cells were eliminated by the addition of 10−5 M Ara-C (Cytosine β-D arabino furanoside; Sigma) 4–6 hours after fusion and then removed after 16 hours. When E14tg2a ES cells or derivatives were used, HAT (20 µM hypoxanthine, 0.08 µM aminopterine and 3.2 µM thymidine; Sigma) was added to the medium 24 hours after fusion.Quantitative RT-PCR AnalysisRNA extraction was performed using RNA-BEE reagent (Tel-Test Inc., Friendswood, TX) and residual DNA was eliminated using the DNA-free kit (Ambion, Austin, TX). 3 µg of total RNA was then reverse transcribed using Superscript First-Strand Synthesis system (Qiagen) with oligo (dT)12-18 (Invitrogen). cDNAs of interest were then quantified using real-time qPCR amplification. Real-time PCR analysis was carried out on a Opticon DNA engine using Opticon Monitor software (MJ Research Inc., Waltham, MA), running the following program: 95°C for 15 min, then 40 cycles of 94°C for 15 sec, 60°C for 30 sec, 72°C for 30 sec, followed by plate-read. PCR reactions included 2× Sybr-Green PCR Mastermix (Qiagen), 300 nM primers and 2 µl of template in a 35 µl reaction volume. Each measurement was performed in triplicate and data normalised according to Gapdh expression. Primers were designed with Primer Express software (Applied Biosystems) and tested for the specific detection of human transcripts (and not mouse). Standard curves were calculated on serial dilutions of positive control cDNA. Primer sequences used for this analysis are indicated in Table S1.Bisulfite Genomic SequencingBisulfite modification of DNA was carried out with the EZDNA methylation kit (Zymogenetics Inc., Orange, CA) according to manufacturer's recommendations. PCR primers that recognise bisulfite-converted human DNA only are listed in Table S1. Amplified products were cloned into pCR2 (Invitrogen) and ten clones were randomly picked and sequenced.Antibodies, Imaging, and FACS AnalysisFor immunofluorescence and FACS analysis, the following antibodies and dilutions were used: mouse monoclonal anti-human Lamin A/C (VP-L550; Vector Laboratories Inc., Burlingame, CA) at 1∶100 dilution; rabbit polyclonal anti-GFP (A11122; Molecular Probes) at 1∶200 dilution; mouse monoclonal anti-human SSEA4 (MC-813-70; Developmental Hybridoma Studies Bank, Iowa City, IA) at 1∶3 dilution; mouse monoclonal anti-human TRA-1-60 and TRA-1-81 (MAB4360 and MAB4381; Chemicon International, Temecula, CA) at 1∶12 and 1∶20 dilutions, respectively; rabbit polyclonal anti-human Nanog and Nestin (ab21624 and ab28944; Abcam Ltd., Cambridge, UK) at 1∶100 dilution; mouse monoclonal anti-Flag (F3165, Sigma) at 1∶1000 dilution. Secondary antibodies conjugated with fluorochromes were purchased from Molecular Probes and used at 1∶400 dilution. Immunofluorescence was performed as previously described [13]. Mouse and human nuclei were distinguished in the resulting heterokaryons by counterstaining with 4,6-diamidino-2-phenylindole (DAPI) and human Lamin A/C staining. Individual cells were delineated by F-actin staining (Phalloidin; A12380, Molecular Probes). For alkaline phosphatase assays, hybrid colonies 8 days after cell fusion were stained with alkaline phosphatase assay kit (Sigma). All slides were analyzed on a Leica TCS SP5 confocal microscope and processed with Leica software and Adobe Photoshop. Images of live GFP fluorescent hybrid colonies and alkaline phosphatase staining were collected using a Leica DM IRE2 microscope running Metamorph software. For FACS analysis a FACScalibur (BD Biosciences) with CellQuest software was used. FACS purification was performed using a FACSAria cell sorter. Western blot analysis was performed as previously described [55] using a goat anti-Oct3/4 polyclonal antibody (sc-8628; Santa Cruz Biotechnology Inc., Santa Cruz, CA) or a mouse anti-Flag monoclonal antibody. As a loading control, blots were incubated with anti-Lamin B polyclonal antibody (sc-6216; Santa Cruz Biotechnology Inc.). Each lane contained 20 ìg total protein.Supporting InformationFigure S1Characterisation of heterokaryon reprogramming of fused hB x mES cells. (A) Human B-lymphocytes (hB) and mouse embryonic stem cells (mES) were respectively labelled with the cell membrane dyes DiI and DiD and fused in the presence of polyethylene glycol (PEG). Fused cells, identified by double-labelling (upper right quadrant), were sorted by FACS and cultured. (B) Mouse and human nuclei were distinguished by FISH using probes specific for mouse γ-satellite DNA (red) or human α-satellite DNA (green), and DAPI counterstained (blue). Confocal sections of human B cells (hB) and mouse ES cells (mES) before and after cell fusion (hB x mES) are shown. Heterokaryons (cells in which parental nuclei share the same cytoplasm but remain discrete, day 1 and 2) were identified up to 2 days after fusion, but by day 3 hybrid formation (where genomes are mixed in the same nucleus, day 3) was detected. Scale bar, 10 µm. (C) Expression of human ES-specific (hOct4, hNanog) and human lymphocyte-specific (hCD20, hCD45) transcripts detected by RT-PCR using human-specific primers. Prior to fusion, hB cells expressed hGapdh, hCD20 and hCD45 but not embryonic stem cell-specific genes. Following heterokaryon formation (hB x mES d2), human pluripotency-associated genes hOct4 and hNanog were expressed (upper panel) and hCD20 and hCD45 were extinguished (lower panel). mES, -RT and H2O were used as negative controls and human embryonic stem cells (hES) as a positive control. hGapdh was used to standardise input. (D) Expression of human hTert transcripts detected by qRT-PCR 0 to 8 days after cell fusion using human-specific primers. Positive (hES-NCL1, black bars) and negative (hB) controls for this analysis were included. Data were normalised to hGapdh expression. Error bars indicate the s.d. of 3 independent experiments.(5.32 MB TIF)Click here for additional data file.Figure S2Differences between human and mouse ES cells and the identification of SSEA4 positive reprogrammed cells. (A) Expression of Fgfr1, Fgfr2, Fgf2, Bmp4, Lifr, and Jak3 was assessed by qRT-PCR in human ES cells (hES, NCL1), mouse ES cells (mES) and human B-lymphocytes (hB). Fgfr1, Fgfr2, and Fgf2 were uniquely expressed by human ES cells. (B) FACS analysis showed that >90% of hES cells (H1 cell line) expressed SSEA4, while hB and mES do not (2.1% and 1.5% respectively). A proportion of heterokaryons showed SSEA4 expression (15.8%) 8 days after cell fusion (hB x mES d8). (C) FACS sorting of SSEA4 positive cells co-purifies reprogrammed cells that express hOct4, hNanog, and hCripto, as assessed by qRT-PCR. Data were normalised to Gapdh expression.(0.74 MB TIF)Click here for additional data file.Figure S3Expression of human-specific embryonic antigens in hybrid cells. Human B cells (hB) and mouse ES cells (mES) were fused and the resulting colonies (hB x mES, day 8) expressed hNanog protein (red) and the human ES-specific antigens SSEA4, TRA-1-81 and TRA-1-60 (green) as assessed by immunofluorescence. Control hB cells did not express any of the markers. DAPI staining is shown in blue. Images are single confocal sections. Scale bar, 50 µm.(5.14 MB TIF)Click here for additional data file.Figure S4Kinetic analysis of Oct4 protein distribution in heterokaryons and the importance of Oct4 for successful reprogramming. (A) Flag-mOct4 ES cells were fused to hB cells and Oct4 protein detected by immunofluorescence at 0, 1, 3, 6, 9, and 12 hours with Oct4 or Flag antibodies (green). Heterokaryons were scored according to the following Oct4 distribution: Oct4 protein not detected (Negative), stronger staining in mES-derived nucleus than hB nucleus (mES>hB), nuclei equally labelled (mES = hB), stronger in the human nucleus (mES<hB). Confocal sections of representative heterokaryons from each of the categories are shown (upper panels). Human nuclei were distinguished from mouse nuclei on basis of diffuse versus punctuate DAPI staining (blue), respectively. Actin labelling (red) delineates the cell membrane. Scale bar, 10 µm. n = 100. (B) The ability of mouse ES cells to fuse to human B cells is unaffected by doxycicline (Dox) treatment. ZHBTc4 and hB cells were labelled (with DiD and DiI, respectively) and PEG-fused. Fusion efficiencies were obtained by FACS, as a percentage of double-labelled cells. (C) ZHBTc4 ES cells expressing Oct4 (black bars), or in which Oct4 expression has been partially or completely ablated (grey and white bars, respectively) were fused to hB-lymphocytes. The activation of human ES-specific genes (hOct4, hNanog, hCripto, hDnmt3b, hSox2, hTle1, hTert, and hRex1) and silencing of lymphocyte-specific genes (hCD19, hCD45, and hCD37) were quantified by qRT-PCR over the period of 3 days after cell fusion. hHprt was added as a control gene. Data were normalised to hGapdh expression. Error bars indicate the s.d. of 2–3 independent experiments.(3.61 MB TIF)Click here for additional data file.Figure S5siRNA-mediated knock-down of mOct4 abolishes reprogramming. (A) E14tg2a ES cells were transfected with either mOct4-siRNA or target-less-siRNA (a negative control siRNA designed to have no expected targets in human and mouse cells) vectors. 48 hours later, transfected cells (GFP+) were FACS sorted and analysed by quantitative RT-PCR analysis. mOct4-siRNA targeted cells showed a >90% reduction in Oct4 transcript levels as compared to cells transfected with target-less-siRNA (control). (B) E14tg2a ES cells expressing mOct4-siRNA or control-siRNA were fused to hB-lymphocytes, and successful reprogramming was assessed by quantifying the abundance of human ES-associated transcripts (hNanog and hCripto) two days after fusion by qRT-PCR. Successful reprogramming judge by the activation of human pluripotency-associated transcripts was abolished by pre-treatment of mES cells with Oct4-siRNAs. Data were normalised to Gapdh expression. Error bars indicate the s.d. of 2 independent experiments.(0.82 MB TIF)Click here for additional data file.Figure S6Kinetic of human lymphocyte reprogramming by mES cells after Sox2 ablation. 2TS22C (black bars), Sox2 depleted cells (grey and white bars; Dox 12 and 24 hours, respectively) and 2O1 cells (red bars; Sox2-deficient mES cells in which mOct4 expression is constitutively up-regulated) were used as fusion partners with hB cells and reprogramming was assessed by quantification of human-ES transcripts (hOct4, hNanog, hCripto, hDnmt3b, hSox2, hTle1, hTert and hRex1) using qRT-PCR over 3 days after cell fusion. hHprt was added as a control gene. Data were normalised to hGapdh expression. Error bars indicate the s.d. of 2–3 independent experiments.(0.59 MB TIF)Click here for additional data file.Figure S7Characterisation of mouse embryonic hybrid cells. (A) Contribution of the lymphocyte genome within hybrid cells was confirmed by detection of a rearranged IgH locus (D–J region). IgH rearrangement was seen in B-lymphocytes (mB), hybrid cells (mES x mB) but not in mES cells. The rearranged DNA can be detected by PCR amplification and visualized on the gel as a 750 bp band. (B) Lymphocyte-specific genes (mCD19, mPax5, and mLy108) were not detected in hybrid cells although ES-specific genes (mOct4, mNanog, mSox2, mRex1, and mUtf1) remain detectable by RT-PCR. (C) Specific detection of Oct4 transgene (Oct4βgeo) by RT-PCR with primers within βgeo cassette, which specifically amplify ZHBTc4-derived Oct4 but not endogenous mOct4. mES and mB cells were included as controls. mGapdh was used to standardise input. (D) Doxycycline (Dox) treatment of ZHBTc4 ES cells results in morphological changes characteristic of trophectoderm differentiation (upper panel). These were not observed in hybrid clones 4 and 12 under the same conditions. GFP protein (Oct4 promoter-driven) remains detectable in hybrid cells throughout the experiment, as assessed by immunofluorescence.(7.86 MB TIF)Click here for additional data file.Table S1Primers used in this study.(0.17 MB DOC)Click here for additional data file.Text S1Supplementary methods.(0.04 MB DOC)Click here for additional data file.\n\nREFERENCES:\n1. HochedlingerKJaenischR\n2002\nMonoclonal mice generated by nuclear transfer from mature B and T donor cells.\nNature\n415\n1035\n1038\n11875572\n2. WilmutISchniekeAEMcWhirJKindAJCampbellKH\n1997\nViable offspring derived from fetal and adult mammalian cells.\nNature\n385\n810\n813\n9039911\n3. GurdonJB\n1962\nAdult frogs derived from the nuclei of single somatic cells.\nDev Biol\n4\n256\n273\n13903027\n4. ByrneJAPedersenDAClepperLLNelsonMSangerWG\n2007\nProducing primate embryonic stem cells by somatic cell nuclear transfer.\nNature\n450\n497\n502\n18004281\n5. DavisRLWeintraubHLassarAB\n1987\nExpression of a single transfected cDNA converts fibroblasts to myoblasts.\nCell\n51\n987\n1000\n3690668\n6. XieHYeMFengRGrafT\n2004\nStepwise reprogramming of B cells into macrophages.\nCell\n117\n663\n676\n15163413\n7. TakahashiKYamanakaS\n2006\nInduction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors.\nCell\n126\n663\n676\n16904174\n8. WernigMMeissnerAForemanRBrambrinkTKuM\n2007\nIn vitro reprogramming of fibroblasts into a pluripotent ES-cell-like state.\nNature\n448\n318\n324\n17554336\n9. YuJVodyanikMASmuga-OttoKAntosiewicz-BourgetJFraneJL\n2007\nInduced pluripotent stem cell lines derived from human somatic cells.\nScience\n318\n1917\n1920\n18029452\n10. TakahashiKTanabeKOhnukiMNaritaMIchisakaT\n2007\nInduction of pluripotent stem cells from adult human fibroblasts by defined factors.\nCell\n131\n861\n872\n18035408\n11. YingQLNicholsJEvansEPSmithAG\n2002\nChanging potency by spontaneous fusion.\nNature\n416\n545\n548\n11932748\n12. WeimannJMJohanssonCBTrejoABlauHM\n2003\nStable reprogrammed heterokaryons form spontaneously in Purkinje neurons after bone marrow transplant.\nNat Cell Biol\n5\n959\n966\n14562057\n13. TerranovaRPereiraCFDu RoureCMerkenschlagerMFisherAG\n2006\nAcquisition and extinction of gene expression programs are separable events in heterokaryon reprogramming.\nJ Cell Sci\n119\n2065\n2072\n16638804\n14. TadaMTadaTLefebvreLBartonSCSuraniMA\n1997\nEmbryonic germ cells induce epigenetic reprogramming of somatic nucleus in hybrid cells.\nEmbo J\n16\n6510\n6520\n9351832\n15. TadaMTakahamaYAbeKNakatsujiNTadaT\n2001\nNuclear reprogramming of somatic cells by in vitro hybridization with ES cells.\nCurr Biol\n11\n1553\n1558\n11591326\n16. Tae DoJHanDWGentileLSobek-KlockeIStehlingM\n2007\nErasure of cellular memory by fusion with pluripotent cells.\nStem Cells\n25\n1013\n1020\n17218392\n17. FlaszaMSheringAFSmithKAndrewsPWTalleyP\n2003\nReprogramming in inter-species embryonal carcinoma-somatic cell hybrids induces expression of pluripotency and differentiation markers.\nCloning Stem Cells\n5\n339\n354\n14733752\n18. CowanCAAtienzaJMeltonDAEgganK\n2005\nNuclear reprogramming of somatic cells after fusion with human embryonic stem cells.\nScience\n309\n1369\n1373\n16123299\n19. KimuraHTadaMNakatsujiNTadaT\n2004\nHistone code modifications on pluripotential nuclei of reprogrammed somatic cells.\nMol Cell Biol\n24\n5710\n5720\n15216876\n20. SuraniMA\n2001\nReprogramming of genome function through epigenetic inheritance.\nNature\n414\n122\n128\n11689958\n21. SimonssonSGurdonJ\n2004\nDNA demethylation is necessary for the epigenetic reprogramming of somatic cell nuclei.\nNat Cell Biol\n6\n984\n990\n15448701\n22. HanDWDoJTGentileLStehlingMLeeHT\n2008\nPluripotential reprogramming of the somatic genome in hybrid cells occurs with the first cell cycle.\nStem Cells\n26\n445\n454\n18065396\n23. FrebergCTDahlJATimoskainenSCollasP\n2007\nEpigenetic reprogramming of OCT4 and NANOG regulatory regions by embryonal carcinoma cell extract.\nMol Biol Cell\n18\n1543\n1553\n17314394\n24. Deb-RinkerPLyDJezierskiASikorskaMWalkerPR\n2005\nSequential DNA methylation of the Nanog and Oct-4 upstream regions in human NT2 cells during neuronal differentiation.\nJ Biol Chem\n280\n6257\n6260\n15615706\n25. UlanerGAYangYHuJFLiTVuTH\n2003\nCTCF binding at the insulin-like growth factor-II (IGF2)/H19 imprinting control region is insufficient to regulate IGF2/H19 expression in human tissues.\nEndocrinology\n144\n4420\n4426\n12960026\n26. StojkovicMLakoMStojkovicPStewartRPrzyborskiS\n2004\nDerivation of human embryonic stem cells from day-8 blastocysts recovered after three-step in vitro culture.\nStem Cells\n22\n790\n797\n15342943\n27. ThomsonJAItskovitz-EldorJShapiroSSWaknitzMASwiergielJJ\n1998\nEmbryonic stem cell lines derived from human blastocysts.\nScience\n282\n1145\n1147\n9804556\n28. Lande-DinerLCedarH\n2005\nSilence of the genes–mechanisms of long-term repression.\nNat Rev Genet\n6\n648\n654\n16012529\n29. JenuweinTAllisCD\n2001\nTranslating the histone code.\nScience\n293\n1074\n1080\n11498575\n30. GilbertDM\n2002\nReplication timing and transcriptional control: beyond cause and effect.\nCurr Opin Cell Biol\n14\n377\n383\n12067662\n31. PeraMFTrounsonAO\n2004\nHuman embryonic stem cells: prospects for development.\nDevelopment\n131\n5515\n5525\n15509763\n32. XuRHPeckRMLiDSFengXLudwigT\n2005\nBasic FGF and suppression of BMP signaling sustain undifferentiated proliferation of human ES cells.\nNat Methods\n2\n185\n190\n15782187\n33. GinisILuoYMiuraTThiesSBrandenbergerR\n2004\nDifferences between human and mouse embryonic stem cells.\nDev Biol\n269\n360\n380\n15110706\n34. SatoNSanjuanIMHekeMUchidaMNaefF\n2003\nMolecular signature of human embryonic stem cells and its comparison with the mouse.\nDev Biol\n260\n404\n413\n12921741\n35. BoyerLALeeTIColeMFJohnstoneSELevineSS\n2005\nCore transcriptional regulatory circuitry in human embryonic stem cells.\nCell\n122\n947\n956\n16153702\n36. NiwaHMiyazakiJSmithAG\n2000\nQuantitative expression of Oct-3/4 defines differentiation, dedifferentiation or self-renewal of ES cells.\nNat Genet\n24\n372\n376\n10742100\n37. MasuiSNakatakeYToyookaYShimosatoDYagiR\n2007\nPluripotency governed by Sox2 via regulation of Oct3/4 expression in mouse embryonic stem cells.\nNat Cell Biol\n9\n625\n635\n17515932\n38. MaheraliNSridharanRXieWUtikalJEminliS\n2007\nDirectly Reprogrammed Fibroblasts Show Global Epigenetic Remodeling and Widespread Tissue Contribution.\nCell Stem Cell\n1\n55\n70\n18371336\n39. BlauHMPavlathGKHardemanECChiuCPSilbersteinL\n1985\nPlasticity of the differentiated state.\nScience\n230\n758\n766\n2414846\n40. KoziolMJGarrettNGurdonJB\n2007\nTpt1 activates transcription of oct4 and nanog in transplanted somatic nuclei.\nCurr Biol\n17\n801\n807\n17442571\n41. WongCCGaspar-MaiaARamalho-SantosMReijo PeraRA\n2008\nHigh-Efficiency Stem Cell Fusion-Mediated Assay Reveals Sall4 as an Enhancer of Reprogramming.\nPLoS ONE\n3\ne1955\n18414659\n42. PralongDMrozikKOcchiodoroFWijesundaraNSumerH\n2005\nA novel method for somatic cell nuclear transfer to mouse embryonic stem cells.\nCloning Stem Cells\n7\n265\n271\n16390262\n43. HorikawaIChiangYJPattersonTFeigenbaumLLeemSH\n2005\nDifferential cis-regulation of human versus mouse TERT gene expression in vivo: identification of a human-specific repressive element.\nProc Natl Acad Sci U S A\n102\n18437\n18442\n16344462\n44. TesarPJChenowethJGBrookFADaviesTJEvansEP\n2007\nNew cell lines from mouse epiblast share defining features with human embryonic stem cells.\nNature\n448\n196\n199\n17597760\n45. BronsIGSmithersLETrotterMWRugg-GunnPSunB\n2007\nDerivation of pluripotent epiblast stem cells from mammalian embryos.\nNature\n448\n191\n195\n17597762\n46. AoiTYaeKNakagawaMIchisakaTOkitaK\n2008\nGeneration of Pluripotent Stem Cells from Adult Mouse Liver and Stomach Cells.\nScience\nDOI: 10.1126\n47. HannaJMarkoulakiSSchorderetPCareyBWBeardC\n2008\nDirect reprogramming of terminally differentiated mature B lymphocytes to pluripotency.\nCell\n133\n250\n264\n18423197\n48. BrambrinkTForemanRWelsteadGGLengnerCJWernigM\n2008\nSequential expression of pluripotency markers during direct reprogramming of mouse somatic cells.\nCell Stem Cell\n2\n151\n159\n18371436\n49. StadtfeldMMaheraliNBreaultDTHochedlingerK\n2008\nDefining Molecular Cornerstones during Fibroblast to iPS Cell Reprogramming in Mouse.\nCell Stem Cell\n2\n230\n240\n18371448\n50. LohYHWuQChewJLVegaVBZhangW\n2006\nThe Oct4 and Nanog transcription network regulates pluripotency in mouse embryonic stem cells.\nNat Genet\n38\n431\n440\n16518401\n51. WangJRaoSChuJShenXLevasseurDN\n2006\nA protein interaction network for pluripotency of embryonic stem cells.\nNature\n444\n364\n368\n17093407\n52. LiangJWanMZhangYGuPXinH\n2008\nNanog and Oct4 associate with unique transcriptional repression complexes in embryonic stem cells.\nNat Cell Biol\n10\n731\n739\n18454139\n53. PalmieriSLPeterWHessHScholerHR\n1994\nOct-4 transcription factor is differentially expressed in the mouse embryo during establishment of the first two extraembryonic cell lineages involved in implantation.\nDev Biol\n166\n259\n267\n7958450\n54. XuCInokumaMSDenhamJGoldsKKunduP\n2001\nFeeder-free growth of undifferentiated human embryonic stem cells.\nNat Biotechnol\n19\n971\n974\n11581665\n55. AzuaraVPerryPSauerSSpivakovMJorgensenHF\n2006\nChromatin signatures of pluripotent cell lines.\nNat Cell Biol\n8\n532\n538\n16570078"
+}

batch_0/PMC2528010.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2528010",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2528010\nAUTHORS: Andrew A Adjei, Henry B Armah, Foster Gbagbo, Isaac Boamah, Clement Adu-Gyamfi, Isaac Asare\n\nABSTRACT:\nBackgroundHuman herpesvirus 8 (HHV-8), cytomegalovirus (CMV) and Epstein-Barr virus (EBV) are prevalent in Africa, but less common elsewhere and the modes of transmission are still subject to debate. Generally, they rarely cause disease in the immunocompetent host but are highly oncogenic when associated with immunosuppression. Although the high prevalence of HHV-8, CMV and EBV has been well documented in Africa, such data are sparse from Ghana.MethodsSerum samples from 3275 HIV-seronegative healthy blood donors and 250 HIV-AIDS patients were tested for antibodies specific for HHV-8, CMV and EBV by IgG ELISA assays. Differences in seropositivity rates by gender and age were evaluated using the Chi-square test with Yates correction.ResultsOf the 3275 HIV-seronegative healthy blood donors tested, 2573 (78.6%) were males and 702 (21.4%) were females, with ages ranging from 18 to 65 years (median 32.6; mean 31.2; mode 30). Of the 250 HIV-AIDS patients tested, 140 (56%) were males and 110 (44%) were females, with ages ranging from 17 to 64 years (median 30.8; mean 30.3; mode 28). Among the HIV-seronegative healthy blood donors, overall seroprevalence of HHV-8, CMV and EBV was 23.7%, 77.6% and 20.0%, respectively. Among the HIV-AIDS patients, overall seroprevalence of HHV-8, CMV and EBV was 65.6%, 59.2% and 87.2%, respectively. The seroprevalence of HHV-8 (p < 0.005) and EBV (p < 0.001) was statistically significantly higher in HIV-AIDS patients compared to HIV-seronegative healthy blood donors. There was no statistically significant difference (p = 0.24) between CMV seroprevalence in HIV-AIDS patients and HIV-seronegative healthy blood donors. Age and gender were not independent determinants (p > 0.05) for all three infections among HIV-seronegative healthy blood donors and HIV-AIDS patients in Ghana.ConclusionThe results presented herein indicate that HHV-8, CMV and EBV infections are hyperendemic in both HIV-seronegative and HIV-seropositive Ghanaians, and suggest primarily a horizontal route of transmission of these three viral infections in Ghana.\n\nBODY:\nBackgroundThere are currently eight known human herpesviruses: cytomegalovirus (CMV), Epstein-Barr virus (EBV), herpes simplex virus 1, herpes simplex virus 2, human herpesvirus 6, human herpesvirus 7, human herpesvirus 8 (HHV-8), and varicella-zoster virus. All eight, except herpesvirus 6 and herpesvirus 7, are known to be pathogenic to humans. HHV-8 is also known as Kaposi's sarcoma-associated herpesvirus (KSHV). HHV-8, CMV, and EBV are lymphotropic herpesviruses and responsible for a wide variety of human diseases, caused either by primary infection or by reactivation under immunosuppressive conditions. The majority (>90%) of the adult human population carries asymptomatic infection of EBV and CMV. Although HHV-8 shares substantial homology with EBV, it has a marked lower (2–30%) seroprevalence rate in the adult human population, with a specific tropism for people of Mediterranean and sub-Sahara African countries [1-4]. HHV-8 and EBV are oncogenic viruses with a long latency period in healthy hosts and will reactivate from dormancy when the hosts are immunosuppressed. Primary infections with these viruses in the immunocompetent host are generally asymptomatic. The neoplastic potentials of these two viruses have been well established, especially within the context of immunosuppressed patients who are undergoing bone-marrow transplantation or are co-infected with the human immunodeficiency virus (HIV) [5].HHV-8 is a γ-herpesvirus that was discovered in 1994 in Kaposi's sarcoma (KS) tissues from a patient with AIDS, thereby establishing a link between HHV-8 infection and the emergence of KS. HHV-8 is now considered to be the etiological agent of all the clinico-epidemiological forms of KS (including AIDS KS, classic KS, endemic KS, and iatrogenic KS), primary effusion lymphoma, body cavity-based lymphoma, and multicentric Castleman's disease. Several studies show high prevalence rates of HHV-8 antibodies among male homosexuals, African children, Brazilian Amerindians, and elderly individuals in certain regions of the Mediterranean basin [4]. Sexual transmission of HHV-8 might play an important role among high-risk group populations, such as homosexual men in Western countries. However, in endemic areas where HHV-8 seroprevalence is high during childhood and adolescence, viral transmission might occur through nonsexual contact. This is particularly evident in African populations where high prevalence rates have been observed in infants and children, with a HHV-8 seroprevalence similar to that observed in adults [4].CMV is a β-herpesvirus and known to be present in saliva, cervical secretions, breast milk, semen, and human lymphocytes. CMV is an ubiquitous agent, and seropositivity rates in the adult population over 40 years of age worldwide are 60 to 100%, possibly due to transmission through breastfeeding, sexual contact and spread from children [6,7]. Transfusion-transmitted CMV infection is a significant cause of morbidity and mortality, particularly in immunocompromised patients (including premature low-birth-weight infants [<1500 g] born to CMV-seronegative mothers, CMV-seronegative recipients of autologous or allogeneic bone marrow or peripheral blood stem cell transplantation, CMV-seronegative solid-organ transplant recipients, and CMV-seronegative patients with AIDS [8]. In all of these at-risk patients, it is appropriate to provide \"CMV-safe\" blood for transfusion.EBV was first discovered in 1964 in Burkitt lymphoma (BL), a B-cell-derived tumor. EBV is ubiquitous in the adult population worldwide, and establishes a life-long persistent infection of B lymphocytes characterized by virus shedding into saliva [9]. African children are infected early in life and most have seroconverted by age 3 years, while in affluent countries, primary infection is delayed until young adult life [10]. EBV is now considered to be etiologically associated with endemic Burkitt's lymphoma (BL), nasopharyngeal carcinoma, classical Hodgkin's lymphoma (HL) and extranodal nasal NK/T-cell lymphoma. EBV is transmitted via saliva in an oral-fecal route of transmission, and it infects B lymphocytes as well as certain epithelial cells.In a recent review of 28 HHV-8 seroepidemiologic studies of adult populations from 16 African countries reported between 1996 and 2002, most African countries (namely Botswana, Cameroon, Democratic Republic of Congo, Egypt, Gambia, Ghana, Ivory Coast, Nigeria, South Africa, Tanzania, Uganda, Zambia, and Zimbabwe) had high seroprevalence rates ranging from 26% to 86%, with the exception of Central African Republic, Eritrea and Senegal which had relatively lower (up to 25%) seroprevalence rates [11]. Another recent review of 7 HHV-8 seroepidemiologic studies of pediatric populations from 7 African countries reported between 1998 and 2003 showed high seroprevalence rates ranging from 30% to 58.1% in most African countries (namely Cameroon, Egypt, Ghana, Tanzania, Uganda, and Zambia), with the exception of Eritrea which had a very low (up to 2%) seroprevalence rate [4]. Although Kaposi's sarcoma is common in Ghana compared to other cutaneous malignancies [12], data on the seroprevalence of HHV-8 in Ghana are scanty. Ablashi and colleagues first reported a HHV-8 seroprevalence rate of 41.9% in healthy Ghanaians aged 13–72 years in 1999 [3]. Subsequently, Nuvor and colleagues reported statistically significant (p < 0.05) difference in HHV-8 seroprevalence rate between HIV-seronegative healthy blood donors (32.3%) and asymptomatic HIV-seropositive individuals (45.5%) in Ghana [13]. Data on the prevalence of CMV and EBV infections in Ghana are even scantier. The reported seroprevalence rate of CMV among healthy Ghanaian blood donors aged 18–60 years was 93.2% [14]. EBV viral DNA was detected in plasma samples of 40% (47% in the malaria-infected and 34% in the non-malaria group) of Ghanaian children aged 6 months to 12 years [15]. The aim of this study was to determine and compare the seroprevalence of HHV-8, CMV, and EBV infections among HIV-AIDS patients and HIV-seronegative healthy blood donors in Ghana, in an effort to further define the seroepidemiology and transmission of these infections in Ghana.MethodsStudy sites and populationsThree thousand two hundred and seventy-five HIV-seronegative serum samples were obtained from 7 of the 10 regional blood banks of the Ghana National Blood Transfusion Service from healthy blood donors who gave their informed consent between 2001 and 2002. In Ghana, blood donors are volunteers (both occasional and periodic repeat non-compensated volunteer donors) and are also sought from family members and friends of patients requiring blood transfusion. They are selected based on the following criteria: age between 18 and 65 years; weight >45 kg; haemoglobin >12.5 g/dl; normal blood pressure [BP], pulse, and body temperature; and not belonging to any high-risk group (homosexually or heterosexually promiscuous, intravenous drug users; history of sexually transmitted diseases; and history of any severe current or chronic illnesses). Donated blood is routinely screened for HIV 1 & 2, HBsAg, anti-HCV and syphilis antibodies in Ghana.Two hundred and fifty HIV-AIDS patients with chronic diarrhea on admission to the Fevers Unit of the Korle-Bu Teaching Hospital, Accra, Ghana, were recruited for the study between February 2001 and June 2002. The Korle-Bu Teaching Hospital is the leading tertiary hospital in Ghana that serves the city of Accra, its surrounding urban population, and the southern part of Ghana. Accra, the capital city of Ghana, is a rapidly expanding city with a population of approximately 3 million. All 250 HIV-AIDS patients (with a single infection from HIV-1) fell within the Centers for Disease Control and Prevention (CDC) clinical staging A1–C3 categories, representing asymptomatic to severe AIDS conditions. The participating HIV-AIDS patients had mean CD4 counts of 288 cells per microliter (95% confidence interval of 237–340 cells per microliter). All participating HIV-AIDS patients reported watery stools lasting from 3–90 days (diarrhea episodes, 3–10 watery stools per day). The HIV testing for antibodies to HIV-1 and HIV-2 in the study HIV-AIDS patients and blood donors was done by the particle agglutination test (Serodia HIV-1 and HIV-2; Serodia Fujirebio Inc., Tokyo, Japan) and confirmed by Western blot analysis (New Lav Blot I and II; Sanofi Diagnostic Pasteur, Marnes-la-Coquette, France). The study received ethical review and approval from the Ethical and Protocol Review Committee of the University of Ghana Medical School, Accra, Ghana, and written informed consent was obtained from all study participants.Serological analysisSera were tested at the Virology Unit, Noguchi Memorial Institute for Medical Research, for the presence of antibodies to HHV-8 (ELISA; IgG; Advanced Biotechnologies, Columbia, Maryland, USA), antibodies to CMV (ELISA; IgG; Diamedix Corporation, Miami, Florida, USA), and antibodies to EBV (ELISA; IgG antibody to viral capsid antigen; Advanced Biotechnologies, Columbia, Maryland, USA), in accordance with the respective manufacturer's instructions. Positive and negative standard sera, accompanying the kit were included in each assay.Statistical analysisThe Statistical Analysis System (SAS Institute, Cary, NC, USA) version 9.1 was used to complete all data analyses. Seropositivity rates were calculated and compared by gender and among different 10-year interval age groups. Differences were evaluated using the Chi-square test with Yates correction. A P value of < 0.05 was considered statistically significant.ResultsOf the 3275 HIV-seronegative healthy blood donors tested, 2573 (78.6%) were males and 702 (21.4%) were females, with ages ranging from 18 to 65 years (median 32.6; mean 31.2; mode 30). Of the 250 HIV-AIDS patients tested, 140 (56%) were males and 110 (44%) were females, with ages ranging from 17 to 64 years (median 30.8; mean 30.3; mode 28).Table 1 shows HHV-8, CMV and EBV seropositivity according to age and gender among the 3275 HIV-seronegative healthy blood donors in Ghana. Among the 3275 HIV-seronegative healthy blood donors, overall seroprevalence of HHV-8, CMV and EBV was 23.7%, 77.6% and 20.0%, respectively (Table 1). There was no statistically significant (p > 0.05) difference in the overall seroprevalence of HHV-8 between male and female HIV-seronegative healthy blood donors. Additionally, there was no statistically significant (p > 0.05) difference in the overall seroprevalence of CMV between male and female HIV-seronegative healthy blood donors. Finally, there was no statistically significant (p > 0.05) difference in the overall seroprevalence of EBV between male and female HIV-seronegative healthy blood donors [Table 1]. Hence, gender was not an independent determinant (p > 0.05) for all three infections among HIV-seronegative healthy blood donors in Ghana.Table 1HHV-8, CMV and EBV seropositivity according to age and gender among the 3275 HIV-seronegative healthy blood donors in GhanaAge group, yearsProportion (%) of HIV-seronegative blood donors with HHV8 seropositivityProportion (%) of HIV-seronegative blood donors with CMV seropositivityProportion (%) of HIV-seronegative blood donors with EBV seropositivityMaleFemaleAllMaleFemaleAllMaleFemaleAll16–25104/681 (15.3)65/335 (19.4)169/1016 (16.6)558/681 (81.9)232/335 (69.3)790/1016 (77.8)108/681 (15.9)74/335 (22.1)182/1016 (17.9)26–35241/1120 (21.5)88/220 (40.0)329/1340 (24.6)831/1120 (74.2)158/220 (71.8)989/1340 (73.8)35/1120 (3.1)28/220 (12.7)63/1340 (4.7)36–45118/478 (24.7)29/74 (39.2)147/552 (26.6)423/478 (88.5)57/74 (77.0)480/552 (87.0)223/478 (46.7)36/74 (48.6)259/552 (46.9)46–5548/158 (30.4)12/42 (28.6)60/200 (30.0)101/158 (63.9)37/42 (88.1)138/200 (69.0)76/158 (48.1)18/42 (42.9)94/200 (47.0)56–6556/136 (41.2)16/31 (51.6)72/167 (43.1)120/136 (88.2)25/31 (80.6)145/167 (86.8)40/136 (29.4)18/31 (58.1)58/167 (34.7)Total567/2573 (22.0)210/702 (29.9)777/3275 (23.7)2033/2573 (79.0)509/702 (72.5)2542/3275 (77.6)482/2573 (18.7)174/702 (24.8)656/3275 (20.0)Table 2 shows HHV-8, CMV and EBV seropositivity according to age and gender among the 250 HIV-AIDS patients in Ghana. Among the 250 HIV-AIDS patients, overall seroprevalence of HHV-8, CMV and EBV was 65.6%, 59.2% and 87.2%, respectively (Table 2). There was no statistically significant (p > 0.05) difference in the overall seroprevalence of HHV-8 between male and female HIV-AIDS patients. Additionally, there was no statistically significant (p > 0.05) difference in the overall seroprevalence of CMV between male and HIV-AIDS patients. Finally, there was no statistically significant (p > 0.05) difference in the overall seroprevalence of EBV between male and HIV-AIDS patients [Table 2]. Hence, gender was not an independent determinant (p > 0.05) for all three infections among HIV-AIDS patients in Ghana.Table 2HHV-8, CMV and EBV seropositivity according to age and gender among the 250 HIV-AIDS patients in GhanaAge group, yearsProportion (%) of HIV-AIDS patients with HHV8 seropositivityProportion (%) of HIV-AIDS patients with CMV seropositivityProportion (%) of HIV-AIDS patients with EBV seropositivityMenWomenAllMenWomenAllMenWomenAll16–2511/18 (61.1)13/19 (68.4)24/37 (64.9)12/18 (66.7)10/19 (52.6)22/37 (59.5)16/18 (88.9)14/19 (73.7)30/37 (81.1)26–3536/49 (73.5)29/41 (70.7)65/90 (72.2)31/49 (63.3)28/41 (68.3)59/90 (65.6)45/49 (91.8)35/41 (85.4)80/90 (88.9)36–4524/38 (63.2)21/35 (60.0)45/73 (61.6)22/38 (57.9)19/35 (54.3)41/73 (56.2)35/38 (92.1)30/35 (85.7)65/73 (89.0)46–5512/20 (60.0)9/16(56.3)21/36 (58.3)10/20 (50.0)8/16 (50.0)18/36 (50.0)17/20 (85.0)14/16 (87.5)31/36 (86.1)56–655/8 (62.5)4/6 (66.7)9/14 (64.3)5/8 (62.5)3/6 (50.0)8/14 (57.1)7/8 (87.5)5/6 (83.3)12/14 (85.7)Total88/140 (62.9)76/110 (69.1)164/250 (65.6)80/140 (57.1)68/110 (61.8)148/250 (59.2)122/140 (87.1)96/110 (87.3)218/250 (87.2)Table 3 shows the comparison of seroprevalence of HHV-8, CMV and EBV between HIV-seronegative healthy blood donors and HIV-AIDS patients in Ghana by age and gender. The overall seroprevalence of HHV-8 was statistically significantly (p < 0.005) higher in HIV-AIDS patients (65.6%, 164/250) compared to HIV-seronegative healthy blood donors (23.7%, 777/3275). Additionally, the seroprevalence of HHV-8 was statistically significantly (p < 0.005) higher in male HIV-AIDS patients (62.9%, 88/140) compared to male HIV-seronegative healthy blood donors (22.0%, 567/2573). Furthermore, the seroprevalence of HHV-8 was statistically significantly (p < 0.05) higher in female HIV-AIDS patients (69.1%, 76/110) compared to female HIV seronegative healthy blood donors (29.9%, 210/702). Finally, the seroprevalence of HHV-8 was statistically significantly (p < 0.05) higher in HIV-AIDS patients compared to HIV-seronegative healthy blood donors within each of the five 10-year interval age groups (Table 3).Table 3Comparison of seroprevalence of HHV-8, CMV and EBV between HIV-seronegative healthy blood donors and HIV-AIDS patients in Ghana by age and genderPatient characteristicProportion (%) of HHV-8 seropositivesProportion (%) of CMV seropositivesProportion (%) of EBV seropositivesHIV-HIV+*P valueHIV-HIV+*P valueHIV-HIV+*P valueAge group, years16–25169/1016 (16.6)24/37 (64.9)< 0.001790/1016 (77.8)22/37 (59.5)0.26182/1016 (17.9)30/37 (81.1)< 0.000526–35329/1340 (24.6)65/90 (72.2)< 0.005989/1340 (73.8)59/90 (65.6)0.4563/1340 (4.7)80/90 (88.9)< 0.000136–45147/552 (26.6)45/73 (61.6)< 0.05480/552 (87.0)41/73 (56.2)0.10259/552 (46.9)65/73 (89.0)< 0.0546–5560/200 (30.0)21/36 (58.3)< 0.05138/200 (69.0)18/36 (50.0)0.2594/200 (47.0)31/36 (86.1)< 0.0556–6572/167 (43.1)9/14 (64.3)< 0.05145/167 (86.8)8/14 (57.1)0.1258/167 (34.7)12/14 (85.7)< 0.05GenderMale567/2573 (22.0)88/140 (62.9)< 0.0052033/2573 (79.0)80/140 (57.1)0.19482/2573 (18.7)122/140 (87.1)< 0.0005Female210/702 (29.9)76/110 (69.1)< 0.05509/702 (72.5)68/110 (61.8)0.36174/702 (24.8)96/110 (87.3)< 0.005Total777/3275 (23.7)164/250 (65.6)< 0.0052542/3275 (77.6)148/250 (59.2)0.24656/3275 (20.0)218/250 (87.2)< 0.001*Comparison between HIV-seronegative blood donors and HIV-AIDS patients using the chi-square test with Yates correction.There was no statistically significant (p = 0.24) difference in the overall seroprevalence of CMV between HIV-AIDS patients and HIV-seronegative healthy blood donors. Additionally, the seroprevalence of CMV was not statistically significantly (p = 0.19) different between male HIV-AIDS patients and male HIV-seronegative healthy blood donors. Furthermore, the seroprevalence of CMV was not statistically significantly (p = 0.36) different between female HIV-AIDS patients and female HIV seronegative healthy blood donors. Finally, the seroprevalence of CMV was not statistically significantly (p > 0.05) different between HIV-AIDS patients and HIV-seronegative healthy blood donors within each of the five 10-year interval age groups (Table 3).The overall seroprevalence of EBV was statistically significantly (p < 0.001) higher in HIV-AIDS patients (87.2%, 218/250) compared to HIV-seronegative healthy blood donors (20.0%, 656/3275) [Table 3]. Additionally, the seroprevalence of EBV was statistically significantly (p < 0.0005) higher in male HIV-AIDS patients (87.1%, 122/140) compared to male HIV-seronegative healthy blood donors (18.7%, 482/2573). Furthermore, the seroprevalence of EBV was statistically significantly (p < 0.005) higher in female HIV-AIDS patients (87.3%, 96/110) compared to female HIV seronegative healthy blood donors (24.8%, 174/702). Finally, the seroprevalence of EBV was statistically significantly (p < 0.05) higher in HIV-AIDS patients compared to HIV-seronegative healthy blood donors within each of the five 10-year interval age groups (Table 3).The seroprevalence of HHV-8 among HIV-seronegative healthy blood donors increased with increasing age; with lowest (16.6%) in 16–25 age group, through 24.6% in 26–35 age group, 26.6% in 36–45 age group, 30.0% in 46–55 age group, and highest (43.1%) in 56–65 age group (Tables 1 &3). However, the increasing HHV-8 seropositivity among HIV-seronegative healthy blood donors with increasing age did not reach the level of statistical significance (p for trend > 0.05, data not shown). The seroprevalence of HHV-8 among HIV-AIDS patients was lowest (58.3%) in 46–55 age group and highest (72.2%) in 26–35 age group (Tables 2 &3), and there was no statistically significant difference in HHV-8 seropositivity among HIV-AIDS patients between the different age groups (p for trend > 0.05, data not shown). The seroprevalence of CMV among HIV-seronegative healthy blood donors was lowest (69.0%) in 46–55 age group and highest (87.0%) in 36–45 age group (Tables 1 &3), and there was no statistically significant difference in CMV seropositivity among HIV-seronegative healthy blood donors between the different age groups (p for trend > 0.05, data not shown). The seroprevalence of CMV among HIV-AIDS patients was lowest (50.0%) in 46–55 age group and highest (65.6%) in 26–35 age group (Tables 2 &3), and there was no statistically significant difference in CMV seropositivity among HIV-AIDS patients between the different age groups (p for trend > 0.05, data not shown). The seroprevalence of EBV among HIV-seronegative healthy blood donors was lowest (4.7%) in 26–35 age group and highest (47.0%) in 46–55 age group (Tables 1 &3), and there was no statistically significant difference in EBV seropositivity among HIV-seronegative healthy blood donors between the different age groups (p for trend > 0.05, data not shown). The seroprevalence of EBV among HIV-AIDS patients was lowest (81.1%) in 16–25 age group and highest (89.0%) in 36–45 age group (Tables 2 &3), and there was no statistically significant difference in EBV seropositivity among HIV-AIDS patients between the different age groups (p for trend > 0.05, data not shown). Hence, age was not an independent determinant (p > 0.05) for all three infections among both HIV-seronegative healthy blood donors and HIV-AIDS patients in Ghana.DiscussionSeveral studies have suggested that HHV-8 transmission may differ between endemic and non-endemic countries. In countries where infection is highly endemic, HHV-8 seroprevalence is very low in children under 2 years of age and increases soon after that age [16-19]. These seroepidemiologic studies suggest that HHV-8 is mainly transmitted among family members and close contacts via a horizontal, non-sexual route; transmission during pregnancy and through breastfeeding having a minimal role in propagating the virus [16-21]. Other studies have suggested that sexual transmission also occurs in endemic populations [17,22-24]. Volpi and colleagues [23] recently demonstrated a statistically significant association between HHV-8 and HSV-2 (a prototypic sexually transmitted infection) in Northern Cameroon (a HHV-8 endemic African country), thus suggesting sexual transmission of these two viruses with HSV-2 probably facilitating the sexual transmission of HHV-8 infection in endemic countries. Additionally, Rezza and colleagues [24] demonstrated a statistically significant association between HIV, HHV-8 and EBV in Northern Cameroon, thus suggesting their shared mode of transmission. In non-endemic countries, heterosexual transmission is probably not frequent [25]. In contrast, sexual transmission is more common among men who have sex with men in non-endemic countries [18]. Several studies have demonstrated that saliva is the principal reservoir for HHV-8, whereas the viral load of HHV-8 is consistently lower in peripheral blood, secretions from genital sites, and semen [17,18,26,27]. Therefore, although HHV-8 may be transmitted mainly through saliva in endemic countries like Ghana and Cameroon, sexual transmission may be an important additional mode of transmission in endemic African population [17,22-24].The herein reported high seroprevalence of HHV-8 in both HIV-seronegative healthy blood donors (23.7%) and HIV-AIDS patients (65.6%) confirms that HHV-8 is endemic in Ghana, and is consistent with the range of 32.3–45.5% previously reported in Ghana [3,13]. Additionally, our finding confirms the known endemicity of HHV-8 in the general population of African countries [4,11]. However, the herein reported HHV-8 seroprevalence rate of 23.7% among Ghanaian healthy blood donors is higher than the recently reported HHV-8 seroprevalence rate of 11.5% among blood donors in Burkina Faso [28], the immediate northern neighbour of Ghana. The herein reported prevalence rate of 77.6% for CMV IgG among HIV-seronegative healthy blood donors is comparable to the rate (93.2% for CMV IgG) recently reported among a smaller sample of HIV-seronegative healthy blood donors at one blood bank in Ghana [14]. The high CMV seropositivity rate in Ghana is suggestive of ubiquitous past exposure to infection. The high CMV seropositivity rate in blood donors reported in this study is comparable to the rates reported in Tunisia (97.0%) [29] and India (96.0%) [30], respectively. Additionally, the hyperendemicity of CMV in Ghana may explain the herein reported lack of statistically significant differences in the seroprevalence of CMV in HIV-seronegative healthy blood donors and HIV-AIDS patients between the sexes and the different age groups.The herein reported significantly higher seroprevalence of HHV-8 and EBV in HIV-AIDS patients compared to HIV-seronegative healthy blood donors suggests that sexual transmission might play an important role among high-risk sexual behaviour populations, such as HIV-seropositive individuals. The herein significantly higher seroprevalence of HHV-8 in HIV-AIDS patients compared to HIV-seronegative healthy blood donors is consistent with one previous study in Ghana, which reported statistically significant (p < 0.05) difference in HHV-8 seroprevalence rate between HIV-seronegative healthy blood donors (32.3%) and asymptomatic HIV-seropositive individuals (45.5%) [13]. However, the herein reported comparably high seroprevalence of HHV-8, CMV and EBV during both adolescence and adulthood suggests that their transmission might occur primarily through horizontal, non-sexual, contact. Indeed, this is particularly evident in African populations where high prevalence rates have been observed in infants and children, with seroprevalence rates similar to that observed in adults [4,11,20,21]. This large seroepidemiology study supports the view that these three viral infections are primarily transmitted non-sexually in Ghana. Therefore, non-sexual transmission mainly through close interpersonal (especially between mother and child and among siblings) contact of non-intact skin or mucous membranes with blood containing secretions or saliva, may be the primary mode of transmission of HHV-8 in Ghana, similar to that suggested in previous reports from endemic areas [4,11,20,21]. However, the relatively smaller number of HIV-AIDS patients compared to HIV-seronegative healthy blood donors in this study may be a limitation. Therefore, we suggest that further epidemiological studies should be carried out in Ghana in order to understand the relationship between HIV and HHV-8 infection in association with KS among the general population and HIV-infected individuals.An important issue that has major public health implications is the possibility of transmission of HHV-8, CMV and EBV through blood transfusion [7,8,31,32], especially in hyperendemic countries such as Ghana. Of these three viruses, cytomegalovirus (CMV) is the only one that has assumed very significant importance in blood transfusion [32]. The American Association of Blood Banks recommends the transfusion of \"CMV-safe\" (CMV-seronegative or leukocyte-reduced) blood to at-risk individuals, and this has been the standard of care in most developed countries since the late 1980s. These guidelines have helped in drastically minimizing transfusion-transmitted CMV infection in immunosuppressed recipients [32-34]. The observed high seroprevalence of CMV among Ghanaian blood donors does not justify pre-donation blood donor screening for this virus in Ghana because CMV serology is just a proxy of viremia, blood donor screening for CMV would be an obstacle to blood supply in Ghana, and CMV-seronegative blood is recommended only for organ recipients or other immunosuppressed patients. However, it does justify post-donation testing of donated blood in Ghana for CMV in order to identify the very few CMV-seronegative blood donors, motivate these CMV-seronegatives to become periodic repeat non-compensated volunteer donors, maintain a database of the epidemiological and contact information of these CMV-seronegatives to enable their rapid recall in times of need, and educate and counsel these CMV-seronegatives on how to maintain their status and the importance of their status for themselves and the increasing immunosuppressed population in Ghana. Additionally, the above proposed post-donation testing of donated blood for CMV and the subsequent determination of the actual titres of neutralization antibodies in the numerous CMV-seropositives will ensure the identification of those CMV-seropositives with very high neutralizing antibody titres from whom immunoglobulins can be obtained to treat CMV infections in at-risk individuals, and who will be followed-up and recalled when necessary in the same manner described above for the few CMV-seronegatives. However, the maintenance of CMV-seropositive and CMV-seronegative \"dual inventories\" in blood banks is expensive, and some countries with high CMV seroprevalence have found it difficult to maintain adequate supplies of CMV-seronegative products [35], as would be the case for a developing country such as Ghana. Thus, alternate methods for the provision of \"CMV safe\" blood products have been pursued, including the use of leukocyte-reduced blood products. The question whether the use of CMV-seronegative versus leukocyte-reduced blood components is equally efficacious in preventing transfusion-acquired CMV infection remains unresolved in the literature [36,37]. Bowden and colleagues reported that the use of leukocyte-reduced blood products was comparable to the use of CMV-seronegative blood products for the prevention of transfusion-transmitted CMV infection after marrow transplant [36]. However, a recent meta-analysis of the available controlled studies indicated that CMV-seronegative blood components were more efficacious than leukocyte-reduced blood components in preventing transfusion-acquired CMV infection [37].ConclusionThe high seroprevalence of the three viruses among both HIV-positive and HIV-negative individuals suggests endemicity and predominant horizontal, non-sexual, transmission of the infections in Ghana. The higher seroprevalence of HHV-8 and EBV among HIV-AIDS patients compared to healthy blood donors suggests an additional role of sexual transmission.AbbreviationsAIDS: acquired immunodeficiency syndrome; CMV: cytomegalovirus; EBV: Epstein Barr virus; HHV-8: human herpes virus 8; HIV: human immunodeficiency virus; KSHV: Kaposi's sarcoma-associated herpesvirus; OR: odds ratio; 95% CI: 95% confidence intervalCompeting interestsThe authors declare that they have no competing interests.Authors' contributionsAAA conceived study, provided guidance to all aspects of study, and revised manuscript for important intellectual content. HBA performed quality assessment of data, data analysis, data preparation, and drafted manuscript. AAA, HBA, FG, IB, CA and IA participated in design and coordination of study, data and sample collection, and performed and supervised immunoassays. All authors read and approved final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2528061.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2528061",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2528061\nDescemet's Membrane of Diabetic Goto-Kakizaki Rats\nand Its Suppression by Antidiabetic Agents\n\nAUTHORS: Yoshihiro Akimoto, Hajime Sawada, Mica Ohara-Imaizumi, Shinya Nagamatsu, Hayato Kawakami\n\nABSTRACT:\nWe examined changes in the ultrastructure and localization of major extracellular matrix components, including 5 types of collagen (type I, III, IV, VI, and VIII), laminin, fibronectin, and heparan sulfate proteoglycan in Descemet's membrane of the cornea of diabetic GK rats. In the cornea of diabetic GK rats, more long-spacing collagen fibrils were observed in Descemet's membrane than in the membrane of the nondiabetic Wistar rats. Both GK and Wistar rats showed an age-dependent increase in the density of the long-spacing collagen. Immunoelectron microscopy showed that type VIII collagen was localized in the internodal region of the long-spacing collagen, which was not labelled by any of the other antibodies used. The antidiabetic agents nateglinide and glibenclamide significantly suppressed the formation of the long-spacing collagen in the diabetic rats. Long-spacing collagen would thus be a useful indicator for studying diabetic changes in the cornea and the effect of antidiabetic agents.\n\nBODY:\n1. INTRODUCTIONCorneal keratopathy is one of the\ndiabetic complications. Clinically, the diabetic cornea often shows superficial\npunctate keratopathy and persistent epithelial defects and recurrent epithelial\nerosion that are considered to be a form of diabetic keratoepitheliopathy.\nVarious degrees of epithelial disturbance take place in the diabetic cornea.\nThickening of epithelial basement membrane in the diabetic cornea has been\nextensively studied [1, 2].\nIn a previous study, we showed that hemidesmosomes in the epithelial basal\ncells were decreased in number in the diabetic rats and that the basement\nmembrane detached from the epithelial basal cells [3]. Also, the\ncontent of O-GlcNAc-modified proteins was found to be increased in the corneal\nepithelium as well as in the nerves, kidneys, and pancreas of diabetic rats [3–5].The diabetic\ncorneal endothelium has been shown by speculum-aided microscopy to have morphological\nabnormalities such as polymorphism [6]. The endothelial cells\nvary in cell shape and in cell area in the diabetic rat and human cornea [7, 8]. At the posterior side of the cornea, a thick\nbasement membrane called Descemet's membrane is located adjacent to the\nendothelium. In the normal human cornea, long-spacing collagen, which is\ncross-striated fiber bundle, is located only in the anterior-banded zone of\nDescemet's membrane [9, 10]. However, in the cornea of normal\nSprague-Dawley rats, there is no such collagen in Descemet's membrane [11]. Although the normal rat\nand human Descemet's membranes differ in this regard, in the corneas of both diabetic\nhuman patients and Streptozotocin-induced diabetic rats, unusual long-spacing\ncollagen was observed scattered in Descemet's membrane [11, 12].The spontaneously\ndiabetic Goto-Kakizaki (GK) rat is a nonobese model of type 2 diabetes that was\ndeveloped by the selective breeding of glucose-intolerant Wistar rats [13–15]. In the eyes of GK rats, various\nabnormalities have been reported, including decreased retinal microcirculation [15], elevated levels of vascular endothelial growth factor [16], nitric oxide synthase activity in the retina [17], delayed\nwound closure, as well as phenotypic changes in the corneal epithelium [18]. However, little attention\nhas been paid to Descemet's membrane in GK rats. To investigate Descemet's\nmembrane in terms of the pathogenesis of diabetes mellitus, in the present\nstudy we examined the ultrastructural morphology, and immunohistochemically\ndetermined the composition of Descemet's membrane in the cornea of diabetic GK\nrats in comparison with normal Wistar rats. Furthermore, we examined if the\nmorphological change detected could be prevented by antidiabetic agents. Our\nfindings revealed that unusual long-spacing collagen appeared and increased in\ncontent rapidly with aging in the Descemet's membrane of the diabetic rat cornea,\nand that its appearance could be suppressed by the antidiabetic agents.2. MATERIALS AND METHODS2.1. Animals and tissuesAll experimental procedures using\nlaboratory animals were approved by the Animal Care and Use Committee of Kyorin\nUniversity School of Medicine. The corneas of 15-, 33-, and 62-week-old male\n(n = 6 for each age) Goto-Kakizaki rats and Wistar rats (as normal controls),\nobtained from Kurea (Tokyo, Japan), were\nused in the present study. Rats were\nhoused under 12-hour light: 12-hour dark cycle and given free access to food and water. Serum\nglucose levels in Wistar and GK rats, which were measured after an overnight\nfast, were, respectively, 158.0 ± 12.0 and 375.9 ± 11.6 (mean ± SEM) mg/dL at 15 weeks, 118.5 ± 10.5 and 333.8 ± 22.4 mg/dL at 33 weeks, and\n167.0 ± 28.6 and 314.9 ± 48.7 mg/dL at 62 weeks. As\nreported previously [5], serum insulin levels were also\nhigher in the GK rats.2.2. AntibodiesA monoclonal antibody (clone\n9H3) against Type VIII collagen and a monoclonal antibody (clone 15B6) against\nType VI collagen were prepared and characterized as described previously [19, 20]. Polyclonal antibody against laminin and type IV collagen were purchased from EY laboratory (San Meteo, Calif, USA) and LSL (Tokyo, Japan),\nrespectively. Monoclonal antibody against heparan sulfate proteoglycan and\nfibronectin were purchased from Upstate Biotechnology (Lake Placid, NY, USA) and Chemicon International (Temecula, Calif, USA), respectively. Alexa568-conjugated donkey anti-rabbit or mouse IgG and\nSYBR-Green I were obtained from Molecular Probes (Eugene, Ore, USA).2.3. Immunohistochemical localizationImmunofluorescence\nobservation for localization of components of extracellular matrix was\nperformed as described earlier [21]. Corneas were fixed in 4%\nparaformaldehyde in 0.1 M phosphate buffer (pH7.3) for 1 hour at 4°C. After\nhaving been washed with PBS, the specimens were embedded in OCT compound\n(Miles; Elkhart, Ill, USA).\nFrozen sections (4 μm thick) were made, washed with PBS, and incubated for 10\nminutes in 5% BSA in PBS. The sections were then incubated with\nanti-extra-cellular matrix component antibody for 1 hour at room temperature,\nwashed with PBS, and subsequently incubated with Alexa 568-conjugated donkey\nanti-rabbit or mouse IgG antibody (1:200). Nuclei were stained with SYBR-Green\nI (1:500). After a final wash with PBS, the specimens were mounted in 90%\nglycerol-0.1 M Tris-HCl buffer (pH8.5) containing 0.5 mM p-phenylene diamine,\nand observed under a laser scanning confocal microscope (LSM510, Zeiss, Mass, USA). For a control\nexperiment, the specimens were incubated with normal rabbit or mouse IgG or\nwith 0.1% BSA-PBS alone instead of the primary antibodies. No positive staining\nwas observed in the control experiment (data not shown).2.4. Electron microscopyCorneas were fixed in\nphosphate-buffered 2.5% glutaraldehyde (pH7.4). Strips of cornea were taken\nfrom the central part of the cornea, and were postfixed in 1% OsO4 in 0.1 M phosphate buffer (pH7.4), and dehydrated with graded\nalcohols. After immersion in propylene\noxide, the specimens were embedded in Epon 812. \nUltrathin sections were cut perpendicular to the epithelium, doubly\nstained with uranyl acetate and lead citrate, and examined with a transmission\nelectron microscope, TEM-1010 (JEOL, Tokyo, Japan).2.5. Immunoelectron microscopyImmunoelectron microscopic\nobservation for localization of components of extracellular matrix was\nperformed as described earlier [22]. Fixation was carried out in the same way as for light\nmicroscopy. Ultrathin frozen-sections were cut at −90 to −100°C. The sections were washed with PBS and\npretreated with 1% BSA in PBS for 10 minutes. After a PBS rinse, they were\nincubated with the desired antibodies for 1 hour, washed with PBS, and\nincubated with colloidal gold-conjugated goat anti-rabbit or mouse IgG\nantibodies for 1 hour. After another wash with PBS, the sections were refixed\nin 2% glutaraldehyde-0.1 M phosphate buffer, pH7.4, and embedded in a mixture\nof methylcellulose, polyethyleneglycol, and uranyl acetate.2.6. Antidiabetic agents and experimental designThe administration of antidiabetic drugs\nwas started at 8 weeks old. Nateglinide (50 mg/kg) or glibenclamide (2 mg/kg)\nwas suspended in 0.5% methylcellulose and administered to GK rats via a stomach\ntube in volume of 10 mL/kg [23–25]. These doses of the antidiabetic\nagents were chosen from the data on their suppressive effects on the peak blood\nglucose levels after oral sucrose or glucose loading of fasted normal rats for\n15 weeks [23, 26]. GK rats were fed twice daily (9:00 and 16:00)\nfor 1 hour and were given nateglinide or glibenclamide orally just before each\nmeal. Control rats were treated with 0.5% methylcellulose alone (the vehicle).2.7. Statistical analysisResults were expressed as the mean ± standard deviation, and the statistical analysis was done by the use of the unpaired\nStudent's t-test. Differences were defined as significant at P < .05.3. RESULTS3.1. Changes in ultrastructural morphology of Descemet's membrane in the cornea of diabetic and normal rats with agingThe ultrastructure of Descemet's membrane\nin the cornea in 15-, 33-, and 62-week-old diabetic and nondiabetic rat corneas\nwas examined by electron microscopy. The thickness of the membrane remained\nunchanged in both groups. In the 15-week-old rat cornea, abnormal collagen\nfibril bundles (long-spacing collagen) were frequently observed in Descemet's\nmembrane of diabetic rats, whereas they were observed less so in that of the\nnormal rats. Figure 1 shows electron microscopic images of the long-spacing\ncollagen. The banding pattern was wide, averaging 110–120 nm. A\nrod-like structure was observed in the internodal region of typical long-spacing\ncollagen (Figure 1). The size and number of long-spacing collagen molecules increased with aging in both the normal and diabetic Descemet's membrane (Figures 2 and 3). At\n62 weeks, the average length of the long-spacing collagen was 0.25–1.0 μm in\nnormal rats and 0.5–2.15 μm in the diabetic ones. In the diabetic cornea, the\nnumber of long-spacing collagen fibrils in the membrane increased more than in\nthe normal cornea with aging. The\ndensity of these collagen fibrils was higher adjacent to the endothelium and\nlower toward the stroma (Figure 2).3.2. Immunohistochemical localization of type VIII collagenIt was earlier shown that type VIII\ncollagen is localized in Descemet's membrane [19]. So we\nexamined the localization of type VIII collagen in the rat cornea by laser\nconfocal scanning microscopy (Figure 4). Whereas weak staining was observed in Descemet's\nmembrane of the normal cornea, more intense punctate staining was observed in that\nof the diabetic one (Figure 4).3.3. Immunoelectron microscopic localization of type VIII collagenLocalization of type VIII collagen in\nDescemet's membrane was examined immunoelectron-microscopically by using the\ncolloidal-gold labeling method (Figure 5). \nIn the normal cornea, the colloidal gold was detected diffusely in\nDescemet's membrane (Figure 5(a)). In the diabetic cornea, however, it was found in\nthe region between bands of the long-spacing collagen (Figure 5(b)). This\nlocalization of type VIII collagen in the diabetic rat is consistent with that\nreported previously for the diseased human Descemet's membrane [9, 10].3.4. Localization of other components of extracellular matrixNext we examined the localization of the\nother components of the extracellular matrix, that is, laminin and type I, III,\nIV, and VI collagens. There is a report that type VI collagen becomes localized\nin the long-spacing collagen in the corneoscleral meshwork of human eyes [27]. So we examined immunoelectron-microscopically whether type VI\ncollagen could also be detected in the long-spacing collagen of Descemet's\nmembrane. In both the normal and diabetic cornea, the colloidal gold label was\nmostly seen in the corneal stroma; but none was detected in either Descemet's\nmembrane or in the long-spacing collagen of this membrane (Figure 6). Collagens\ntype I and III showed the same distribution as the type VI (data not shown). Laminin\n(Figure 7) and type IV collagen (data not shown) were localized in the\namorphous material of Descemet's membrane, but were not present in the\nlong-spacing collagen in neither\nthe normal nor\ndiabetic cornea.3.5. Effect of antidiabetic agents on the formation of long-spacing collagenThe morphological change in long-spacing\ncollagen is thought to be one of the complications of diabetes. Thus we\nexamined whether the abnormal formation of long-spacing collagen could be suppressed\nby the antidiabetic agents nateglinide and glibenclamide. As shown in our\nprevious study [25], during the period of antidiabetic agent\ntreatments, there was no difference in fasting blood glucose levels among\nvehicle-treated, nateglinide-treated, and glibenclamide-treated GK rat. Nateglinide administration reduced blood\nglucose levels 1 hour after feeding, whereas glibenclamide reduced blood\nglucose levels 2 and 3 hours after feeding, but not 1 hour after feeding [25]. Although the degree of the antidiabetic effect was different\nbetween the 2 agents, both of them significantly (P < .05) inhibited the formation of the long-spacing collagen in\nthe diabetic GK rats (Figure 8). Glibenclamide treatment was more effective than\nnateglinide treatment (Figure 8).4. DISCUSSIONOur\npresent study revealed that a diabetes-associated increase in the number of\nlong-spacing collagen fibrils occurred in Descemet's membrane in the cornea of\ntype II diabetes model GK rats. This result is consistent with the findings\nmade in the corneas of diabetic humans and type I diabetes model Streptozotocin-induced\ndiabetic rats [11, 12].\nThe long-spacing collagen was much more abundant in the diabetic cornea than in\nthe nondiabetic one, and it increased with aging in both the nondiabetic Wistar\nand diabetic GK rats (Figures 2 and 3). These results reveal that the age-associated\nmorphological change in Descemet's membrane was accelerated by diabetes.Immunoelectron\nmicroscopy showed that the long-spacing collagen contained type VIII collagen\nmolecules (Figure 5). However, antibodies against\ntype I, III, IV, or VI collagen did not bind to the long-spacing collagen.\nThese results are consistent with those obtained from human and bovine Descemet's\nmembrane [28, 29]. Type VIII collagen\nwas found as a product of rabbit corneal and bovine aortic endothelial cells [30, 31]. Type VIII collagen is a major\nconstituent of the hexagonal lattice of Descemet's membrane [19]. Descemet's lattice collagen can assemble into other long-spacing fibrils\nwith a longer periodicity [19]. It is thought that the function\nof type VIII collagen is to provide an open, porous structure that can\nwithstand compressive force [19]. Under some special condition of\nthe diabetic state, type VIII collagen may contribute to the assembly of these\nunusual long-spacing collagen fibrils. It was earlier postulated that the\naggregates of wide-spacing collagen fibrils may reflect an excessive\nglycosylation in diabetes [11, 12].Long-spacing\ncollagen is also observed in the eyes from patients with\niridocorneal-endothelial syndrome, primary open-angle glaucoma, age-related\nmacular degeneration, and Fuchs' endothelial corneal dystrophy [9, 10, 32–34]. In\nFuchs' corneal dystrophy, Descemet's membrane thickens abnormally; and many\nlong-spacing collagen fibrils\nare formed in its posterior layer [10, 35]. A recent\nstudy showed that Fuchs' corneal dystrophy results from a mutation in the gene\nencoding alpha 2 chain of type VIII collagen [36, 37]. Lack of\ntype VIII collagen results in dysgenesis of the anterior segment of the eye, in\nwhich Descemet's membrane is markedly thinned [38]. Whereas\nhuman and mouse corneas have an anterior banded layer and a posterior unbanded\nlayer in their Descemet's membrane, in the present study these 2 layers could\nnot be distinguished clearly in the rat cornea. \nThe long-spacing collagen tended to be localized in the posterior side\nof Descemet's membrane in the diabetic rat cornea (Figure 2). This localization\nis consistent with that observed in Fuch's corneal dystrophy.Diabetes\ninduces the dysfunction of corneal endothelium [39–42]. High glucose levels in\ndiabetes cause the increase in sorbitol accumulation [43],\nadvanced glycation end products [44], and O-GlcNAc-modified\nproteins [3] in the cornea. These changes may cause\ndysfunction of corneal endothelial cells and induce the morphological change of\nDescemet's membrane as shown in the present study. Kaji et al. [44] reported\nthat advanced glycation end products in Descemet's membrane may be responsible\nfor the corneal endothelial abnormalities in diabetes.D-Phenylalanine derivative drug nateglinide and sulfonylurea drug\nglibenclamide are antidiabetic agents that increase insulin secretion. When we\nexamined the effect of nateglinide and glibenclamide on the morphological\nchanges in Descemet's membrane of GK rats, we found that the abnormal formation\nof the long-spacing collagen was significantly suppressed by the administration\nof either nateglinide or glibenclamide (Figure 8). These results suggest that\ncontrol of postprandial hyperglycemia is essential to prevent the abnormal\nformation of long-spacing collagen in type-2 diabetes.In the present study, glibenclamide treatment was more effective\nthan nategelinide treatment (Figure 8). The different effects of these two\nantidiabetic agents may be due to their different action mechanisms.\nNateglinide and glibenclamide display different effects on insulin secretion in\nbeta cell. Our previous study showed that decreased first-phase insulin release\nwas partially recovered when GK rats were treated with nateglinide, whereas no\nfirst-phase release occurred with glibenclamide treatment [25].\nNateglinide administration reduced blood glucose 1 hour after feeding, whereas\nglibenclamide administration reduced the blood glucose level 2 and 3 hours\nafter feeding [25]. Glibenclamide treatment is more effective\nthan nateglinide treatment in the dysfunction of second-phase insulin release.\nThe present study suggests that glibenclamide treatment might be more effective\nin the inhibition of long-spacing collagen formation by recovering second-phase\ninsulin release.In\nsummary, more long-spacing collagen fibrils were observed in Descemet's\nmembrane of diabetic GK rats than in the membrane of the nondiabetic Wistar\nrats. Type VIII collagen was localized\nin the internodal region of the long-spacing collagen. Further studies are needed to elucidate the\nrole of type VIII collagen in the formation of long-spacing collagen. Antidiabetic\nagents nateglinide and glibenclamide significantly suppressed the formation of\nthe long-spacing collagen in the diabetic rats. The\nlong-spacing collagen of the cornea would appear to be a useful indicator for\nstudying diabetic changes in the cornea and the effect of antidiabetic agents.\n\nREFERENCES:\n1. TaylorHRKimseyRACorneal epithelial basement membrane changes in diabetesInvestigative Ophthalmology & Visual Science19812045485537216670\n2. EngermanRLColquhounPJEpithelial and mesothelial basement membranes in diabetic patients and dogsDiabetologia19822365215247152162\n3. AkimotoYKawakamiHYamamotoKMunetomoEHidaTHiranoHElevated expression of O-GlcNAc-modified proteins and O-GlcNAc transferase in corneas of diabetic Goto-Kakizaki ratsInvestigative Ophthalmology & Visual Science20034493802380912939295\n4. AkimotoYYamamotoKMunetomoEElevated post-translational modification of proteins by O-linked N-acetylglucosamine in various tissues of diabetic Goto-Kakizaki rats accompanied by diabetic complicationsActa Histochemica et Cytochemica2005382131142\n5. AkimotoYHartGWWellsLElevation of the post-translational modification of proteins by O-linked N-acetylglucosamine leads to deterioration of the glucose-stimulated insulin secretion in the pancreas of diabetic Goto-Kakizaki ratsGlycobiology200717212714017095531\n6. OhguroNMatsudaMOhashiYFukudaMTopical aldose reductase inhibitor for correcting corneal endothelial changes in diabetic patientsBritish Journal of Ophthalmology19957912107410778562538\n7. MeyerLAUbelsJLEdelhauserHFCorneal endothelial morphology in the rat. Effects of aging, diabetes, and topical aldose reductase inhibitor treatmentInvestigative Ophthalmology & Visual Science19882969409482967259\n8. SchultzROMatsudaMYeeRWEdelhauserHFSchultzKJCorneal endothelial changes in type I and type II diabetes mellitusAmerican Journal of Ophthalmology19849844014106486211\n9. LevySGMcCartneyACESawadaHDopping-HepenstalPJCAlexanderRAMossJDescemet's membrane in the iridocorneal-endothelial syndrome: morphology and compositionExperimental Eye Research19956133233337556496\n10. LevySGMossJSawadaHDopping-HepenstalPJCMcCartneyACEThe composition of wide-spaced collagen in normal and diseased Descemet's membraneCurrent Eye Research199615145528631203\n11. RehanyUIshiiYLahavMRumeltSCollagen pleomorphism in Descemet's membrane of streptozotocin-induced diabetic rats: an electron microscopy studyCornea200019339039210832705\n12. RehanyUIshiiYLahavMRumeltSUltrastructural changes in corneas of diabetic patients: an electron-microscopy studyCornea200019453453810928773\n13. GotoYSuzukiKOnoTSasakiMToyotaTDevelopment of diabetes in the non-obese NIDDM rat (GK rat)Advances in Experimental Medicine and Biology198824629313074659\n14. BisbisSBailbeDTormoM-AInsulin resistance in the GK rat: decreased receptor number but normal kinase activity in liverAmerican Journal of Physiology19932655E807E8138238507\n15. MiyamotoKOguraYNishiwakiHEvaluation of retinal microcirculatory alterations in the Goto-Kakizaki rat. A spontaneous model of non-insulin-dependent diabetesInvestigative Ophthalmology & Visual Science19963758989058603874\n16. SoneHKawakamiYOkudaYOcular vascular endothelial growth factor levels in diabetic rats are elevated before observable retinal proliferative changesDiabetologia19974067267309222654\n17. CarmoACunha-VazJGCarvalhoAPLopesMCNitric oxide synthase activity in retinas from non-insulin-dependent diabetic Goto-Kakizaki rats: correlation with blood-retinal barrier permeabilityNitric Oxide20004659059611139367\n18. WakutaMMorishigeNChikamaT-ISekiKNaganoTNishidaTDelayed wound closure and phenotypic changes in corneal epithelium of the spontaneously diabetic Goto-Kakizaki ratInvestigative Ophthalmology & Visual Science200748259059617251454\n19. SawadaHKonomiHHirosawaKCharacterization of the collagen in the hexagonal lattice of Descemet's membrane: its relation to type VIII collagenThe Journal of Cell Biology199011012192272104858\n20. SawadaHYazamaFType VI collagen in the rat testis: monoclonal antibody, isolation, and localization during developmentBiology of Reproduction19945037027108167242\n21. AkimotoYYamakawaNFurukawaKKimataKKawakamiHHiranoHChanges in distribution of the long form of type XII collagen during chicken corneal developmentJournal of Histochemistry & Cytochemistry200250685186212019301\n22. AkimotoYKreppelLKHiranoHHartGWLocalization of the O-linked N-acetylglucosamine transferase in rat pancreasDiabetes199948122407241310580430\n23. KitaharaYMiuraKTakesueKDecreased blood glucose excursion by nateglinide ameliorated neuropathic changes in Goto-Kakizaki rats, an animal model of non-obese type 2 diabetesMetabolism: Clinical and Experimental200251111452145712404197\n24. MineTMiuraKKitaharaYOkanoAKawamoriRNateglinide suppresses postprandial hypertriglyceridemia in Zucker fatty rats and Goto-Kakizaki rats: comparison with voglibose and glibenclamideBiological & Pharmaceutical Bulletin200225111412141612419950\n25. KawaiJOhara-ImaizumiMNakamichiYInsulin exocytosis in Goto-Kakizaki rat β-cells subjected to long-term glinide or sulfonylurea treatmentBiochemical Journal200841219310118254725\n26. IkenoueTOkazakiKFujitaniSEffect of a new hypoglycemic agent, A-4166 [(-)-N-(trans-4-isopropyl- cyclohexanecarbonyl)-D-phenylalanine], on postprandial blood glucose excursion: comparison with voglibose and glibenclamideBiological & Pharmaceutical Bulletin19972043543599145209\n27. UedaJYueBYJTDistribution of myocilin and extracellular matrix components in the corneoscleral meshwork of human eyesInvestigative Ophthalmology & Visual Science200344114772477914578398\n28. MurataYYoshiokaHIyamaKUsukuGDistribution of type VI collagen in the bovine corneaOphthalmic Research198921167722652026\n29. MarshallGEKonstasAGLeeWRImmunogold fine structural localization of extracellular matrix components in aged human cornea. II. Collagen types V and VIGraefe's Archive for Clinical and Experimental Ophthalmology19912292164171\n30. SageHTrüebBBornsteinPBiosynthetic and structural properties of endothelial cell type VIII collagenThe Journal of Biological Chemistry19832582113391134016630235\n31. BenyaPDPadillaSRIsolation and characterization of type VIII collagen synthesized by cultured rabbit corneal endothelial cells. A conventional structure replaces the interrupted-helix modelThe Journal of Biological Chemistry19862619416041693081516\n32. RohenJWLütjen-DrecollEFlügelCMeyerMGriersonIUltrastructure of the trabecular meshwork in untreated cases of primary open-angle glaucoma (POAG)Experimental Eye Research19935666836928595810\n33. van der SchaftTLde BruijnWCMooyCMKetelaarsDAde JongPTIs basal laminar deposit unique for age-related macular degeneration?Archives of Ophthalmology199110934204252003806\n34. RothSIStockELJutabhaREndothelial viral inclusions in Fuchs' corneal dystrophyHuman Pathology19871843383413493969\n35. BourneWMJohnsonDHCampbellRJThe ultrastructure of Descemet's membrane. III. Fuch's dystrophyArchives of Ophthalmology198210012195219556983339\n36. GottschJDZhangCSundinOHBellWRStarkWJGreenWRFuchs corneal dystrophy: aberrant collagen distribution in an L450W Mutant of the COL8A2 geneInvestigative Ophthalmology & Visual Science200546124504451116303941\n37. GottschJDSundinOHLiuSHInheritance of a novel COL8A2 mutation defines a distinct early-onset subtype of Fuchs corneal dystrophyInvestigative Ophthalmology & Visual Science20054661934193915914606\n38. HopferUFukaiNHopferHTargeted disruption of Col8a1 and Col8a2 genes in mice leads to anterior segment abnormalities in the eyeThe FASEB Journal200519101232124416051690\n39. RavalicoGTognettoDPalombaMCalderiniSVattovaniOCorneal endothelial function in diabetes: a fluorophotometric studyOphthalmologica199420841791847970543\n40. LarssonL-IBourneWMPachJMBrubakerRFStructure and function of the corneal endothelium in diabetes mellitus type I and type IIArchives of Ophthalmology199611419148540858\n41. McNamaraNABrandRJPolseKABourneWMCorneal function during normal and high serum glucose levels in diabetesInvestigative Ophthalmology & Visual Science19983913179430539\n42. ZiadiMMoirouxPd'AthisPBronABrunJ-MCreuzot-GarcherCAssessment of induced corneal hypoxia in diabetic patientsCornea200221545345712072718\n43. Cisarik-FredenburgPDiscoveries in research on diabetic keratopathyOptometry2001721169170412363257\n44. KajiYAmanoSUsuiTAdvanced glycation end products in Descemet's membrane and their effect on corneal endothelial cellCurrent Eye Research200123646947712045898"
+}

batch_0/PMC2528309.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2528309",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2528309\nAUTHORS: W.R. Parulekar, M. McKenzie, K.N. Chi, L. Klotz, C. Catton, M. Brundage, K. Ding, A. Hiltz, R. Meyer, F. Saad\n\nABSTRACT:\nThe designation “clinically localized prostate cancer” comprises a group of biologically heterogeneous tumours with different growth rates and risks of relapse. Because prostate cancer is primarily a disease of older men, treatment selection must take into account the prognosis of the tumour, patient age, comorbidities, side effects of treatment, and patient preferences. Clinical trials must identify the various prognostic groups and test the appropriate treatment strategies within these subgroups.\n\nBODY:\n1. INTRODUCTIONProstate cancer is the most common malignancy in Canadian men and ranks third behind lung and colon cancer in terms of cancer-related mortality. However, from 1994 to 2003, mortality from prostate cancer declined at a rate of 2.7% annually. That decline is attributed both to the widespread use of testing for prostate-specific antigen (psa), which has led to a shift in stage and grade at diagnosis, and to the existence of effective therapies for clinically localized disease.In 2007, estimates placed new cases of prostate cancer at 22,300 and deaths from the disease at 4300 1. Those statistics highlight some important facts about prostate cancer:In most cases, prostate cancer is not a fatal condition.Current treatment options still fail to cure or control disease in a significant proportion of cases, and approximately 20% of patients die from their prostate cancer.Not surprisingly, the treatment strategies under evaluation in ongoing clinical trials in early prostate cancer reflect the biologic heterogeneity of the disease. They include such diverse therapies as active surveillance for good-prognosis disease and the addition of cytotoxic chemotherapy to radical radiation or prostatectomy for disease with high risk of relapse. The present article reviews ongoing studies in localized prostate cancer conducted by the National Cancer Institute of Canada (ncic) Clinical Trials Group (ctg).2. NCIC CTG PR.11: A PHASE III STUDY OF ACTIVE SURVEILLANCE THERAPY AGAINST RADICAL TREATMENT IN PATIENTS DIAGNOSED WITH FAVOURABLE-RISK PROSTATE CANCER (START)2.1 BackgroundTumours detected because of psa testing comprise most of the localized prostate cancer cases diagnosed today. Although testing may allow for diagnosis and the use of curative therapy at an earlier stage in a potentially life-threatening disease, it also clearly identifies a group of patients with biologically indolent tumours in whom radical therapy may be unnecessary and detrimental because of its associated morbidity and costs 2,3. Previous nonrandomized studies have identified the prognostic significance of stage and grade in patients treated with conservative therapy or observation, thus identifying a patient population for whom curative therapy can potentially be withheld without compromise to long-term outcome 4–6. An extension of the concept of observation is that of “active surveillance,” which entails close follow-up of disease and intervention with curative intent triggered by early signs of disease progression.The strategy of active surveillance for patients with favourable-risk prostate cancer was evaluated in a large phase ii study by Klotz. Active surveillance was applied in 331 patients with favourable-risk disease (defined as psa below 15 ng/mL, Gleason score of 7 or less, and tumour stage less than T2B), following them until criteria of early disease progression [defined by biochemical, histologic (grade), or clinical progression] were met. Of those patients, 80% had a Gleason score of 6 or less, and the same proportion had a psa below 10 ng/mL. With a median follow-up of 72 months, 34% of patients discontinued active surveillance. Biochemical progression led to discontinuation in 15%; clinical progression, in 3%; histologic progression, in 4%; and patient preference, in 12%. With a median follow-up of 7 years, overall survival was 85%, and disease-specific survival was 99%. The median psa doubling time for the entire cohort was 7.0 years; a psa doubling time of less than 2 years was associated with a high risk of local progression for patients who underwent radical prostatectomy. At January 2007, 134 patients remained on active surveillance 7.2.2 Study DesignThe PR.11 trial is an Intergroup study led by the ncic ctg (study chair, Dr. Laurence Klotz) that compares active surveillance with radical therapy (prostatectomy or radiotherapy depending on physician and patient choice) at the time of diagnosis in a randomized phase iii setting. Eligible patients are those with a life expectancy of more than 10 years and favourable-risk prostate cancer [defined as clinical stage T1B, T1C, T2A, or T2B at the time of diagnosis; clinical (diagnostic biopsy) Gleason score of 6 or less; psa 10.0 ng/mL or lower]. Patients randomized to the active surveillance arm will undergo radical intervention (again, prostatectomy or radiotherapy depending on physician and patient choice) at the time one or more of the following pre-specified criteria are met:Biochemical progression—psa doubling time less than 3 years, based on at least 5 separate consecutive measurements over a minimum of 12 months from the date of the baseline measurement or from the date that the psa reached a value greater than or equal to the psa before initiation of androgen deprivation therapy (if applicable), as assessed by the local investigator.Histologic or grade progression—Gleason pattern predominant 4 or greater (that is, a Gleason score of 7 (4+3) or higher) in re-biopsy of the prostate.Clinical progression—”local progression” defined as local progression of prostate cancer resulting in urinary retention, gross hematuria, or hydronephrosis; or “distant metastasis” defined by radiology, cytology, or histology (or a combination) at sites remote from the prostate and regional lymph nodes.Using a non-inferiority design to rule out a greater than 5% difference in 15-year survival between the radical treatment and active surveillance groups, 2130 patients will be accrued over a 5-year period. The primary endpoint is disease-specific survival. That endpoint, rather than overall survival, was selected because of the need to determine the effect of the active surveillance strategy specifically on prostate cancer mortality. Secondary endpoints include overall survival, quality of life, distant disease-free survival, psa relapse or progression after radical intervention, initiation of androgen deprivation therapy, proportion of patients on the active surveillance arm receiving radical intervention, prognostic significance of psa doubling time before diagnosis, and prognostic significance of molecular biomarkers. Quality of life is an important part of the study, and the Expanded Prostate Cancer Index, rand SF-12, and State–Trait Anxiety Inventory will be used to provide a comprehensive examination of the various components of patient-reported outcomes on study. The feasibility phase has commenced in designated centres of participating cooperative groups. If the results of the feasibility phase indicate sufficient patient and physician willingness to participate in the randomization process, then accrual will be opened widely.3. NCIC CTG PR.12: A PHASE III STUDY OF NEOADJUVANT DOCETAXEL AND ANDROGEN SUPPRESSION PLUS RADIATION THERAPY VERSUS ANDROGEN SUPPRESSION ALONE PLUS RADIATION THERAPY FOR HIGH-RISK LOCALIZED ADENOCARCINOMA OF THE PROSTATE (DART)3.1 BackgroundRadical radiotherapy and long-term androgen suppression constitute an accepted treatment option for localized but high-risk disease as defined by clinical stage (T3) and high Gleason score (8 or higher) or high psa (20 ng/mL or more), or both. Results from previously conducted randomized studies are consistent with 5-year disease-free survival rates of 46%–74% with combined therapy 8–10, thus providing the rationale for continued evaluation of therapies to improve outcome by control of micrometastatic disease.Docetaxel is a good candidate drug. The mechanism of action of this agent involves disruption of the microtubular network critical for mitotic and inter-phase cellular functions. Doses of 75–100 mg/m2 intravenously (IV) administered are well tolerated, with neutropenia, alopecia, cutaneous reactions, gastro-intestinal effects (nausea, diarrhea), neurotoxicity, and edema being among the most frequently reported adverse events. Severe hypersensitivity reactions characterized by respiratory or circulatory instability or generalized rash or erythema occur in fewer that 5% of patients, although lesser grades are more common 11. Using overall survival as the primary endpoint, two pivotal studies have demonstrated the efficacy of docetaxel in advanced hormone-refractory prostate cancer 12,13. Efficacy and adverse event data support the use of the every-three-weeks docetaxel schedule, and that schedule has been widely adopted for use in this patient population.Using changes in psa as a marker of antitumour effect, studies have shown that docetaxel is also active against hormone-sensitive prostate cancer 14–17. Furthermore, based on preclinical data that suggest that docetaxel may result in phosphorylation and inactivation of the anti-apoptotic protein Bcl-2 (which is upregulated with androgen suppression), combination therapy with docetaxel and androgen suppression may lead to greater antitumour effect 18–21.Timing of therapy appears to be important. Eigl et al. implanted LNCaP human prostate cancer and Shionogi mouse mammary carcinoma cell lines into mice, and followed up with treatment using one of these three regimens: castration with paclitaxel on progression, paclitaxel with castration on progression, or concurrent castration and paclitaxel 22. As compared with sequential castration followed by paclitaxel, concurrent therapy resulted in significantly longer time-to-progression and time-to-sacrifice in the mice. Notably, a marked lack of response to castration in the mice treated initially with paclitaxel was seen.Clinical studies in patients with locally advanced prostate cancer have demonstrated the feasibility and tolerability of combined therapy with docetaxel and androgen suppression in the neoadjuvant setting before prostatectomy 23 or radiotherapy 24. In 54 men with high-risk prostate cancer, McKenzie et al. used one of two neoadjuvant treatment schedules before radical radiotherapy: 6 months of androgen suppression, plus 2 cycles of docetaxel 35 mg/m2 IV weekly for 6 weeks out of 8; or 5 months of androgen suppression, plus 4 cycles of docetaxel 75mg/m2 IV every 3 weeks. Androgen suppression was continued after completion of radiotherapy for a total duration of 3 years. The primary endpoint was unacceptable toxicity. Eight patients (14.8%) developed unacceptable toxicity: 5 in the weekly docetaxel regimen [grade 3 acute genitourinary radiotherapy-related adverse events (n = 3), grade 3 docetaxel hypersensitivity (n = 1), grade 3 fatigue lasting more than 2 weeks (n = 1)] and 3 in the every-three-weeks arm [febrile neutropenia (n = 1), grade 4 neutropenia lasting more than 7 days, grade 3 acute genitourinary radiotherapy-related adverse event (n = 1)]. Compliance with the radiotherapy was excellent, and all patients completed planned treatment. Long-term follow-up continues. The neoadjuvant regimen containing androgen suppression and every-three-weeks docetaxel was chosen for further study based on the promising results of this pilot and the proven efficacy and tolerability of every-three-weeks docetaxel dosing in the advanced-disease setting.3.2 Study DesignThe ncic ctg PR.12 trial is a phase iii study comparing the every-three-weeks docetaxel and neoadjuvant androgen suppression regimen piloted by McKenzie et al. to androgen suppression alone in addition to radical radiotherapy (three-dimensional conformal radiotherapy, 46 Gy in 23 fractions, with 24–28 Gy in 12–14 fractions). Study chairs are Drs. Michael McKenzie and Kim Chi. In both treatment arms, androgen suppression will be given for a total duration of 3 years. Patients with high-risk disease (defined as at least clinical stage T3 or T4, Gleason score of 8 or higher, or psa above 20 ng/mL) are eligible for the study. The primary endpoint is disease-free survival. The sample size for this study is estimated based on detecting an estimated 33.3% risk reduction in disease progression favouring the experimental arm [hazard ratio (hr): 0.667], using a 1-sided log-rank test at the 2.5% significance level and 90% power. An estimated 530 patients (assuming a 14.8% loss to follow-up) will be accrued over 4.5 years, with an additional 5 years of follow-up. Secondary endpoints include overall survival, time to biochemical disease progression, time to local disease progression, time to distant disease progression, time to next anticancer therapy, progression-free survival, degree of psa suppression before radiotherapy, quality of life, and adverse events. Centres will be credentialed by ncic ctg for delivery of radiotherapy before randomization of the first patient. Tumour and biologic specimens will be collected during the study to determine the prognostic role of cytokines and insulin-like growth factor axis markers. In addition, cytokine levels and changes in levels over time will be correlated with fatigue (as measured by the Common Terminology Criteria, version 3.0) and quality of life.4. NCIC CTG PR.13: RADIOTHERAPY AND ANDROGEN DEPRIVATION IN COMBINATION AFTER LOCAL SURGERY (RADICALS)4.1 BackgroundThe PR.13 study represents a collaborative effort between the Medical Research Council Clinical Trials Unit (United Kingdom) and the ncic ctg (Canadian study chairs: Drs. Charles Catton and Fred Saad). This large pragmatic study is addressing two fundamental issues in the postoperative management of patients with resectable prostate cancer: What is the optimal timing of radiotherapy in these patients? And what role, if any, does androgen suppression play in determining outcome? The relevance of the study to current practice is underscored by the fact that prostatectomy is a standard of care in men presenting with operable prostate cancer. In Ontario alone, the number of radical prostatectomies between 1993–1994 and 2003–2004 rose by 171% 25.The role of postoperative radiotherapy has been addressed in three randomized studies:In eortc 22911, 1005 patients with pT3 disease post radical prostatectomy were randomized to either observation or adjuvant radiotherapy 26. The primary endpoint, local control, was modified to clinical progression-free survival and later to biochemical progression-free survival. After a median follow-up of 5 years, biochemical progression-free survival was significantly improved in the irradiated group [74.0%; 98% confidence interval (ci): 68.7 to 79.3] as compared with the observation group (52.6%; 98% ci: 46.6 to 58.5; p < 0.0001). Clinical progression-free survival was significantly better with adjuvant radiation (hr: 0.61; 98% ci: 0.43 to 0.87; p = 0.0009). No difference in overall survival was detected. The rate of 5-year grade 3 or higher toxic effects was 2.6% in the no-further-treatment group and 4.2% in the postoperative irradiation group (p = 0.0726). The incidence of grade 3 urethral stricture and incontinence was 1.4% (6 patients) in each group.A similar design was used in swog 8794 (ncic ctg PR.2), which randomized 425 men with pT3 disease to observation or to radiotherapy to the prostate bed 27. The primary endpoint was metastasis-free survival, defined as the time from randomization to first evidence of metastatic disease or death from any cause. With a median follow-up of 10.6 years, the metastases-free survival was not significantly different between the two arms (hr: 0.75; 95% ci: 0.55 to 1.02; p = 0.06). Overall survival favoured the adjuvant radiotherapy arm, but did not reach statistical significance (hr: 0.80; 95% ci: 0.58 to 1.09; p = 0.16). The rate of biochemical relapse was significantly lower in men with an undetectable psa level post prostatectomy (n = 249) treated with adjuvant radiotherapy (hr: 0.43; 95% ci: 0.31 to 0.58; p < 0.001), as was recurrence-free survival [defined as survival without evidence of measurable or evaluable disease, excluding psa relapse (hr: 0.62; 95% ci: 0.46 to 0.82; p = 0.001)]. Approximately one third of patients randomized to the observation arm ultimately received pelvic radiotherapy. Rectal complications (3.3% vs. 0%, p = 0.02), urethral stricture [17.8% VS. 9.5%; risk ratio (rr): 1.9; 95% ci: 1.1 to 3.1; p = 0.02), and urinary incontinence (6.5% vs. 2.8%; rr: 2.3; 95% ci: 0.9 to 5.9; p = 0.11) were more frequent in the adjuvant radiotherapy arm.Results from the aro 96–02 study were reported at the 2007 meeting of the American Society of Clinical Oncology 28. That study randomized patients with pT3 disease to adjuvant radiotherapy or a “wait-and-see policy.” Those who failed to achieve an undetectable psa level postoperatively on either arm were given a designation of progressive disease and offered radiotherapy. The primary endpoint, biochemical control, was significantly improved in the adjuvant radiotherapy arm (hr: 0.53, p = 0.0015).Taken together, the results of the foregoing trials fail to fully inform physicians and patients about the role of post-prostatectomy radiotherapy in current practice because of differences in outcome definitions used in the trials, lack of consistent effect of adjuvant radiotherapy on clinical (non-psa) endpoints, variable use of late radiotherapy in patients randomized to the observation arm, and current use of assays for psa testing that are more sensitive than those used during the studies.The situation regarding the use of hormone therapy in this group of patients is even less clear. No randomized controlled trials have reported addressing the role and optimal duration of hormone therapy in men receiving post-prostatectomy radiotherapy. The uncertainty among clinicians regarding the role of adjuvant radiotherapy and hormone therapy is reflected in recent surveys of urologists and oncologists, indicating a wide variation in use of these therapies in the post-prostatectomy patient population 29,30.4.1 Study DesignThe radicals trial is designed to address the issues of radiotherapy timing (immediate vs. early salvage) and of hormone therapy duration (none vs. short-term vs. long-term). The primary endpoint is disease-specific survival. It is estimated that the radiotherapy timing randomization will have to recruit 2600 patients and the hormone-duration randomization, 3500 patients. Many patients will be in both randomizations. The trial is planned to address these questions over 12–13 years with 5.5 years of accrual and 7 years of further follow-up. Secondary endpoints include freedom from treatment failure, clinical progression-free survival, overall survival, non-protocol hormone therapy, treatment toxicity, and patient-reported outcomes.The radiotherapy timing randomization involves immediate radiotherapy to the prostate bed versus a salvage radiotherapy policy at the time of psa failure. The radiotherapy will use standard techniques and dose fractionation schedules: 66 Gy in 33 fractions over 6.5 weeks or 52.5 Gy in 20 fractions over 4 weeks. The hormone duration randomization involves no hormone therapy with radiotherapy, compared with short-term (6 months) hormone therapy beginning shortly before radiotherapy, compared with long-term (24 months) hormone therapy beginning shortly before radiotherapy. Patients who decide not to enter the three-way randomization will be able to choose randomization between two of the three arms: 0 as compared with 6 months of hormone therapy if they do not want to be randomized to a long duration of treatment, or 6 as compared with 24 months of hormone therapy if they do not want to be randomized to the no-hormonetherapy treatment arm.Key eligibility criteria for the radiotherapy timing randomization include a postoperative serum psa below 0.4 ng/mL within 3 months after radical prostatectomy, and uncertainty in the opinion of the clinician and patient regarding the need for immediate postoperative radiotherapy. For the hormone duration randomization, patients must be expected to receive radiotherapy (adjuvant or salvage) and must have a psa of 10 ng/mL or more at the time of randomization. In an 18-month feasibility stage, radicals will carefully assess randomization rates and the trial as a whole. Continuation of the trial beyond the feasibility stage will be conditional on satisfactory patient accrual.5. SUMMARYOngoing studies at the ncic ctg are addressing fundamental questions regarding the management of localized prostate cancer.The randomized phase iii Intergroup study PR.11 led by ncic ctg is asking the single most important question regarding the management of favourable-risk prostate cancer: Is active surveillance with a radical intervention based on signs of disease progression as good as radical intervention at diagnosis? The results of this study, whether positive or negative, have the potential to define the management of low-risk prostate cancer globally and to clarify the role of psa doubling time in decision-making.The hypothesis being tested in PR.12 is whether the addition of docetaxel to standard treatment with androgen suppression combined with radiotherapy improves outcome in a high-risk prostate cancer population. This study builds on preclinical data demonstrating the interaction between taxanes, androgen suppression, and development of androgen resistance, and also the extensive literature demonstrating activity of docetaxel in prostate cancer.Finally, PR.13 is a large study that seeks to clarify the roles of post-prostatectomy radiotherapy timing (adjuvant vs. relapse) and the optimal duration of hormone therapy (0 months vs. 6 months vs. 24 months) in patients already treated with prostatectomy.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2528561.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2528561",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2528561\nAUTHORS: E.J. Maher\n\nABSTRACT:\nEstablished in 1993 after a 2-year consultation between professionals and cancer patients, the Lynda Jackson Macmillan Centre (ljmc) has been a catalyst for change in the United Kingdom. The Centre began with a small core staff in a purpose-built building next to a cancer centre, networking with outreach workers in 12 surrounding hospitals, with a mission to improve information, communication, and support for cancer patients. Since 1996, the ljmc model has been adopted and developed by the charity Macmillan Cancer Support and has been spread to more than 60 units across the United Kingdom and Australia.Introducing complementary therapies (cams) to a cancer centre was a particular early challenge. Establishing a shared understanding of the role of complementary therapies and developing nationally accredited written information about them, credible recruitment and governance procedures for therapy practitioners, agreed outcome measures, and peer-reviewed evaluation and research have all been important in engaging cancer physicians and managers; however, charitable funding is still required to support free access to most complementary therapies.An integrated supportive care service for cancer patients begins with a shift in the culture of cancer treatment organizations, moving from a professional-centred to a patient-centred agenda. Real reach and impact requires “new” ideas and services to be integrated into the routine practice of the cancer care delivery organizations. A key lesson learned over the last 15 years is that an integrated support centre must continually adapt to be viable. Sustaining meaningful user guidance is a particular challenge. Support for self-management and the testing and development of cam services are growing parts of the portfolio.\n\nBODY:\nNo Body Content\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2528563.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2528563",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2528563\nAUTHORS: M.J. Verhoef, H.S. Boon, S.A. Page\n\nABSTRACT:\nBackgroundTo ensure the safety and effectiveness of cancer management, it is important for physicians treating cancer patients to know whether their patients are using complementary and alternative medicine (cam) and if so, why.ObjectiveHere, we discuss the ethical and legal obligations of physicians to discuss cam use in an oncology setting, and we provide practical advice on how patient–provider communication about cam can be improved.ResultsPhysicians have both ethical and legal obligations to their patients, including the obligation to respect patient autonomy. This latter obligation extends to use of cam by patients and needs to be addressed beginning early in the patient–provider relationship. Because lack of education in this field and lack of time during patient consultations are barriers to talking with patients about cam, we provide resources to facilitate such discussions. These resources include suggestions on how to discuss the topic of cam and a wide range of information sources.ConclusionsDiscussing cam with patients is the physician’s responsibility, and such discussion will facilitate evidence-based, patient-centred cancer care.\n\nBODY:\n1. INTRODUCTIONStudies report that most patients undergoing cancer treatment also choose to use selected forms of complementary and alternative medicine (cam), including natural health products 1,2. Reasons cited by cancer patients for cam use include treating the cancer, managing treatment side effects, enhancing quality of life and well-being, boosting the immune system, maintaining hope, and having more control over their cancer care 2–4.It is important that physicians treating cancer patients know whether their patients are using cam and, if so, why. First, physicians need to know because of the possibility of direct adverse events associated with the use of cam. Second, interaction effects between conventional medicine and cam are possible, and the harms and benefits of cam may be misattributed to conventional treatment and thus may complicate treatment regimens. Third, patients may delay the use of conventional treatment when using cam. Finally, knowing why patients are using cam may provide important information about beliefs, values, expectations, and hopes on the part of the patient and will facilitate building a trusting relationship that will enhance the delivery of patient-centred cancer care.However, research has shown that 40%–77% of patients who use cam therapies do not disclose their use of or interest in cam, or their desire to use cam, to their physicians because of concerns that the physicians will react negatively or will dismiss their questions 5–7. Patients may also think that physicians do not need to know that they are using cam, because the patients may believe that cam therapies are natural, completely safe, and not within the physicians’ scope of practice. Finally, patients often do not tell, simply because physicians do not ask about cam use.In this paper, we discuss the ethical and legal obligations of physicians to discuss cam use in an oncology setting, and we provide practical advice on how patient–provider communication about cam can be improved.2. DISCUSSION2.1 Obligation of Physicians to Discuss CAMPhysicians have both ethical and legal obligations to their patients, including the obligation to respect patient autonomy. Operationally, respecting patient autonomy in the context of treatment decision-making means allowing patients to make choices 8. However, choice is meaningless if it is not made in light of all relevant information and advice 9. Although patients are the ones who must make the treatment decisions, physicians have the duty to inform patients about all therapeutic options, including cam. This means that physicians must be prepared to provide information and advice about benefits and likely outcomes of treatment, risks involved in treatment, possibility and probability of complications, and side effects and alternative treatment options.This information is needed to meet the traditional ethical principles of non-maleficence (do no harm) and beneficence (offer a benefit) 10. Specifically for cam, the obligation to provide information and advice means that where risks are unknown and benefits are uncertain, it is necessary to highlight this absence of information.Canadian courts have been very liberal and expansive in interpreting the disclosure obligations of physicians. In the context of informed consent, physicians are legally and ethically required to provide patients with detailed information about the evidence (for and against) the possible efficacy of treatments and to discuss costs, risks, and how a given treatment compares with other therapeutic options 11,12. The informed consent approach applies whether the recommended treatment is labelled biomedical or cam, and it raises the question of whether physicians must discuss cam treatment options to fulfil the informed consent requirement to compare a given treatment option with other therapeutic options.It has been argued that an exploration of cam treatment options with patients is necessary, especially when information about cam options will be material—that is, significant—to a decision about a conventional treatment. Thus, if a patient’s decision about pursuing a particular biomedical treatment is likely to be influenced by knowing about specific cam treatments (including evidence of their safety, efficacy, and cost), then a physician has an obligation to include a discussion of these cam options as part of the informed consent process 12,13. Given the widespread use of cam among Canadians diagnosed with cancer, it appears reasonable to assume that cam options will be material to many patients. In addition, perhaps the only way to determine if cam options are relevant factors in the decision-making processes of patients is to open a dialogue about the issues.There is increasing consensus in the literature about the importance of fully disclosing detailed information about the risks and benefits of cam interventions, including clear explanations of what is not known about them 11–13. The focus appears to be on protecting patients from physicians that promote cam products and therapies beyond what is believed by others to be supported by scientific evidence. Legally, physicians are required to practice in accordance with the “standard of care,” which generally refers to “the level of care the average and prudent health care professional in a given community would provide” 13. This standard changes over time and across cultural contexts. Although therapies with scientific evidence of safety and efficacy are unlikely to be judged outside the standard of care, scientific evidence is not the only criterion upon which such judgments are made.In today’s culture of evidence-based practice, scientific evidence is becoming increasingly important 11, but clinical judgment and patient values are important, too 14. Thus, physicians who provide or recommend cam therapies for which there is little evidence could leave themselves open to charges of medical malpractice 11,13. In contrast, requirements to fully discuss cam options as “alternative” treatment options when recommending a conventional biomedical treatment may soon become standard practice.Adams et al.\n15 identified a wide range of patient-and physician-related factors that affect decision-making and subsequent use of cam, including severity of the illness, curability with conventional treatment, side effects of conventional treatment, quality of evidence of safety and efficacy of cam, degree of understanding of risks and benefits, knowing and voluntary acceptance of risks by the patient, and commitment to cam use by the patient.Clearly, the need for physicians to assist patients in treatment decision-making is high and requires more than being informed about cam. Physicians also need to have effective (and non-judgmental) communication skills to manage the discussion. Patient–physician communication plays a crucial role, because these issues are best resolved by means of shared decision-making between patient and physician, where sufficient information is exchanged to create a consensus approach to deciding on the optimal clinical course.2.2 How to Discuss CAM with PatientsIt is becoming increasingly clear that patients have many legitimate needs and concerns that are not being met by conventional medicine. By adhering to the ethical criteria for informed decision-making and by honouring patient autonomy, physicians should be able to engage in open discussions with their patients about cam and to enable their patients to make sound decisions 6.Not communicating with patients about cam may not only result in decreased trust within the therapeutic relationship, but also in selection by the patients of harmful, ineffective, and costly cam therapies. Patients who use cam may have unidentified needs or may be dissatisfied with the conventional care they are receiving. Once the issue of cam use is raised, the unidentified needs or dissatisfaction may come to the forefront and be addressed. However, the fact that relatively few physicians talk with their patients about cam suggests that these conversations are not easy for physicians. Lack of training in cam, limited training in communication skills, limited knowledge about cam, lack of scientific evidence about the risks and benefits of cam, and skepticism towards cam all appear to prevent physician engagement in such discussions.Tasaki 16 found that patients identified these major barriers to successful discussions of cam: perceived indifference or opposition towards cam by the physician, emphasis on scientific evidence by the physician, and anticipation on the part of the patient of a negative response from the physician.Initiating communication about cam is crucial. Table I summarizes a number of suggestions on how to encourage patients to talk about cam. Foley 19 underscores the need for such conversations by talking about “the need for us as oncology professionals to ‘Seek first to understand,’ to be open and to learn from our patients to serve them better.”Eisenberg 20 was one of the first to propose a step-by-step strategy that conventionally trained physicians could use to proactively discuss cam use. This strategy involves a formal discussion of the treatment preferences and expectations of patients, the maintenance of symptom diaries, and follow-up visits to monitor for potentially harmful situations. In the absence of medical and legal guidelines, the proposed management plan emphasizes patient safety, the need for documentation in patient records, and the importance of shared decision-making. Although this strategy may be impractical and cumbersome in practice, Eisenberg should be credited for highlighting essential elements in physician–patient communication and for focusing on documentation and follow-up.Informally, the five major steps to intervention—“ask, advise, assess, assist, and arrange” (5 A’s) 21—have been mentioned as important guidelines for communicating about cam, but a more focused approach has been presented by Cohen et al.\n13. They suggest exploration (of the patient’s main issues), validation (acknowledge and commend the patient for seeking to resolve symptoms and improve health), empathy (with the patient’s desire to do everything possible), evaluation (consult with colleagues and other experts, and consult reputable sources of information), communication (share findings with the patient), and documentation of the conversations with the patient and of the patient’s progress as previously highlighted by Eisenberg 20.Yet another perspective is provided by Frenkel et al.\n22. These authors suggested that, to help cancer patients be truly informed and autonomous, physicians need to identify the patient’s beliefs, fears, hopes, and expectations; learn which conventional treatments have been tried, have failed, or have been rejected and why; make sure the patient understands the prognostic factors associated with his or her stage of disease, plus the potential benefits and harms of conventional medicine; acknowledge the patient’s spiritual and religious values and beliefs to understand how these affect health care choices; and assess the level of support that the patient has from friends, family, and community.These frameworks approach the issue of communication with patients from slightly different angles, yet all are important, and physicians will most likely use elements of all approaches depending on the particular situation.Implicitly, all models suggest that cam use should be inquired about from the beginning of contact with the patient, ideally before the patient starts using cam. This emphasis suggests that cam use should be made a regular part of history-taking. Physicians therefore need effective communication skills to fulfil a variety of roles, including collecting medical histories, answering patients’ questions, developing interpersonal relations, and suggesting treatment 16. Although this need seems obvious, it is yet another demand on physicians working in often busy and stressful situations.Lastly, it is important to consider that, although disclosure of cam is essential, successful communication hinges on supporting patient autonomy even when the patient is making the decision to use a therapy of which a physician does not approve.2.3 Current and Future Trends in Patient–Provider CAM DiscussionThe importance of talking with patients about cam therapies is currently receiving much attention. Recently, the National Center for Complementary and Alternative Medicine (nccam) in the United States started the Time to Talk campaign 23, urging health care providers to talk about cam. The nccam Web site also includes tips on how to talk with patients. In addition, the British Medical Journal recently published a challenging editorial, “Wham, bam, thank you cam” 24, highlighting the need to discuss cam; however, the author’s question, “Alternative medicine is wildly popular ... but what are we supposed to do about it?” raises the challenge of finding relevant, evidence-based information.Uncovering evidence-based information is especially difficult given the large number and heterogeneity of cam interventions. As a result, only a limited number of interventions have been adequately tested. Limited time in which to learn about cam and to discuss cam-related issues in patient consultations that are often already too short to address all patients’ concerns poses yet another challenge for physicians.Currently, no single resource contains comprehensive summaries of the evidence base of all cam treatments relevant for patients diagnosed with cancer. In addition, the available evidence changes almost daily. However, helpful information can be found in a number of places. Table II includes a list of evidence-based cam resources that may help physicians when talking to patients about cam. Most are Web sites, because these are much easier to access (and are updated more regularly) than are books and articles. We recommend that physicians track the sources they access and find helpful, so that those sources are readily available when needed. Information on how to evaluate the wide range of cam information sources on the Web is available from nccam in the United States 25.Because patients may see cam practitioners for cancer and cancer-related symptoms, it is also beneficial to be informed about the cam practitioners in the local area who are seeing cancer patients. The Prince of Wales Foundation for Integrated Health in the United Kingdom has published several guides for patients using cam\n26. These guides include helpful information on finding cam practitioners and asking the right questions about those practitioners. An important aspect to assess is whether a given cam profession is regulated and whether a specific practitioner has adequate credentials.It will be important in the future to ensure that cam is a topic in medical education, because all graduating physicians will encounter this issue in their practice. The Canadian cam in ume (undergraduate medical education) Web site provides useful resources for those involved in teaching medical students about cam\n17. For physicians already in practice, continuing medical education may be a solution; however, most important is what can be learned from talking with patients: not only what they use, but also what their questions are.3. CONCLUSIONSMany people have already been using cam before a cancer diagnosis, and they consider it to be part of their health care. It is important to note that it is not possible to provide evidence-based, patient-centred care without engaging in a discussion of cam, including an exploration of the patient’s beliefs. Patients may be reluctant to discuss cam because of a fear of rejection or because of their beliefs about the complete safety of cam. It is therefore important that physicians initiate discussion of the topic—ideally, early in the relationship (for example, during initial history-taking), before patients have made any decisions about cam treatments. It truly is time to talk about cam with patients.TABLE ICommunicating with patients: keeping the door open17,18What follows is a list of possible questions and issues that physicians can raise. Obviously, it is neither possible nor necessary to address all areas; however, even one good question may be the key to either opening the door to discussing cam or keeping the door open.1. Always ask about complementary and alternative medicine (cam) use—for example, “What else are you doing to take care of your cancer?” Ask in an open, non-judgmental way, and avoid using labels such as quackery, unscientific, and so on.2. Watch for “non disclosing” clues: “You have read a lot about this. Have you seen other types of practitioners?”3. Give permission for the patient to raise the topic by asking, “Many of my patients are interested in trying complementary therapies. Have you used any other therapies for this problem?”4. Check with patients about their explanatory models: “What do you think is causing your symptoms [or cancer (because many patients have strong opinions of causes of cancer)]?”5. Seek more information from patients and other sources: “Do you have any articles you can share with me?” a. Be prepared for patients doing their own research. b. Be aware of what they are being told about cam.6. Explore why patients are using cam, and learn about their beliefs and values. It is important to consider that a. a great deal more than evidence goes into a patient’s decision to use cam. b. for many patients, care (enhancing well-being, easing suffering) is as important as cure.7. Discuss the patient’s treatment preferences and expectations.8. Review issues of efficiency and safety with respect to cam.9. Be frank about your level of understanding or knowledge. It is okay not to know everything about cam.10. Support the patient in efforts to obtain answers to important questions about risk and benefit. Ask yourself: a. Is the cam therapy really dangerous? b. Does it prohibit necessary medical care? c. Can you work within the patient’s belief system to provide good care?  If the answer to the last question is yes, the next steps include negotiation and education. If the answer is no, the next step would be to arrive at a mutually acceptable course of action.11. Discussing cam use does not mean that you are endorsing or promoting cam use.TABLE IIResource books and Web sitesThe Desktop Guide to Complementary and Alternative Medicine: An Evidence-based Approach Ernst E, Pittler MH, Wider B, editors. 2nd edition, 2006 480 pages, paperbackIntegrative Medicine: Principles for Practice (Chapter 23, pp. 535–549) Kligler B, Lee R.2004 700 pages, hardcoverComplementary and Alternative Medicine Secrets: Q&As about Integrating CAM Therapies into Clinical Practice (Chapters 54 and 55, pp. 363–388) Kohatsu W. 2002 456 pages, paperbackIntegrative Medicine(Section 13, pp. 809–899; evidence for all treatments is rated) Rakel D. 2nd edition, 2007 1238 pages, hardcoverThe Oxford Handbook of Complementary MedicineErnst E, Pittler MH, Wider B, Boddy K.  2008 512 pages, paperbackCAMline www.camline.caCenter for Health and Healing (a service of Beth Israel Medical Center in New York)  www.healthandhealingny.orgNational Center for Complementary and Alternative Medicine (nccam) nccam.nih.gov/healthNatural Medicines Comprehensive Database www.naturaldatabase.comNatural Medicines Comprehensive Database—Clinical Management Series www.naturaldatabase.com/(S(st2arzb2hbi2v355rtipno2p))/nd/ClinicalMngt.aspx?cs=&s=NDNatural Standard Database www.naturalstandard.comTurning Research into Practice (trip) databasewww.tripdatabase.comThe University of Texas MD Anderson Cancer CenterComplementary/Integrative Medicine Education Resources www.mdanderson.org/departments/cimerMemorial Sloan–Kettering Cancer Center www.mskcc.org/mskcc/html/44.cfm\n\nREFERENCES:\n1. RichardsonMASanderTPalmerJLGreisingerASingletarySEComplementary/alternative medicine use in a comprehensive cancer center and the implications for oncologyJ Clin Oncol20001825051410893280\n2. SparberABauerLCurtGUse of complementary medicine by adult patients participating in cancer clinical trialsOncol Nurs Forum2000276233010833691\n3. BoonHStewartMKennardMAThe use of complementary and alternative medicine by breast cancer survivors in Ontario: prevalence and perceptionsJ Clin Oncol20001825152110893281\n4. VerhoefMBalneavesLBoonHVroegindeweyAReasons for and characteristics associated with complementary and alternative medicine use among adult cancer patients: a systematic reviewIntegr Cancer Ther200542748616282504\n5. RobinsonAMcGrailMRDisclosure of cam use to medical practitioners: a review of qualitative and quantitative studiesComplement Ther Med20041290815561518\n6. PappasSPerlmanAComplementary and alternative medicine: the importance of doctor–patient communicationMed Clin North Am20028611011795082\n7. AdlerSFosketJDisclosing complementary and alternative medicine use in the medical encounter: a qualitative study in women with breast cancerJ Fam Pract199948453810386489\n8. WeirMObligation to advise of options for treatment—medical doctors and complementary and alternative medicine practitionersJ Law Med20031029630712650001\n9. BrophyEDoes a doctor have a duty to provide information and advice about complementary and alternative medicine?J Law Med2003102718412649999\n10. CohenMLegal and ethical issues relating to use of complementary therapies in pediatric hematology/oncologyJ Pediatr Hematol Oncol200628190316679948\n11. CaulfieldTFeasbyCPotions, promises and paradoxes: complementary medicine and alternative medicine and malpractice law in CanadaHealth Law J2001918320312141221\n12. ErnstECohenMHInformed consent in complementary and alternative medicineArch Intern Med200116122889211606143\n13. CohenLCohenLKirkwoodCRussellNDiscussing complementary therapies in an oncology settingJ Soc Integr Oncol20075182417309810\n14. SackettDStraussSRichardsonWRosenbergWHaynesREvidence-based Medicine. How to Practice and Teach EBMTorontoChurchill Livingstone2000\n15. AdamsKECohenMHEisenbergDJonsenAREthical considerations of complementary and alternative medical therapies in conventional medical settingsAnn Intern Med2002137660412379066\n16. TasakiKMaskarinecGShumayDTatsumuraYKakaiHCommunication between physicians and cancer patients about complementary and alternative medicine: exploring patients’ perspectivesPsychooncology2002112122012112481\n17. Community Health Sciences, University of CalgaryComplementary and Alternative Medicine Issues in Undergraduate Medical Education [Web resource]CalgaryUniversity of Calgary2003[Available at: www.caminume.ca; cited March 24, 2008]\n18. CoulehanJLBlockMRThe Medical Interview. Mastering Skills for Clinical Practice5th edPhiladelphiaF.A. Davis Company200630118\n19. FoleyGThe quest for understanding: a challenge in cancer careCancer Pract20021017312100100\n20. EisenbergDMAdvising patients who seek alternative medical therapiesAnn Intern Med19971276199214254\n21. DoshSAHoltropJTorresTArnoldABaumannJWhiteLChanging organizational constructs into functional tools: an assessment of the 5 A’s in primary care practicesAnn Fam Med20053suppl 2S50216049088\n22. FrenkelMBen-AryeEBaldwinCSierpinaVApproach to communicating with patients about the use of nutritional supplements in cancer careSouth Med J2005982899415813155\n23. National Institutes of Health, National Centre for Complementary and Alternative Medicine (nccam)Time to Talk [Web page]BethesdaNCCAMn.d. [Available at: nccam.nih.gov/timetotalk; cited: March 24, 2008]\n24. KamerowDWham, bam, thank you camBMJ200733564717901515\n25. National Institutes of Health, National Centre for Complementary and Alternative Medicine (nccam)10 Things to Know About Evaluating Medical Resources on the Web [Web page]BethesdaNCCAM[available at: nccam.nih.gov/health/webresources; cited: July 2, 2008]\n26. PinderMPedroLTheodorouGTreacyKMillerWComplementary Healthcare: A Guide for PatientsLondon, U.K.The Prince of Wales’s Foundation for Integrated Health2005[Available online at: www.fih.org.uk/document.rm?id=19; cited March 24, 2008]"
+}

batch_0/PMC2528937.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2528937",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2528937\nAUTHORS: Marco Cassano, Stefano Biressi, Amanda Finan, Laura Benedetti, Claudia Omes, Renata Boratto, Frank Martin, Marcello Allegretti, Vania Broccoli, Gabriella Cusella De Angelis, Paolo M. Comoglio, Cristina Basilico, Yvan Torrente, Paolo Michieli, Giulio Cossu, Maurilio Sampaolesi\n\nABSTRACT:\nBackgroundHepatocyte Growth Factor (HGF) is a pleiotropic cytokine of mesenchymal origin that mediates a characteristic array of biological activities including cell proliferation, survival, motility and morphogenesis. Its high affinity receptor, the tyrosine kinase Met, is expressed by a wide range of tissues and can be activated by either paracrine or autocrine stimulation. Adult myogenic precursor cells, the so called satellite cells, express both HGF and Met. Following muscle injury, autocrine HGF-Met stimulation plays a key role in promoting activation and early division of satellite cells, but is shut off in a second phase to allow myogenic differentiation. In culture, HGF stimulation promotes proliferation of muscle precursors thereby inhibiting their differentiation.Methodology/Principal FindingsMagic-Factor 1 (Met-Activating Genetically Improved Chimeric Factor-1 or Magic-F1) is an HGF-derived, engineered protein that contains two Met-binding domains repeated in tandem. It has a reduced affinity for Met and, in contrast to HGF it elicits activation of the AKT but not the ERK signaling pathway. As a result, Magic-F1 is not mitogenic but conserves the ability to promote cell survival. Here we show that Magic-F1 protects myogenic precursors against apoptosis, thus increasing their fusion ability and enhancing muscular differentiation. Electrotransfer of Magic-F1 gene into adult mice promoted muscular hypertrophy and decreased myocyte apoptosis. Magic-F1 transgenic mice displayed constitutive muscular hypertrophy, improved running performance and accelerated muscle regeneration following injury. Crossing of Magic-F1 transgenic mice with α-sarcoglycan knock-out mice –a mouse model of muscular dystrophy– or adenovirus-mediated Magic-F1 gene delivery resulted in amelioration of the dystrophic phenotype as measured by both anatomical/histological analysis and functional tests.Conclusions/SignificanceBecause of these features Magic-F1 represents a novel molecular tool to counteract muscle wasting in major muscular diseases such as cachexia or muscular dystrophy.\n\nBODY:\nIntroductionHepatocyte Growth Factor (HGF), also known as Scatter Factor (SF), is a pleiotropic cytokine of mesenchymal origin that mediates a characteristic array of biological activities including cell proliferation, survival, motility and morphogenesis [1]–[3]. Its high affinity receptor, the tyrosine kinase Met, is expressed by a wide range of tissues including epithelial, endothelial, hematopoietic, neuronal and muscular cells [4], [5]. Embryonic muscle precursor cells express Met and migrate following HGF gradients during embryo development [6]–[10]. Genetic impairment of HGF-Met signaling in mice leads to abnormal muscle development in the limbs, thorax and tongue [11]–[13], and newborns -which are ataxic and have breathing problems- die a few hours later because they cannot suck mother's milk [14]. In the adult, the HGF-Met pathway is involved in muscle regeneration following injury. Muscle satellite cells, which reside in the stroma of muscular tissues and express both HGF and Met [15], represent a pool of muscle precursors that are activated and stimulated to divide when muscle regeneration or adaptive growth is needed [16], [17]. Autocrine HGF-Met stimulation plays a key role in mediating activation and early division of satellite cells, but is shut off in a second phase in order to allow the cells to exit the cell cycle and to enter the differentiation process [18], [19]. HGF stimulation of cultured satellite cells promotes cell proliferation and inhibits myogenic differentiation [20].Magic Factor-1 (Met-Activating Genetically Improved Chimeric Factor-1 or Magic-F1) is an HGF-derived, engineered protein that contains two Met-binding domains repeated in tandem. It has a reduced affinity for Met and, in contrast to HGF, it elicits activation of the AKT but not the ERK signaling pathway. As a result of its partial ability to activate Met signaling, Magic-F1 is not mitogenic but conserves the ability to protect cells against apoptosis. We have analyzed the effects of Magic-F1 on muscular cells both in vitro and in mice. We show that Magic-F1 protects myogenic precursors against apoptosis and thus enhances the differentiation process, which is naturally accompanied by cell death. This pro-differentiative effect is observed both in cultured myogenic cell systems and in two different in vivo models. Remarkably, constitutive or transient expression of Magic-F1 in a mouse model of muscular dystrophy partially rescues the dystrophic phenotype and allows animals to perform better in a classic tread mill functional test. These features make Magic-F1 a novel, potential molecular tool to counteract muscle wasting in major muscular diseases including cachexia and muscular dystrophy.ResultsEngineering of Magic-F1, a bivalent Met ligandMature HGF is a dimeric molecule consisting of a α- and a β-chain joint by a disulphide bridge [21]. The α-chain contains a leader peptide for secretion, an N-domain similar to the activation domain of plasminogen, and four kringle domains (K 1–4) typical of the blood clotting cascade proteases [22]. In functional terms, HGF is a bivalent molecule containing two distinct Met binding sites, one in the α-chain high affinity; [23] and one at in the β-chain low affinity; [24]. Isolated HGF domains containing only one receptor binding site (HGF NK1, HGF NK2, HGF α-chain, HGF β-chain) can bind to the Met receptor but do not activate it [22]–[25], thus suggesting that a bivalent molecule is necessary to achieve receptor activation. Consistent with this idea, some monovalent scatter factor subdomains (HGF NK1, HGF NK2) display a partial agonistic activity when they are stabilized in a dimeric form by extracellular matrix proteoglycans [26]. To generate new recombinant proteins capable of inducing specific patterns of biological responses, we engineered several artificial molecules containing different HGF domain in various combination. Magic-F1, the prototype of this series, contains the signal peptide plus the N-domain and the first two kringles repeated in tandem and joint by a linker (Fig. 1A). A poly-histidine tag was engineered at the C-terminal end to facilitate protein purification. Since the high affinity Met binding site lies within the N and K1 domains [23], Magic-F1 is a bivalent ligand. Magic-F1 recombinant protein was produced using both transiently and stably transfected CHO cells, and was purified by affinity chromatography as described in the Experimental Protocol section (Fig. 1B). The affinity of Magic-F1 for Met was measured in a ELISA binding assay using a recombinant chimera between Met and the Fc portion of a human immunoglobulin Fc-Met; [27]. Fc-Met was absorbed in solid phase and exposed to increasing concentrations of Magic-F1 or HGF in liquid phase. Binding was revealed using biotinylated anti-HGF antibodies. This analysis revealed that Magic-F1 has an affinity for Met that is approximately 7–8 times lower than that of HGF (i.e. 0.8 nM; Fig. 1C). These data are consistent with previous measurements that determined the affinity of different subdomains of HGF for Met [23].10.1371/journal.pone.0003223.g001Figure 1Magic-F1 elicits partial activation of the Met pathway in C2C12 myogenic cells.(A) Schematic representation of the Magic-F1 molecule. The indicated restriction sites refers to the corresponding cDNA map. BamH1 and Sal1 were destroyed and the DNA fragment was cloned into the EcoRV restriction site of the pIRES-neo plasmid. (B) Purification of Magic-F1 by metal-chelate affinity chromatography. Following elution, fractions (F1-6) were resolved by SDS-PAGE in non-reducing conditions along with bovine serum albumine (BSA) standards. Proteins were revealed by Coomassie staining. MW, molecular weight; kDa, kilo Dalton units (C) ELISA binding assay. A fixed amount (100 ng/well) of Fc-Met chimera was absorbed in solid phase and exposed to increasing concentrations of HGF or Magic-F1 in liquid phase. Binding was revealed using biotinylated anti-HGF antibodies. (D) Met phosphorylation analysis in C2C12 cells. Cells were stimulated with no factor (Ctrl), 5 nM Magic-F1 (M) or 5 nM HGF (HGF), and Met phosphorylation was determined using anti-phosphotyrosine antibodies (IB). The same blots were reprobed with anti-Met antibodies to normalize the amount of receptor immunoprecipitated (IP). (E) Growth curves of C2C12 cells treated with the indicated concentrations of Magic-F1, HGF or no factor. (F) Signal transduction analysis. Cells were stimulated with no factor (Ctrl), 5 nM HGF, 5 nM NK2, 5 nM Magic-F1 (M), or 5 nM Magic-F1 and 5 nM HGF (M+HGF). Cell lysates were analyzed by Western blotting using antibodies against ERK or AKT (total) as well as antibodies against the phosphorylated forms of these signal transducers (ph).Magic-F1 does not induce myoblast proliferationSince HGF has been shown to affect satellite cell proliferation and differentiation, the action of the Magic-F1 on these biological processes was investigated by different approaches. We first subjected the myogenic cell line C2C12 [28] to different biological and biochemical assays in the presence of recombinant Magic-F1. Myoblast proliferation was evaluated by culturing C2C12 cells with Magic-F1, HGF or no factor as control. While HGF induced myoblast proliferation in a dose-dependent manner, Magic-F1 did not affect proliferation even at high concentrations as well as NK2 (Fig. 1E). As phosphorylation of Met is necessary for the activation of the HGF signaling cascade [1], we tested whether Magic-F1 could induce Met receptor phosphorylation. Immunoprecipitation analysis of Met followed by Western blot analysis using anti-phosphotyrosine antibodies revealed that both HGF and Magic-F1 induce phosphorylation of Met in C2C12 cells (Fig. 1D), indicating that the inability of Magic-F1 to affect myoblasts proliferation is not due to defective receptor activation. Since HGF is able to promote cell proliferation through the ERK pathway and to prevent apoptosis through AKT signaling [29], we next tested the ability of Magic-F1 to activate these two distinct pathways. While HGF induced phosphorylation of both MAPK and AKT. Magic-F1, differently form NK2, induced phosphorylation of AKT. Moreover, consistent with the idea that HGF and Magic-F1 compete for the same binding site on Met, Magic-F1 inhibited HGF-mediated MAPK phosphorylation (Fig. 1F).Magic-F1 promotes myoblast differentiation and survivalNext, we generated several stable clones of C2C12 myoblasts expressing Magic-F1 (Fig. 2A). Surprisingly, C2C12 cells expressing Magic-F1 differentiated at a faster rate compared to controls. In fact, they started to express myosin heavy chain, a marker of terminal differentiation, only one day following switch to differentiation medium (Fig. 2B). Consistent with accelerated differentiation, the myogenic markers MyoD and Myf5 were up-regulated while the Pax3 protein was down-regulated (Fig. 2D). Moreover, Magic-F1 increased the expression of 30 out of 36 genes known to be upregulated during C2C12 differentiation [30]; Figure S2. Magic-F1-expressing C2C12 cells fused into myotubes containing on average more nuclei than controls, while HGF did not affect myoblast fusion (Fig. 2C). Interestingly, in stable clones expressing Magic-F1, myostatin expression but not follistatin or IGF1 expression was down-regulated earlier compared to controls (Fig. 2E). This is in agreement with previous data showing promotion of myoblast differentiation and muscle hypertropy following myostatin ablation [31], [32]. Finally, cells expressing Magic-F1 displayed a marked reduction in the expression of several pro-apoptotic genes, including Bad, Bax and p53 (Fig. 2F) suggesting that the anti-apoptotic properties documented for HGF [29] are conserved in Magic-F1. Thus, Magic-F1 is an engineered, HGF-derived protein that elicits a selective pattern of biological responses on myoblasts. Firstly, it is a partial agonist of Met that activates the AKT pathway but not the ERK pathway. Secondly, it conserves the anti-apoptotic activity of HGF but not its mitogenic properties. Thirdly, it significantly enhances the differentiation potential of myoblasts without affecting their proliferation. The latter property is likely to be due to its inability to activate the ERK pathway.10.1371/journal.pone.0003223.g002Figure 2Morphological analysis of C2C12 cells expressing Magic-F1.(A) Magic-F1 detection in the culture media of stably transfected clones by Western blot analysis. (B) Immunofluorescence analysis for myosin heavy chain expression on Magic-F1 expressing clones (right panels) or control clones (left panels). Cells were analyzed after 1 day (1 d), 3 days (3 d) and 5 days (5 d). Nuclei were stained with DAPI. (C) Fusion index of C2C12 cells stably transfected with Magic-F1, HGF or mock-transfected (Ctrl). Fusion index is the ratio between the number of myocites with two or more nuclei versus the total number of myocites. (D) RT-PCR analysis of myogenic transcription factors (Myf-5, Pax3 and MyoD) on stably transfected clones (M) or control cells (C). GAPDH is used as internal control. (E) RT-PCR analysis of myostatin, follistatin and IGF1 expression in proliferating (P) versus differentiating (D) C2C12 cells. C, control cells; M, cells expressing Magic-F1. Skeletal muscle tissue (sk Mus) and a mouse embryo at 10.5 days (E10.5) were also used as controls. GAPDH was used as an internal control. (F) RT-PCR analysis of pro-apoptotic genes Bad, Bax and p53 in C2C12 clones stably transfected with Magic-F1 (M) or control cells (Ctrl). GAPDH was used as an internal control.Electro-enhanced Magic-F1 DNA transfer in vivo promotes muscle hypertrophy and protects myocites against apoptosisEfficient secretion of therapeutic proteins can be induced into skeletal muscle through electro-enhanced DNA transfer [33]. Using this technology, we tested the activity of Magic-F1 on mouse skeletal muscles in vivo. A plasmid encoding Magic-F1 was co-electroporated with a plasmid expressing β-galactosidase into the tibialis anterior and quadriceps muscles of juvenile mice (postnatal day 10) as described.A vector encoding HGF and an empty vector without insert were used as controls. Histological analysis using X-gal staining showed that β-galactosidase was widely expressed one week after intra-muscular DNA electrotransfer but rapidly declined afterwards (Fig. 3A). Expression of the foreign genes also reached its maximum one week post-transfer and lasted for up to three weeks, as determined by RT-PCR analysis (Fig. 3B). Morphometric analysis performed on the tibialis anterior and quadriceps (9 mice for each group and 300–460 fibres for each sample were analyzed) revealed a significant increase of the cross-sectional area of Magic-F1-electrotransferred muscles compared to the control muscles starting two weeks after electrotransfer (Fig. 3C) as well as an increase in fiber perimeter (not shown). Representative images of electroporated quadriceps stained with hematoxylin and eosin are shown in Fig. 3D. Next, we evaluated whether Magic-F1 could protect muscle cells against apoptosis. To this end, we performed a TUNEL analysis of muscle sections one week after in vivo electrotransfer. This analysis indeed showed a decreased number of apoptotic nuclei (TUNEL positive) in muscles treated with either Magic-F1 or HGF (Fig. 3E). Taken together, the in vitro and in vivo data presented here suggest that Magic-F1 induces hypertrophy in the developing skeletal muscle by enhancing the differentiation and fusion ability of myogenic cells and by protecting them against apoptosis.10.1371/journal.pone.0003223.g003Figure 3In vivo electrotransfer-mediated delivery of Magic-F1 to muscles.10-day-old mice were subjected to in vivo electroporation of a combination of two plasmids expressing beta-galactosidase (β-gal) and HGF or Magic-F1, respectively. (A) Percent of β-gal-positive fibers following X-gal staining 4 weeks after electroporation. Representative images of electroporated muscles 1 week (1 w) and 4 weeks (4 w) after electrotransfer are shown in the upper panel. (B) RT-PCR analysis of electroporated muscles expressing HGF or Magic-F1 (data refers to samples obtained from quadriceps). (C) Histograms of morphometric analysis performed on tibialis anterior. Nine mice per group were analyzed. For each mouse, 300–460 fibers were examined (p<0.01). (D) H&E staining of quadriceps electroporated muscles. Note the larger fibers formed after 4 weeks in the Magic-F1 group relative to the control (Ctrl) or HGF group. (E) Tunel analysis of quadriceps 1 week after in vivo electrotransfer.Magic-F1 transgenic mice display hypertrophic fast-twitch fibers and improved running abilityTo further investigate the ability of Magic-F1 to promote muscle hypertrophy, we generated transgenic mice expressing Magic-F1 under the control of the skeletal muscle-specific regulatory elements of the rat myosin light chain MLC1F gene locus [34]; Fig. 4A. MLC1F/Magic-F1 transgenic lines were identified by genotyping PCR with primers specific for the Magic-F1 coding sequence (Fig. 4B). Expression of the Magic-F1 transgene in adult mice was detected by RT-PCR in all muscles analyzed; on the contrary, no signal was detected in the liver of transgenic mice or in any organ of wild-type animals (Fig. 4C). Protein expression in fast transgenic muscles was also confirmed by Western blotting analysis (Fig. 4D). Embryonic and post-natal development of MLC1F/Magic-F1 transgenic animals occurred without overt differences compared to control mice. Skeletal muscle hypertrophy became apparent at around 5 weeks of age, consistent with the in vivo electrotransfer results. Morphometric analysis of the fast tibialis anterior muscles in transgenic mice showed a statistical significant increment of myofiber cross-sectional areas compared to age-matched wild-type controls (Fig. 4E, left histogram). Interestingly, morphometric analysis of slow-twitch soleus muscles unveiled no difference between transgenic and control animals (Fig. 4E, right histogram), even though transgene expression was detected in the soleus muscle (see Fig. 4C). A treadmill test was performed in order to evaluate the effect of Magic-F1 on muscular performance. This in vivo motility assay revealed that MLC1F/Magic-F1 transgenic mice cover on average a longer distance in comparison to their wild-type counterparts (Fig. 4F), thus demonstrating that Magic-F1-induced muscle hypertrophy results in increased muscular performance.10.1371/journal.pone.0003223.g004Figure 4Morphological and functional analysis of Magic-F1 transgenic mice.(A) Transgenic construct used for two different microinjections into ES cells. (B) Representative example of tail genotyping by PCR. The M13newborn is positive for Magic-F1 integration. In two different microinjections we obtained 2 founders out of 14, that generated two different transgenic colonies. (C) RT-PCR analysis of Magic-F1 expression in transgenic (MLC1F/Magic-F1) and wild-type muscles (WT). No signal was detected in the liver of transgenic mice (Liv) or in any organ of wild-type animals. (D) Western blot analysis of Magic-F1 expression using anti-HGF antibodies or anti-GAPDH antibodies as control. A 60 kDa band appeared only in muscles from transgenic mice. (E) Morphometric analysis of tibialis anterior (left histogram) and soleus (right histogram) muscles. For each sample; 300–400 fibers were analyzed. (F) Distance performed by transgenic and control mice on a treadmill test. For more information, please refer to the Materials and Methods section.Magic-F1 transgenic mice display enhanced muscle regenerative capacityIn muscular dystrophy disorders, fiber degeneration is only partially counterbalanced by regeneration of new fibers by satellite cells [35]. Hypertrophic factors represent a potential therapeutic approach against muscle wasting. We therefore analyzed the effect of Magic-F1 on muscle regeneration. Muscle damage was induced in the tibialis anterior muscles of adult MLC1F/Magic-F1 transgenic or wild-type mice by a single intramuscular injection of cardiotoxin. MLC1F/Magic-F1 transgenic animals responded to muscle crush by rapidly activating the regenerative program. Three days after cardiotoxin injection, an enhanced number of centrally-nucleated regenerating myofibers and an increased expression of the regeneration hallmark protein, embryonic myosin heavy chain (MyHC), was observed in damaged muscles of transgenic mice compared to those of age-matched wild-type animals (Fig. 5A). Furthermore, one week post-injury, muscle fibers of MLC1F/Magic-F1 transgenic mice were characterized by enhanced peripherycal localization of nuclei and by the down-regulation of embryonic MyHC, indicating successful completion of the regeneration program. In contrast, in wild-type animals, regeneration persisted for a few more days (Fig. 5A). Interestingly, also regenerating centrally-nucleated fibers in the MLC1F/Magic-F1 transgenic mice appeared to have a greater cross-sectional area in comparison to wild-type animals after 3 days of injury (Figure S3). Consistent with these observations, satellite cells collected from MLC1F/Magic-F1 transgenic showed enhanced differentiation potential in vitro compared to satellite cells from wild-type mice. Furthermore, satellite cells from MLC1F/Magic-F1 transgenic mice were more differentiation-prone as revealed by smaller clone size and accelerated appearance of differentiated myotubes (Fig. 5B). Moreover, cardiotoxin induced a rapid apoptotic response in injected areas, which appeared to be strongly reduced in MLC1F/Magic-F1 transgenic animals (Fig. 5C and D). Rapid and efficient muscle regeneration in transgenic muscles subjected to cardiotoxin treatment is also explained by earlier and increased expression of the muscle master genes MyoD and Myf5 (Fig. 5E). This resulted in reduction of central nucleated fibers at 10 days following cardiotoxin treatment (Fig. 5F) and in greater cross-sectional area of regenerated transgenic fibers compared to wild-type animals (Figure S3B).10.1371/journal.pone.0003223.g005Figure 5Magic-F1 promotes muscular regeneration.(A) Immunofluorescence analysis of muscle fibers using antibodies against embryonic myosin heavy chain (red) or laminin (green) in the tibialis anterior of transgenic and wild-type mice. Nuclei were stained with DAPI. (B) Immunofluorescence analysis for desmin (middle panels, in green) and myosin heavy chain (lower panels, in red) of satellite cells isolated from tibialis anterior of Magic-F1 transgenic mice (M) and wild-type (WT) mice subjected to cardiotoxin treatment. Nuclei are stained with DAPI (in blue). The upper panels show a phase contrast image of satellite cell clones, 3 days after low density seeding. (C) TUNEL analysis of tibialis anterior after 3, 7 and 14 days after cardiotoxin treatment. (D) Quantification of apoptotic nuclei (ap nuclei) relative to the experiment described in C. Red line, transgenic mice; blue line, wild-type mice. (E) RT-PCR analysis of myogenic transcription factor expression (MyoD and Myf5) conducted on tibialis anterior from transgenic (M) or wild-type (WT) mice. (F) Representative images of tibialis anterior muscles stained with H&E extracted from Magic-F1 transgenic mice and wild-type mice 10 days after cardiotoxin treatment. Note the larger size of fibers in the Magic-F1 group (M) compared to the control group (WT).Magic-F1 partially rescues the dystrophic phenotype of alpha-sarcoglycan knock-out miceThe therapeutic potential of Magic-F1 was tested in alpha-sarcoglycan (α-SG) knock-out mice, which represent an established animal model of muscular dystrophy. Due to their genetic defect, these mice display persistent degeneration and regeneration areas in skeletal muscles [36]. To achieve Magic-F1 expression in these mice, we undertook two different approaches. Firstly, we crossed Magic-F1 transgenic mice with α-SG knock-out animals, thus generating α-SG knock-out mice expressing Magic-F1 in their muscles (Fig. 6A). Secondly, we engineered an adenoviral vector [37] expressing Magic-F1 and administered it by intramuscular injection to 45 day-old α-SG knock-out female mice under immunosuppressive conditions [38]. Morphological analysis of the tibialis anterior of α-SG knock-out/Magic-F1 transgenic mice revealed significant muscular hypertrophy compared to α-SG knock-out controls, which persisted until at least 6 months of age (Fig. 6B). Consisted with this, α-SG knock-out/Magic-F1 transgenic mice performed much better than control α-SG knock-out mice in a classic treadmill test (Fig. 6C). Adenovirus-mediated delivery of Magic-F1 also ameliorated the dystrophic phenotype of α-SG knock-out mice, although to a reduced extent compared to α-SG knock-out/Magic-F1 transgenic mice (Fig. 6C). This may be due to the lower expression levels of Magic-F1 achieved by adenoviral transduction (see Western blot analysis in Fig. 6A). In any case, the values obtained were statistically significant compared to dystrophic animals treated with a control adenovirus (Fig. 6C).10.1371/journal.pone.0003223.g006Figure 6Magic-F1 increases muscle strength in α-SG knock-out mice.(A) After performance of an exhaustion treadmill tests, mice were sacrificed and Magic-F1 expression was evaluated by RT-PCR (left panel, upper lanes) and Western blot analysis (right panel, upper lanes) on tibialis anterior muscles of the indicated mice. Lane 1, α-SG knock-out dystrophic mice injected with an adenovirus expressing Magic-F1; lane 2, Magic-F1 transgenic mice; lane 3, α-SG knock-out/Magic-F1 transgenic mice; lane 4, α-SG knock-out dystrophic mice injected with an empty adenovirus. GAPDH was used as internal control (bottom panel). (B) H&E staining of tibialis anterior (left upper panel) and quadriceps (left lower panel) muscles from 3 month-old Magic-F1 mice, tibialis anterior of double transgenic α-SG knoc-out/Magic-F1 transgenic mice (upper middle panel), α-SG knock-out mice (right upper panel), α-SG knock-out mice injected with Ad-Magic-F1 (middle lower panel) or with Ad-Mock (right lower panel). Mice were sacrificed 14 days after the exhaustion treadmill tests. The bar is a 50 µm marker. (C) Exhaustion treadmill tests carried on wild-type mice (blue label), α-SG knock-out/Magic-F1 transgenic mice (red label), α-SG knock-out mice injected with Ad-Magic-F1 (red-black label) or Ad-Mock (black label). Note that mice expressing Magic-F1 showed increased running performance compared to Ad-Mock-injected mice. Muscle strength of treated and control animals was measured as time (left histogram) or distance to exhaustion (right histogram; n = 4, p<0.05).DiscussionProtein engineering allows creating recombinant factors displaying selective biological functions. This is particularly useful for pleiotropic factors eliciting several different biological responses like HGF. Magic-F1, an engineered protein derived from HGF, maintains the ability to protect cells against apoptosis and to promote myoblast differentiation, but is devoid of any mitogenic activity typical of its parental factor. This results in remarkable enhancement of skeletal muscle regeneration without induction of cell proliferation, a crucial feature for its potential therapeutic application. Notably, muscle hypertrophy was induced in normal and regenerating muscle both when Magic-F1 was present as a transgene and when it was delivered to post-natal muscles, as it would occur in a cell or gene therapy context.The potential relevance of inducing muscle hypertrophy to the treatment of muscle disorders in humans has been suggested by studies involving mdx mice, which carry a mutation in the dystrophin gene and therefore serve as a genetic model of Duchenne's muscular dystrophy [39]. For example, mdx mice lacking myostatin were found not only to be stronger and more muscular than their mdx counterparts with normal myostatin, but also to have reduced fibrosis and fat deposition, suggesting sustained muscle regeneration [40]. Furthermore, injection of neutralizing monoclonal antibodies directed against myostatin into either wild-type or mdx mice increases muscle mass and specific force, suggesting that myostatin plays an important role in regulating muscle growth in adult animals [41]. Magic-F1 is a molecule with a potential clinical application as it can induce muscle hypertrophy by both down-regulating myostatin and directly activating MyoD, Myf5 and several anti-apoptotic pathways. Interestingly, no side effects have been observed in skeletal muscles following electro-enhanced Magic-F1 DNA transfer or in transgenic mice expressing the Magic-F1 under the control of a muscle-specific promoter.Our data showing the inability of Magic-F1 to induce the ERK pathway together with an inhibitory interference with HGF-induced ERK activation are particular relevant to a potential therapeutic use of this engineered factor. In fact, several tissues other than myocytes and satellite cells express the Met receptor, including epithelial cells of kidney, liver, lung, skin, breast and the whole gastrointestinal tract, as well as neurons, endothelial cells and hematopoietic precursors [1], [5]. Furthermore, Met overexpression is a very frequent event in human cancer [42]. This raises the concern that stimulating the proliferation of Met-expressing cells may lead to tumor formation or progression [43]. In this regard, the lack of any mitogenic activity makes Magic-F1 a potentially safe cytokine for cell therapy.Because of its potent and selective effect on myoblast survival and differentiation, Magic-F1 promoted muscular hypertrophy in all mouse models analyzed. This biological activity, revealed by in vitro experiments, was extensively confirmed by the analysis of muscles treated by electro-enhanced DNA transfer or derived from transgenic mice expressing Magic-F1 under the control of a muscle-specific promoter. Interestingly, a statistically significant increase of myofiber cross-sectional areas was observed in the tibialis anterior muscles but not in slow-twitch soleus muscles. This can be attributed to the specificity of the promoter, active in fast twitch fibers, and to the fact that the amount of circulating Magic-F1 (escaping from fast-twitch muscles) is not enough to induce muscle hypertrophy. Moreover, following cardiotoxin treatment, regenerating centrally-nucleated fibers in the MLC1F/Magic-F1 transgenic mice appeared to have a greater cross-sectional area compared to wild-type animals. This can be explained by the enhanced differentiation potential of satellite cells, which indeed displayed an earlier differentiation program in vitro compared to cells isolated from wild-type mice. We previously reported the presence of myogenic precursors, named mesoangioblasts, in the skeletal muscles of mice [44], dogs [45] and humans [46]. These cells could also be positively affected by Magic-F1 and we cannot exclude their participation in the regeneration of skeletal muscle tissues. On the other hand, the rapid apoptotic response in cardiotoxin-treated muscles is strongly reduced in MLC1F/Magic-F1 transgenic mice. This results in a more evident muscular hypertrophy of transgenic muscles.Several authors have reported that HGF inhibits muscle differentiation both in vitro and in vivo\n[19], [20]. Recently, it has been reported that HGF gene therapy improves LV remodeling and dysfunction post-infarction through promotion of cardiomyocyte hypertrophy, and that HGF plays a role in the induction of stem cell commitment to the cardiomyocyte lineage [47]–[49]. Magic-F1 exhibits biological effects in the renewal of skeletal muscles tissues similar though not identical to those observed for HGF in cardiac tissue regeneration. Further studies are necessary to elucidate the different potential effects of HGF in this context and –in this sense– supplementary studies on Magic-F1 signal transduction could provide useful information.Successful adenovirus-mediated gene delivery under immunosuppressive conditions in adult muscles was previously demonstrated [50], [51]. In the present study, we transduced muscle fibers of juvenile α-SG knock-out mice with adenoviral vectors carrying Magic-F1 cDNA. All injected mice showed a physiological benefit and performed much better compared to mock-treated dystrophic animals in treadmill tests. As discussed, the less efficient rescue of the dystrophic phenotype by adenovirus-mediated Magic-F1 delivery compared to the crossing with Magic-F1 transgenic mice is conceivably due to incomplete muscle transduction. Importantly, in those mice in which all dystrophic fibers were transduced, the treadmill test performance was similar to that covered by control, non-dystrophic animals (not shown).In conclusion, Magic-F1 is a soluble, engineered factor that displays marked anti-apoptotic and pro-differentiative clues on muscle precursors. Its ability to promote and enhance muscle regeneration makes it a potential candidate molecule for regenerative medicine, particularly for muscular dystrophy syndromes and other muscle degenerative disorders. Given the small size of its cDNA (approximately 1 kb), Magic-F1 may be used alone in a gene therapy setting or inserted as a second adjuvant gene in a vector already encoding a therapeutic gene, for example encoding a deacetylase inhibitor [52]. The lack of mitogenic activity allows a safe use of Magic-F1 as a therapeutic cytokine, promoting muscle regeneration without the potential risk of stimulating uncontrolled proliferation.Materials and MethodsMagic-F1 Factor engineering and purificationMagic-F1 is an engineered factor containing two HGF NK2 domains joint by a linker. The exact amino acidic sequence of Magic-F1 corresponds to: residues 1–285 of human HGF (Gene Bank # M73239); a linker with the sequence (GGGGS)3; residues 30–285 of human HGF; a poly-histidine tag with the sequence DDDKHHHHHH. Factors were produced in a CHO cell line (ATTC, Rockville, Maryland). Purification was performed by dual-step affinity chromatography using a heparin-Sepharose column and a Ni2+-chelate column (Amersham Pharmacia, Uppsala, Sweden). Activated human recombinant HGF was purchased from R&D Systems (Minneapolis, Minnesota) while Metron Factor-1 (Metr. in Figure S1) recombinant protein [53] was produced at Dompé Pharmaceutical Company S.p.A. (L'Aquila, Italy).ELISA binding assayHGF and FcMET (a chimera consisting of the extracellular domain of MET fused to the Fc region of a human IgG1) were purchased from R&D Systems. Binding of Magic-F1 and HGF to Fc-Met was measured by ELISA using the receptor in solid phase and the ligands in liquid phase. A fixed concentration (100 ng/well) of Fc-Met was adsorbed to 96-well ELISA plates and incubated with increasing concentrations of ligands. Binding was revealed using biotinylated anti-HGF antibodies (R&D). Binding data were analyzed and fit using Prism software (Graph Pad Software, San Diego, California).ImmunoreagentsThe antibodies used in this study were obtained as follows: anti-human HGF for both Western blotting and immunoprecipitation, Santa Cruz Biotechnology (Santa Cruz, California); anti-human Met for Western blotting, Santa Cruz; anti-human Met for immunoprecipitation, as described [5]; anti-mouse Met, Santa Cruz; anti-AKT and anti-phospho-AKT, New England Biolabs (Beverly, Massachusetts); anti-MAPK (p42–44/ERK) and anti-phospho-MAPK, Promega (Madison, Winsconsin); anti-laminin polyclonal rabbit antibodies, Sigma (St. Louis, Missouri); anti-desmin rabbit polyclonal antibody, Sigma; MF20 and embryonic myosin monoclonal antibody, Developmental Studies Hybridoma Bank (Iowa City, Iowa).Receptor activation and signal transductionFor receptor activation analysis, quiescent cells plated on collagen-coated 100 mm plates (Becton Dickinson, Franklin Lakes, New Jersey) were stimulated with 5 nM HGF or Magic-F1 for 30 min at 37°C and then lysed as described [54]. Lysates were immunoprecipitated with anti-Met and analyzed by Western blotting using anti-phosphotyrosine antibodies. For signal-transduction analysis, cells were stimulated as above for different times and then lysed. For MAPK and AKT activation, lysates were directly analyzed by Western blotting using antibodies specific for the activated forms of the signaling molecules. Quantification of enhanced chemiluminescence signal was performed using a STORM apparatus and Image Quant software (Molecular Dynamics, Amersham Biosciences, Sunnyvale, California).Cell cultures and bioassaysMouse myogenic cell line C2C12 was maintained in DMEM supplemented with 2 mM glutamine, 100 IU/ml penicillin, 100 µg/ml streptomycin and 10% FBS. C2C12 cells were induced to differentiate into myotubes by replacing 10% FBS with 2% horse serum (HS). Differentiation was completed in 7–8 days. All cultures were performed at 37°C in a humidified incubator with 5% CO2 and 95% air. Satellite cells were prepared as previously described [46]. Briefly, muscle fragments were digested with 2% collagenase II (Invitrogen, Carlsbad, California) for 60 min at 37°C. Digested cells were discarded and fragments were incubated again with 0.05% trypsin (Invitrogen) for 15 min at 37°C with gentle agitation. After the incubation, isolated cells were collected and fragments were incubated again until the whole tissue was digested (usually three times). Isolated cells were pooled, centrifuged and resuspended in DMEM supplements with 20% pre-screened FCS, 1% gentamycin, and plated onto collagen coated dishes at a density of 104 cells×cm2. Contamination by non-myogenic cell was reduced by pre-plating the cell suspension onto plastic dishes where fibroblasts tend to adhere more rapidly. Differentiation was induced shifting the medium to DMEM supplemented with 2% horse serum. Cell morphology was examined daily with a phase-contrast microscope connected to an image analyzer. Cells were trypsinized daily and counted on a hemocytometer. Cell viability was determined by trypan blue dye exclusion assay. Cell cytotoxicity was performed using an XTT-based in vitro toxicology assay kit (Sigma) according to manufacturer's protocol. Incubation medium was collected after 3 hours and read spectrophotometrically at a wavelength of 450 nm. Background signals, obtained from plates without cells, were subtracted from sample readings. Apoptosis was quantified using an ApopTag Fluorescein In situ Apoptosis detection kit (Chemicon, Temecula, California) according to the manufacturer's protocol. Cell differentiation was carried out for 8 days. Cells were grown on 6 cm Petri dishes until sub-confluent, washed with PBS, fixed with 4% paraformaldehyde at room temperature for 10 minutes and then permeabilized with 0.1% Triton X-100 in PBS for 5 minutes. After incubation with PBS containing 10% normal serum, samples were incubated overnight at 4°C with anti-GFP at 1∶200 dilution, anti myosin heavy chain (MF20) antibody at 1∶2 dilution. After incubation, cells were washed three times in PBS and incubated with the appropriate FITC- or TRITC-conjugated secondary antibodies for 1 hour at room temperature. After washing in PBS, cells were analyzed under a fluorescent microscope and photographed. As a control for the immunofluorescence method, we omitted the primary antibody and no staining was detected under these conditions. Cell nuclei were counterstained with DAPI.Plasmids and DNA preparationMagic-F1 was cloned into pIRESneo (Clontech, Italy) for C2C12 transfection experiments whereas it was cloned into pcDNA3 (Invitrogen) containing the cytomegalovirus (CMV) promoter for electrotransfer experiments (pCMV-Magic-F1); a pCMV-bgal plasmid coding for beta-galactosidase and a pCMV-hHGF plasmid coding for human hepatocyte growth factor were also used. Plasmids were prepared by using standard procedures. All plasmid preparations was obtained using a GenElute™ HP Endotoxin-Free Plasmid Maxiprep Kit (Sigma) and contained a high percentage of supercoiled DNA (70–80%). No RNA was detectable by gel electrophoresis.DNA electro-transfer and animal handlingMouse experiments were performed in the San Raffaele Hospital SPF Animal Care Facilities according to international ethical guidelines (EEC Council Directive 86/609; NIH Guide for the Care and Use of Laboratory Animals, 1985). Authorization for animal experimentation was obtained from the Italian Ministry of Health. Gene transfer into skeletal muscle mediated by electric pulse was performed as previously reported [33]. Briefly, 20 µg of DNA in 10 µl of PBS was injected into the tibialis anterior or in the quadriceps muscle of anesthetized, 10 day-old C57Bl/6 mice (Iffa Credo, St. Germain sur l'Arbresle, France) with a Hamilton syringe. There were 10 muscles included in each experimental group. Five minutes after DNA injection trans-cutaneous electric pulses were applied by two stainless steel plate electrodes placed 3.8–4.3 mm apart, at each side of the leg. Electrical contact with the leg skin was ensured by shaving each leg and applying a conductive gel. Square-wave electric pulses (eight pulses; 200 V/cm; 20 ms per pulse; 1 Hz) were generated by a digital Stimulator (Panlab 3100, Biological Instruments, Varese, Italy).Muscular regeneration analysisAcute skeletal muscle damage was induced in male and female MLC1F/Magic-F1transgenic mice and control mice (7 animals/group) by i.m. injection of 10 nM cardiotoxin (Gentaur, Brussels, Belgium) in physiologic solution (0.9% w/v NaCl). Control mice were injected with physiologic solution alone. At 3, 7, and 14 days after drug injection, mice were sacrified and subjected to histological evaluation and morphometric analysis of tibialis anterior. After excision, muscles were sectioned (4–6 µm) and processed for immunofluorescence analysis using the primary antibodies listed above. All sections were washed three times in PBS and incubated with 10% donkey serum for 30 min at RT before the addition of the appropriate Alexa 488-, Alexa 594- or Alexa 647-conjugated donkey secondary antibodies. Alternatively, some sections were stained with hematoxylin and eosin and examined by an independent histopathologist not informed of sample identity to determine muscle fiber sizes using Scion Image software (Scion, Frederick, Maryland).Biochemical and molecular analysisWestern blot analysis of cells or tissues was performed as described [41], 55. Total RNA from control or treated cells was extracted using Trizol reagent (Invitrogen) and analysed by PCR after reverse transcription with random hexamers. RT-PCR analysis has been performed using the following primers:BaxFw 5′-TGTTTGCTGATGGCAACTTC-3′\nRv 5′-GATCAGCTCGGGCACTTTAG-3′\nBcl-2Fw 5′-GGGATGCCTTTGTGGAACTA-3′\nRv 5′-CTCACTTGTGGCCCAGGTAT-3′\nP53Fw 5′-GGATGCCCGTGCTGCCGAGGAG-3′\nRv 5′-AGTGAAGGGAC TAGCATTGTC-3′\nMagic-F1Fw 5′-TTCAGAAGTTGAATGCATGACCTG-3′\nRv 5′-TCTTCTTTTCCTTTGTCCCTCTAG-3′\nGAPDHFw 5′-TTCACCACCATGGAGAAGGC-3′\nRv 5′-GGCATGGACTGTGGTCATGA-3′\nMyoDFw 5′-TGCACTTCCACCAACCCCAACCAGC-3′\nRv 5′-CCTGGACTCGCGCACCGCCTCACT-3′\nMetFw 5′-AGAAATTCATCAGGCTGTGAAGCGCG-3′\nRv 5′-TTCCTCCGATCGCACACATTTGTCG-3′\nPAX3Fw 5′-AGGAGGCGGATCTAGAAAGGAAG-3′\nRv 5′-TGTGGAATAGACGTGGGCTGGTA-3′\nMyf5Fw 5′-GAGCTGCTGAGGGAACAGGTGGAGA-3′\nRv 5′-GTTCTTTCGGGACCAGACAGGGCTG-3′\nIGF1Fw 5′-CTGTGCCCCACTGAAGCCTA-3′\nRv 5′-GGACTTCTGAGTCTTGGGCATG-3′\nMyostatinFw 5′-AGTGACGGCTCTTTGGAAGATG-3′\nRv 5′-AGTCAGACTCGGTAGGCATGGT-3′\nFollistatinFw 5′-CTGTACAAGACCGAACTGAGC-3′\nRv 5′-TCCACAGTCCACGTTCTCACA-3′\nGeneration of Magic-F1 and α-SG knock-out/Magic-F1 transgenic miceWe constructed the transgene by inserting the Magic-F1 construct into the pMex plasmid containing the 1,500-bp fragment of the MLC promoter, an 840-bp fragment of SV40 poly(A), and a 900-bp fragment from the 3′ end of the MLC1f/3f gene, which acts as an enhancer [34]; provided by Dr. Antonio Musarò, University of Rome, Italy. We microinjected the transgene into the male pronucleus of fertilized eggs from FVB mice (Jackson Laboratories, Bar Harbor, Maine) that were implanted into pseudopregnant foster mothers. We identified positive transgenic mice by PCR. For PCR detection, sense and antisense primers specific respectively for the MLC1F promoter and the linker region of Magic-F1 were used. Transgenic founders were mated with wild-type FVB mice to generate F1 offspring. After obtaining MLC1F/Magic-F1 mice we mated them with α-SG knock-outs [36] to generate α-SG knock-out/Magic-F1 transgenic mice. The animals were housed in a temperature controlled (22°C) room with a 12∶12 hours light-dark cycle. All studies have been performed using Tg:MLC1F/Magic-F1 hemizygous mice, following the protocols approved by the Animal Care and Use Committee of the San Raffaele Institute (IACUC 264) and communicated to the Ministry of the Health and local authorities according to Italian law.Adenovirus preparation and administrationpAd/CMV-Magic-F1/V5-DEST was engineered using the ViraPower Adenoviral Expression System from Invitrogen. The Adenoviral vector was linearized with PacI restriction enzyme and transfected into 293A cells. Cells were grown in Iscove Medium supplemented with 10% heat-inactivated FBS, 2 mM l-glutamine, 50 units/ml penicillin, 50 µg/ml streptomycin (Sigma). After complete detachment of cells, the supernatant was used to superinfect 293A cells. The purification of Adenoviral particles was performed with Vivapure AdenoPACK 100TM (Sartorius, Goettingen, Germany) starting from 200 ml of cell culture. Juvenile α-SG knock-out mice (8 weeks old) were anesthetized with an intraperitoneal injection of avertin (0.2 ml/10 g bodyweight of a 1.2% solution), hair was shaved from the skin and pAd-Magic-F1 suspension (2.5×109 pfu diluited in 30 µl of PBS containing 100 ng of VEGF) were injected i.m. with a 30-gauge needle in the center of gastrocnemius, quadriceps and tibialis anterior. To prevent an immune-mediated clearance of adeno-infected fibers, all mice were immunosuppressed with FK506 (5 mg/day/Kg, subcutaneously). The immunosuppressive treatment was started on the day before the Adenoviral injection and continued until mice were sacrificed.Treadmill analysesTreadmill analyses were carried out using a six-lane motorized treadmill (Exer 3/6 Treadmill; Columbus Instruments, Columbus, Ohio) supplied with shocker plates. The first trial was performed at low intensity and for short duration to accustom the mice to the exercise (5 m/min for 5 minutes, after which the speed was increased 1 m/min every 2 minutes until it reached 9 m/min). After the first trial, the treadmill was run at an inclination of 0° at 5 m/min for 5 minutes, after which the speed was increased 1 m/min every 1 minute. The test was stopped when the mouse remained on the shocker plate for more than 20 s without attempting to re-engage the treadmill, and the time to exhaustion was determined.Supporting InformationFigure S1Production of Magic-F1 in eucaryotic systems. (A) After transient transfection with pIRES-neo-Magic-F1 plasmid, cells were washed and incubated with fresh serum-free medium; aliquots of medium after 3, 6 and 18 hours (lane 1, 2 and 3, respectively) were concentrated 100 times and subjected to Western blot analysis using anti-HGF antibodies. Cell lysate (lane 5) and mock conditioned medium (lane 4) were also analyzed. (B) Growth curves of CHO cells expressing Magic-F1 (#PM21) or transfected with an empty vector (CHO). (C) Western blot analysis of conditioned media from different CHO clones expressing Magic-F1 at different levels. (D) Quantification of Magic-F1 protein in the conditioned medium of clone # 21; a related recombinant protein, Metron Factor-1, was used as standard (Metr). (E) Immunoprecipitation of Magic-F1 protein from the conditioned medium of clone # 5, 7, 21 and 25. In some case (# 7f), following storage at −20°C, Magic-F1 could not be immunoprecipitated any more. In each lane we loaded 10 µl of medium concentrated 100 times.(0.73 MB TIF)Click here for additional data file.Figure S2Quantitative RT-PCR analysis of Magic-F1-transfected C2C12 cells. The expression of 36 genes involved in myogenic differentiation was evaluated one day after transfection of C2C12 using quantitative real time PCR analysis. The results indicate that 30 out of the 36 analyzed genes were upregulated in C2C12 expressing Magic-F1 compared to controls, confirming an enhanced rate of differentiation induced by Magic-F1 expression.(0.54 MB TIF)Click here for additional data file.Figure S3Magic-F1 enhances muscle regeneration. (A) Morphometric analysis on a tibialis anterior section shows a marked increase of the fiber area in Magic-F1 transgenic mice (red bar) relative to wild-type mice (blue bar). Note that the number of fibers with a larger cross sectional area is higher in Magic-F1 transgenic mice when compared to wild-type mice. This effect is evident in both regenerated and regenerating fibers (centrally nucleated) as showed in (B), where statistical analysis is reported.(1.11 MB TIF)Click here for additional data file.\n\nREFERENCES:\n1. TrusolinoLComoglioPM\n2002\nScatter-factor and semaphorin receptors: cell signalling for invasive growth. Nature Rev.\nCancer\n4\n289\n300\n2. TamagnoneLComoglioPM\n1997\nControl of invasive growth by hepatocyte growth factor (HGF) and related scatter factors.\nCytokine Growth Factor Rev\n8\n129\n142\n9244408\n3. ZarnegarRMichalopoulosG\n1995\nThe many faces of hepatocyte growth factor: from hepatopoiesis to hematopoiesis.\nJ Cell Biol\n129\n1177\n1180\n7775566\n4. RubinJSBottaroDPAaronsonSA\n1993\nHepatocyte growth factor/scatter factor and its receptor, the c-met proto-oncogene product.\nBiochim Biophys Acta\n1155\n357\n371\n8268192\n5. BirchmeierCBirchmeierWGherardiEVande WoudeGF\n2003\nMet, metastasis, motility, and more.\nNature Rev Mol Cell Biol\n4\n915\n925\n14685170\n6. WoolfASKolatsi-JoannouMHardmanPAndermarcherEMoorbyC\n1995\nRoles of hepatocyte growth factor/scatter factor and the met receptor in the early development of the metanephros.\nJ Cell Biol\n128\n171\n184\n7822413\n7. TakayamaHLa RochelleWJAnverMBockmanDEMerlinoG\n1996\nScatter factor/hepatocyte growth factor as a regulator of skeletal muscle and neural crest development.\nProc Natl Acad Sci USA\n93\n5866\n5871\n8650184\n8. AndermarcherESuraniMGherardiE\n1996\nCo-expression of the HGF/SF and c-met genes during early mouse embryogenesis precedes reciprocal expression in adjacent tissues during organogenesis.\nDev Genet\n18\n254\n266\n8631159\n9. MainaFHiltonMCPonzettoCDaviesAMKleinR\n1997\nMet receptor signaling is required for sensory nerve development and HGF promotes axonal growth and survival of sensory neurons.\nGenes Dev\n11\n3341\n3350\n9407027\n10. KosLAronzonATakayamaHMainaFPonzettoC\n1999\nHepatocyte growth factor/scatter factor-MET signaling in neural crest-derived melanocyte development.\nPigment Cell Res\n12\n13\n21\n10193678\n11. SchmidtCBladtFGoedeckeSBrinkmannVZschiescheW\n1995\nScatter factor/hepatocyte growth factor is essential for liver development.\nNature\n373\n699\n702\n7854452\n12. UeharaYMinowaOMoriCShiotaKKunoJ\n1995\nPlacental defect and embryonal lethality in mice lacking hepathocyte growth factor/scatter factor.\nNature\n373\n702\n705\n7854453\n13. BladtFRiethmacherDIsenmannSAguzziABirchmeierC\n1995\nEssential role for the c-met receptor in the migration of myogenic precursor cells into the limb bud.\nNature\n376\n768\n771\n7651534\n14. MainaFCasagrandaFAuderoESimeoneAComoglioPM\n1996\nUncoupling of Grb2 from the Met receptor in vivo reveals complex roles in muscle development.\nCell\n87\n531\n42\n8898205\n15. SheehanSMTatsumiRTemm-GroveCJAllenRE\n2000\nHGF is an autocrine growth factor for skeletal muscle satellite cells in vitro.\nMuscle Nerve\n23\n239\n245\n10639617\n16. MillerKJThaloorDMattesonSPavlathGK\n2000\nHepatocyte growth factor affects satellite cell activation and differentiation in regenerating skeletal muscle.\nAm J Physiol Cell Physiol\n278\nC174\nC181\n10644525\n17. TatsumiRAndersonJENevoretCJHalevyOAllenRE\n1998\nHGF/SF is present in normal adult skeletal muscle and is capable of activating satellite cells.\nDev Biol\n194\n114\n128\n9473336\n18. AnastasiSGiordanoSSthandierOGambarottaGMaioneR\n1997\nA natural hepatocyte growth factor/scatter factor autocrine loop in myoblast cells and the effect of the constitutive Met kinase activation on myogenic differentiation.\nJ Cell Biol\n137\n1057\n1068\n9166406\n19. Gal-LeviRLeshemYAokiSNakamuraTHalevyO\n1998\nHepatocyte growth factor plays a dual role in regulating skeletal muscle satellite cell proliferation and differentiation.\nBiochim Biophys Acta\n1402\n39\n51\n9551084\n20. LeshemYSpicerDBGal-LeviRHalevyO\n2000\nHepatocyte growth factor (HGF) inhibits skeletal muscle cell differentiation: a role for the bHLH protein twist and the cdk inhibitor p27.\nJ Cell Physiol\n184\n101\n109\n10825239\n21. NakamuraTNishizawaTHagiyaMSekiTShimonishiM\n1989\nMolecular cloning and expression of human hepatocyte growth factor.\nNature\n342\n440\n443\n2531289\n22. HartmannGNaldiniLWeidnerKMSachsMVignaE\n1992\nA functional domain in the heavy chain of scatter factor/hepatocyte growth factor binds the c-Met receptor and induces cell dissociation but not mitogenesis.\nProc Natl Acad Sci USA\n89\n11574\n11578\n1280830\n23. LokkerNAMarkMRLuisEABennettGLRobbinsKA\n1992\nStructure-function analysis of hepatocyte growth factor: identification of variants that lack mitogenic activity yet retain high affinity receptor binding.\nEMBO J\n11\n2503\n2510\n1321034\n24. MatsumotoKKataokaHDateKNakamuraT\n1998\nCooperative interaction between alpha- and beta-chains of hepatocyte growth factor on c-Met receptor confers ligand-induced receptor tyrosine phosphorylation and multiple biological responses.\nJ Biol Chem\n273\n22913\n22920\n9722511\n25. CioceVCsakyKGChanAMBottaroDPTaylorWG\n1996\nHepatocyte growth factor (HGF)/NK1 is a naturally occurring HGF/scatter factor variant with partial agonist/antagonist activity.\nJ Biol Chem\n271\n13110\n13115\n8662798\n26. SchwallRHChangLYGodowskiPJKahnDWHillanKJ\n1996\nHeparin induces dimerization and confers proliferative activity onto the hepatocyte growth factor antagonists NK1 and NK2.\nJ Cell Biol\n133\n709\n718\n8636243\n27. MarkMRLokkerNAZioncheckTFLuisEAGodowskiPJ\n1992\nExpression and characterization of hepatocyte growth factor receptor-IgG fusion proteins. Effects of mutations in the potential proteolytic cleavage site on processing and ligand binding.\nJ Biol Chem\n267\n26166\n26171\n1334493\n28. BlauHMChiuCPWebsterC\n1983\nCytoplasmic activation of human nuclear genes in stable heterocaryons.\nCell\n32\n1171\n1180\n6839359\n29. XiaoGHJeffersMBellacosaAMitsuuchiYVande WoudeGF\n2001\nAnti-apoptotic signaling by hepatocyte growth factor/Met via the phosphatidylinositol 3-kinase/Akt and mitogen-activated protein kinase pathways.\nProc Natl Acad Sci USA\n98\n247\n252\n11134526\n30. TomczakKKMarinescuVDRamoniMFSanoudouDMontanaroF\n2004\nExpression profiling and identification of novel genes involved in myogenic differentiation.\nFASEB J\n18\n403\n405\n14688207\n31. McPherronACLeeSJ\n1997\nDouble muscling in cattle due to mutations in the myostatin gene.\nProc Natl Acad Sci USA\n94\n12457\n12461\n9356471\n32. McPherronACLawlerAMLeeSJ\n1997\nRegulation of skeletal muscle mass in mice by a new TGF-beta superfamily member.\nNature\n387\n83\n90\n9139826\n33. MirLMBureauMFGehlJRangaraRRouyD\n1999\nHigh-efficiency gene transfer into skeletal muscle mediated by electric pulses.\nProc Natl Acad Sci USA\n96\n4262\n4267\n10200250\n34. MusaròAMcCullaghKPaulAHoughtonLDobrowolnyG\n2001\nLocalized Igf-1 transgene expression sustains hypertrophy and regeneration in senescent skeletal muscle.\nNat Genet\n27\n195\n200\n11175789\n35. CossuGSampaolesiM\n2004\nNew therapies for muscular dystrophy: cautious optimism.\nTrends Mol Med\n10\n516\n520\n15464452\n36. DuclosFStraubVMooreSAVenzkeDPHrstkaRF\n1998\nProgressive muscular dystrophy in alpha-sarcoglycan-deficient mice.\nJ Cell Biol\n142\n1461\n1471\n9744877\n37. FeeroWGRosenblattJDHuardJWatkinsSCEpperlyM\n1997\nViral gene delivery to skeletal muscle: insights on maturation-dependent loss of fiber infectivity for adenovirus and herpes simplex type 1 viral vectors.\nHum Gene Ther\n8\n371\n380\n9054512\n38. VilquinJTGueretteBKinoshitaIRoyBGouletM\n1995\nFK506 immunosuppression to control the immune reactions triggered by first-generation adenovirus-mediated gene transfer.\nHum Gene Ther\n6\n1391\n1401\n8573612\n39. BulfieldGSillerWGWightPAMooreKJ\n1984\nX chromosome-linked muscular dystrophy (mdx) in the mouse.\nProc Natl Acad Sci USA\n81\n1189\n1192\n6583703\n40. WagnerKRMcPherronACWinikNLeeSJ\n2002\nLoss of myostatin attenuates severity of muscular dystrophy in mdx mice.\nAnn Neurol\n52\n832\n836\n12447939\n41. BogdanovichSKragTOBartonERMorrisLDWhittemoreLA\n2002\nFunctional improvement of dystrophic muscle by myostatin blockade.\nNature\n420\n418\n421\n12459784\n42. CorsoSComoglioPMGiordanoS\n2005\nCancer therapy: can the challenge be MET?\nTrends Mol Med\n11\n284\n292\n15949770\n43. TakaharaTXueFMazzoneMYataYNonomeK\n2008\nMetron Factor-1 prevents liver injury without promoting tumor angiogenesis and metastasis.\nHepatology\n47\n2010\n2025\n18506889\n44. SampaolesiMTorrenteYInnocenziATonlorenziRD'AntonaG\n2003\nCell therapy of alpha-sarcoglycan null dystrophic mice through intra-arterial delivery of mesoangioblasts.\nScience\n301\n487\n492\n12855815\n45. SampaolesiMBlotSD'AntonaGGrangerNTonlorenziR\n2006\nMesoangioblast stem cells ameliorate muscle function in dystrophic dogs.\nNature\n444\n574\n579\n17108972\n46. DellavalleASMTonlorenziRTagliaficoESacchettiBPeraniL\n2007\nPericytes of human post-natal skeletal muscle are myogenic precursors distinct from satellite cells.\nNat Cell Biol\n9\n255\n267\n17293855\n47. FiaccaventoRCarotenutoFMinieriMFantiniCForteG\n2005\nStem cell activation sustains hereditary hypertrophy in hamster cardiomyopathy.\nJ Pathol\n205\n397\n407\n15682436\n48. ForteGMinieriMCossaPAntenucciDSalaM\n2006\nHepatocyte growth factor effects on mesenchymal stem cells: proliferation, migration, and differentiation.\nStem Cells\n24\n23\n33\n16100005\n49. LiYTakemuraGKosaiKYugeKNaganoS\n2003\nPostinfarction treatment with an adenoviral vector expressing hepatocyte growth factor relieves chronic left ventricular remodeling and dysfunction in mice.\nCirculation\n107\n2499\n2506\n12695295\n50. LochmullerHPetrofBJPariGLarochelleNDodeletV\n1996\nTransient immunosuppression by FK506 permits a sustained high-level dystrophin expression after adenovirus-mediated dystrophin minigene transfer to skeletal muscles of adult dystrophic (mdx) mice.\nGene Ther\n3\n706\n716\n8854096\n51. YangLLochmullerHLuoJMassieBNalbantogluJ\n1998\nAdenovirus-mediated dystrophin minigene transfer improves muscle strength in adult dystrophic (MDX) mice.\nGene Ther\n5\n369\n379\n9614557\n52. MinettiGCColussiCAdamiRSerraCMozzettaC\n2006\nFunctional and morphological recovery of dystrophic muscles in mice treated with deacetylase inhibitors.\nNat Med\n12\n1147\n1150\n16980968\n53. MichieliPCavassaSBasilicoCDe LucaAMazzoneM\n2002\nAn HGF-MSP chimera disassociates the trophic properties of scatter factors from their pro-invasive activity.\nNat Biotechnol\n20\n488\n495\n11981563\n54. NaldiniLTamagnoneLVignaESachsMHartmannG\n1992\nExtracellular proteolytic cleavage by urokinase is required for activation of hepatocyte growth factor/scatter factor.\nEmbo J\n11\n4825\n4833\n1334458"
+}

batch_0/PMC2529261.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529261",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529261\nAUTHORS: Charles A Willis-Owen, Jas S Daurka, Alvin Chen, Angus Lewis\n\nABSTRACT:\nWe describe a case of bilateral femoral neck fractures secondary to transient osteoporosis of pregnancy, which were diagnosed after delivery due to the desire to avoid ionising radiation. These fractures were presumed to be secondary to transient osteoporosis of pregnancy and were treated successfully with internal fixation despite delayed presentation. We discuss the role of MRI in the evaluation of hip pain in pregnancy.\n\nBODY:\nIntroductionTransient osteoporosis of pregnancy (TOP) is a rare, idiopathic self-limiting condition typically associated with the third trimester of pregnancy. It almost always affects a single hip although bilateral presentation and involvement of the knee have been reported [1-3].TOP usually presents with a sudden, quite severe onset of unilateral groin pain with no history of trauma. The patient may be unable to walk, or may have an antalgic gait. Pain is elicited by hip rotation, although a full range of motion is common. Radiographs are avoided in pregnancy where possible, and are a poor investigation for demonstrating early osteopaenia. Magnetic Resonance Imaging (MRI) reveals low signal intensity of bone marrow on T1 weighted images, and high signal on T2 weighted images suggestive of bone marrow oedema[4]. The natural history is of resolution of symptoms over the course of 3 to 6 monthsHip fracture secondary to TOP is very rare with only 12 reported patients in the literature to date; in two cases the hip fractures were bilateral[2,3,5-8]. The majority of these fractures were caused by a traumatic event. Atraumatic hip fractures secondary to TOP are even more unusual and are easily overlooked and hence may present to the orthopaedic surgeon at a late stage, making management more challenging.We report of a case of bilateral TOP leading bilateral atraumatic femoral neck fractures that were diagnosed post partum. Despite the delay in presentation internal fixation was successfully carried out. We highlight the importance of adequate investigation of hip pain during pregnancy and discuss the role of MRI.Case presentationA 34 year old Persian woman, gravida 1, para 0, presented at 22 weeks of pregnancy with a two week history of left hip pain with no apparent precipitating event. Her past medical history included mild Multiple Sclerosis from which she was asymptomatic. She did not smoke or drink alcohol, had no history of corticosteroid, anticonvulsant or anticoagulant use and was not on any other medication. Clinical examination was unremarkable and no investigations were deemed appropriate. The working diagnosis at this stage was non-specific hip pain related to pregnancy and supportive measures were instituted.Over the following 12 weeks her hip pain worsened, and she started to experience pain in the contra lateral hip. Again there was no history of a traumatic event. Because of her pregnancy imaging of her hips was avoided. By 36 weeks of pregnancy she was unable to weight bear and became wheelchair bound. Pain in her hips and limitation of motion meant that a normal vaginal delivery was impossible; hence she underwent a caesarean delivery of a healthy baby at full term.She was brought to the attention of the orthopaedic team when plain radiographs (see figure 1) following delivery revealed a displaced intracapsular femoral neck fracture on the left and a valgus impacted right intracapsular femoral neck fracture on the right. The radiographs also revealed considerable osteopaenia. MRI (see figure 2) revealed these fractures, with reduced signal on T1 and increased signal on T2 in the femoral necks in keeping with TOP.Figure 1Antero-posterior radiograph of the pelvis post partum.Figure 2T1 weighted coronal MRI scan of the pelvis post partum.She underwent closed reduction and internal fixation of the left hip. The right hip was internally fixed in situ. Two hole 135 degree dynamic hip screws were used in order to provide sufficient stability to allow immediate mobilisation despite bilateral fractures. Difficulty was encountered in ensuring that the threads of the dynamic hip screw had crossed the fracture site in the left hip as the level of the fracture was high in the femoral neck, consequently the tip of the implant had to be implanted close to the subchondral plate (see figure 3).Figure 3Antero-posterior radiograph of the pelvis post fixation with dynamic hip screws.The post-operative course was uncomplicated and the hip pain significantly improved immediately. Full weight bearing on the right, and partial weight bearing on the left was initiated on the first postoperative day, and maintained for the first 12 weeks. Check radiographs at 3 months showed no loss of fixation and the fractures appeared to be uniting in an adequate position. At six months she was pain free with no evidence of avascular necrosis or implant failure.DiscussionMusculoskeletal complaints are very common in pregnancy. The position and weight of the gravid uterus alters the centre of gravity and loading patterns of the axial and appendicular skeleton, whilst hormonal changes lead to joint laxity, and fluid retention may cause neural compression[9]. The majority of musculoskeletal complaints are not serious, and are managed conservatively without a specific diagnosis.Pregnant women frequently complain of hip or pelvic pain. The differential diagnosis includes some serious problems that need to be excluded, namely transient osteoporosis, osteonecrosis and pubic symphysiolysis.Conventionally ionising radiation is avoided during pregnancy although Brodell et al. suggested that in the third trimester of pregnancy the benefits of adequate investigation of hip pain may outweigh the minimal risks[5]. There is no conclusive evidence that MRI has deleterious effects, however the safety of MRI has yet to be definitively proven[10]. It is in common use in the third trimester of pregnancy where clinically indicated[11] and is generally considered to be safe[12]. MRI has a high sensitivity for diagnosis of occult hip fracture[13] and can reliably distinguish between osteonecrosis transient osteoporosis[4], making it the investigation of choice for hip pain in the third trimester of pregnancy.Displaced intracapsular fractures have a high incidence of non-union and avascular necrosis[14]. It has however been shown that the risk of non-union is independent of bone quality[15] therefore in young patients with high value hips internal fixation should be the goal.ConclusionThis case report highlights the need for vigilance in the assessment of musculoskeletal complaints in pregnancy, and demonstrates that the more conservative approach of internal fixation is viable. We suggest MRI should be considered for women presenting with significant hip pain in the third trimester of pregnancy.ConsentWritten informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsAC assessed the patient at first presentation, JD and AL carried out the surgery, CWO was involved throughout admission and follow-up and was the major contributor in writing the manuscript. All authors contributed to, read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2529279.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529279",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529279\nAUTHORS: Swapna Mahurkar, Seema Bhaskar, D Nageshwar Reddy, Swami Prakash, G Venkat Rao, Shivaram Prasad Singh, Varghese Thomas, Giriraj Ratan Chandak\n\nABSTRACT:\nBackgroundTropical calcific pancreatitis (TCP) is a type of chronic pancreatitis unique to developing countries in tropical regions and one of its important features is invariable progression to diabetes, a condition called fibro-calculous pancreatic diabetes (FCPD), but the nature of diabetes in TCP is controversial. We analysed the recently reported type 2 diabetes (T2D) associated polymorphisms in the TCF7L2 gene using a case-control approach, under the hypothesis that TCF7L2 variants should show similar association if diabetes in FCPD is similar to T2D. We also investigated the interaction between the TCF7L2 variants and N34S SPINK1 and L26V CTSB mutations, since they are strong predictors of risk for TCP.MethodsTwo polymorphisms rs7903146 and rs12255372 in the TCF7L2 gene were analyzed by direct sequencing in 478 well-characterized TCP patients and 661 healthy controls of Dravidian and Indo-European ethnicities. Their association with TCP with diabetes (FCPD) and without diabetes was tested in both populations independently using chi-square test. Finally, a meta analysis was performed on all the cases and controls for assessing the overall significance irrespective of ethnicity. We dichotomized the whole cohort based on the presence or absence of N34S SPINK1 and L26V CTSB mutations and further subdivided them into TCP and FCPD patients and compared the distribution of TCF7L2 variants between them.ResultsThe allelic and genotypic frequencies for both TCF7L2 polymorphisms, did not differ significantly between TCP patients and controls belonging to either of the ethnic groups or taken together. No statistically significant association of the SNPs was observed with TCP or FCPD or between carriers and non-carriers of N34S SPINK1 and L26V CTSB mutations. The minor allele frequency for rs7903146 was different between TCP and FCPD patients carrying the N34S SPINK1 variant but did not reach statistical significance (OR = 1.59, 95% CI = 0.93–2.70, P = 0.09), while, TCF7L2variant showed a statistically significant association between TCP and FCPD patients carrying the 26V allele (OR = 1.69, 95% CI = 1.11–2.56, P = 0.013).ConclusionType 2 diabetes associated TCF7L2 variants are not associated with diabetes in TCP. Since, TCF7L2 is a major susceptibility gene for T2D, it may be hypothesized that the diabetes in TCP patients may not be similar to T2D. Our data also suggests that co-existence of TCF7L2 variants and the SPINK1 and CTSB mutations, that predict susceptibility to exocrine damage, may interact to determine the onset of diabetes in TCP patients.\n\nBODY:\nBackgroundPancreatitis is generally believed to be a disease where pancreas is injured by enzymes that are normally secreted by the acinar cells. Chronic Pancreatitis (CP) is a continuing or relapsing inflammatory process of pancreas leading to exocrine and/or endocrine insufficiency. Tropical calcific pancreatitis (TCP) is a type of CP unique to developing countries in tropical region [1]. An important feature of TCP is its consistent progression to diabetes, commonly known as fibro-calculous pancreatic diabetes (FCPD) [1,2]. FCPD is thought to be a type of diabetes secondary to TCP, resulting from destruction of beta-cell mass in the pancreas [3]. However, several studies have shown partial preservation of beta-cell mass [3] and evidence of insulin resistance to a similar degree as seen in type 2 diabetes (T2D) patients [4], suggesting that the diabetes in FCPD could be type T2D, while others have not found insulin resistance to be a major factor in FCPD [5]. It was believed earlier that diabetes specific complications do not occur in FCPD [6], but prevalence of retinopathy, [7] nephropathy and neuropathy [8] in FCPD patients has been reported to be no different from matched group of patients with T2D. Similarly, although the diabetes is severe and insulin requiring in both FCPD and type 1 diabetes (T1D), FCPD patients rarely develop ketoacidosis, in contrast to the T1D patients, who are ketosis prone [9,10].Mutations in the serine protease inhibitor, Kazal type 1 (SPINK1) [11-15], cystic fibrosis transmembrane regulator (CFTR) [16], cathepsin B (CTSB) [17] and recently, chymotrypsin C (CTRC) [18] genes have been identified to be associated with TCP but they mostly associate with pancreatic exocrine dysfunction. No study has yet investigated the genetic basis of diabetes in TCP and FCPD. Based on the suggestive linkage of T2D to chromosome 10q, a microsatellite, DG10S478, within intron 3 of transcription factor 7-like 2 (TCF7L2) gene was found to be strongly associated with T2D [19]. An association of a variant of the gene, rs7903146 along with other SNPs in linkage disequilibrium with this polymorphism was first reported in Islandic, Danish and in the US cohort [19]. Subsequently this association was replicated in other populations like Indian [20,21], French [22], U.K [23] and Finnish populations [24], and these variants account for the highest T2D risk confirmed to date [25]. TCF7L2 gene variants have also been proposed to play important role in T1D because of its effects on blood glucose homeostasis [26]; however a recent study failed to find any association and age-of-onset effect of T1D with rs7903146 SNP in TCF7L2 gene [27]. This suggested that a T2D mechanism mediated by polymorphisms in TCF7L2 does not participate in the etiology of T1D, thus susceptibility factors for T2D could be different from those involved in T1D. Hence, investigating a known susceptibility factor for T1D or T2D can help in understanding the type and mechanism of diabetes in FCPD patients.As the type of diabetes in FCPD is not clearly understood, we used association of TCF7L2 variants with T2D as a marker to decipher the type of diabetes in FCPD. Since there are suggestions that TCP is the pre-diabetic stage of FCPD, we also analyzed the association of TCF7L2 polymorphisms in TCP patients. We hypothesized that we would observe an association of variants in the TCF7L2 gene with FCPD, if diabetes in these patients is T2D. Since diabetes in TCP is also thought to be due to destruction of endocrine pancreatic cells secondary to destruction of exocrine pancreas, we investigated the interaction between the TCF7L2 variants and N34S SPINK1 and L26V CTSB mutations and explored whether presence of TCF7L2 variants in patients with these mutations predisposes them to FCPD.MethodsPatients and controls478 unrelated individuals (320 males and 158 females) were diagnosed as TCP (n = 286) or FCPD (n = 192) patients based on the established WHO criteria [28]. Of these, 333 patients were of Dravidian ethnicity and 145 belonged to Indo-European ethnicity. Six hundred and sixty one age matched individuals (332 males and 329 females) comprising of 259 Dravidians and 402 Indo-Europeans without any complaints and evidence of pancreatitis were included as controls [17,20]. Both patients and the controls filled a detailed questionnaire and signed a written informed consent for genetic analysis. The Institutional Ethics Committee of all the institutes approved the study following the Indian Council of Medical Research guidelines for research on human subjects.Genetic analysisGenomic DNA from all the patients and healthy volunteers were utilized for this study. Primers, amplifying segments of TCF7L2 gene harboring SNPs rs7903146 and rs12255372 were adopted from our earlier study [20]. PCR products were purified and sequenced individually on both the strands using Big-dye terminator cycle sequencing ready kit (Applied Biosystems, Foster City, CA) on an ABI3730 Genetic Analyzer (Applied Biosystems Foster City, CA). In case of unclear sequence data, we repeated direct sequencing under various conditions until the genotype was determined correctly. Ten percent of the genotyping results were validated on tetra primer based analysis for the 2 SNPs [20] and no discrepancy was observed.Statistical analysisThe allele and genotype frequencies were calculated for the SNPs (table 1) in cohorts of both ethnicities separately as well as together and to analyze deviation from the Hardy-Weinberg equilibrium, observed and expected genotype frequencies were compared by Markov simulation based goodness of fit test [29]. Chi-square test was used to analyze the statistical significance of the difference in allelic distribution of various polymorphisms in patients and controls (DeFinitte; ). For assessing the overall significance irrespective of ethnicities, the meta-analysis statistic was used and the forest plots were generated under the fixed effect model using Comprehensive Meta Analysis  software version 2.2.046 and the Q test was used to test for homogeneity of groupings [30]. The whole cohort was dichotomized initially based on the presence or absence of N34S SPINK1 and L26V CTSB mutations and then the two groups were subdivided into TCP and FCPD patients and distribution of TCF7L2 variants was compared between them. Unless indicated specifically, a p-value of 0.05 was considered significant in all the analyses. This study with random selection of patients and controls has 80% power to detect an effect with an OR as low as 1.3 at α = 0.05 and 95% power at the OR of 1.46, which was identified in our earlier study on T2D subjects [20].Table 1Allelic and genotypic frequencies for the TCF7L2 variants in TCP patients and controls of different ethnic groupsSNP (NCBI 36.2&)AlleleDravidianIndo-EuropeansTotalGenotypeDravidian@Indo-Europeans@Total@PatientsControlsPatientsControlsPatientsControlsPatientsControlsPatientsControlsPatientsControlsn = 333n = 259n = 145n = 402n = 478n = 661n = 333n = 259n = 145n = 402n = 478n = 661rs7903146 (114748339)C0.710.700.720.710.730.71CC175 (52.6)130 (50.2)78 (53.8)207 (51.5)253 (53.0)337 (51.0)CT126 (37.8)104 (40.2)53 (36.6)160 (39.8)179 (37.3)264 (39.9)T0.290.300.280.290.270.29TT32 (9.6)25 (9.7)14 (9.7)35 (8.7)46 (9.6)60 (9.1)n = 332n = 180n = 144n = 402n = 476n = 582n = 332n = 180n = 144n = 402n = 476n = 582rs12255372 (114798892)G0.770.780.780.780.770.78GG201 (60.5)108 (60.0)88 (61.1)244 (60.7)289 (60.7)352 (60.5)GT110 (33.1)64 (35.6)48 (33.3)135 (33.6)158 (33.2)199 (34.2)T0.230.220.220.220.230.22TT21 (6.3)8 (4.4)8 (5.6)23 (5.7)29 (6.1)31 (5.3)SNP, single nucleotide polymorphism; n, number of individuals; &, Chromosome position according to National Centre for Biotechnology Information (NCBI), Build 36.2, contig accession number NT 030059.12; @, values in the parentheses indicate percentage genotype frequencyResults and discussionThe two polymorphisms rs7903146 and rs12255372 in the TCF7L2 gene, reported to be most strongly associated with T2D, were analyzed in a cohort of TCP and FCPD patients and controls belonging to Dravidian and Indo-European ethnicities. It is believed by most workers in the field that FCPD is the logical end point of TCP and enough evidence exists to suggest that TCP is the pre-diabetic stage of FCPD [3]. Thus, we also analyzed the association of TCF7L2 variants in the entire cohort irrespective of their diabetic status. In addition, clinical presentation is known to be variable for FCPD [1,2]. Most of the patients present with pain abdomen and evidence of pancreatitis and subsequently develop diabetes at a later stage; a small proportion present with diabetes and are detected to have pancreatic stones and calcification on subsequent investigations [1,2]. It may be surmised that additional diabetes susceptibility gene may account for the earlier phenotype of diabetes. Hence, an attempt was also made to dichotomize the cohort into TCP and FCPD to investigate whether FCPD patients have an additional risk due to TCF7L2 polymorphisms. We also investigated whether co-inheritance of TCF7L2 variants with the SPINK1 and CTSB mutations predisposes these patients to develop diabetes.Both the polymorphisms followed Hardy-Weinberg equilibrium (p > 0.05) and on comparing the allele frequencies within the ethnic groups, Dravidian patients vs Dravidian controls and Indo-European patients vs Indo-European controls, no significant differences were seen (table 1), neither did the genotype relative risk differ significantly between patients and controls (table 2). A meta-analysis of all cases and control subjects from both ethnicities, showed similar results for both SNPs in TCF7L2 gene [(95%CI, 0.95–1.16; P = 0.63, Cochran's Q = 0.0092, P = 0.92 for rs7903146) and (95%CI, 1.01–1.29; P = 0.92, Cochran's Q = 0.0482, P = 0.83 for rs12255372)] (fig 1). In order to explore the possibility of association of TCF7L2 variants with FCPD, the analysis was carried out in TCP patients with diabetes (FCPD) and those without diabetes separately. Allele and genotypic frequencies did not differ significantly, between TCP patients and controls, FCPD patients and controls and between TCP and FCPD patients, suggesting lack of statistically significant association of TCF7L2 polymorphisms with FCPD (table 3).Table 2Estimates of the genotype and allele relative risks for the TCF7L2 variants in the cases and controls based on ethnicitySNP*Het OR (95% CI)PHom OR (95% CI)P&OR (95% CI)PAll cases vs all controlsrs79031460.90 (0.70–1.15)0.401.02 (0.67–1.55)0.920.96 (0.80–1.16)0.70rs122553721.03 (0.80–1.34)0.800.88 (0.52–1.49)0.620.98 (0.80–1.21)0.88DR cases vs DR controlsrs79031460.90 (0.64–1.27)0.550.95 (0.54–1.68)0.860.94 (0.73–1.21)0.65rs122553720.92 (0.63–1.36)0.691.41 (0.60–3.30)0.421.04 (0.76–1.41)0.81IE cases vs IE controlsrs79031460.86 (0.57–1.30)0.481.06 (0.54–2.08)0.860.96 (0.71–1.29)0.79rs122553720.99 (0.65–1.48)0.950.96 (0.42–2.23)0.930.98 (0.71–1.36)0.92SNP, single nucleotide polymorphism; *baseline genotype at rs7903146-CC, rs12255372-GG; Het OR & Hom OR, genotype relative risk (GRR) for heterozygotes and homozygotes respectively (GRR was calculated by comparing with the baseline genotype); &, allelic OR; P, P value; DR, Dravidians; IE, Indo-EuropeanTable 3Estimates of the genotype and allele relative risks for the TCF7L2 variants in the cases and controls based on clinical diagnosisSNP*Het OR (95% CI)PHom OR (95% CI)P&OR (95% CI)PTCP (n = 286) vs Controlsrs79031460.95 (0.71–1.28)0.740.98 (0.59–1.61)0.940.97 (0.78–1.21)0.81rs122553720.92 (0.68–1.26)0.621.31 (0.72–2.36)0.381.03 (0.81–1.31)0.79FCPD (n = 192) vs Controlsrs79031460.82 (0.58–1.16)0.271.08 (0.62–1.87)0.780.95 (0.74–1.22)0.69rs122553721.03 (0.73–1.46)0.860.89 (0.41–1.92)0.760.99 (0.75–1.31)0.94TCP vs FCPDrs79031460.86 (0.58–1.28)0.471.10 (0.58–2.08)0.760.98 (0.73–1.30)0.87rs122553721.11 (0.75–1.65)0.590.68 (0.30–1.55)0.360.96 (0.70–1.31)0.79SNP, single nucleotide polymorphism; TCP, tropical calcific pancreatitis; FCPD, fibro-calculous pancreatic diabetes; *baseline genotype at rs7903146-CC, rs12255372-GG; Het OR & Hom OR, genotype relative risk (GRR) for heterozygotes and homozygotes respectively (GRR was calculated by comparing with the baseline genotype); &, allelic OR; P, P valueFigure 1Meta-analysis for association of TCF7L2 variants with TCP. Forest plot showing results of meta analysis. Odds ratio for each study is represented by a block bounded by its confidence interval; Combined effect for the two studies has been calculated using fixed effect model; A, rs7903146; B, rs12255372.Association analysis of the total cohort after dichotomization based on N34S SPINK1 and L26V CTSB mutation status showed comparable allele and genotype frequencies for rs7903146 in both groups, indicating that co-existence of these variants does not increase the risk of developing diabetes in these patients (table 4). However, the minor allele frequency for rs7903146 was different between TCP and FCPD patients carrying the N34S SPINK1 variant but did not reach statistical significance (OR = 1.59, 95% CI = 0.93–2.70, P = 0.09). Interestingly, similar analysis using L26V CTSB variant showed a statistically significant association between TCP and FCPD patients carrying the mutant allele compared to those without the variant (OR = 1.69, 95% CI = 1.11–2.56, P = 0.013) (table 4). Similar results were obtained on analysis of the rs12255372 variant in TCF7L2 gene (data not presented). This suggests that co-existence of TCF7L2 variants and the variants predicting susceptibility to exocrine damage may interact to determine the onset of diabetes in TCP patients. However, this may need to be replicated in larger sample size since there is a possibility of a chance association due to small sample size.Table 4Association of TCF7L2 variant rs7903146 on dichotomization of the patient cohort based on N34S SPINK1 and L26V CTSB mutation statusN34S SPINK1L26V CTSBWildMutantWildMutantrs7903146MAFMAFOR (95% CI)PMAFMAFOR (95% CI)PTotal Cohort(n = 307)(n = 148)(n = 165)(n = 236)0.290.280.95 (0.70–1.30)0.750.280.281.01 (0.74–1.38)0.95FCPD(n = 129)(n = 49)(n = 65)(n = 78)0.310.351.13 (0.69–1.85)0.630.280.361.42 (0.86–2.36)0.17TCP(n = 178)(n = 99)(n = 100)(n = 158)0.270.240.87 (0.58–1.29)0.480.280.240.83 (0.55–1.24)0.36FCPD vs TCPOR1.22 (0.85–1.73)1.59 (0.93–2.70)0.98 (0.60–1.61)1.69 (1.11–2.56)P0.270.090.950.013*TCP, tropical calcific pancreatitis; FCPD, fibro-calculous pancreatic diabetes; MAF, Minor allele frequency; OR, odds ratio; CI, confidence interval; P, p value;We and others have earlier replicated the strong association of TCF7L2 variants with T2D in Indian population and provided evidence of its likely role in the pathogenesis of T2D by influencing both insulin secretion and insulin resistance [20]. According to accelerator hypothesis, T1D and T2D may share a common etiology of hyperglycemia-induced beta cell damage but T1D may have the added effects of autoimmunity [31]. However, the lack of association of TCF7L2 with T1D, as shown by Field et al., does not support the model of a shared major causal pathway in T2D and T1D [32]. Hence the genes that determine susceptibility to T1D must be different from the susceptibility genes of T2D. This has important implications for diabetes in TCP since overlapping features of T1D and T2D are observed in FCPD. As the overall evidence for association of TCF7L2 gene variants exceeds genome-wide significance criteria (10-5) and clearly establishes TCF7L2 as a T2D susceptibility gene of substantial importance in majority of populations world-wide [33] including Indian population [20], it is less likely that T2D may have a susceptibility factor stronger than TCF7L2. A lack of association of TCF7L2 with TCP or FCPD observed in our study, may suggest a role for genes other than TCF7L2 to be predictive of susceptibility to T2D. Since there is debate about the type of diabetes in TCP and FCPD, the lack of association with TCF7L2, the gene most strongly associated with T2D may suggest that the diabetes in TCP patients does not have similar features as T2D.ConclusionAs TCF7L2 is a major susceptibility gene for T2D, a lack of association of TCF7L2 variants with TCP or FCPD observed in our study suggests that T2D associated TCF7L2 variants are not associated with diabetes in TCP or the diabetes in TCP patients may not be similar to T2D. Thus, although the variations in TCF7L2 increase the risk for T2D and may affect insulin secretion, they do not alter susceptibility to FCPD, the diabetes in TCP patients. However, co-inheritance of the TCF7L2 variants with the pancreatitis associated susceptibility variants in SPINK1 and CTSB genes may predict the development of diabetes in these patients, but these observations need to be confirmed independently.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsSM did all the genotyping, statistical analysis and wrote the first draft of the manuscript. SB and SP assisted in the genotyping and statistical analysis whereas DNR, GVR, SPS and VT were involved in the recruitment of the patients and controls. GRC conceptualized the study, supervised the results and finalized the manuscript. All the authors have gone through the manuscript and have consented to the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2529292.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529292",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529292\nAUTHORS: Marije Bosch, Rob Dijkstra, Michel Wensing, Trudy van der Weijden, Richard Grol\n\nABSTRACT:\nBackgroundRedesigning care has been proposed as a lever for improving chronic illness care. Within primary care, diabetes care is the most widespread example of restructured integrated care. Our goal was to assess to what extent important aspects of restructured care such as multidisciplinary teamwork and different types of organizational culture are associated with high quality diabetes care in small office-based general practices.MethodsWe conducted cross-sectional analyses of data from 83 health care professionals involved in diabetes care from 30 primary care practices in the Netherlands, with a total of 752 diabetes mellitus type II patients participating in an improvement study. We used self-reported measures of team climate (Team Climate Inventory) and organizational culture (Competing Values Framework), and measures of quality of diabetes care and clinical patient characteristics from medical records and self-report. We conducted multivariate analyses of the relationship between culture, climate and HbA1c, total cholesterol, systolic blood pressure and a sum score on process indicators for the quality of diabetes care, adjusting for potential patient- and practice level confounders and practice-level clustering.ResultsA strong group culture was negatively associated to the quality of diabetes care provided to patients (β = -0.04; p = 0.04), whereas a more 'balanced culture' was positively associated to diabetes care quality (β = 5.97; p = 0.03). No associations were found between organizational culture, team climate and clinical patient outcomes.ConclusionAlthough some significant associations were found between high quality diabetes care in general practice and different organizational cultures, relations were rather marginal. Variation in clinical patient outcomes could not be attributed to organizational culture or teamwork. This study therefore contributes to the discussion about the legitimacy of the widespread idea that aspects of redesigning care such as teamwork and culture can contribute to higher quality of care. Future research should preferably combine quantitative and qualitative methods, focus on possible mediating or moderating factors and explore the use of instruments more sensitive to measure such complex constructs in small office-based practices.\n\nBODY:\nBackgroundConsistently, studies show that patients with chronic illnesses do not receive optimal treatment [1,2]. Redesigning primary care by separating acute care from planned management of chronic conditions has been proposed to close the quality chasm between current practices and optimal standards [3]. Of all chronic conditions, care for diabetic patients is probably the most manifest and widely spread example of primary care development [4,5]. In the Netherlands, 85% of patients with Diabetes Mellitus type 2 are treated within primary care [6].The creation of practice teams with a clear division of labour is an important aspect within this context [7]. Nurses and nurse assistants both are generally involved in management of patients with diabetes. Therefore, key elements of teamwork, such as sharing clear goals, division of labour, training and communication [8] are suspected to potentially improve care for these patients [7,9]. Studies showed positive associations between higher levels of teamwork and such outcomes as clinical performance [10], absence of hospital physicians due to sickness [11], job satisfaction [12], and patient outcomes such as satisfaction of patients with their care [12-15]. A related construct that is increasingly described in quality improvement research is organizational culture. This interest is based on the increasing recognition that cultural changes are needed alongside the structural changes to secure gains in quality [16]. Some studies showed that organizational cultures that support teamwork and quality improvement may contribute to achieving high quality care [17-20]. However, it has also been shown that a mix of cultures was associated with higher levels of team effectiveness [21], whereas several other studies failed to find associations between culture and performance [22,23].In most countries, primary care practices are small, office-based organizations, usually consisting of no more than a handful of people. Although evidence for the possible relevance of teamwork and culture is growing, most evidence for these-intuitively appealing-concepts is based on studies in hospital settings. In this study we therefore investigate whether higher levels of teamwork and specific types of organizational culture are associated to diabetes care in small office-based general practices.MethodsDesign and populationThe present cross sectional study was embedded in an intervention study, in which 350 practices in three regions in the middle and south of the Netherlands were invited to participate. Forty general practices agreed to participate (response rate 11.4%), and they were paired on stratification criteria and randomly allocated to intervention or control group [6]. A researcher visited intervention practices at the beginning of the intervention period, in February to April 2003, to discuss the current practice procedures for diabetes care with the staff. Situations in which various staff members shared tasks was a special topic of discussion. Then a diabetes passport was introduced, a patient-held booklet with important personal information that can be used to track results, record treatment targets and give (educational) information. The professionals discussed how the passport could best fit in the practice routines and work processes. The researcher summarized the various responsibilities involved in diabetes care and the use of the passport on a desk-top card. In the first three months, patients received their passport. Three months later, a researcher visited the practice to discuss the progress of the project and to see whether the division of tasks was being maintained as planned. After 6 months, all patients completed a short questionnaire on the use of the diabetes passport, after which each practice received benchmarked feedback on the introduction and use of the passports [6]. At post-intervention, in May to July 2004, all practice members in the 40 practices who indicated to be actively involved in medical care for patients with diabetes type II (general practitioners, nurse practitioners, and practice assistants) were invited to complete questionnaires on team climate and organizational culture. Team and culture measures were combined with data of diabetes mellitus type II patients younger than 80. The study was approved by the ethics committee Arnhem-Nijmegen. Written, informed consent was received from all study participants.MeasuresClinical outcomes were HbA1c level, systolic blood pressure and total cholesterol levels. A fourth outcome was clinical performance which was measured with a sum score of 10 process indicators of diabetes care quality, based on national guidelines on diabetes care [24] (see Figure 1; measured at the level of the individual patients, Chronbach's alpha 0.86). A patient could be given a score between 0 and 10, because each indicator was scored either done (1) or not done (0). All outcomes were derived by scrutinizing the electronic medical record systems (EMR) by trained research personnel at post-intervention in July 2004.Figure 1Clinical performance measure: diabetes guideline recommendations.Independent factorsTo measure organizational culture, we used the 'Competing Values Framework' (CVF) in which respondents were asked to distributed 100 points across four sets of organizational statements according to the description that best fits their own organization in five questions [25]. This approach recognizes that no organization exhibits only one culture or set of values, but that multiple cultures and values coexist simultaneously and compete for attention. The framework distinguishes two dimensions: 'internally oriented' versus 'externally oriented', and 'stability' versus 'flexibility and change', resulting in four ideal types of culture. The group culture emphasizes teamwork, cohesiveness, and participation. The developmental culture is characterized by the promotion of innovation and risk-taking, and is oriented towards growth. The rational culture emphasizes achievement and meeting objectives; people are rewarded to achieve organizational goals and working efficiently. Finally, the hierarchical culture emphasizes stability, rules, regulations and coordination. The statements reflect the four culture types. For each question, non blank respondent errors (i.e. the allocation of more or less than 100 points) were corrected by proportionally adjusting the responses to sum up to 100. For each practice, we determined the mean scores on the four types of culture. Internal consistency reliability for the four culture types, using Cronbach's alpha, were 0.64 for group culture, 0.51 for developmental culture, 0.55 for hierarchical culture, and 0.46 for rational culture. In addition, we calculated how well the scores for the different organizational types of culture were in balance, using the Blau index that has been described in previous studies [21,22]. The hypothesis underlying this measure is that it is the relative balance among the four culture types that is associated with team effectiveness. Higher scores on this index indicate a more even distribution of points among the four culture types, so practices that distributed their points in a 25/25/25/25 pattern had the highest score on 'culture balance' (1), whereas practices with more points for one or the other culture type had lower balance scores (< 1).Teamwork was measured with the 14 item short version of the 'Team Climate Inventory' (TCI) [26,27], answered on 5-point Likert scales. The underlying theory argues that group innovations often result from team activities which are characterized by 1) focusing on clear and realistic objectives in which the team members are committed (vision), 2) interaction between team members in a participative and inter-personally non-threatening climate (participative safety), 3) commitment to high standards of performance and, thus, preparedness for basic questions and appraisal of weaknesses (task orientation), and finally, 4) enacted support for innovation attempts including, e.g. cooperation to develop and apply new ideas (support for innovation). For each scale, mean scores were calculated per individual and then averaged to practice-level scores. Chronbach's alphas were 0.81, 0.79, 0.78, and 0.82 respectively, and correlations (r) ranged from 0.49 to 0.53. We finally combined these to one single score [15]. Overall Chronbach's alpha for the 14 questions was 0.91. Correlations between scales and the overall measure ranged from 0.75 to 0.84.We translated both the team and culture instruments into Dutch according to guidelines for cross-cultural translation [28]. Analysis of variance tests verified that individual level responses to the culture and team climate instrument could be validly aggregated to the level of the teams for all but one scale. The within-team variability of responses was less than the between-team variability (F values ranging from 2.29 to 3.90 (p < 0.005)). This test was not significant for the hierarchical culture scale (F value 1.3; p = 0.19).The following-possibly confounding-factors were included: whether the practice had special diabetes consulting hours, and whether it was an intervention or control practice, measured by a checklist that was completed by a member of each practice personnel at the start of the project. Finally, age and gender of the patients were included, derived from mailed patient questionnaires, and the baseline measures of the four outcomes derived from the EMR.AnalysisWe performed multi level regression analyses (mixed models) with patients (level 1) nested within the practices (level 2). We examined bivariate correlations to check for high correlations (Pearson's correlation and cross tabulations with χ2 test and studied single relationships between the outcomes and all predictors before adding the control variables. Since we were interested in the effect of each of our variables of interest separately (different types of organizational culture and team climate), we used separate models to study one of these variables at a time. Thus, for each outcome, six models were conducted; four different models examined the four organizational cultures, one examined the balance among these culture types, and one examined team climate. Each model controlled for patient age, sex, and the baseline measure on the particular outcome, whether the practice had special diabetes consulting hours, and whether it was an intervention or control practice. All analyses were performed using SPSS version 12.0.1.ResultsPractice characteristicsIn total, 146 practice members in 40 practices were invited to complete the questionnaires. We obtained team climate and culture data from 92 respondents, 46 general practitioners (response rate 71%), 8 practice nurses (response rate 73%) and 38 practice assistants (response rate 54%) working in 39 practices (overall response rate: 63%). The analysis on organizational culture and team climate was restricted to the practices in which at least two practice members returned the questionnaires. Therefore, we excluded 9 practice members in 10 practices in which this was not the case. The mean number of appointed members per practice was 3.7 (SD 1.0) and did not differ significantly for excluded practices as compared to included practices (3.4, range 2 to 5 and 3.8, range 2 to 6 respectively, p = 0.2). Also, excluded practices were as often single handed practices as included practices (p = 0.3).Table 1 shows the characteristics of the practices. Single handed practices were underrepresented in our sample as compared to the national mean (40% versus 60%) [29]. Among the four types of culture, group culture by far received most of the points (mean across practices = 51.6), followed by hierarchical (19.7), developmental (16.9) and finally rational culture (11.8). The balance among these values of culture was 0.60 on average. We also explored the data for the dominant culture [17] (the culture scoring highest in each practice; data not shown). In only 3 practices, hierarchical culture received the highest amount of points. All the other practices had a dominant group culture. The overall mean score on team climate was 1.94. Scores on the four scales were 1.84 for vision, 1.83 for participative safety, 1.96 for task orientation and 2.16 for support for innovation; data not shown).Table 1Characteristics of practices (N = 30)%/Mean (SD)Type of practice (% Single handed)40%Special diabetes consulting hours36.7%Group culture (0 – 100)51.6 (13.2)Developmental culture (0 – 100)16.9 (7.4)Hierarchical culture (0 – 100)19.7 (8.0)Rational culture (0 – 100)11.8 (5.6)Cultural balance (0 – 1)0.60 (0.10)Team climate (1 – 5)1.94 (0.39)Patient characteristicsIn 40 practices, 2106 patients received questionnaires. Response rates were 68% for the first, and 69% for the second questionnaire, which resulted in data from 993 patients. Since we excluded 10 practices, 241 patients were excluded, leaving 752 patients for this study. Excluded patients did not differ significantly from included patients with respect to age, sex, and our outcomes.Inspection of Table 2 learns that the mean age of the patients was 63 years, and 48.7% was male. Mean systolic blood pressure was 144.2; mean total cholesterol was 81.5 and mean HbA1c was 7.0. Scores on diabetes care quality differed considerably, and varied from 0 to 9, with a mean score of 5.82.Table 2Characteristics of patients (N = 752)N%/Mean (SD)Gender, % male75248.7%Age, years (SD)75263.0 (9.7)Systolic blood pressure (SD)716144.2 (19.4)Total Cholesterol (SD)71681.5 (9.6)HbA1c6967.0 (1.2)Quality of diabetes care (0 – 10)7525.82 (2.8)Table 3 shows that none of the selected clinical patient outcomes (HbA1c, systolic blood pressure and total cholesterol) showed significant associations with team climate or culture. However, we did find significant relations with clinical performance. A higher score on group culture was associated with lower scores on diabetes care quality (p = 0.04) with a coefficient of -0.04. This means that every 10-unit change on the group culture score (e.g. from 20 to 30 points) resulted in a 0.4 lower score on the diabetes care quality indicator. In theory, if a practice would move from the lowest group culture score to the highest (a difference of 55.6 points in this sample), the score on the quality indicator would decrease by 5.6 * 0.4 = 2.24 points. Since the range in the mean scores for the quality indicator was from 0 to 9 points, 2.24 points therefore represents a maximum decrease of 24.9%. In total, 15.6% of the variation in the quality indicator outcome was determined by our model that included group culture of which 2.7% was accounted for by group culture. On the other hand, maintaining a balance between the different culture types was positively associated with quality (β = 5.97, p = 0.03), representing a maximum 27.6% of the nine point practice range in our quality indicator. A 0.1-unit change in the balance score (e.g. from 0.6 to 0.7) resulted in a 0.6 higher score on the quality indicator. Our model including cultural balance explained 16.2% of the variation in the quality indicator, of which 3.5% was explained by cultural balance.Table 3Associations between team climate, organizational culture and HbA1c, systolic blood pressure, total cholesterol and the aggregated diabetes process quality indicator, measured at patient level (N = 752).HbA1cSystolic blood pressureTotal cholesterolClinical performanceβ95% CIβ95% CIβ95% CIβ95% CIGroup culture-0.01-0.02, 0.00-0.08-0.25, 0.100.00-0.01, 0.00-0.04-0.08, 0.00 *Developmental culture0.00-0.02, 0.010.11-0.16, 0.390.01-0.01, 0.020.04-0.03, 0.11Hierarchical culture0.010.00, 0.020.10-0.14, 0.340.00-0.01, 0.010.03-0.03, 0.09Rational culture0.020.00, 0.03-0.11-0.44, 0.230.00-0.01, 0.020.04-0.05, 0.12Cultural balance1.35-0.03, 2.729.70-14.53, 33.930.65-0.42, 1.725.970.66, 11.28 *Team climate-0.22-0.50, 0.052.06-2.53, 6.640.09-0.13, 0.30-0.57-1.76, 0.76* sign < 0.05DiscussionOverall, we found that high group culture scores were negatively correlated with adherence to diabetes guidelines in primary care practice (β = -0.04), whereas maintaining a balance among the different types of culture on the other hand was positively correlated to managing diabetes care well (β = 5.97). None of our variables of interest showed associations with our clinical patient outcomes.Comparison with other studiesThis study confirmed results of recent studies in primary care in the UK, using the CVF, by showing that primary care organizations primarily have group cultures [22,30]. In one of those studies managers of primary care trusts pointed out the possible disadvantages of group cultures, such as a tendency to be 'inward looking'. They expected quality improvement to be hard to achieve unless practices change their culture to one that valued greater collaboration and sharing of expertise, and a willingness to be more flexible in the way that they operated [31]. In our study, high scores on the group culture variable were negatively correlated with indicators for managing care well. This might be explained in light of the suggestion that different culture types are related to those aspects of performance that are valued by that specific dominant culture type [16]. In other words, for example for changing routines (in quality improvement projects), a more team-focused and developmental culture type with a focus on flexibility might be helpful in attaining good results, whereas for performing routine tasks, such as inspecting feet every 3 months, aspects valued in the more control orientated rational or hierarchical culture types, with a focus on policies, procedures and production might be needed. Therefore, one could also argue that -to reach and sustain high quality care for chronic diseases such as diabetes-teams need to find the balance between flexible and control oriented culture types since continuous measuring and improvement, good teamwork, a drive to gain better results, and working according to protocols are equally important. This might be in line with the fact that we found that a high balance between the different types of culture was positively correlated to high quality diabetes care. An earlier study on the role of perceived team effectiveness in improving chronic illness care reasoned that it would be the relative balance among culture values of participation, achievement, openness to innovation and adherence to rules that is most likely to be associated with perceived team effectiveness. Indeed, this study showed an association between a culture balance and team effectiveness, although it was rather marginal [21]. A recent study in primary care hypothesized that a high score on cultural balance would be associated with high levels of team climate, which was not confirmed by the data [22].Although previous studies suggested the relevance of teamwork in diabetes care [9,13,15], we failed to find significant associations between team climate and our outcomes, as did a recent UK study [22]. Again, the type of outcome might shed some light on this topic, since studies that did find associations often included outcomes such as work satisfaction [12], absence from work due to sickness [11] and satisfaction by patients with their care [12-15]. Interestingly, climate scores were also quite low as compared to other studies [32,33]. This might point to the fact that different practice members involved in diabetes care may not experience their relationships as a 'true' team when it comes to diabetes care [23,34]. The varied nature of clinical problems in primary care practice make team building especially challenging as compared to 'single specialty practices' [8].Our study failed to find associations between our organizational factors of interest and intermediate clinical patient outcomes. These findings are consistent with recent findings in studying and reviewing the link between safety-factors and risk-adjusted patient outcomes [35,36]. Although the selection of a clinical outcome is recommended, the selection of such a specific variable may just be too narrow to reflect the complexity of modern patient care [37].Strengths and limitations of this studyTo gain better insight on organizational factors influencing health care quality, it has been suggested that studies should preferably focus on factors on different levels (e.g. organizational as well as team), include patient outcomes and use multi level data analyses to correct for clustering effects [38]. In the current study, we have taken these suggestions into account. However, some limitations need to be addressed.First, the relative small sample size in our study may have limited the power to find associations. Since general practices are generally small office-based organizations, the number of participants who returned our questionnaire on organizational culture and team climate was relatively low (varying from 2 to 4). Previous studies using the TCI excluded practices if less than 30% of respondents completed questionnaires [12,22]. However, the number of GPs and other care providers per practice seems generally somewhat lower in the Netherlands than in -for instance-UK practices [13,29,39]. In this study, we also excluded the practices in which only one person returned our questionnaire. The low numbers of respondents could impact the validity of our culture and team climate measures. Low Cronbach's alphas for the culture measures for instance, and the low F-value for the aggregation of the scores on the hierarchical culture scale might point to that. In addition, the fact that primary care practices-both in our study and in the UK [22,30] – tend to have predominantly group cultures raises questions about the sensitivity of the CVF in this setting, especially if culture is analyzed as categorical variable. We have taken this point partly into account by using continuous culture variables in the analyses, however, this cannot fully clear away some concerns about the appropriateness of use of this particular instrument in small practices. Although this instrument has some clear advantages over others, such as the fact that it has been used in several other studies in varying settings, and the fact that it measures 'culture typologies' rather than simple variables [16], the factors measured may have a different meaning in different health care settings.Also, and partly related to our previous point, since climate and culture are considered to be shared attributes, individual measures are aggregated to practice level. Yet, this ignores the fact that different subgroups may have different opinions (for instance general practitioners may experience the culture differently from the practice nurses or assistants) [12,16]. Especially in very small practices (for instance with only one general practitioner and two practice assistants), it is debatable whether the aggregated score is a valid measure of the reality. However, for subgroup analysis researchers would need much bigger samples of respondents, which raises questions about the feasibility of survey based methods in measuring these complicated constructs.Further, our process measure was assessed by scrutinising the EMR. However, a considerable gap may exist between what the practice members record, and what they actually do in practice. Especially preventive or counselling activities, such as advising physical exercise, have been found to be under recorded [40]. Also, the guideline indicated that smoking behaviour should be discussed with all patients on a yearly basis, even if they are non-smokers. Therefore, we may have underestimated the scores on the quality indicator. However, it is likely that this holds for all practices to the same extent since they all used an EMR. We cannot rule out the possibility though that other confounding factors may have played a role, such as whether or not a physician received feedback or reminders in the EMR, which may have prompted these GPs to perform and register particular preventive activities. At the time of the study, no specific arrangements with insurance companies existed that may have influenced diabetes management. Some practices had a practice nurse who performed tasks related to care for patients with chronic diseases, however, the availability of practice nurses was equal for all regions in the Netherlands. Single handed practices were underrepresented in our study. However, previous research showed no difference in delegation of preventive tasks and treatment of chronic diseases between GPs in single handed practices compared to GPs in group practices [41] so we can assume that our sample is representative for Dutch practices.Finally, it is important to note that it is not possible to conclude we showed causal linkages between culture and our outcomes, since the results were based on cross sectional data. We therefore do not know whether high scores on group culture lead to poor diabetes management, or -the other way around-practices in which quality of care is managed in a certain way develop certain types of culture, or culture and performance emerge together in a reciprocal and reinforcing manner [16].ConclusionThis study contributes to the discussion around the evidence for intuitively appealing features such as culture and teamwork that have been suggested as a lever for health care improvement. We did find some significant associations between culture and high quality diabetes care, but the relations were rather marginal. On the one hand, one could argue that if organizational culture would have only limited influence on many aspects of care during a long period of time, the resultant of that might still add up to a substantial level. On the other hand, feasibility of current measurements of constructs such as climate and culture is still debatable-especially in primary care settings-, given the fact that response rates are low, and scores are aggregated, which causes power reduction and loss of information. Further, we failed to find any associations with our clinical outcomes, which begs the question if and exactly how these constructs can contribute to evidence based care, and -eventually-healthier patients.Future studies in primary care should preferably combine quantitative and qualitative research methods and use more complex designs to get a better insight into these complex constructs and possibly mediating or moderating factors. Also, it would be worth exploring possible associations between culture and climate and changes in health care quality, as well as the use of other measurement instruments and methods that are more sensitive to -for instance-different subcultures that might exist within organizations, especially in primary care practices where people work in very small teams and deal with a big variety of clinical problems.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsMB, RD, MW, TvdW and RG designed the study. MB performed the data collection and data analyses, and all other authors contributed to interpreting the data. MB wrote the first draft, which was critically revised by RD and then by all others. All authors have read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2529294.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529294",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529294\nAUTHORS: JM Fleming, EL Long, E Ginsburg, D Gerscovich, PS Meltzer, BK Vonderhaar\n\nABSTRACT:\nBackgroundThe normal growth and function of mammary epithelial cells depend on interactions with the supportive stroma. Alterations in this communication can lead to the progression or expansion of malignant growth. The human mammary gland contains two distinctive types of fibroblasts within the stroma. The epithelial cells are surrounded by loosely connected intralobular fibroblasts, which are subsequently surrounded by the more compacted interlobular fibroblasts. The different proximity of these fibroblasts to the epithelial cells suggests distinctive functions for these two subtypes. In this report, we compared the gene expression profiles between the two stromal subtypes.MethodsFresh normal breast tissue was collected from reduction mammoplasty patients and immediately placed into embedding medium and frozen on dry ice. Tissue sections were subjected to laser capture microscopy to isolate the interlobular from the intralobular fibroblasts. RNA was prepared and subjected to microarray analysis using the Affymetrix Human Genome U133 GeneChip®. Data was analyzed using the Affy and Limma packages available from Bioconductor. Findings from the microarray analysis were validated by RT-PCR and immunohistochemistry.ResultsNo statistically significant difference was detected between the gene expression profiles of the interlobular and intralobular fibroblasts by microarray analysis and RT-PCR. However, for some of the genes tested, the protein expression patterns between the two subtypes of fibroblasts were significantly different.ConclusionThis study is the first to report the gene expression profiles of the two distinct fibroblast populations within the human mammary gland. While there was no significant difference in the gene expression profiles between the groups, there was an obvious difference in the expression pattern of several proteins tested. This report also highlights the importance of studying gene regulation at both the transcriptional and post-translational level.\n\nBODY:\nBackgroundBreast cancer is the most commonly diagnosed cancer, and is the second leading cause of cancer mortality in women in the U.S. [1]. Metastasis of the tumor is the primary cause of morbidity and mortality. In late-stage breast cancer, tumor metastasis can be found in several tissues, including bone, lung, lymph node, and liver [2]. Because metastasis is a major challenge in cancer management, a better understanding of the metastatic progression is required. Tumor progression and metastasis are both regulated by the surrounding microenvironment, i.e. the local stroma. Therefore, studies targeted towards understanding the function of normal breast stroma will facilitate the development of methods for preventing breast cancer metastasis.Normal growth, function, and homeostasis of breast epithelial cells depend on intricate interactions between the numerous stromal cells within the mammary gland. The stromal cells are composed of a diverse assortment of cell types including the vasculature, adipocytes, resident immune cells, and fibroblasts. These cells secrete multiple cellular products, such as growth factors and extracellular matrix components, which have profound effects the behavior of the breast epithelial cells. Alterations in the regular communications between these cells can lead to the progression or expansion of malignant growth.It is now well documented that stromal cells have a striking effect on the behavior of mammary epithelial cells in culture [3-7] as well as on the formation, growth, and metastasis of epithelial-derived tumors in vivo [8-11]. Both in vitro and in vivo studies have shown that epithelial cell contact with tumor-derived or normal fibroblasts can either promote or inhibit tumorigenic cell growth, respectively [11,12]. In agreement with these reports, one study using microarray analyses demonstrated that the gene expression profiles of cancer-derived fibroblasts had a distinctive gene expression pattern that differentiated them from normal breast stroma [13]. Furthermore, breast cancer stroma differs morphologically from the stroma found in normal breast tissue. For example, in ductal carcinomas in situ (DCIS), and most invasive breast carcinomas, the stroma exhibits enhanced accumulation of fibroblasts and a modified collagenized extracellular matrix compared to its normal counterpart [3,14-19]. Understanding the mechanisms of the interactions between cancerous or normal epithelial cells and the stroma might lead to novel methods for cancer therapies that target the function of the resident stromal cells.Most models of breast cancer development are studied using mouse in vivo models. However, the stroma within the human mammary gland is fundamentally different from that in the mouse [20]. These differences make it difficult to ascertain the tumor/stromal interactions that would occur in the human breast when epithelial cells are implanted into the mouse mammary fat pad. Compared to the human breast, the mouse mammary gland contains large depots of adipose laced with small amounts of interspersed connective tissue. The functional lobular units of the mouse gland are embedded within the fat pad, and have a considerable amount of space between the minimally branched ducts. In contrast, the functional lobular units of the human mammary gland are surrounded by loose intralobular connective tissue, consisting primarily of fibroblasts. This intralobular stroma is subsequently surrounded by a more compact interlobular stroma, which detaches the lobules and intralobular stroma from any substantial direct contact with the adipose tissue [21]. Stemming from the observations that these stroma subtypes differ in their physical location in relation to the functional epithelial lobules, and that epithelial/stromal interactions can promote or inhibit tumorigenesis, we investigated the differences between the two distinct stromas.MethodsSample CollectionThis study was performed in accordance with the guidelines of the National Cancer Institute Review Board, protocol 02-C-0144. All patients provided written informed consent. Fresh human mammary tissue was collected from four (two Caucasian, one African-American, and one Hispanic) female, pre-menopausal, reduction mammoplasty patients, ages ranging from 18 to 40 years old. The tissue was embedded in Tissue-Tek O.C.T. embedding medium (Sankura Finetek Inc., Torrance CA) and frozen on dry ice immediately after surgery. Eight – 10 micron sections of tissue were cut using a Leica 2800 Frigocut-E cryostat (Bannockburn, IL). Every tenth section was subjected to hematoxylin and eosin staining. For each patient sample, sections with distinctive intralobular and interlobular regions were selected for laser capture and microarray analysis.Laser capture microdissection and microarraysLaser capture microdissection (LCM) and microarrays were performed by Cogenics, Inc. (Morrisville, NC). Briefly, selected intralobular and interlobular stroma sections of frozen tissue were subjected to an AutoPix™ automated LCM system from Arcturus, using static image settings. RNA was isolated from each specimen, pooled, and then evaluated by spectrophotometry and by using an Agilent Bioanalyzer before proceeding to sample amplification. For each sample, 50 ng of total RNA was amplified using Affymetrix Two-Cycle Target Labeling kit (Santa Clara, CA). Ten micrograms of biotinylated cRNA spiked with bioB, bioC, bioD, and cre as a control was hybridized to the Affymetrix Human Genome U133 GeneChip® for 16 h at 45°C. Following hybridization, arrays were washed and stained with Affymetrix GeneChip Fluidics Station. Stained arrays were scanned with an Affymetrix GeneChip Scanner 3000. Quality check and preliminary data analysis were carried out using Affymetrix GeneChip Operating Software and Quality Reporter.Microarray analysisMicroarray data were analyzed using the Affy package available at the Bioconductor website . The raw data were first background-corrected by the Robust Multichip Average (RMA) method [22] and then normalized by an invariant set method. Unsupervised hierarchical clustering analysis was performed on 1,115 most variable genes. The difference of gene expression between the inter- and intra-stromal samples was analyzed by the Limma package available at the Bioconductor website. P-values obtained from the multiple comparison tests were corrected by false discovery rates. The microarray data has been deposited in the public repository, Gene Expression Omnibus, accession number GSE12306.ImmunohistochemistryAll reagents were obtained from Sigma (St. Louis, MO) unless otherwise indicated. Ten-micron-thick sections of frozen tissue were fixed in 1:1 methanol:acetone for 10 min and washed in 1× phosphate buffered saline, pH 7.4 (PBS). Endogenous peroxidase activity was blocked by a 10 min incubation in 3% hydrogen peroxide followed by a 10 min wash in 1× PBS. Immunostaining was carried out using the Vectastain ABC kit (Vector, Burlingame, CA) according to the manufacturer's instruction. Color was developed with diaminobenzidine peroxidase substrate kit (Vector), and sections were counterstained with hematoxylin. Antibodies were obtained from the following sources: Met, Cell Signaling Technologies (Boston, MA); SOS2, Santa Cruz Biotechnology (Santa Cruz, CA); Tenascin-C, Invitrogen (Gaithersburg, MD); CD44, BD Biosciences (San Jose, CA); CD13, Novocastra (Visions Biosystems Bannockburn IL); CD26, Abcam Inc. (Cambridge, MA). Collagen staining with Sirius red was performed as previously described [23]. A positive and negative control was included in each experiment to validate the specificity of each antibody. A breast tissue sample that had been previously determined to express high levels of the protein of interest was used as a positive control. A serial section of each sample that received all staining steps, with the exception of the primary antibody, was used as a negative control.Reverse Transcription and PCRReverse transcription (RT) reactions were performed with 0.5 μg of total RNA isolated from LCM using Moloney murine leukemia virus reverse transcriptase (Invitrogen) primed with oligo-dT and random hexamers in a final volume of 25 μL. PCR was performed on 2.5 μL RT product using PCR Master Mix (Roche, Indianapolis, IN) with 0.2–0.4 μM of each primer. Primer sequences can be found in Table 1. Conditions for each PCR reaction were as follows: 94°C for 3 min for one cycle, followed by 94°C for 1 min, 60°C for 1 min, 72°C for 2 min, with a final extension at 72°C for 10 min. For a given experiment, PCR was performed using a predetermined number of cycles that spanned the linear range for the samples tested (20–30 cycles). RT-PCR products were resolved by agarose gel electrophoresis, visualized with ethidium bromide, quantified using NIH Image, and normalized to the respective level of GAPDH mRNA. Semi-quantitative RT-PCR analyses were conducted on a minimum of three patient samples. Appropriate negative controls were included for each RT-PCR.Table 1Primer sequencesPrimerPrimer sequences (listed 5' – 3')GAPDHForward CATGTGGGCCATGAGGTCCACCACReverse TGAAGGTCGGTGTGAACGGATTTGGCc-MetForward ACCTGCTGAAATTGAACAGCGAGCReverse ACACTTCGGGCACTTACAAGCCTASOS2Forward TAGAGAAAGGCGAGCAGCCAATCAReverse AGGGTGAGATTTGTGGTATGGCGACD44Forward GCCTGGCGCAGATCGATTTGAATAReverse CCCTGTGTTGTTTGCTGCACAGATTenascin-CForward AGATGTCACAGACACCACTGCCTTReverse TGTGGCTTGTTGGCTCTTTGGAACCD13Forward TCCACACCTTTGCCTACCAGAACAReverse TGCCTGATGTGCTGAAGAGATCGTCD26Forward TGGAGGCATTCCTACACAGCTTCAReverse ACAGCTCCTGCCTTTGGATATGGAResults and discussionMicroarray analysis was employed to identify genes that were differentially regulated between the intralobular and interlobular stromal subtypes. Normal human mammary tissue was obtained from healthy, pre-menopausal, reduction mammoplasty patients with no incidence of neoplasia. For each sample collected, tissue sections with distinctive intralobular and interlobular regions were selected for laser capture and microarray analysis (Fig. 1). RNA was extracted from the samples, checked for quality, and then hybridized with the Affymetrix Human Genome133A GeneChips, containing over 22,000 oligonucleotide probes. Surprisingly, no significant difference in gene expression was found between the two stroma subtypes. The microarray data demonstrated a wide range of expression values and a 45-degree straight line in each pair of samples, indicating the microarray assay and the normalization procedure were valid. Despite the small sample number, the scatter plots showed limited spread of the off-diagonal lines, suggesting that any differential expression between samples is subtle, and not significant (Fig. 2A). A heatmap with dendrograms was also generated from the data (Fig. 2B). At first glance, genes from sample numbers 13L, 27L, and 29R appeared to separate as different clusters with respect to interlobular and intralobular samples. However, further in-depth analysis using hierarchical clustering of the samples, based on the 1,115 most variable genes, did not reveal a distinct expression pattern between the intralobular and interlobular stromal tissues, further indicating there was no significant difference in terms of gene expression at the transcriptional level. Six of the genes with the largest difference in expression levels between intralobular in interlobular stroma are listed in Table 2, along with the fold-change and p-value. The lowest p-value was found to be 0.4726, which is far from statistically significant. In order to validate the microarray data, RT-PCR was performed on three of the top six genes listed in Table 2. The RT-PCR products from three patient samples as well as the quantitation of the products are shown in Figure (3A&3B). The expression levels of c-Met, SOS2, and CD44 reflect the findings of the microarray data; there was no significant difference between the intralobular and interlobular stroma.Figure 1Identification of intralobular and interlobular stroma in normal human breast tissue. Hemotoxylin and eosin staining of 8–10 micron sections of normal mammary tissue. The intralobular stroma isolated for laser capture microscopy is outlined in green while the interlobular stroma is outlined in black. Scale bar = 200 μM.Figure 2Scattered plots of normalized data and unsupervised hierarchical clustering of the samples and genes. A. Raw intensity data was background-corrected and normalized as described in Materials and Methods. The normalized data from seven samples were plotted against one sample (SB13L-Intra). B. 1,115 most variable genes were used for hierarchical clustering among samples. The gene expression values were scaled by row and shown in the heat map.Figure 3c-Met, SOS2, and CD44 expression levels in intralobular and interlobular normal human breast stroma. A. RT-PCR analysis of the indicated genes expression. B. Mean ± SD gained by densitometric examination of RT-PCR product from three independent samples. C. Tissues were subjected to immunohistochemical analysis with the specific antibody indicated (left panels) or corresponding negative controls (right panels). Scale bar = 200 μM.Table 2Top six genes with the highest p-valueGene nameAccession numberLog fold changeP-valueC-MetAA005141-0.48910.4726SOS2AI276593-0.43450.7845CPT1ABC000185-0.44650.7845PDLIM7AW206786-0.37220.7845TSC22D2AF201292-0.44550.7845CD44AW851559-0.47040.7845Examples of protein levels reflecting the gene expression levels between intralobular and interlobular stromaWhile the microarray and supportive RT-PCR analysis revealed no significant difference between gene expression levels, previous reports have documented a distinctive immunohistochemical difference between intralobular and interlobular stroma [23-27]. Therefore, we investigated whether we could observe a similar phenomenon using the same patients tissue samples utilized in the microarray analysis. We first investigated the protein expression of c-Met, the gene with the smallest p-value (0.4726). Fixed preparations of human mammary tissue from the four patients used in the microarray analysis, as well as additional samples, were immunoassayed using a specific antibody for c-Met. As shown in Figure 3C, all of the stroma uniformly stained positive for c-Met protein expression, with no detectable difference between the two stroma subtypes. The c-Met gene encodes the tyrosine kinase receptor for the hepatocyte growth factor/scatter factor (HGF/SF). c-Met/HGF signaling is required for mammalian embryogenesis and is important in cell migration, morphogenic differentiation, cell growth and angiogenesis. In normal breast tissue, c-Met was reported to be associated with ductal cells, and involved in ductal branching [28]. Additionally, the overexpression of c-Met has been shown to contribute to the development and progression of different human malignancies including lung, prostate, colorectal, gastric, and breast cancer [29]. Recently, it was reported that c-Met protein is overexpressed in inflammatory breast cancer compared to non-inflammatory breast cancer, and that an imbalance of c-Met protein expression between tumor and surrounding normal tissue is associated with an aggressive DCIS phenotype [30,31]. In the present study, c-Met protein was easily detectable and uniformly distributed throughout the normal breast.We next investigated the protein expression pattern of SOS2 (Son of Sevenless), the gene with the second lowest p-value (0.7845). SOS2 is a Ras-specific nucleotide-exchange factor that is involved in the receptor tyrosine kinase-Ras-ERK cascade [32]. This cascade has been implicated in the control of diverse biological processes including cell proliferation, differentiation, and survival. All tissues immunoassayed for SOS2 showed sparse, weak, staining in the stroma with no detectable difference in the staining pattern between the stroma subtypes (Fig. 3C). The only significant positive staining was found in the luminal epithelium of each sample.CD44, a protein which has recently gained much attention in breast cancer [33-37], is a ubiquitously expressed, multifunctional cell surface adhesion molecule involved in cell-cell and cell-matrix interactions, cell trafficking, and transmission of numerous growth signals [38]. The primary ligand for CD44 is hyaluronic acid, which is an important component of the extracellular matrix. However, other CD44 ligands include collagen, fibronectin, laminin, and chondroitin sulfate. Stromal hyaluronic acid levels are a strong, independent, negative predictor for patient survival in breast cancer [39,40]. Additionally, many cancer cells overexpress CD44 or express CD44 variants [41]. Mouse models of breast cancer tumorigenicity have suggested that CD44 expression is a cell surface marker that differentiates tumor initiating from non-tumorigenic breast cancer cells in immuno-compromised mice [42]. Furthermore, injection of reagents interfering with CD44-ligand interaction, such as CD44-specific antibodies, has been shown to inhibit local tumor growth and metastatic spread in mouse models of human cancer [38]. These findings suggest that CD44 may confer a growth advantage on some neoplastic cells and, therefore, could be used as a target for cancer therapy. In the present study, CD44 was one of the top genes with differential regulation between intralobular and interlobular stroma, although the p-value was 0.7845 and not significant. Immunohistochemical analysis reflected the microarray and RT-PCR results. There was uniform staining of the stroma, and the highest immuno-reactivity for CD44 was found in the epithelial cells (Fig. 3C).Examples of proteins differentially regulated between the intralobular and interlobular stroma, with no significant change in gene expressionAs previously stated, several reports illustrate a difference in protein expression between the intralobular and interlobular stroma. Atherton et al. (1998) reported that immuno-localization of type XIV collagen/undulin in the human mammary gland revealed greater deposition in the interlobular stroma than in the intralobular stroma [25]. Fibroblasts isolated from the interlobular stroma synthesized 3- to 5-fold more type XIV collagen/undulin than intralobular fibroblasts, but synthesized type I and type IV collagens in similar amounts. The authors suggest this protein is a way to separate the two types of distinct stroma for analysis. Collagen fibers have also been reported to be more abundant and densely packed throughout interlobular stroma compared to intralobular stroma in the bovine mammary gland [23]. Thus, we examined the collagen fiber deposition in the tissue samples from the patients used in the microarray data. Using Sirius Red, a pan stain for collagen fibers, there was a clear visible difference in the deposition of collagen fibers between the two types of stroma (Fig. 4A). In our microarray data, the fold change and p-value for type XIV collagen/undulin were -0.199 and 0.785, respectively. Undulin had the best p-value compared to all other types of collagen, but again, no values for any of the collagen genes were significant. RT-PCR analysis of the patients samples used in the microarray revealed an inconsistent expression of undulin between samples, resulting in no significant change between intralobular and interlobular expression (Fig. 4B&4C).Figure 4Localization and expression levels of collagen fibrils, CD13, Tenascin-C, and CD26 in intralobular and interlobular normal human breast stroma. A. Tissues were stained with Sirius Red alone (top panel) or with Fast Green counterstain (bottom panel). B. RT-PCR analysis of the indicated genes expression. C. Mean ± SD gained by densitometric examination of RT-PCR product from three independent samples. D. Tissues were subjected to immunohistochemical analysis with the specific antibody indicated (left panels) or corresponding negative controls (right panels). Note large quantities of intensely stained interlobular stroma (asterisks) compared to the paler-staining intralobular stroma (arrow). Scale bar = 200 μM.Our current data illustrated that the interlobular stroma has increased stromal collagen compared to the intralobular stromal. Mammographically dense breast tissue is one of the greatest risk factors for developing breast carcinoma, and regions of high breast density are associated with increased stromal collagen [43-45]. A recent report investigating the effects of collagen density on mammary tumor formation and progression utilized a bi-transgenic tumor model with increased stromal collagen in mouse mammary tissue [46]. This increased stromal collagen significantly increased tumor formation and resulted in a significantly more invasive phenotype, with increased lung metastasis. This study provided the first data causally linking increased stromal collagen to mammary tumor formation and metastasis, and demonstrated that fundamental differences arise and persist in epithelial tumor cells that progressed within collagen-dense microenvironments. It could be hypothesized that a change in the protein expression of the intralobular stroma to mimic the collagen expression of the interlobular stroma would enhance breast cancer progression. Studying the mechanisms, which lead to the differential levels in collagen deposition between these two stromal subtypes, could facilitate in understanding the physiology of breast density and the resultant influences on mammary epithelial cell function.Tenascin-C has also been reported to be expressed in the intralobular stroma as well as in the basement and sub-basement membrane zone of normal breast tissue [47]. Tenascin-C is a member of the tenascin family of modular and multifunctional extracellular matrix glycoproteins. These molecules are expressed in the adult during normal processes such as wound healing and tissue involution, and in pathological states including vascular disease, tumorigenesis, and metastasis [48]. In the present study, there was substantially more immuno-staining in the intralobular stroma compared to the interlobular stroma (Fig. 4D). Both the sub-basement membrane as well as the stroma had higher immuno-reactivity compared to the interlobular stroma. The microarray data reported a -0.119 fold change and a p-value of 0.852 for tenascin-C. RT-PCR was also performed on the same patient samples, and similar to the microarray data, showed no significant difference in tenascin-C expression (Fig. 4B&4C). Tenascin-C has been reported to be overexpressed in the extracellular matrix of the stroma in many solid tumors, including breast tumors [49,50]. Additionally, expression of tenascin-C in DCIS has been demonstrated to predict invasion, and high expression has been related to poor prognosis, as well as local and distant reoccurrence in breast cancer patients. [51-55]. Interestingly, both the distribution and quantity of tenascin-C changes in the breast during the menstrual cycle [47], which may explain the variations in tenascin staining in normal tissue, as well as hormone-dependent and independent tumors. Although intralobular stroma was reported to undergo cyclic changes during the menstrual cycle [56], there was no measurable difference in protein levels of the estrogen and progesterone receptor status within the two types of stroma (data not shown). Similar to the results seen with the collagen deposition, this is another example of gene expression levels that do not reflect the abundance of the protein between the two types of stroma.Atherton et al. (1994) have reported a unique regulation of ectoenzymes between the intralobular and interlobular stroma [26]. In normal breast tissue, aminopeptidase N (CD13) was reported to be uniformly expressed in all stroma, while dipeptidyl peptidase IV (CD26) was absent in the intralobular stroma, but present in the interlobular stroma. The two subpopulations of stromal cells were isolated by enzymatic digestion and cell culture, and then analyzed via flow cytometry and immunohistochemistry. Interestingly, after several passages on tissue plastic culture dishes, the intralobular stroma lost their expression of CD26 and became phenotypically similar to the interlobular fibroblasts. This suggests that growth on tissue culture plastic causes a reversion of the stroma subpopulations to one phenotype. We subjected the patient samples from the microarray data to both immunohistochemistry and RT-PCR for both CD13 and CD26. In contrast to Atherton's report, immunohistochemical analysis of CD13 illustrated a predominately intralobular stroma staining in all patient samples tested (Fig. 4D). However, similar to the microarray data, RT-PCR analysis showed inconsistent expression levels between patients (Fig 4B&4C). Of the four patients used in the microarray analysis, two samples had higher CD13 expression in the intralobular stroma, while the other two had higher expression in the interlobular stroma. A larger sample size is necessary to determine whether the RT-PCR results were significant. Additionally, the intensity of the immunoreactivity may be attributed to the density of the stromal cells between the stroma subtypes, and the overall stromal density of each patient may influence the immunohistochemical analysis and a larger sample size is required for absolute conclusion.As with CD13, CD26 demonstrated inconsistent staining between samples, without specific staining to the intralobular or interlobular stroma (Fig. 4D). In some patient samples, CD26 demonstrated a slightly greater deposition in interlobular stroma than intralobular stroma, while in other samples CD26 was ubiquitously expressed throughout all stroma. The RT-PCR results reflected the inconsistency of CD26 protein expression, and similar to the microarray data, quantitation of the samples resulted in no significant difference in expression (Fig. 4B&4C).ConclusionRecently it was reported that the gene expression signatures of cancer-adjacent and breast reduction-normal tissues were essentially homogeneous and not distinguishable [57]. The stroma used in this microarray study was exclusively interlobular stroma, and specifically excluded any intralobular stroma. The authors state this was the most complete study to date of gene expression in normal breast tissue, and that normal tissue adjacent to breast carcinomas has not undergone significant gene expression changes. However, the present study highlights the importance of post-transcriptional or post-translational regulation of proteins. Since surgery is a common procedure performed on tissue with potential for tumor progression, the alterations in the adjacent stroma could have important clinical implications. This study emphasizes the importance of using techniques other than gene expression levels to investigate protein regulation within the stroma.A recent report utilizing two-dimensional gel electrophoresis supports the present study and shows that carcinoma-associated fibroblasts, tumor-adjacent fibroblasts (cells 2 cm away from the tumor margin), and normal breast fibroblasts have different proteome profiles, with many different proteins differentially expressed among these cells [58]. Interestingly, the carcinoma-associated fibroblasts and tumor-adjacent fibroblasts expressed high levels of the cancer marker survivin and consequently exhibited high resistance to the chemotherapeutic agent cisplatin and UV light. Furthermore, the tumor-adjacent fibroblasts, although histologically normal and not in contact with the tumor cells, contained genetic changes that were distinct from the normal fibroblasts and the carcinoma-associated fibroblasts. It was hypothesized that the carcinoma-associated fibroblasts, as well as their corresponding tumor-adjacent fibroblasts, acquired tumor-like changes that are necessary for tumor growth. The authors further speculated that certain genes are up-regulated early during carcinogenesis and have a promoting role during cancer development. It would be of interest to investigate, based on the data obtained from the present study, whether the carcinoma-associated fibroblasts and tumor-adjacent fibroblasts arise from the intralobular or the interlobular stroma, and what effects tumorigenic changes in either subtype have on each other and the progression of the cancer.The failure to grow normal or premalignant human mammary epithelial cells in vivo had previously hindered any possibility of a model for human breast cancer progression using human cells. Recently, Kuperwasser et al. [12] successfully developed a dynamic in vivo model which recapitulates human breast epithelial morphogenesis. In this model, human mammary fibroblasts are injected into the gland and allowed to grown into the gland and \"humanize\" the mouse fat pad, prior to injection of the epithelial cells. This model demonstrated that stroma promoted the normal or premalignant to malignant growth of the epithelial cells, depending on the type of fibroblasts used. It may be informative to observe the differences in the normal outgrowth or tumorigenesis of epithelial cells when either intralobular or interlobular fibroblasts are chosen to humanize the gland. Furthermore, future studies isolating the differences between these two stromal subtypes may bring further insight into the tumor/stroma environment as well as normal mammary development.AbbreviationsLCM: Laser capture microscopy; RT-PCR: reverse transcription polymerase chain reaction; HGF/SF: hepatocyte growth factor/scatter factor; SOS2: Son of Sevenless; CD13: aminopeptidase N; CD26: dipeptidyl peptidase IVCompeting interestsThe authors declare that they have no competing interests.Authors' contributionsJF performed immunohistochemistry, RT-PCR, and drafted the manuscript. LL performed the analysis on the microarray data. EG participated in the design of the study, collection and distribution of patient samples, and helped edit the manuscript. DG prepared the samples for laser capture microscopy, PM and BKV designed the experiments and helped interpret the results. All authors read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2529301.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529301",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529301\nAUTHORS: Bianca Dumitrescu, Svenjhalmar van Helden, Rene ten Broeke, Arie Nieuwenhuijzen-Kruseman, Caroline Wyers, Gabriela Udrea, Sjef van der Linden, Piet Geusens\n\nABSTRACT:\nThe aetiology of osteoporotic fractures is multifactorial, but little is known about the way to evaluate patients with a recent clinical fracture for the presence of secondary osteoporosis.The purpose of this study was to determine the prevalence of contributors to secondary osteoporosis in patients presenting with a clinical vertebral or non-vertebral fracture. Identifying and correcting these contributors will enhance treatment effect aimed at reducing the risk of subsequent fractures.In a multidisciplinary approach, including evaluation of bone and fall-related risk factors, 100 consecutive women (n = 73) and men (n = 27) older than 50 years presenting with a clinical vertebral or non-vertebral fracture and having osteoporosis (T-score ≤-2.5) were further evaluated clinically and by laboratory testing for the presence of contributors to secondary osteoporosis.In 27 patients, 34 contributors were previously known, in 50 patients 52 new contributors were diagnosed (mainly vitamin D deficiency in 42) and 14 needed further exploration because of laboratory abnormalities (mainly abnormal thyroid stimulating hormone in 9). The 57 patients with contributors were older (71 vs. 64 yrs, p < 0.01), had more vertebral deformities (67% vs. 44%, p < 0.05) and a higher calculated absolute 10-year risk for major (16.5 vs. 9.9%, p < 0.01) and for hip fracture (6.9 vs. 2.4%, p < 0.01) than patients without contributors. The presence of contributors was similar between women and men and between patients with fractures associated with a low or high-energy trauma.We conclude that more than one in two patients presenting with a clinical vertebral or non-vertebral fracture and BMD-osteoporosis have secondary contributors to osteoporosis, most of which were correctable. Identifying and correcting these associated disorders will enhance treatment effect aimed at reducing the risk of subsequent fractures in patients older than 50 years.\n\nBODY:\nBackgroundClinical vertebral and non-vertebral fractures are the most frequent fractures in patients presenting to the emergency ward of the hospital with a fracture [1]. After such fracture, patients are at increased risk for subsequent fracture and guidelines on osteoporosis advocate to evaluate patients presenting with a fracture in order to consider treament to reduce the risk of subsequent fractures [2].One aspect of care identified within the management of fracture patients is the existence of contributors to secondary causes of bone loss [3]. Effective therapy requires that these contributors be recognised and when present managed appropriately [3,4]. If these conditions, however, are not recognized, treatment may be suboptimal or ineffective [5,6].Apart from bone mineral density (BMD)-osteoporosis (T-score less than or equal to -2.5) [4,7,8] many risk factors are related to fracture risk, independently of BMD, such as clinical risk factors [9], fall risks [10], prevalent morphometric vertebral fractures (MVF) [11] and secondary osteoporosis [12]. There is increasing evidence that secondary osteoporosis is more prevalent than initially thought, not only in males, but also in females [13], but the true prevalence of contributors to secondary osteoporosis is unknown and no consensus regarding its evaluation is available [14].Published data from referral centres for evaluation of osteoporosis indicate that 32 to 37% of women with low BMD have a history of other diseases or medications known to contribute to osteoporosis [3,15]. From 20% up to 64% of patients had previously unknown secondary causes of osteoporosis that were only identified by laboratory testing [5]. In a study of patients with a clinical fracture, a high prevalence of contributors to secondary osteoporosis (77%) was reported, but the study included only a limited number of patients with measured low BMD [16]. In a study of patients with a hip fracture, 80% had secondary causes of bone loss, mainly related to disturbed calcium and vitamin D homeostasis [6]. To date, we lack studies on the prevalence of contributors to secondary osteoporosis in other fracture populations.The purpose of this study was to determine the prevalence of contributors to secondary osteoporosis, in the context of other bone- and fall-related fracture risks in patients presenting with a clinical vertebral or non-vertebral fracture and with a low BMD. Identifying and correcting contributors will enhance treatment effect aimed at reducing the risk of subsequent fractures.MethodsIn this prospective observational study, 100 consecutive and consenting patients older than 50 years, who presented between April 2005 and April 2006 with a clinical fracture at Maastricht University Hospital in the Netherlands, were included. After receiving medical treatment for the fracture, patients had a consultation with the fracture nurse. The fracture nurse provided information about the study and invited the patients to the Fracture and Osteoporosis Outpatient Clinic. Patients already on osteoporosis treatment (44/1246, 4% of all) or with pathological fractures due to malignancy or Paget's disease of bone were excluded from the analysis. Patients who agreed to participate were further referred to the program. Patients with osteoporosis according to World Health Organization (WHO) criteria for BMD [4] and in whom all laboratory data were available were included in the present study (Figure 1). This group was part of the evaluation of all consecutive patients presenting with a clinical fracture, of whom 35% had osteoporosis and 44% had osteopenia [1]. The medical ethical committee of the University Hospital Maastricht approved the study (MEC 03-194-5).Figure 1Flow chart of patients included in the study (see text for details) in one year.BMD in the left or right hip and the lumbar spine was determined using dual X-ray absorptiometry (DXA) with Hologic QDR 4500 Elite. Diagnosis of osteoporosis was based on the WHO criteria for BMD [4], as provided by the manufacturer for women and men. Patients were classified according to the lowest value of T-score in either total hip or spine.All patients were interviewed for bone-related risk factors for fracture (previous non-vertebral and vertebral fractures, mother with fracture, body weight <60 kg, severe immobility, use of glucocorticoids) and fall-risk factors (falls in the past 12 months, use of assistive devices, sedative medication, activities of daily living, mobility, impaired vision, articular complaints, urine incontinency), according to the Dutch guidelines on osteoporosis [2] and fall prevention [10]. Additionally, data about vitamin D status (regular sun exposure, dietary intake and supplements), calcium intake [17,18], height, history of height loss [19] and a description of the circumstances leading to the fracture (with specification of fall from a standing height or other trauma) were recorded.Patients with T-scores ≤-2.5 were given a pre-planned set of laboratory tests that included erythrocyte sedimentation rate (ESR), haemoglobin, leucocytes and serum levels of creatinine, calcium, albumin, alkaline phosphatase, 25-OHD3 and TSH. Calcium and creatinine were measured in a 24-hour urine collection. All laboratory analyses were performed in the same laboratory. Patients with osteoporosis and having the full set of evaluation were sent for a consultation with either a rheumatologist or an endocrinologist. The specialist decided further investigation and treatment. When clinically appropriate, additional tests were performed. The diagnosis of contributors to secondary osteoporosis was based on all data from the medical files. Renal insufficiency was defined with the cut-off value of creatinine clearance ≤40 using the Cockroft Gault formula [20]. Vitamin D status was defined as severely deficient when values were ≤30 nmol/L, deficient when between 30 and ≤50 nmol/L, insufficient when values were between 50 and ≤75 nmol/L [21,22] and abnormally high when above 220 nmol/L [23]. Exploration for hypogonadism in men was considered when a morning serum testosterone level was below 12 nmol/L [24] and for thyroid disorders when TSH were outside the reference ranges (0.4–3.5 mU/L). Hyperparathyroidism was diagnosed when serum parathyroid hormone (PTH) levels were above 5.5 pmol/l. Further exploration for hypercalciuria was considered when the total urinary calcium in a 24 hours collection exceeded 7 mmol/d and creatinuria indicated an appropriate collection (between 4.5 – 13.3 mmol/hour) [25]. According to clinical judgement, patients suspected to have lactose intolerance had a lactose tolerance test [26].Vertebral fracture assessment (VFA) [27,28] by single X-ray absorptiometry on the lateral spine images was performed to identify the presence of morphometric vertebral deformities (MVD). Images were saved in a digital format. Physician Viewer software (Hologic, USA) provided the tools necessary to perform quantitative vertebral morphometry. Visual assessment and measurements of the anterior, posterior and mid heights from T4 to L4 were performed twice, by a trained rheumatologist (BD). These assessments were inputted into a database. The observer was blinded from the results of the first measurements. The intra-observer coefficient of variation (ICC) at vertebral level for heights was 0.917 (95% confidence interval (CI): 0.905–0.930, Cronbach's alpha 0.959). The arithmetic mean heights of the two measurements were used for calculation. The anterior-posterior ratio, the middle-posterior ratio, the posterior-posterior ratio were calculated. Prevalent morphometric vertebral deformities (MVD) were defined according to the Genant grading [29]. Vertebral deformities were classified into three types (wedge, biconcavity, crush) and three grades (mild (any ratio <20%), moderate (any ratio between 25–40%), and severe (any ratio >40%)).The WHO Fracture risk assessment tool (FRAX) was used to calculate the absolute 10-year risk for major and for hip fractures in women and men [12].Statistical analyses were performed using SPSS version 12.01. Categorical variables and proportions were analyzed using chi-square statistic. Odds ratio (OR) with 95% confidence intervals (95% CI) were calculated based on the chi squares. One way Anova and chi-square statistics were used to analyze differences in continuous variables between subgroups. Observations were considered significant if two-sided p-values were < 0.05.ResultsOf the 100 patients, 73 were women and were 27 men. Mean age was 68 years (standard deviation: 10 years, range: 50 to 90 years). Demographic data are summarized in Table 1. The majority of patients were Caucasian (97%). Fractures were found at the following locations: clinical vertebral fractures (n = 4), clavicle (n = 3), pelvis (n = 2), humerus (n = 10), radius and/or ulna (n = 24), hand (n = 6), hip (n = 17), tibia/fibula/patella (n = 8) and foot (n = 21). Five patients had multiple simultaneous fractures and 80 patients had fractures after a fall from standing height.Table 1Characteristics of the patient population (N = 100)Variable Median+/-SDAll patientsWomenMenFragility fractureHigh energy traumaWith contributorsWithout contributorsNumber100732780205743Women/men (n)73/27nana66/147/13**40/1733/10***Caucasian ethnicity (n)97942779185542Age (years)68 ± 1070 ± 962 ± 8*69,1 ± 963.3 ± 9**71 ± 1064 ± 7***Weight (kg)66 ± 1363 ± 1373 ± 11*65,4 ± 1468.8 ± 1164 ± 1469 ± 11Spine T-score-2.88 ± .91-2.94 ± 0.79-2.73 ± 1.17-2.9 ± 0.92-2.8 ± 0.83-2.85 ± 0.97-2.93 ± 0.82Hip T-score-1.92 ± 0.9-2.13 ± 0.92-1.37 ± 0.70*-2 ± 0.87-1.6 ± 1.09-2.1 ± 1.00-0.66 ± 0.63***Hip Z-score-1.92 ± 0.8-0.57 ± 0.85-0.89 ± 0.73-0.62 ± 0.85-0.83 ± 0.72-0.66 ± 0.97-0.66 ± 0.63BMD spine(g/sq cm)0.772+/- 0.1000.756 ± 0.0850.810 ± 0.130*0.765 ± 0.1030.795 ± 0.0940.776 ± 0.1090.766 ± 0.091BMD hip(g/sq cm)0.718+/- 0.3950.676 ± 0.1150.825 ± 0.107*0.700 ± 0.1160.779 ± 0.160**0.695 ± 0.1430.749 ± 0.106Calcium intake(mg/day)852 ± 432828 ± 448915 ± 389851 ± 467854 ± 266744 ± 343993 ± 497***Serum 25OH vitamin D (nmol/L)66 ± 5367 ± 6063 ± 2863 ± 5775 ± 30420Creatinine clearance (ml/min)67 ± 2362 ± 2065 ± 2365 ± 2374 ± 2645 ± 1882 ± 24***Fracture after fall from standing height (n)806614nana4535N of contributors (n)8662247016nanaN with contributors (n)5740174512nanaN with bone-related fracture risks54421244103321N with fall-related fracture risks79572265144633Time Go Up and Go (min)8.6 ± 7.98.4 ± 8.08.9 ± 8.08.2 ± 8.09.8+/-8.08.1+/-8.69.2+/-7.1N with MVD <0.80 (n/n measured)53/9335/6618/2742/7311/2036/5417/39***N with MVD <0.752922724111910*p < 0.05 women vs. men, ** p < 0.05 between fragility fracture and high-energy trauma, *** p < 0.05 between group with and without contributor, Na: not applicableA total of 86 contributors to secondary osteoporosis were diagnosed in 57 patients (Table 1 and 2). Contributors consisted of known medical conditions (34 in 27 patients) or newly diagnosed (52 in 50 patients). Seven patients had only known contributors, 20 had known plus a newly diagnosed contributor and 30 had only newly diagnosed contributors. One contributor was found in 32 (of whom 24 were vitamin D deficient), more than one contributor in 25 and 43 had none.Table 2Contributors to secondary osteoporosis identified in men and women >50 years with a recent clinical fracture (N = 100)ContributorsTotalKnownNewly diagnosedFragility Fracture (N = 80)High-energy trauma (N = 20)Endocrine disorders Serum 25-OHD3 s ≤50 nmol/l42042375 Hyperparathyroidism secondary to low calcium intake20211 Hyperthyroidism33030 Hypogonadism (in men)11001 Anorexia nervosa (in women)22020 Diabetes mellitus55041Gastrointestinal disorders Lactose intolerance10101Connective tissue disorders Rheumatoid arthritis22011 Giant-cell arteritis11010Renal disordersRenal insufficiency without secondary hyperparathyroidism115674Renal insufficiency with secondary hyperparathyroidism33030Miscellaneous Severe immobility33030 Pulmonary diseases55041Medication and life style Exogenous hyperthyroidism10110 Alcohol abuse44022Total number of contributors (N = 86)8634526917Total number of patients with contributors to osteoporosis (N = 57)5727504512Based on serum levels of 25-OHD3, 11 patients had severe deficiency, 31 were deficient and 31 had insufficient serum values. All were newly diagnosed. Serum levels of 25-OHD3 could not be predicted by any of questions on vitamin D or by the sum of those questions. Calcium intake below 1200 mg was reported in 86 patients. Only three patients had both a calcium intake above 1200 mg and a serum 25-OHD3 level above 75 nmol/L (Figure 2).Figure 2Calcium intake and serum serum levels of 25OHD3. Only 3 patients had sufficient intake of calcium and normal serum levels of 25-OHD3.Five patients had secondary hyperparathyroidism, of which four were newly diagnosed (Table 2). Hyperparathyroidism was secondary to renal insufficiency in three cases and to low calcium intake in two cases. We found 14 patients with renal insufficiency, of which 6 were newly diagnosed. Three patients were known with hyperthyroidism, 1 new case of exogenous hyperthyroidism and 1 new case of hypothyroidism was detected. One new case of lactose intolerance was diagnosed. Further contributors included anorexia nervosa in 2 women, documented hypogonadism in one men, pulmonary diseases in 5 patients (chronic obstructive lung disease and asthma), alcohol abuse in 4 men, inflammatory rheumatic diseases in 3 patients (2 with rheumatoid arthritis and 1 with giant cell arteritis) and 3 with severe immobility. Most of these patients did not receive preventive measures for osteoporosis prior to the fracture and were thus not recognized as having a contributor to secondary osteoporosis before the fracture occurred.Other laboratory abnormalities that required further exploration were found in 14 patients (18 abnormalities in total), including exogenous hypervitaminosis D (n = 1), hypercalciuria (n = 3), TSH outside normal ranges (n = 13) and low serum testosterone in one men (Table 3). Among the 9 patients being treated for hypothyroidism, one was over-treated while three were under-treated based on abnormal serum TSH levels. Among the 3 patients being treated for hyperthyroidism, two were under-treated while one was over-treated.Table 3Laboratory abnormalities that required further exploration in men and women more than 50 years of age with a recent clinical fracture (N = 100)Laboratory abnormalityTotalExogenous hypervitaminosis D (>220 nmol/l)1Hypercalciuria in 24 hours urine3TSH 0.4–3.5 mU/L13 - >3.5 mU/L10 - treated hypothyroidism9  -TSH <0.4 mU/L1  -TSH >3.5 mU/L3 - treated hyperthyroidism3  -TSH >3.5 mU/L1 -TSH <0.4 mU/L2Serum testosterone in men <12 nmol/L (one measurement)1Total number of patients14According to the Dutch guideline for osteoporosis 54 patients had clinical bone-related risk factors for fractures in addition to their current fracture (Table 4). A history of an additional clinical fracture after the age of 50 was present in 31 patients (two with a previous clinical spine fracture). Additionally, 12 had a mother that had suffered one or more fractures, 3 were severely immobilised, and 23 had a low body weight (60 ≤kg). One bone-related risk factor was found in 41 patients, 2 in 12 and 3 in one patient. According to the Dutch guideline for fall prevention, we found fall related risk factors in 79 of the patients: 22 patients had one risk factor, 21 had two risk factors, and 36 had more than two risk factors. An overlap between clinical bone related risk factors and fall related risk factors was present in 45 patients. The prevalence of clinical bone-related and fall-related risk factors was similar between patients that had documented contributors secondary osteoporosis and those who did not (50% versus 59% for clinical bone related risk factors and 79% versus 84% for fall-related risk factors for fractures).Table 4Clinical risks for fractures recorded in patients with a recent clinical fracture according to the Dutch guidelinesCLINICAL RISKS OF FRACTURETotalWomenMenFragility fractureHigh traumaContributorsNo contributorsNumbers of patients100732780205743BONE RELATED RISK FACTORS54421244103321History of clinical fracture after 50 years312382561912History of clinical vertebral fracture2101010Mother with fracture129310284Low body weight (<60 kg)23203185158Severe immobility3303030Glucocorticosteroids user0000000FALL RELATED RISK FACTORS79572265144633Mobility: Time Get up and Go test242132041212Previous falls: 2 or more falls in the previous year272252341512Medication use (benzodiazepines, antiepileptics)16142142133Low activities of daily living4938114316*34**15Osteoarthritis48408444*2523Snellen score-visual acuity less than 0.41688142133Urinary incontinence19172190136*p < 0.05 between fragility fracture and high-energy traum fracture, **p < 0.05 between groups with and without contributorsVFA could be performed for 93 patients. Lateral spine images were not available in 7 cases due to severe scoliosis or other technical difficulties such as positioning patients with humerus fracture on the DXA table. On VFA, 57% of patients had a MVD, 31% had more than one MVD and 31% had moderate and severe MVD.The 57 patients with contributors to secondary osteoporosis were older (71 versus 64 yrs, p < 0.01) (Table 1). They had more of some fall risks (multi-medication use (13 versus 3, p < 0.05), restricted activities of daily living (34 versus 15, p < 0.05) and disturbed vision (13 versus 3, p < 0.05) (Table 4)). They had lower calcium intake (744 versus 993 mg, p < 0.05) (Table 1) and more MVD (67% versus 44%, p < 0.05, OR: = 2.6, 95% CI: = 1.1–6.0) (Figure 3).Figure 3Prevalence of MVD defined according to the grading of Genant et al. in patients with contributors to secondary osteoporosis and in patients without contributors.In contrast, the proportions of women (70 versus 77%) and of patients with fragility fractures (79 versus 81%) were similar between patients with and without contributors. There were also no differences in the prevalence of bone-related clinical risks (59 versus 50%).Based on the FRAX tool, patients with contributors had a higher calculated absolute 10-year risk for major (16.5 vs. 9.9%, p < 0.01) and for hip fractures (6.9 vs. 2.4%, p < 0.01).Compared to patients with a high-energy trauma, patients with fragility fractures were older (69 versus 63 years), had better activities of daily living (43 versus 16 patients), and more osteoarthritis (44 versus 4 patients) (Table 1 and 3).DiscussionIn the 100 patients older than 50 years presenting with a recent clinical fracture and osteoporosis and referred by the surgeons to the rheumatologists in collaboration with the endocrinologists the prevalence of contributors to secondary osteoporosis was high: almost two out of three patients had one or more contributors most of which were correctable.Our results show that many patients (27%) had known contributors to secondary osteoporosis, a percentage similar to that of Tannenbaum et al. in women with postmenopausal osteoporosis who were seen in an osteoporosis referral centre (32%) [3]. The categories of known contributors to secondary osteoporosis were globally similar as reported by Tannennbaum [3] (endocrine, gastrointestinal and inflammatory rheumatic and pulmonary diseases, severe immobility, alcohol abuse). One exception was glucocorticoid users who are presumably frequently referred to an osteoporosis clinic, but were not represented in our group of patients. In contrast to Tannenbaum, we performed the laboratory test set also in patients with already known contributors and 20 additional contributors were diagnosed in 20 patients (mainly low 25-OHD3, n = 14). Presumably none of the patients with known contributors had received attention in the context of osteoporosis, as none had osteoporosis treatment or calcium and vitamin D supplements.In the other, presumably healthy patients without known contributors, laboratory testing identified newly diagnosed contributors to secondary osteoporosis in 30 more patients, mostly vitamin D deficiency and renal disorders. The number of patients with newly diagnosed contributors (50%) was higher than reported by Tannenbaum et al. (33%) and concerned mainly vitamin D deficiency, secondary hyperparathyroidism (to renal insufficiency and to low calcium intake), malabsorption and exogenous hyperthyroidism [3]. In contrast to Tannenbaum, we performed TSH not only in patients with a history of thyroid diseases, but in all patients, and were able to diagnose one new case of hypothyroidism.Vitamin D status could not be identified by history, despite including four specific questions regarding vitamin D intake. It has been shown that there is only a modest relation between reported vitamin D intake from an extensive dietary questionnaire and serum levels of 25-OHD3 [30]. In our study a wide spectrum of levels of serum 25-OHD3 were found, from severely deficient to normal. There is still no consensus about how much vitamin D supplements are required to normalise serum levels. Some propose a unique dose of 800 IU/day together with 1000–1200 mg calcium/day to achieve 50 nmol/L [31,32]. Others state that a unique dose of 800–1600 IU/day would normalize serum levels to >75 nmol/L [33,21]. As patients with low serum levels of 25-OHD3 require, at least temporarily, high doses of vitamin D supplements while those with normal levels require less or none [34], measuring serum 25-OHD3 levels is helpful in patients with osteoporosis in order to decide about appropriate vitamin D supplementation [3].The calcium homeostasis was further compromised by the low calcium intake (<1200 mg/day) in most patients, resulting in secondary hyperparathyroidism in 2, and only 3% of the patients had both adequate calcium intake and vitamin D status. Correcting these combined deficiencies has been demonstrated to reduce fracture risk, at least in institutionalized elderly women [35] and to reduce the risk of falls [36]. Calcium and vitamin D supplementation are thus needed in most patients presenting with a fracture and osteoporosis. However, supplementation with calcium and vitamin D alone is an insufficient measure in patients with osteoporosis, as drug therapy for osteoporosis has been shown to reduce the risk of fractures on top of correcting such deficiencies. Our data, together with those of Edwards et al. [6] indicate that calcium and vitamin D deficiency is frequently present in patients presenting with a fracture, and that these deficiencies need to be identified and corrected.Interestingly, the presence of contributors was similar between women and men, and between patients with fractures associated with low or high-energy trauma, suggesting that evaluation for secondary contributors is indicated in women and men and after low or high-energetic trauma.An additional 14 patients had laboratory abnormalities that required further investigation, mainly hypercalciuria, uncontrolled treatment of thyroid disorders and low testosterone (in one man). Hyperthyroidism, whether endogenous or exogenous, can increases bone turnover and contributes to secondary osteoporosis [37,38]. Hypothyroidism on the other hand increases the risk of fractures through low bone turnover if untreated or high bone turnover if over treated [39]. Thus fine-tuning thyroid treatment is indicated. The same is probably true for patients with hypercalciuria in whom thiazides are indicated [40], and for hypogonadism in men that can be treated with testosterone supplementation [41], although no fracture prevention data are available in these conditions.Therefore, measuring serum 25-OHD3, calcium in 24 hours urine, serum creatinine, TSH, PTH as proposed by Tannenbaum et al. is indicated in patients with osteoporosis and a recent clinical fracture, and enabled us to identify 47 (96%) newly diagnosed contributors and 13 of the 14 laboratory abnormalities [3]. As many patients had endocrine diseases, collaboration with endocrinologists appeared to be highly valuable for diagnosis and treatment.The prevalence of clinical bone-related fracture risks in postmenopausal women, as evaluated by the Dutch guidelines, was similar between patients with and without documented contributors to and it contributed to further specify the risk for fractures.Nearly 80% of patients had fall-related risk factors for fractures, as reported by others [16]. Although it has not been shown until now that fall prevention strategies itself can prevent fractures, they reduce the risk of falls. [42] A multidisciplinary, multifactorial intervention program reduces postoperative falls and injuries after femoral neck fracture and are therefore applied in our ongoing prevention program [43].An interesting finding was the prevalence of MVD which was more than twice as high among patients with documented contributors for secondary osteoporosis compared to those without contributors, in spite of similar low BMD in both groups. MVDs, that are related to future fracture risk independent of BMD [11], reflect bone failure independently of BMD and thus indicate other mechanisms of bone's decreased resistance to fracture than low BMD, such as changes in the bone turnover, alterations in micro architecture of bone and deficient mineralization, especially in the context of the high prevalence of calcium and vitamin D deficiency.In several recent publications differential diagnosis and search for contributors to secondary osteoporosis is advocated [44,45]. Only limited data are available about collaboration between surgeons and internists in taking care for osteoporosis in patients presenting with a fracture. Some initiatives were very successful [46], but in most instances the collaboration is failing [47]. This study indicates that such collaborations add to better treatment of patients with a clinical fracture.This study has several limitations. Smoking history, which is part of the WHO FRAX tool, was not recorded as it is not part of the Dutch guideline. The sample size was relatively small, but the strength of the study was that consecutive patients were evaluated showing that even in a small group many contributors to secondary osteoporosis could be diagnosed. Some laboratory abnormalities needed further exploration, but were not followed up and so no definite diagnosis could be reported in these patients. VFA has several limitations. Not all vertebrae could be measured, mainly at the upper thoracic level. Identifying patients with MVD by VFA requires additional X-rays to differentiate deformities due to other conditions, such as Scheuerman's disease, degenerative changes or non-osteoporotic short vertebral height. However, the method has a high negative predictive value in predicting the absence of vertebral fractures on X-rays [27]. Another limitation is that only patients with BMD-osteoporosis were included. Most patients with a fracture have no BMD-osteoporosis. The results of our study suggest that documentation of the prevalence of contributors to secondary osteoporosis should also be studied in patients with a clinical fracture without BMD osteoporosis.ConclusionWe conclude that more than one in two patients presenting with a clinical vertebral or non-vertebral fracture and BMD-osteoporosis have secondary contributors to osteoporosis, most of which were correctable. Identifying and correcting these associated disorders will enhance treatment effect aimed at reducing the risk of subsequent fractures in patients older than 50 years.AbbreviationsBD: Bianca Dumitrescu; BMD: Bone mineral density; CI: confidence interval; DXA: Dual X-Ray absortiometry; ESR: Erythrocyte sedimentation rate; EULAR: European League against Rheumatism; FRAX: Fracture risk assessment tool; MEC: Medical Ethical Committee; MVD: Morphometric vertebral deformity; MVF: Morphometric vertebral fracture; OR: Odds ratio; PTH: Parathormone; TSH: Thyroid stimulating hormone; VFA: Vertebral fracture assessment; WHO: World Health Initiative.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsBD analyzed clinical and laboratory data for the diagnosis of contributors to secondary osteoporosis, performed vertebral fracture assessment, statistical analyses and wrote the manuscript. SvH implicated in the coordination of the study, involved in the treatment of patients included in the study, participated to sequence alignment and data presentation. RtB involved in the coordination of the study, involved in the treatment of patients included in the study. AN–K analyzed clinical and laboratory data for the diagnosis of contributors to secondary osteoporosis, coordinated data presentation. CW gathered laboratory and clinical data, performed statistical analysis. GU participated in the sequence alignment. SvdL analyzed clinical and laboratory data for the diagnose of contributors to secondary osteoporosis, coordinated data presentation. PG conceived the study, participated in the design of the study, coordinated the study, analyzed the data for correct diagnosis and drafted the manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2529308.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529308",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529308\nAUTHORS: Gyaneshwer Chaubey, Monika Karmin, Ene Metspalu, Mait Metspalu, Deepa Selvi-Rani, Vijay Kumar Singh, Jüri Parik, Anu Solnik, B Prathap Naidu, Ajay Kumar, Niharika Adarsh, Chandana Basu Mallick, Bhargav Trivedi, Swami Prakash, Ramesh Reddy, Parul Shukla, Sanjana Bhagat, Swati Verma, Samiksha Vasnik, Imran Khan, Anshu Barwa, Dipti Sahoo, Archana Sharma, Mamoon Rashid, Vishal Chandra, Alla G Reddy, Antonio Torroni, Robert A Foley, Kumarasamy Thangaraj, Lalji Singh, Toomas Kivisild, Richard Villems\n\nABSTRACT:\nBackgroundHuman genetic diversity observed in Indian subcontinent is second only to that of Africa. This implies an early settlement and demographic growth soon after the first 'Out-of-Africa' dispersal of anatomically modern humans in Late Pleistocene. In contrast to this perspective, linguistic diversity in India has been thought to derive from more recent population movements and episodes of contact. With the exception of Dravidian, which origin and relatedness to other language phyla is obscure, all the language families in India can be linked to language families spoken in different regions of Eurasia. Mitochondrial DNA and Y chromosome evidence has supported largely local evolution of the genetic lineages of the majority of Dravidian and Indo-European speaking populations, but there is no consensus yet on the question of whether the Munda (Austro-Asiatic) speaking populations originated in India or derive from a relatively recent migration from further East.ResultsHere, we report the analysis of 35 novel complete mtDNA sequences from India which refine the structure of Indian-specific varieties of haplogroup R. Detailed analysis of haplogroup R7, coupled with a survey of ~12,000 mtDNAs from caste and tribal groups over the entire Indian subcontinent, reveals that one of its more recently derived branches (R7a1), is particularly frequent among Munda-speaking tribal groups. This branch is nested within diverse R7 lineages found among Dravidian and Indo-European speakers of India. We have inferred from this that a subset of Munda-speaking groups have acquired R7 relatively recently. Furthermore, we find that the distribution of R7a1 within the Munda-speakers is largely restricted to one of the sub-branches (Kherwari) of northern Munda languages. This evidence does not support the hypothesis that the Austro-Asiatic speakers are the primary source of the R7 variation. Statistical analyses suggest a significant correlation between genetic variation and geography, rather than between genes and languages.ConclusionOur high-resolution phylogeographic study, involving diverse linguistic groups in India, suggests that the high frequency of mtDNA haplogroup R7 among Munda speaking populations of India can be explained best by gene flow from linguistically different populations of Indian subcontinent. The conclusion is based on the observation that among Indo-Europeans, and particularly in Dravidians, the haplogroup is, despite its lower frequency, phylogenetically more divergent, while among the Munda speakers only one sub-clade of R7, i.e. R7a1, can be observed. It is noteworthy that though R7 is autochthonous to India, and arises from the root of hg R, its distribution and phylogeography in India is not uniform. This suggests the more ancient establishment of an autochthonous matrilineal genetic structure, and that isolation in the Pleistocene, lineage loss through drift, and endogamy of prehistoric and historic groups have greatly inhibited genetic homogenization and geographical uniformity.\n\nBODY:\nBackgroundMore than one sixth of humanity currently lives on the Indian subcontinent. This population is spread across up to 40,000 endogamous and semi-endogamous culturally, linguistically, and socially differentiated groups [1]. The majority of these groups or populations are castes, but they also include nearly 500 'scheduled tribes' [2] and ca. 500 'scheduled castes' [3]. Thus, the Indian subcontinent is an ideal region for studying the relationships between culture, geography and genes, and for developing interdisciplinary models concerning the demographic history of Homo sapiens or anatomically modern humans (AMH). Moreover, the large number of deep-rooting mtDNA lineages emerging from the basal nodes of both superhaplogroup M and N (including R) [4-11] indicate that the Indian subcontinent was probably the first major outcome of the dispersals of AMH from Africa. Furthermore, these deep-rooted mtDNA haplogroups generally cross cultural and social boundaries; this suggests a common origin to the highly diverse peoples of the Indian sub-continent, with indigenous or autochthonous diversification of the maternal gene pool [12-16].These results have been generally corroborated by data from the Y chromosome [17,18] and autosomal DNA [13,19,20]. The only exception, for mtDNA, are the Tibeto-Burman speakers of north-eastern India, who share about half of their maternal genetic heritage with populations living further east of India [14,21]. It has been argued, that following the initial colonization of Indian subcontinent, maternal gene flow from the west has been rather limited and largely restricted to the western states of contemporary India and Pakistan [14,15,22]. Consequently, the haplogroup richness of the Indian subcontinent appears to have formed in situ, and date back to some point in the later Pleistocene, most probably between 40 Ka and 60 Ka ago. Furthermore, this high level of genetic diversity may also be linked to the possibility that the South Asian population in the Pleistocene was demographically large in global terms. Comparisons of relative regional population sizes through time, deduced by Bayesian coalescent inference methods applied to global mtDNA complete sequence data, indicate that between approximately 45 Ka and 20 Ka ago most of humanity lived in Southern Asia [23].Two language families, Indo-European and Dravidian, account for the majority of linguistic diversity in India. However, apart from a number of linguistic isolates, there are two other major families – Tibeto-Burman and Austro-Asiatic (AA). The origin of the Austro-Asiatic language family is a highly debated issue. Building on archaeological and linguistic evidence, and the assumption that rice domestication was a single event, the currently preferred hypothesis places the origin of this language family in Southeast Asia [24-26]. The alternative model, based on genetic evidence (that shows multiple domestications of rice varieties [27]), and comparative phonology, advocates an East Indian cradle for the AA language group [28]. The AA language family tree has two basic branches – Munda and Mon-Khmer. The former is distributed exclusively in the Indian subcontinent; the latter is predominantly Southeast Asian, although there are a few Indian representatives (Khasian and Nicobarese) [26].The genetic origin(s) of extant AA speakers, however, may or may not coincide with the origin of the language group. Studies of mtDNA diversity have shown that the AA speakers from Southeast Asia and the Indian subcontinent carry mtDNAs of different sources [7,9,14,16,29,30]. Although the data on Southeast Asian populations, which speak languages of the Mon-Khmer branch of the AA tree, are still somewhat limited, it seems safe to conclude, that their mtDNA characteristics are similar to those of the surrounding Southeast Asian populations, and distinct from AA tribes of India (Munda-speakers) [30]. Similarly, the Indian tribes speaking different Munda languages show generally the same mtDNA haplogroup composition as the Indo European and Dravidic groups of India [9,14,16,29]. In contrast, the Y chromosomes of Indian and Southeast Asian AA speaking populations share a common marker, M95, which defines a single branch (O2a) in the overwhelmingly East Asian specific tree of haplogroup O. This evidence provides a strong basis for proposing a Southeast Asian origin of the paternal lineages of the Munda speaking populations of India [13,17,18].The AA speaking populations of Myanmar, which is a likely dispersal route, or original location, for the ancestral populations of Munda speakers of India, have not yet been sampled for their mtDNA. It is still possible that some of the mtDNA clades present among the AA speakers of India (and in their neighbours) could, in fact, be due to gene flow to India from further east. In an attempt to identify mtDNA lineages that would reveal a phylogeographic distribution similar to that of the Y chromosome marker M95, we analyzed mtDNA samples representing all the major linguistic groups of India, with a particular focus to haplogroup R derived lineages.The first thorough study of complete mtDNA sequences from India [4] identified numerous indigenous clades emerging directly from the roots of superhaplogroups N, R and U, such as N5, R5-R8, R30, R31, U2a-d and U7. West Eurasian specific haplogroups HV, JT, N1, and U (xU2a-d, U7) occur at lower frequencies, suggesting limited but phylogeographically well detectable gene flow into the Indian subcontinent, most probably from west and northwest Eurasia [14]. Here we have now extended the complete mtDNA sequencing by determining 35 new complete sequences, in order to further refine the phylogeny of the Indian subcontinent-specific segment of haplogroup R. Furthermore, to explore the correlations between genes, languages and geography in Indian subcontinent, we have carried out high resolution genotyping and phylogeographic detailed analyses on R7, which occurs at high frequency among the Austro-Asiatic (Munda) speaking groups of India.Results and DiscussionThe inclusion of our 35 novel sequences (Table 1) into the phylogeny of haplogroup R allows the recognition of eight new subclades within six haplogroup R branches unique to the Indian subcontinent (Fig. 1, [see Additional file 1]). We refine here the internal topology of haplogroups R5, R6 and R8, and describe two novel sub-clades of hg R7, to be discussed below in detail. Subclade R5a is defined by a deletion at nucleotide positions (np) 522–523 and one control region mutation at np16266. R6a is defined by two control region substitutions (at sites 16129 and 16266). In haplogroup R7, two new subclades R7a and R7b can be identified (for details see further down). A new subclade of R8, called R8a, is defined by a single coding region substitution at np 5510. Haplogroup R30 splits into two subclades R30a and R30b, the former supported by ten coding region substitutions and the latter by 24 coding and control region mutations. Similarly, in haplogroup R31 a new subclade R31a can be distinguished by 17 control and coding region mutations. Coalescent estimates suggest an ancient branching pattern in hgs R30 and R31, dating back almost to the earliest diversification of the superhaplogroup R itself. This most probably occurred soon after the out of Africa dispersals into the Indian subcontinent [see Additional file 1].Table 1Geographical, Linguistic and Haplogroup Affiliations of Completely Sequenced mtDNAs.Si No.Sample codeHaplogroupPopulationLocationLingustic affiliation1Kol77R5a1KoliGujaratIndo-European2Ben46R5a1aBengalWest BengalIndo-European3Up41R5a1aMiddle casteUttar PradeshIndo-European4Kall43R5a2bKallarTamil NaduDravidian5K35R5a2bKotaTamil NaduDravidian6Ori74R5a2b2OraonOrissaDravidian7Mo38R5a2b3MoorSri LankaDravidian8Gu35R5a2b3GujaratGujaratIndo-European9Pn32R5a2b4PaniyaKeralaDravidian10Mal33R5a2b4MalayanKeralaDravidian11Ko 5R6a1aKoyaAndhra PradeshDravidian12Ko31R6a1aKoyaAndhra PradeshDravidian13Lam43R7a1LambadiAndhra-PradeshDravidian14As426R7a1AsurJharkhandAustro-Asiatic15Mw1R7a1aMawasiChhattisgarhAustro-Asiatic16Tor45R7a1aSindhiPakistanIndo-European17Ho433R7a1b1HoJharkhandAustro-Asiatic18Ori7R7a1b1OraonJharkhandDravidian19Ori37R7b1aOraonOrissaDravidian20A474R7a1b2OraonJharkhandDravidian21G39R7a1b2SanthalBiharAustro-Asiatic22G19R7a1b2KanwarMadhya-PradeshIndo-European23KO18R7bKoyaAndhra-PradeshDravidian24KO55R7b1aKoyaAndhra-PradeshDravidian25G66R7b1aGondMadhya-PradeshDravidian26Ko74R8aKoyaAndhra PradeshDravidian27Lam10R8a1a1LambadiAndhra PradeshDravidian28Ko30R8a1a2KoyaAndhra PradeshDravidian29Ko37R8a1a2KoyaAndhra PradeshDravidian30CoB41R8a1bKonkanastha BrahminMaharashtraIndo-European31CoB23R30Konkanastha BrahminMaharashtraIndo-European32Sin49R30aSinhaleseSri LankaIndo-European33Pun47R30bPunjabPunjabIndo-European34Raj25R31a1RajputRajasthanIndo-European35Raj48R31a1RajputRajasthanIndo-EuropeanFigure 1The most parsimonious tree of haplogroup R7 complete mtDNA sequences observed in the Indian subcontinent. This tree was redrawn manually from the output of median joining/reduced network obtained using NETWORK program (version 4.1) [34]http://www.fluxus-engineering.com. The samples were selected through a preliminary sequence analysis of the control region in order to include the widest possible range of R7 variation, language and geographical groups. Coalescent times were calculated by a calibration method described elsewhere [32]. 16182C, 16183C and 16519 polymorphisms were omitted. Suffixes A, C, G, and T indicate transversions, recurrent mutations are underlined. Synonymous (s) and non-synonymous (ns) mutations are distinguished. DRA-Dravidian, AA-Austro-Asiatic, IE-Indo-European. The ethnic affiliation of the samples is as follows: Lam, Lambadi; As, Asur; Mw, Mawasi; Tor45, Pakistan; Ho, Ho; Ori&A, Oraon; G19, Kanwar; G39, Santhal; G66, Gond; KO, Koya. Two sequences, T35 (Thogataveera) and C35 (Brahmin), were taken from the literature [4].Comparison of patterns of haplogroup distribution in relation to linguistic groups reveals that the frequency of the R7 clade is several times greater among AA (Munda) speakers than among Dravidian and Indo-European speaking populations (Table 2, [see Additional file 2]). Geographically, the distribution of R7 in India is centered on the AA \"heartland\" (Bihar, Jharkhand, and Chhattisgarh) [see Additional file 3]. Similar to R7, haplogroup R6 is significantly more frequent among the AA speakers than among other linguistic groups (Table 2, [see Additional file 2]). PC analysis based on frequency data of the hg R subclades confirms that the majority of Munda speaking populations cluster separately from others mainly because of higher hg R7 frequency (Fig. 2). However, only 50.6% of the variation can be explained by the first two principal components. Interestingly, hg R6 is placed within the main cluster, which is comprised of populations from all language groups. Based on these preliminary results we focused on R7 as a potential AA-associated marker.Table 2Frequency of Autochthonous R Subgroups Among Different Language Groups of India.R5R6R7R8R30R31Total SamplesAustro-Asiatic1.12%4.27%5.90%2.64%0.61%0.00%983Indo-European3.62%1.70%0.58%1.61%2.63%0.85%2240Dravidian3.65%1.69%1.37%1.64%2.15%0.32%2190Tibeto-Burman1.74%0.00%0.00%0.58%0.58%0.00%172Figure 2Principal component (PC) analysis of R5-8, R30 and R31 lineages in Indian populations. Munda group and a few Indo-European/Dravidian populations collected from Bihar, Jharkhand and Chhattisgarh states, predominantly cluster with haplogroup R7. Haplogroup frequencies were obtained from published sources [14] and our unpublished data.In general, the elevated frequency of hg R7 among the AA speakers of India can be explained by two alternative scenarios. Firstly, one may consider a possible origin of R7 among AA (Munda) speakers, possibly already outside India. Under this scenario the presence of R7 in some Dravidian and Indo-European speaking communities would be explained by its later introgression from the Munda communities, or by language shift of some Munda speaking groups into Dravidian/Indo-European languages. Secondly, an origin of R7 may lie among non-AA populations of India, with the presently observable higher frequency of R7 among AA resulting from founder effect(s) due to random genetic drift. To test these two scenarios, we carried out a detailed analysis of R7 mtDNAs in populations speaking different subgroups of AA languages, as well as among IE and Dravidian-speaking populations of Indian subcontinent.Complete mtDNA sequence-based topology of hg R7 divulges two deep-rooted subclades (Fig. 1). R7a is defined by four and R7b by six coding region mutations and, in addition, by two control-region substitutions (146 and 16311). We calculated the time to the most recent common ancestor (MRCA) for all R7 major sub-clades (Fig. 1 and 3, Table 3), applying different calibration methods [30,31]. All the AA individuals coalesce to the founder R7a1 that dates back to between approximately 3 Ka and 7 Ka ago, depending on the mutation rate used. The coalescent times of R7 variation among Dravidians and Indo-Europeans are older. In other words, the only R7 lineage found by us in AA speakers of India – R7a1 – is nested within the R7 lineages found among Dravidian and Indo-European speakers of India (Table 3).Figure 3The reduced-median network of 152 mtDNAs belonging to haplogroup R7. Each sample represented on the diagram has been sequenced for the HVS-I region and genotyped for the coding region mutations that are indicated. Circle sizes are proportional to the number of mtDNAs with that haplotype. Recurrent mutations are underlined.Table 3Coalescent times of hg R7 subclades estimated from HVS-I data.CladeNumber of SamplesMotif (Coding region)rho (ρ)sigma (σ)Time (SD)R71521442-6248-7870-9051-9110-10289-13105-138300.7960.3116.064 (6.260)R7(Austro-Asiatic)471442-6248-7870-9051-9110-10289-13105-138300.2340.1024.723 (2.059)R7(Indo-European)291442-6248-7870-9051-9110-10289-13105-138300.7930.30616.005 (6.185)R7(Dravidian)761442-6248-7870-9051-9110-10289-13105-138301.1450.53623.101 (10.822)R7a(Overall)10710143-10915-13404-153460.3890.1027.848 (2.064)R7a(Dravidian)3710143-10915-13404-153460.5140.15110.363 (3.037)R7a(Indo-European)2410143-10915-13404-153460.50.2410.090 (4.757)R7b(Overall)451804-2282-8557-12432-14064-159420.7970.29216.052 (5.891)R7b(Dravidian)391804-2282-8557-12432-14064-159420.7440.26815.006 (5.402)R7a1(Overall)8612406-136740.3370.1156.805 (2.311)R7a1(Austro-Asiatic)4712406-136740.2340.1024.723 (2.059)R7a1(Indo-European)2312406-136740.5220.24610.529 (4.663)R7a1(Dravidian)1612406-136740.3750.1537.568 (3.090)Geographically, the distribution of R7a frequency is concentrated towards Bihar, Jharkhand and Chhattisgarh States, while R7b has its frequency peak in Andhra-Pradesh (Fig. 4a and 4b). The frequency of R7a is higher among AA (Munda) speakers, while R7b is most common among Dravidian speakers from Andhra-Pradesh, although the overall frequency of R7b is much lower than that of R7a (Fig. 4c and 4d). A Mantel test showed a significant correlation between genes and geography for the Indian R sub-clades, but no such correlation for the relationship between genes and languages (Table 4). The spatial autocorrelation analysis favoured a clinal pattern for the distribution of hg R7 [see Additional file 4]. At the local (i.e. district) level, R7 is present in Bihar, Jharkhand, Chhattisgarh, Madhya-Pradesh and the northern districts of Andhra-Pradesh (Adilabad, Warangal and Khammam), whereas elsewhere in India it is virtually absent, including among other AA groups inhabiting Orissa and Maharashtra states [see Additional file 5].Figure 4The frequency distribution of R7a and R7b clades in Indian subcontinent. The upper panel (a, b) shows the spatial distribution (%) of these clades in Indian populations. Isofrequency maps were generated by using Surfer7 of Golden Software (Golden Software Inc., Golden, Colorado), following the Kriging procedure. These isofrequency maps illustrate the geographic spread of the respective mtDNA haplogroups. It should be cautioned, however, that these illustrative maps should not be used to predict the frequency of the clade in geographical areas with missing data. The lower panel (c, d) depicts the frequencies of R7a and R7b in different social and language groups. DRA-Dravidian, AA-Austro-Asiatic, IE-Indo-European.Table 4Mantel correlation test of Autochthonous R Subgroups to assess the significance of correlations between gene and geography, or language.HaplogroupGene vs GeographypGene vs LanguagepR50.12760.04750.17480.2R60.26540.0370.132480.19R70.2990.0230.2190.225R80.2114960.017530.232480.31R300.1899170.1270.13480.28R310.1720.18730.1410.25The overall higher than average frequency of R7 among the AA speakers of India may superficially be seen as supporting the model that places the origin of this haplogroup among AA speakers, possibly even outside India, assuming the language phylum would have arisen elsewhere. Indirectly, such a scenario would be also supported by the Y chromosome evidence (haplogroup O2a, for details, see Introduction). However, the much higher diversity of R7a and R7b sub-clades among non Austro-Asiatic populations of India suggests that the source of haplogroup R7 is not among the maternal ancestors of all Austro-Asiatic tribal groups, but that they acquired this haplogroup via local admixture, together with the rest of the South Asian mtDNA lineages that make up their extant maternal lineage pool. Furthermore, the presence of only a single recent founder branch of R7, i.e. R7a1, among widely dispersed AA populations of India supports the founder event scenario by introgression of this lineage from the local non-AA populations before the range expansion of Munda speaking populations within India. If indeed R7 did have its origin among some so far unsampled populations of the present-day Myanmar or Cambodia, we would then expect to see different sub-divided AA populations losing by drift different sub-branches of R7a and R7b (to explain their reduced diversity), and the admixed Dravidian and Indo-European speaking populations would be expected to have obtained a subset of the R7 variation observed in AA speakers, which is not the case. While the occurrence of R7a1 among Dravidian and Indo-European-speaking populations living close to the AA populations (Fig. 3) could be explained by language shift or secondary admixture with AA speakers, sub-haplogroup R7b appears to be restricted to Dravidian-speakers of the southern part of India (Fig. 4b and 4d). Nevertheless, this haplogroup is also reported in two Indo-European populations (Kolcha and Rathwa) whose local tradition speaks about their ancient split from the Gond (Gondi subfamily of Dravidian language group) population of Central India and further migration to Gujarat. Thus, from the data and analyses shown here, it is most parsimonious to conjecture that R7 originated in India among non-AA, possibly in Dravidian speaking populations.To test further the two hypotheses, a Dravidian origin for R7 with admixture and founder effects, versus an external AA origin of R7, we examined whether the spread of R7 among the different Munda sub-groups in India, as defined by the language trees [26,27], is uniform. This would be expected if R7 was present among the ancestral AA speakers prior to the diversification of the language family into numerous branches. Consistent with the non-AA origin of R7, we found the distribution of R7a1 among AA populations to be profoundly skewed towards the Kherwari sub-branch of the North Munda languages which accounts for ~90% of the AA R7 samples (Fig. 5). Conversely, R7 is very rare in the South Munda group. It is completely absent in Koraput Munda speakers and marginally present only in the Kharia tribe of Madhya Pradesh (in total 3 out of 431 South Munda samples) (Fig. 5). This finding yet again strengthens the argument that only a subset of Indian AA groups has acquired one sublineage of R7a1 in situ after their arrival to Indian subcontinent from local non-AA groups through admixture. Thus, we fail to find from the evidence of the extant maternal lineage pool of the Austro Asiatic speakers of India any major lineages that show signs of potential origin outside India. Overall, the enigma of the origins and demographic past of the AA speakers in India remains, for while the East Asian contribution to their paternal gene pool seems evident, the maternal side of their genetic heritage appears to be autochthonous to Indian subcontinent. This suggests that introduction and spread of AA speakers into India involved a complex and sex-differentiated demography, involving both exogenous males and local females.Figure 5The frequency distribution of haplogroup R7 in different branches of the Austro-Asiatic language family of India[26].In brief, our high-resolution study of haplogroup R7 suggests that this haplogroup originated in India among non-AA population most probably Dravidian, and that the Munda (mainly Kherwari group) speaking populations have acquired a subset of it only relatively recently. The highest frequency of haplogroup R7 among Austro-Asiatic tribal groups can be explained, thus, by their regional admixture with other local Indian subcontinental populations followed by random genetic drift, rather than being a genetic marker of their own. The spread of R7 as well as other ancient sub-clades of haplogroup R in India follows predominantly the geographic rather than linguistic landscape of the subcontinent. The geographic correlations are further manifested in the distribution patterns of the sub-clades: R7a being more common in northern India while R7b is more frequent in the southern parts of the subcontinent. Because Dravidian speakers harbour all the twigs of R7 identified so far, the haplogroup may have arisen among the matrilineal ancestry of the present day Dravidian speakers. However, it is important to caution that autochthonous basal mtDNA lineages in South as well as Southeast and East Asia appear to be significantly more ancient than any linguistic reconstruction offers to present day language families. This would imply that linguistically significant relationships among Indian populations may be superimposed on, and masking, demographic events of much greater antiquity. Our results also remind us, once again, that phylogenetically established within-haplogroup diversity is more informative than mere frequency in establishing the direction of gene flow between populations, language groups and geographically defined regions.MethodsTo refine the phylogeny of superhaplogroup R we sequenced complete mitochondrial genomes of 35 samples selected from different regions and language groups of India (Table 1). The results were incorporated into a phylogenetic tree [see Additional file 1]; for detailed tree for hg R7 see Fig. 1) together with previously published complete mtDNA sequence data from India [4]. For haplogroup R7 we performed a high-resolution survey of phylogenetically diagnostic markers, using information from complete mtDNA sequences. We studied ~12,000 samples collected from all over Indian subcontinent [see Additional file 6]. These samples cover all the language groups and most of the Indian states and union territories. The samples were screened for the presence of R7 mtDNAs based on HVS-I information (motif: 16260-16261-16319-16362). Previously this motif has, together with the restriction enzyme AluI cutting site polymorphism at np. 10143, been used to define haplogroup R20 [14]. However, with the support of new complete mtDNA sequences information the lineage with this HVSI motif was subsequently named R7 [4] and we follow this update of the nomenclature. Further, the identified R7 samples were analyzed for coding region markers by sequencing. Sequencing was carried out in ABI 3730 and 3730XL DNA Analyzers (Applied Biosystems, USA) and mutations were scored against the rCRS [33]. To minimize errors, both strands were double-sequenced. Principal component analysis (PCA) of R subgroups was performed using POPSTR, kindly provided by H. Harpending. Median-joining and reduced median networks were reconstructed with NETWORK program (version 4.1) [34]http://www.fluxus-engineering.com. Reduced median and median-joining procedures were applied sequentially. Coalescence time has been calculated between nucleotide positions 16090–16365 (HVS-I) considering one transition equals to 20,180 years [31], while for the coding region estimates we employed the rate calibrated by Kivisild et al. [32] considering substitution rate estimate for protein-coding synonymous changes of 3.5 × 10-8, which gives 6,764 years per synonymous transition. Standard deviation of the rho estimate (σ) was calculated as in Saillard et al. [35]. Haplogroup isofrequency maps were generated by using Surfer 7 of Golden Software (Golden Software Inc., Golden, Colorado), following the Kriging procedure. To determine whether language or geography has the strongest impact on genetic differentiation, spatial autocorrelation, SAAP [36] and Mantel [37] tests were performed using ARLEQUIN version 2.0 [38]. For Mantel test genetic distance matrixes were generated from ARLEQUIN, and geographic distance calculated from latitude and longitude information. For language groups linguistic distances (ranging from 10–100) assigned manually to each branch, based on published linguistic information and vocabulary match [26-28,39,40].Electronic database informationAccession numbers for data presented herein are as follows (for the complete mtDNA sequence accession numbers FJ004804-FJ004838 and for the HVS-I region sequence accession numbers FJ010662- FJ010785).Authors' contributionsGC, MK, EM, DS–R, VKS, AS, BPN, AK, NA, CBM, BT, SP, RR, PS, SB, SVe, SVa, IK, AB, DS, AS, MR, VC and AGR carried out the mtDNA genotyping. GC, MK, EM, DS–R, VKS, AS, and BPN carried out the mtDNA sequencing analysis. AT, KT and LS contributed to the analysis and interpretation of the data. AT provided complete sequence information of Sindhi sample. GC, MK, EM, MM and TK analyzed the data. TK, GC, MM and RV were responsible for conceiving and designing the study. GC, MK, MM, TK, RV, AT, RF, KT and LS wrote the paper. All authors read and approved the final manuscript.Supplementary MaterialAdditional file 1Phylogenetic tree of 22 Indian complete mtDNA sequences of superhaplogroup R. The tree includes data reported [[4] and references there in] Suffixes A, C, G, and T indicate transversions, \"d\" signifies a deletion; recurrent mutations are underlined. 16182C, 16183C and 16519 polymorphisms are omitted in phylogenetic reconstruction. The sample code, geographic and linguistic affiliations are described in Table 1. The sub-tree of haplogroup R7 sequences is displayed in Fig. 2.Click here for fileAdditional file 2Haplogroup R5-8, R30 and R31 frequency plots with 95% credible regions. Data calculated from the posterior distribution of the proportion of a haplogroup/sub-haplogroup in the population. Linguistic affiliations of the populations are indicated by colors.Click here for fileAdditional file 3Map of Indian subcontinent depicting the spatial frequency distribution of mtDNA haplogroup R7. Isofrequency maps were generated by using Surfer7 Golden software (Golden Software Inc., Golden, Colorado), following the Kriging procedure. The spread of R7 in India is centered around the AA \"heartland\" (Bihar, Jharkhand, and Chhattisgarh). Dots indicate the sampling locations.Click here for fileAdditional file 4Spatial Autocorrelation Analyses Correlograms of haplogroup R7 in Indian subcontinent. The Moran's I coefficient was calculated with five distance classes in binary weight matrix. Significant values are shown as black (p < .05) whereas nonsignificant values as blank circles. Distances are given in Kilometers (KM's).Click here for fileAdditional file 5Map of India showing the frequency distribution (%) of haplogroup R7 at the district level. Only 2,200 samples were available at this resolution. Nevertheless, it is still evident that the frequency peak of R7 is observed in Bihar, Jharkhand, Chhattisgarh, Madhya-Pradesh and the northern districts of Andhra-Pradesh (Adilabad, Warangal and Khammam).Click here for fileAdditional file 6Details of the samples studied for hg R7 in the present study. Data shown are from the present work and from literature: [4,12,14,21,22,41-46].Click here for file\n\nREFERENCES:\n1. PapihaSS\"Genetic variation in India\"Hum Biol1996685607288908794\n2. SinghKSEdThe Scheduled TribesPeople of India1997Oxford, Oxford University Press\n3. SinghKSEdThe Scheduled CastesPeople of India2002New Dehli, Oxford University Press\n4. PalanichamyMGSunCAgrawalSBandeltHJKongQPKhanFWangCYChaudhuriTKPallaVZhangYPPhylogeny of mitochondrial DNA macrohaplogroup N in India, based on complete sequencing: implications for the peopling of Indian subcontinentAm J Hum Genet200475696697815467980\n5. ForsterPMatsumuraSEvolution. Did early humans go north or south?Science2005308572496596615890867\n6. MacaulayVHillCAchilliARengoCClarkeDMeehanWBlackburnJSeminoOScozzariRCrucianiFTahaAShaariNKRajaJMIsmailPZainuddinZGoodwinWBulbeckDBandeltHJOppenheimerSTorroniARichardsMSingle, rapid coastal settlement of Asia revealed by analysis of complete mitochondrial genomesScience200530857241034103615890885\n7. ThangarajKSridharVKivisildTReddyAGChaubeyGSinghVKKaurSAgarawalPRaiAGuptaJMallickCBKumarNVelavanTPSuganthanRUdaykumarDKumarRMishraRKhanAAnnapurnaCSinghLDifferent population histories of the Mundari- and Mon-Khmer-speaking Austro-Asiatic tribes inferred from the mtDNA 9-bp deletion/insertion polymorphism in Indian populationsHum Genet2005116650751715772853\n8. ThangarajKChaubeyGSinghVKVanniarajanAThanseemIReddyAGSinghLIn situ origin of deep rooting lineages of mitochondrial Macrohaplogroup 'M' in IndiaBMC Genomics2006715116776823\n9. ThangarajKChaubeyGKivisildTSelvi RaniDSinghVKIsmailTCarvalho-SilvaDMetspaluMBhaskarLVReddyAGChandraSPandeVPrathap NaiduBAdarshNVermaAJyothiIAMallickCBShrivastavaNDevasenaRKumariBSinghAKDwivediSKSinghSRaoGGuptaPSonvaneVKumariKBashaABhargaviKRLalremruataAGuptaAKKaurGReddyKKRaoAPVillemsRTyler-SmithCSinghLMaternal footprints of Southeast Asians in North IndiaHum Hered20086611918223312\n10. SunCKongQPPalanichamyMGAgrawalSBandeltHJYaoYGKhanFZhuCLChaudhuriTKZhangYPThe dazzling array of basal branches in the mtDNA macrohaplogroup M from India as inferred from complete genomesMol Biol Evol200623368369016361303\n11. PetragliaMAllchinBPetraglia M, Allchin BHuman evolution and culture change in the Indian subcontinentThe Evolution and History of Human Populations in Indian subcontinent2007Springer, Netherlands393443\n12. KivisildTBamshadMJKaldmaKMetspaluMMetspaluEReidlaMLaosSParikJWatkinsWSDixonMEPapihaSSMastanaSSMirMRFerakVVillemsRDeep common ancestry of indian and western-Eurasian mitochondrial DNA lineagesCurr Biol19999221331133410574762\n13. KivisildTRootsiSMetspaluMMastanaSKaldmaKParikJMetspaluEAdojaanMTolkHVStepanovVGolgeMUsangaEPapihaSSCinniogluCKingRCavalli-SforzaLUnderhillPAVillemsRThe genetic heritage of the earliest settlers persists both in Indian tribal and caste populationsAm J Hum Genet200372231333212536373\n14. MetspaluMKivisildTMetspaluEParikJHudjashovGKaldmaKSerkPKarminMBeharDMGilbertMTEndicottPMastanaSPapihaSSSkoreckiKTorroniAVillemsRMost of the extant mtDNA boundaries in south and southwest Asia were likely shaped during the initial settlement of Eurasia by anatomically modern humansBMC Genet200452615339343\n15. ChaubeyGMetspaluMKivisildTVillemsRPeopling of Indian subcontinent: investigating the caste-tribe continuum in IndiaBioessays20072919110017187379\n16. ChaubeyGMetspaluMKarminMThangarajKRootsiSParikJSolnikASelvi RaniDSinghVKNaiduBPReddyAGMetspaluESinghLKivisildTVillemsRLanguage shift by indigenous population: A Model Genetic Study in Indian subcontinentInt J Hum Genet200884150\n17. SahooSSinghAHimabinduGBanerjeeJSitalaximiTGaikwadSTrivediREndicottPKivisildTMetspaluMVillemsRKashyapVKA prehistory of Indian Y chromosomes: evaluating demic diffusion scenariosProc Natl Acad Sci USA2006103484384816415161\n18. SenguptaSZhivotovskyLAKingRMehdiSQEdmondsCAChowCELinAAMitraMSilSKRameshAUsha RaniMVThakurCMCavalli-SforzaLLMajumderPPUnderhillPAPolarity and temporality of high-resolution y-chromosome distributions in India identify both indigenous and exogenous expansions and reveal minor genetic influence of Central Asian pastoralistsAm J Hum Genet200678220222116400607\n19. WatkinsWSPrasadBVNaiduJMRaoBBBhanuBARamachandranBDasPKGaiPBReddyPCReddyPGSethuramanMBamshadMJJordeLBDiversity and divergence among the tribal populations of IndiaAnn Hum Genet200569Pt 668069216266407\n20. RosenbergNAMahajanSGonzalez-QuevedoCBlumMGNino-RosalesLNinisVDasPHegdeMMolinariLZapataGWeberJLBelmontJWPatelPILow levels of genetic divergence across geographically and linguistically diverse populations from IndiaPLoS Genet2006212e21517194221\n21. CordauxRSahaNBentleyGRAungerRSirajuddinSMStonekingMMitochondrial DNA analysis reveals diverse histories of tribal populations from IndiaEur J Hum Genet200311325326412678055\n22. Quintana-MurciLChaixRWellsRSBeharDMSayarHScozzariRRengoCAl-ZaheryNSeminoOSantachiara-BenerecettiASCoppaAAyubQMohyuddinATyler-SmithCQasim MehdiSTorroniAMcElreaveyKWhere west meets east: the complex mtDNA landscape of the southwest and Central Asian corridorAm J Hum Genet200474582784515077202\n23. AtkinsonQDGrayRDDrummondAJmtDNA variation predicts population size in humans and reveals a major Southern Asian chapter in human prehistoryMol Biol Evol200825246847418093996\n24. HighamCFWThosaratRPrehistoric Thailand: From early settlement to Sukothai1998River Books, Bangkok\n25. DiamondJBellwoodPFarmers and their languages: the first expansionsScience200330059760312714734\n26. DifflothGAustroasiatic languagesEncyclopedia Britannica, online edition2005http://www.britannica.com/eb/article-9109792\n27. FullerDQPetraglia M, Allchin BNon-human genetics, agricultural origins and historical linguistics in Indian subcontinentThe Evolution and History of Human Populations in Indian subcontinent2007Springer, Netherlands393443\n28. WitzelMOsada TCentral Asian roots and acculturation in Indian subcontinent: linguistic and archaeological evidence from Western Central Asia, the Hindukush and northwestern Indian subcontinent for early Indo-Aryan language and religionLiguistics, Archaeology and the Human Past2005Research Institute for Humanity and Nature, Kyoto87211\n29. BasuAMukherjeeNRoySSenguptaSBanerjeeSChakrabortyMDeyBRoyMRoyBBhattacharyyaNPRoychoudhurySMajumderPPEthnic India: a genomic view, with special reference to peopling and structureGenome Res200313102277229014525929\n30. BlackMLDufallKWiseCSullivanSBittlesAHGenetic ancestries in northwest CambodiaAnn Hum Biol2006335–662062717381059\n31. ForsterPHardingRTorroniABandeltHJOrigin and evolution of Native American mtDNA variation: a reappraisalAm J Hum Genet19965949359458808611\n32. KivisildTShenPWallDPDoBSungRDavisKPassarinoGUnderhillPAScharfeCTorroniAScozzariRModianoDCoppaAde KnijffPFeldmanMCavalli-SforzaLLOefnerPJThe role of selection in the evolution of human mitochondrial genomesGenetics2006172137338716172508\n33. AndrewsRMKubackaIChinneryPFLightowlersRNTurnbullDMHowellNReanalysis and revision of the Cambridge reference sequence for human mitochondrial DNANat Genet199923214710508508\n34. BandeltHJForsterPRohlAMedian-joining networks for inferring intraspecific phylogeniesMol Biol Evol1999161374810331250\n35. SaillardJForsterPLynnerupNBandeltHJNorbySmtDNA variation among Greenland Eskimos: the edge of the Beringian expansionAm J Hum Genet200067371872610924403\n36. SokalRROdenNLSpatial autocorrelation in biologyBiol J Linn Soc197810199249\n37. MantelNAThe detection of disease clustering and a generalized regression approachcer Res196727209220\n38. SchneiderSRoessliDExcoffierLArlequin ver. 2.000: A software for population genetics data analysis2000Genetics and Biometry Laboratory, University of Geneva, Geneva, Switzerland\n39. KrishnamurtiBThe Dravidian Languages2003Cambridge University Press1574\n40. DriemGLanguages of the Himalayasan ethnolinguistic handbook Leiden1997New York ; Köln : Brill\n41. BamshadMKivisildTWatkinsWSDixonMERickerCERaoBBNaiduJMPrasadBVReddyPGRasanayagamAPapihaSSVillemsRReddAJHammerMFNguyenSVCarrollMLBatzerMAJordeLBGenetic evidence on the origins of Indian caste populationsGenome Res2001116994100411381027\n42. RajkumarRKashyapVKHaplotype diversity in mitochondrial DNA hypervariable regions I and II in three communities of Southern IndiaForensic Sci Int20031361–3798212969624\n43. RoychoudhurySRoySBasuABanerjeeRVishwanathanHUsha RaniMVSilSKMitraMMajumderPPGenomic structures and population histories of linguistically distinct tribal groups of IndiaHum Genet2001109333935011702215\n44. BaigMMKhanAAKulkarniKMMitochondrial DNA diversity in tribal and caste groups of Maharashtra (India) and its implication on their genetic originsAnn Hum Genet200468Pt 545346015469422\n45. BanerjeeJTrivediRKashyapVKMitochondrial DNA control region sequence polymorphism in four indigenous tribes of Chotanagpur plateau, IndiaForensic Sci Int20051492–327127415749372\n46. ThanseemIThangarajKChaubeyGSinghVKBhaskarLVReddyBMReddyAGSinghLGenetic affinities among the lower castes and tribal groups of India: inference from Y chromosome and mitochondrial DNABMC Genet200674216893451"
+}

batch_0/PMC2529319.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529319",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529319\nAUTHORS: Joëlle Michaud, Ken M Simpson, Robert Escher, Karine Buchet-Poyau, Tim Beissbarth, Catherine Carmichael, Matthew E Ritchie, Frédéric Schütz, Ping Cannon, Marjorie Liu, Xiaofeng Shen, Yoshiaki Ito, Wendy H Raskind, Marshall S Horwitz, Motomi Osato, David R Turner, Terence P Speed, Maria Kavallaris, Gordon K Smyth, Hamish S Scott\n\nABSTRACT:\nBackgroundThe RUNX1 transcription factor gene is frequently mutated in sporadic myeloid and lymphoid leukemia through translocation, point mutation or amplification. It is also responsible for a familial platelet disorder with predisposition to acute myeloid leukemia (FPD-AML). The disruption of the largely unknown biological pathways controlled by RUNX1 is likely to be responsible for the development of leukemia. We have used multiple microarray platforms and bioinformatic techniques to help identify these biological pathways to aid in the understanding of why RUNX1 mutations lead to leukemia.ResultsHere we report genes regulated either directly or indirectly by RUNX1 based on the study of gene expression profiles generated from 3 different human and mouse platforms. The platforms used were global gene expression profiling of: 1) cell lines with RUNX1 mutations from FPD-AML patients, 2) over-expression of RUNX1 and CBFβ, and 3) Runx1 knockout mouse embryos using either cDNA or Affymetrix microarrays. We observe that our datasets (lists of differentially expressed genes) significantly correlate with published microarray data from sporadic AML patients with mutations in either RUNX1 or its cofactor, CBFβ. A number of biological processes were identified among the differentially expressed genes and functional assays suggest that heterozygous RUNX1 point mutations in patients with FPD-AML impair cell proliferation, microtubule dynamics and possibly genetic stability. In addition, analysis of the regulatory regions of the differentially expressed genes has for the first time systematically identified numerous potential novel RUNX1 target genes.ConclusionThis work is the first large-scale study attempting to identify the genetic networks regulated by RUNX1, a master regulator in the development of the hematopoietic system and leukemia. The biological pathways and target genes controlled by RUNX1 will have considerable importance in disease progression in both familial and sporadic leukemia as well as therapeutic implications.\n\nBODY:\nBackgroundThe Core Binding Factor (CBF) is a transcriptional regulator complex, which is composed of two sub-units [1]. Mammals have three genes coding for the α-subunits, RUNX1, RUNX2 and RUNX3 [2], and one coding for the β-subunit, CBFβ . The α-subunits recognize a specific sequence (TGT/cGGT) in the regulatory regions of their target genes in order to bind DNA directly, while the β-subunit heterodimerizes with the α-subunits but does not interact directly with the DNA. The interaction with CBFβ stabilizes the RUNX-DNA complex [3,4] and protects the RUNX proteins from degradation [5].In humans, the CBF complex containing RUNX1 as the α-subunit is one of the most frequent targets of chromosomal and genetic alterations in leukemia. Chromosomal rearrangements involving RUNX1 or CBFβ [6], somatic point mutations in RUNX1 [7] and amplification of RUNX1 [8] have all been described in acute leukemia. In addition to somatic alterations, germ-line point mutations in RUNX1 are responsible for an autosomal dominant platelet disorder with a propensity to develop leukemia (FPD-AML, OMIM 601399) [9,10]. Interestingly, the dosage of RUNX1 protein seems to play a role in the determination of the leukemic phenotype. Indeed, low dosage of RUNX1, resulting from haploinsufficient or dominant negative mutations, lead to the development of myeloid leukemia [9-11], whereas amplification of RUNX1 gene is more often observed in lymphoid leukemia, particularly pediatric ALL [12]. A number of observations also suggest that although RUNX1 is involved in the first steps of leukemia development, additional somatic mutations are necessary and probably determinant for the leukemic phenotype: 1) The predisposition to develop leukemia in FPD-AML patients shows that germline RUNX1 mutations are not sufficient for the development of the disease [10]. 2) Somatic translocations are not able to induce leukemia in mouse cells on their own [13]. 3) The translocation t(12;21), which fuses ETV6 (TEL) to RUNX1, can arise in utero but does not trigger leukemia until later in childhood, with as much as nine years latency [14]. These additional mutations are likely to occur in molecules involved in the same biological pathways as RUNX1, as hemizygous loss of several molecules in the same biological pathway (e.g. RUNX1 and SPI1) is thought to be almost as tumorigenic as homozygous loss of one molecule (e.g. homozygous RUNX1 mutation in AML-M0) [15]. Therefore the identification of downstream targets of RUNX1, with care to the model systems including species and cell type of origin, is of great interest in order to identify novel candidate molecules involved in leukemogenesis.The identification of the biological pathways regulated by RUNX1 is also of importance to shed light on its in vivo function and role in leukemia development. The observation that Runx1 knockout mice show a lack of definitive hematopoietic maturation and die at embryonic stage 12 from hemorrhages in the central nervous system demonstrates that RUNX1 plays a critical role during development of the hematopoietic system [16,17]. In addition, RUNX1 might also play a role in other systems as it is expressed in many other embryonic tissues [18-20] and in epithelial cells [19,20]. It is furthermore overexpressed in endometrioid carcinoma [21] and down-regulated in gastric cancer [22]. The in vivo function of RUNX1 is therefore yet to be fully understood.Here we describe the combination of a number of genomic and bioinformatic approaches to identify biological pathways downstream of RUNX1, and report on a number of processes in which RUNX1 is likely to be involved. We also took advantage of the integration of these approaches in order to identify novel RUNX1 target genes.ResultsGene expression profiling of cells harboring different levels of RUNX1Three different model systems were used to identify the biological pathways regulated by the RUNX1 transcription factor. These were haploinsufficiency using FPD-AML patient B cell lines (FPD), overexpression of CBF complex (CBF) in HeLa cells and Runx1 deficiency in mouse embryos (E8.5 and E12) (Figure 1).Figure 1Gene expression profiles and overlaps. The three platforms used in this study are indicated. The number of up-, down- or all differentially expressed genes (DEGs) are indicated below each platform.Lymphoblastic cells derived from FPD patients heterozygous for a RUNX1 frameshift mutation (R135fs) were first analyzed. This mutation results in haploinsufficiency of RUNX1, as the mutant protein has lost its capacity to bind DNA and to transactivate the expression of the target genes [9]. Quantitative RT-PCR on these non-leukemic lymphoblastic cells showed that affected individuals express approximately 55% of the transcript level observed in unaffected individuals (see Additional File 1 :Figure S1). The genes differentially expressed between two affected and two non-affected cell lines are therefore largely the result of a low dosage of RUNX1 protein. Using human cDNA microarrays with the Hs8k cDNA clone library from Research Genetics and a selection of control spots, 366 genes were identified as differentially expressed, of which 52% (192/366) were down-regulated in affected individuals (Figure 1 and see Additional File 2).For overexpression studies, HeLa epithelial cells were transduced using adenoviral vectors. FACS analysis showed that over 90% of HeLa cells were transduced by a EGFP-expressing adenovirus (data not shown). This system results in a highly homogenous cell population in which small changes of expression can be identified. The wild type CBF complex α-subunit, RUNX1, was overexpressed together with the β-subunit, CBFβ (see Additional File 1: Figure S2) and seven hybridizations were performed. Following overexpression of the CBF complex, 721 genes were differentially expressed including the up-regulation of 42% of the genes (300/721; Figure 1 and see Additional File 2).Finally, we compared the expression profiles of two wild type and two Runx1 knockout mouse embryo propers at each embryonic stages E8.5 and E12 using Affymetrix chips. Despite the heterogeneity of the samples, 931 and 297 genes were differentially expressed at embryonic stages E8.5 and E12, respectively. Of these genes, 57% (533/931) and 72% (214/297) were down-regulated in the knockout embryos (Figure 1 and see Additional File 3). These differences in expression are likely to reflect the lack of hematopoiesis and the premature death, respectively, observed in the Runx1 embryos.We then compared the different datasets using a mean-rank gene set enrichment test (MR-GSE) in order to determine the level of connection between the 3 approaches (FPD cell lines, CBF overexpression and Runx1 knockout mouse embryos), disregarding the cell type and the organism. High correspondence was observed between the two human datasets. The correspondence between the human and the mouse datasets was not as good, although still significant. This might partially be explained by the difficulties of matching human and mouse platforms (see Additional File 1: Figure S3).Correlation with clinical AML samplesIt was first necessary to determine whether the genes identified in nonmyeloid cells in this study may play a role in myeloid leukemia development. We therefore compared our data to previously published microarray data obtained from 285 AML and 8 healthy samples [23], using the MR-GSE test. The high correspondence between the FPD-AML and CBF datasets had already suggested that a large number of downstream genes were similar between epithelial and lymphocytic cells. Therefore we used each approach as representative of the RUNX1 gene dosage, regardless of the cell type. The AML samples used in the comparison include 22 patients with a t(8;21) translocation, which fuses RUNX1 to ETO, and 18 patients with inv(16), which fuses the co-factor CBFβ to MYH11. The other samples include a range of common alterations or no identified mutations. RUNX1 activation targets should be positively correlated with RUNX1 expression whereas repression targets should be negatively correlated. Therefore we ranked all the probes-sets on the microarrays according to their correlation with RUNX1 across the 293 AML and normal samples (Figure 2A). MR-GSE tests demonstrated that genes up-regulated in the FPD-AML patients (likely to represent genes repressed by RUNX1), had an expression trend opposite to RUNX1 in the AML patients, suggesting indeed that these genes are repressed in vivo in the presence of RUNX1 (p = 7 × 10-6; Figure 2B). On the other hand, the down-regulated genes do not show any statistically significant trend (Figure 2C). Similarly, the genes activated by the exogenous CBF complex had an expression pattern similar to RUNX1 across the clinical samples (p = 1 × 10-4; Figure 2D), whereas genes repressed by the CBF complex had an expression pattern opposite to RUNX1 (p = 2 × 10-5; Figure 2E).Figure 2Correlation with clinical AML data. A. Published microarray data on 285 AML patients [23] were ordered using Gene Recommender according to the expression pattern of the 11 probe sets for RUNX1. The patients with t(8;21) are marked in orange and those with inv(16) in red. Probes co-regulated with RUNX1 are highly ranked (yellow bar), whereas probes showing an expression pattern the least similar to RUNX1 are ranked lowest (blue bar). B-C. Random permutations were performed to compare the rank of the genes differentially expressed in FPD platform and random set of genes. The histograms show the percentage of up- or down-regulated genes in FPD relative to their rank with \"0\" being the probes co-regulated with RUNX1 (yellow) and \"1\" being the probes the least similar to RUNX1 (blue). The trends observed in the histograms are represented as triangles or rectangle. D-E. Similar histograms showing percentage of up- or down-regulated genes in CBF relative to their rank.MR-GSE tests also showed that genes differentially expressed in the B cell lines derived from FPD-AML patients tended to be differentially expressed in the blasts and mononuclear cells of 22 clinical patients with a t(8;21) translocation (p = 10-10) and of 18 patients with the inv(16) abnormality (p = 3.5 × 10-9). For example, the top 14 differentially expressed genes in the FPD-AML dataset that are also differentially expressed in the clinical samples are shown in Additional File 1 (Table S3). As a whole, these results demonstrate that the genes identified in our study are likely to play an important role in the development of the disease.Biological processes regulated by RUNX1: bioinformatic approachesBioinformatics tools taking into account all differentially expressed genes (direct and indirect RUNX1 targets) were used to systematically identify the biological processes in which RUNX1 may be involved. A number of gene ontology (GO) annotations were significantly enriched in each dataset (Table 1). Some were identified in more than one dataset such as \"cadmium ion binding\" and \"immune response\". Other significantly represented processes were identified through the use of Ingenuity Pathways Analysis (Ingenuity Systems, ) (Figure 3). These include cancer related genes as well as genes involved in hematological disorders. To complete this analysis, a MR-GSE was also performed using a number of published gene sets related to thrombocytopenia, leukemia and cancer (Figure 4, see Additional File 1: Table S4 and Additional File 4). Significant correlation was obtained between the microarray datasets and a number of these sets of genes, including genes involved in megakaryopoiesis and cytokinesis, genes differentially expressed following irradiation of lymphoblasts, and genes consistently differentially expressed in solid-tissue tumors.Figure 3Processes identified by Ingenuity Pathways Analysis. Evidence that each dataset is involved in the given function as determined by the use of Ingenuity Pathways Analysis (Ingenuity Systems, ). The threshold for the significance is indicated by a vertical bar and represents a p-value of 0.05.Table 1Gene ontology enrichmentFPDCBFE8.5E12GO: Biological processesImmune response p = 6.5 × 10-5 36 genesMacromolecular complex assembly p = 0.02 47 genesBlood vessel development p = 0.06 15 genesResponse to external stimulus* p = 0.0003 18 genesNegative regulation of apoptosis p = 0.002 16 genesCell growth p = 0.02 21 genesBehavior p = 0.0003 14 genesResponse to biotic stimulus p = 0.002 19 genesImmune system process p = 0.0006 18 genesCell proliferation p = 0.01 36 genesGO: Molecular functionsCadmium ion binding p = 0.002 4 genesRNA binding p = 0.03 50 genesIgG binding p = 0.006 3 genesCadmium ion binding p = 0.03 4 genesFerric-chelate reductase activity p = 0.03 2 genesPolysaccharide binding p = 0.03 6 genesGO: cellular componentSpindle p = 0.06 11 genesCell junction p = 0.06 14 genesCell surface p = 0.05 9 genesExtracellular space p = 0.06 37 genesInterPro motifs (FatiGo)Vertebrate metallothionein p = 0.0001Vertebrate metallothionein p = 0.02Tubulin p = 0.04The most significant gene ontology annotations are indicated for each dataset as identified through GOStat in April 2007. InterPro motifs were identified through the FatiGo program. The p-values are corrected for multiple testing (False discovery rate, Benjamini).Figure 4MR-GSE test. Representation of the p-values (corrected for multiple testing) resulting from the MR-GSE test for each dataset and 10 gene sets specified in Additional File 1 (Table S4). In brief they are gene sets Mekagaryocyte differentiation, Identification of genes involved in the differentiation of megakaryocytes. DEGs between stem cells and differentiated megakaryocytes; Platelets, Transcription profiling of human blood platelet; ; Normal megakaryocytes, Genes highly expressed in megakaryocytes; ET megakaryocytes, Genes highly expressed in essential thrombocytopenia megakaryocytes; Cytokinesis proteome, Identification of proteins present in the midbody during cytokinesis; Spindle checkpoint, Review ; DNA repair, Review; Lymphoblast irradiation; high dose, Effect of ionising radiation on lymphoblasts; Lymphoblast irradiation; low dose, Effect of ionising radiation on lymphoblasts; Genes DE in cancer, Meta-analysis of cancer microarray data to identify genes consistently DE in tumours. This represents whether the genes present in the published gene sets are also differentially expressed in our expression profiles. For example, the genes expressed in normal or diseased megakaryocytes (lines 3 and 4) are significantly represented in the differentially expressed genes identified in the FPD and CBF approaches.Biological processes regulated by RUNX1: in vivo confirmationsWe designed a series of assays that were performed on either cell lines, or directly on samples from FPD-AML patients with RUNX1 mutations, to confirm the disturbance of several interesting biological processes identified by the above approaches.Heterozygous RUNX1 point mutations affect proliferationRUNX1 is thought to be involved in the balance between cell proliferation and differentiation, whose disruption leads to leukemia development. However, the molecular mechanisms behind this regulation are not known. We observed that genes participating in cellular proliferation were significantly enriched in both FPD and CBF datasets (Table 1 and Figure 3). The genes responsible for this enrichment are indicated in Additional File 1 (Table S5). We therefore performed a BrdU proliferation assay in order to determine whether a subtle proliferation defect was present when RUNX1 level was lower in FPD-AML patients. A slower proliferation was indeed observed in FPD-AML lymphoblasts derived from two independent families compared to unaffected cells (Figure 5A, p < 0.001).Figure 5Functional assays on FPD-AML cell lines. A. The results of a BrdU proliferation assay are indicated for each cell line. Dark bars indicate affected individuals. The standard errors of two independent replicates are shown. A two-way ANOVA resulted in a significant p-value (p < 0.001) between affected and unaffected individuals. B. Examples of the tubulin polymerization assay for an affected and an unaffected individuals in each family. s:soluble tubulin; p:polymerized tubulin. C. The percentage of polymerized tubulin is shown for each cell line. Dark bars indicate affected individuals. The standard errors of three independent replicates are indicated. A two-way ANOVA resulted in a significant p-value (p < 0.002) between affected and unaffected individuals. D. Percentage of polymerized tubulin in the same cell lines before (darker left bars) and after (second bars) induction of polymerization by Taxol. A significant smaller induction is observed in affected individuals (dark bars) as demonstrated by an ANOVA (p < 0.0003). E. Glycophorin A assay. The numbers of N0 (loss of the M allele), NN (mutation changing M to N allele) or total mutant (both N0 and NN) cells are indicated for each individual. The standard errors of three to five technical replicates are indicated. Dark bars represent affected individuals (A1-A2). The control C5 is the unaffected sister of patient A1. ANOVAs were performed for each kind of mutation and the p-values are indicated.RUNX1 modulates microtubule stabilityA significant enrichment of molecules containing a common tubulin motif was observed following overexpression of the CBF complex (Table 1). Five tubulin isoforms were down-regulated following overexpression of the CBF complex. These data led to the observation that CBF overexpression affected the expression of 57 genes associated with cytoskeletal structures according to GO annotation (see Additional File 1: Table S6). This class of genes was not significantly represented in the dataset from the FPD-AML cell lines, however this may be the result of the not complete knock-down of RUNX1 in the affected individuals leading to small changes that are not detected by microarray analysis. Therefore we also tested whether microtubule stability was affected in these cell lines. Significantly higher microtubule polymer levels were observed in the affected patients compared to the unaffected individuals (Figure 5B and 5C; p < 0.002). Furthermore, the microtubules in affected cells could not be stabilized using the drug Taxol to the same extent as the unaffected cells (Figure 5D; p < 0.0003). This might result from the inability of the drug to bind to the microtubule molecule because of the unusual presence of other microtubule stabilizing proteins or from a lack of soluble tubulin molecules in the cellular environment. In any case, these results suggest that RUNX1 is involved in microtubule dynamics.Neither the proliferation nor the tubulin defects are due to the EBV transformation of the cell lines as many independent proliferation and tubulin polymerization assays performed on lymphoblastic cell lines derived from families with predispositions to various haematological malignancies do not show similar familial clustering (data not shown).Genomic instabilityHighly significant correspondence was observed between the FPD, CBF and mouse datasets and the genes switched on after irradiation of lymphoblasts (Figure 4). We used a glycophorin A assay to test whether the FPD-AML patients are more prone to somatic genetic mutations than unaffected individuals. This test assesses the frequency of mutation events occurring at the glycophorin A locus in erythroid progenitors in blood of heterozygous individuals (MN phenotype) [24]. Although more samples would be necessary for corroboration, a significant trend was present between the blood of two affected patients and five unaffected individuals, suggesting that a subtle increase of mutation rate may occur when RUNX1 activity is impaired (Figure 5E; p < 0.01). This increased mutation rate appears to be higher in the assay that would detect deletions (NO), that are the predominate mutations arising due to ionizing irradiation [25].Identification of potential novel RUNX1 target genes – co-expression in human tissues and hematopoietic cell linesWe reasoned that direct RUNX1 target genes must be expressed in the same tissues or cells as RUNX1. Thus, the expression patterns of a number of differentially expressed genes, chosen due to potential functions in leukemia development, were compared to that of RUNX1 (see Additional File 5). The expression of 22 genes in 20 human tissues, 19 hematopoietic cell lines and normal human bone cells was assessed using cDNA panels [26]. 9 of these genes show a high expression in a number of hematopoietic cell lines and all the others show common expression with RUNX1 in various tissues such as liver and peripheral blood leukocytes (PBLs).Identification of potential novel RUNX1 target genes – data overlapsIn order to distinguish between the direct RUNX1 target genes and those effected further downstream by a disregulation of RUNX1 level, we hypothesized that the genes in common in more than one dataset were more likely to be at the top of the genetic pathways regulated by RUNX1 and to be enriched for direct target genes. As suggested by the significant MR-GSE results, we observed statistically significant overlap between each dataset. Among the 366 genes differentially expressed in FPD-AML cell lines, 69 genes were also differentially expressed following overexpression of the CBF complex, while only 32 were expected by chance (Figure 6A). As anticipated when comparing an under- and overexpression system, 61% (42/69) of the genes in this overlap were differentially expressed in the opposite direction. Among these 69 genes 16 were also differentially expressed in at least one embryonic stage of the Runx1 knockout embryos (Table 2, Figure 6A).Table 2Genes differentially expressed in FPD, CBF and in E8.5/E12Gene nameRefSeqRUNX1 BSITM2CNM_030926yGLO1NM_006708NDOGTNM_003605yALAS1NM_000688NDHSPA4LNM_014278yPPIBNM_000942NDCIB1NM_006384NBASP1NM_006317yTACC1NM_006283NDCTSCNM_001814yPBX3NM_006195yTGFBR3NM_003243NDANZA1NM_000700yELF1NM_172373NDIFRD1NM_001550NDMT1GNM_005950NDRUNX1 BS: presence of a RUNX1 binding site in the regulatory region of the gene as determined by oPOSSUM. y stands for the presence of binding site and ND stands for not determined due to the absence of the gene in oPOSSUM.Figure 6A. Overlaps between the datasets and percentage of genes with a RUNX1 binding site in their regulatory regions. The overlaps between the different platforms are represented with arrows. * indicates that the genes differentially expressed in at least one of the mouse datasets are considered for the following overlap. The number of differentially expressed genes (DEGs) containing a conserved RUNX1 binding site (with CBS) in their regulatory regions, as determined by the oPOSSUM program [27], over the number of analyzed genes is indicated for each dataset and overlap. The corresponding percentage is indicated in brackets. B. Luciferase assay for 5 RUNX1 binding sites corresponding to 3 differentially expressed genes. The transactivation activity of RUNX1 over these sites was measured as the fold change of the luciferase activity in the presence of the CBF complex compared to the endogenous activity of each construct. The standard errors of three independent replicates are shown. CASP3 was shown as a negative control as no binding site was found for this gene. The difference in expression for the three genes in each dataset is indicated in the table. 0 means no difference in expression, ↓ stands for down-regulated and ↑ stands for up-regulated.Identification of potential novel RUNX1 target genes – regulatory region analysisIn order to accumulate evidence that some of the genes present in these overlaps are direct target genes, we searched for human RUNX1 binding sites, which were conserved in mouse using the oPOSSUM software ( see Additional File 1) [27]. Many differentially expressed genes contained at least one conserved RUNX1 binding site in their regulatory regions and the overlaps between the datasets show a higher enrichment for such genes as hypothesized above (Figure 6A).The regions flanking five putative conserved binding sites identified in three differentially expressed genes, and one negative control region, were cloned upstream of a luciferase reporter gene and co-transfected together with plasmids expressing RUNX1 and CBFβ. These genes were selected because of their presence in the overlap between the human datasets and/or their interesting functions; ANXA1 (Annexin 1) is involved in cell proliferation and cytoskeleton regulation; ARMET (Arginine-rich, mutated in early stage tumors) is mutated in cancer; CYR61 (Cysteine-rich, angiogenic inducer, 61) promotes proliferation and angiogenesis. An increase in luciferase activity was observed for ANXA1 binding sites and for one of the ARMET binding sites and a diminution of the luciferase activity was observed for the CYR61 binding site (Figure 6B). No modification of the luciferase activity was observed for a sequence derived from the negative control CASP3 regulatory region (where no conserved binding site was identified by the oPOSSUM program). It is likely that a combination of a number of binding sites and the presence of additional co-factors are necessary for a correct and synergistic in vivo regulation of these genes and it might explain the small activity observed for the ARMET binding sites. It might also explain the activation of the ANXA1 site while this gene was repressed by the overexpression of the CBF complex.DiscussionRUNX1 is one of the most frequent targets of somatic mutations in leukemia and is mutated in an autosomal dominant disorder affecting platelets and predisposing to leukemia development. Better characterization of its in vivo function is likely to give insight into the mechanisms leading to the development of leukemia, and will provide new candidate genes for leukemogenesis. We do not believe that as a transcription factor and master regulator of hematological cancers, RUNX1 will alter the function of only one oncogenic molecule, but multiple molecules in the same pathways, and our analyses and functional assays are carefully designed to study these effects. We have described a combination of genomic and bioinformatic approaches to identify the biological pathways and genes regulated by RUNX1, an overview of which is in Figure 7. Each approach independently provides a large source of data to identify RUNX1 targets according to RUNX1 gene dosage. However, the combination of them is powerful because of their convergence. Although the approaches described here are not the ideal models to study myeloid leukemia, each of them has their own advantages and their integration compensates for their limitations: 1) The use of cells derived from patients harbouring a RUNX1 mutation but who have not yet developed leukemia allow us to observe effects, largely due to changes in RUNX1 dosage. However, it should be kept in mind that due to the difficulties of obtaining myeloid cell lines, these studies were performed in lymphoid cells. 2) The overexpression system using HeLa cells provided a highly homogenous cell population, which is necessary to perform gene expression profiling. 3) The knockout mouse embryos represent various cell types, however they give us global information of the complete absence of RUNX1, which is difficult to obtain using cell lines. Efficient and homogenous knockdown levels are indeed difficult to obtain using siRNA especially in hematopoietic cells [28].Figure 7Part of the networks downsteam of RUNX1. Additional data from the literature and our studies were used to update the standard Ingenuity Pathway System (Ingenuity® Systems, ) network analyses. Genes up-regulated (red) or down-regulated (green) in either FPD or CBF are indicated. Selected chosen functions with significant network nodes are shown including all the genes involved in cytoskeleton organization. Grey arrows represent transcriptional regulation, grey lines represent direct interaction, dotted lines represent indirect link. Each kind of molecule is represented by a different symbol (see ).The highly significant correlation observed between the genes identified in the FPD-AML cells and the overexpression system and clinical data on AML samples supports the hypothesis that large number of genes would be broadly regulated by RUNX1 in our various approaches disregarding of the cell type. Genes identified as differentially expressed following disregulation of RUNX1 expression level and/or in these AML samples are good candidates for targets of secondary hits during leukemogenesis downstream of RUNX1 mutation. The various approaches described in this study, including conserved binding sites and co-expression studies, will also help to further prioritize genes that might sustain secondary hits. For example, the gene encoding the Cyclin D3 (CCND3) was differentially expressed following overexpression of the CBF complex and mutations in this gene have been described in acute myeloid leukemia patients [29].In order to generate insights into the in vivo role of RUNX1, we employed bioinformatics tools to identify processes that were changed following alteration of RUNX1 expression level. We have shown that genes involved in megakaryopoiesis tend to be differentially expressed in the FPD and CBF datasets, demonstrating that a large number of the differentially expressed genes may play a role in platelet formation. Enrichment for genes involved in cell proliferation was also observed in both the FPD and CBF datasets, and functional assays on the FPD-AML cell lines showed that heterozygous mutation of RUNX1 reduced proliferation of lymphoblasts. These data validate our integrative approach as they confirm studies in transgenic mice expressing the fusion proteins CBFβ-MYH11 [30] and RUNX1-ETO [13], which both act in a dominant negative fashion over the wild-type protein. These mice show a decrease in both lymphoid and myeloid cell proliferation. This observation also correlates with mouse data showing that Runx1 promotes cell cycle progression from G1 to S phase [31]. An anti-proliferative effect of a RUNX1 mutant protein may have an oncogenic effect due to an improper balance between proliferation and differentiation. For example, overexpression of RUNX1 usually results in ALL while complete or partial loss of RUNX1 results in AML development.Our integrative approach unraveled a novel process that may play an important role in RUNX1 function, involving the cytoskeletal dynamics. Indeed following the finding that an enrichment of microtubule and cytoskeleton related molecules was observed when the CBF complex was overexpressed, functional assay using the FPD-AML cells demonstrated an increase of polymerized microtubules in FPD-AML affected cells compared to cells from unaffected individuals. Microtubules are important in many processes such as cell migration, cell division, cellular transport and signal transduction [32] and microtubule remodeling is essential during the cell cycle, especially during mitosis when a correct microtubule network is essential for proper chromosomal segregation [33]. Interestingly, the fusion protein, CBFβ-MYH11 that results from inv(16), co-localizes with the actin cytoskeleton and disorganizes stress fibers and F-actin structures [34]. A mild microtubule defect might partially explain the platelet defect observed in FPD-AML patients, as microtubules are necessary at several different stages of megakaryopoiesis including endomitosis, production of platelets from mature polyploid megakaryocytes, and release of the content of platelet granules [35]. Moreover, mutations in the actin-binding protein WASP and the myosin heavy chain MYH9 cause the Wiskott-Aldrich [36] and May-Hegglin [37] syndromes of thrombocytopenia, respectively. However, RUNX1 is likely to regulate only specific tubulin isoforms or tissue-specific cytoskeleton-associated proteins as a strong cytoskeleton defect would be more detrimental to the whole organism. In addition, the dosage of normal RUNX1 activity necessary for normal function might differ according to cell type, and some cell types may be more susceptible than others to perturbation in RUNX1 levels. Interestingly, Taxol resistant leukemic cells have been shown to have a reduced total level of tubulin and an increased level of polymerized tubulin [38], similar to the results seen in the FPD-AML cells. Furthermore, a high level of survivin (BIRC5), which was down-regulated following overexpression of the CBF complex, is associated with resistance to Taxol [39]. This is the first evidence demonstrating a relationship between RUNX1 and microtubule dynamics.Finally, we showed that the predisposition of FPD-AML to develop leukemia may be due to an increased rate of mutation in RUNX1 heterozygous cells. Every dataset showed significant correspondence with genes involved in DNA damage response. Although not conclusive, the glycophorin A assay, which measures the frequency of the progeny of mutated erythrocyte precursors in blood, showed a mild increase in mutation frequency in FPD-AML patients compared to unaffected individuals. Recently, it was shown that the RUNX1-ETO fusion protein induces mutations in transfected U937 myeloid cells [40]. This study demonstrated that the fusion protein regulates many genes involved in the base excision repair pathway, which mainly corrects for point mutations. Furthermore, a higher incidence of leukemia in CBFβ-MYH11 chimeras compared to normal chimeras when exposed to ENU mutagenesis has also been observed [41,42]. This demonstrates that alteration of RUNX1 function may increase the rate of mutation and lead to an accumulation of mutated cells.The three processes described here (proliferation, cytoskeleton stability and genomic instability) are tightly interconnected and may explain the phenotype observed in FDP-AML patients. Indeed, a proliferation defect would have an impact on megakaryopoiesis and cytoskeleton remodeling. In turn, a cytoskeleton defect could also affect proliferation and trigger chromosomal aberrations. The necessary threshold level of RUNX1 expression is likely to be cell-specific, explaining why RUNX1 heterozygous mutation affects only hematopoietic cells; nevertheless, our observations could conceivably suggest possible involvement of RUNX1 in solid-tissue tumor.We also identified new potential RUNX1 target genes by analyzing the regulatory regions and the expression pattern of the differentially expressed genes present in the overlaps between the different platforms. Many RUNX1 target genes have already been described in the literature, mainly from in vitro studies and in mouse cells [43,44]. Four of the published target genes, CSF1R, MYB, MPO and TIMP1, were differentially expressed in the Runx1 knockout embryos. In addition, target genes that were described more recently, including CCND3 [45] and IGFBP3 [46], were identified following overexpression of the CBF complex. That there was not more correlation may be due to incomplete microarray platforms, but more importantly is likely to reflect the bias present in the published RUNX1 target genes that were identified because of their primary role in hematopoiesis and these may not represent the most common RUNX1 target genes. Interesing candidates were among the 16 genes differentially expressed in every dataset, such as Annexin I (ANXA1), which was shown to reduce inflammation, by inhibiting neutrophil recruitment [47] and has an anti-proliferative effect by inducing aberrant cytoskeleton formation [48]. This gene is likely to play an important role downstream of RUNX1.ConclusionIn summary, this combination of gene expression profiling platforms allowed prioritization of novel candidate genes for leukemogenesis according to distinct parameters and has shed light on RUNX1 functions by identifying biological pathways downstream of RUNX1 such as microtubule stability and genomic instability and identified a large number of potential novel RUNX1 target genes. Whether or not these are direct RUNX1 targets remains to be demonstrated by further research.MethodsAdenovirus productionRecombinant adenoviruses expressing RUNX1 p49 isoform [49] or CBFβ were generated as described [50], except that VmRL-CMV1 and pSCOT were used as the adenovirus backbone and transfer vector respectively. For details, see Additional File 1.Cell lines and RNA extractionEBV-transformed lymphoblasts generating B cell lines from FPD-AML patients (Pedigree 2, individuals V:1 and V:2;) [9] and related unaffected individuals (Pedigree 2, individuals IV:1 and V:3) were used for the FPD microarray dataset. HeLa cells (4 × 107) were infected with a multiplicity of infection (MOI) of 100 for each adenovirus and incubated for 48 hours. The Qiagen RNeasy maxikit was used for the extraction of total RNA in each case. Runx1 knockout and wild-type embryo propers at embryonic stages E8.5 and E12 were homogenized in Trizol (Invitrogen) and total RNA extracted following the manufacturer's protocol.Mouse samplesRunx1 knockout mice have been previously described [16]. They are maintain on a BalbC genetic background at the Biological Resource Center, (Biopolis, Singapore) and all animal experiments followed the guidelines set by the National Advisory Committee for Laboratory Animal Research. Wild-type and Runx1 knockout mouse embryo propers were harvested at embryonic stages E8.5 and E12.cDNA Microarray hybridizationcDNA microarrays were printed by the Australian Genome Research Facility (AGRF) with the Hs8k cDNA clone library from Research Genetics and a selection of control spots. In total there were 8132 EST probes printed in duplicate. The array also contained 12 copies of the Lucidea Universal ScoreCard controls (Amersham). Labeling, hybridization, and washing were performed as described [51]. In the case of the FPD dataset, four hybridizations were performed comparing two affected individuals against two unaffected individuals of pedigree 2. For the overexpression system, 2 different RNA samples from HeLa cells overexpressing EGFP were used as reference and 2 different RNA samples from HeLa cells overexpressing RUNX1 and CBFβ were used as experimental RNAs. Seven hybridizations (including 3 dyeswaps) were performed. The data were filtered for genes whose difference in expression was due to EGFP, using four hybridizations between EGFP expressing cells and normal HeLa cells.Affymetrix genechip hybridizationLabelling, hybridization and washing were performed by the AGRF following the Affymetrix protocol (701725 rev5). Briefly, total RNA (100 ng) was amplified using T7-oligo dT and the Megascript T7 kit (Ambion). A second round of cDNA synthesis was performed using the total amount of the amplified RNA. Biotin-labeled RNA was subsequently synthesized using the GeneChip IVT Labeling Kit. Labelled RNA (15 μg) was fragmented and the mouse genome 430 2.0 arrays were hybridized overnight and washed as described before being scanned using a GeneChip scanner 3000 (Affymetrix). Two biological replicates were used for each condition.Microarray analysisThe cDNA microarray images were analyzed using SPOT software [52]. Spots were assigned quality weights based on their segmented pixel areas and the log-ratios were print-tip loess normalized [53]. Duplicate printings of each probe on each array were combined using the common correlation method of [51]. For the mouse Affymetrix GeneChips, the intensities for each probe set were normalized and summarized using the Robust Multi-array Analysis algorithm [54]. Differential expression was assessed using empirical Bayes moderated t- and F-statistics from the LIMMA package [55]. Recognizing that p-value calculations make normality and other distributional assumptions, which are hard to verify for microarray data, we decided to use control probes and appropriate plots to guide our criteria for differential expression as far as possible. For the cDNA data, conservative threshold values for differential expression were chosen to minimize the false-positive and false-negative rates estimated from Scorecard control probes printed on the arrays. This resulted in a threshold value of |t|>4 for the FPD data. Of 204 calibration control probes printed on the arrays, none reached this cutoff for statistical significance, suggesting a false discovery rate less than 1/204, without relying on any distributional assumptions. For the mouse Affymetrix data, a threshold of |t|>3 was chosen from a q-q plot of the moderated t-statistics.For the overexpression system arrays, a combination of criteria was used to assess differential expression. These arrays were analyzed as part of a larger microarray study using the same overexpression system to study a range of AML related genes. Genes with |t|>4 were initially assigned as differentially expression, with only one calibration control probe reaching this threshold. A series of nested F-tests (with p-value cutoff 1e-5) was also performed using the larger dataset in order to get an improved estimate of the number of genes significantly differentially expressed in more than one condition simultaneously. This increased the number of differentially expressed genes by a third. Finally, genes were removed from the differentially expressed list if their response to RUNX1/CBFβ transduction was not significantly greater than their response to the adenovirus alone.All the analyzed datasets have been deposited at the NCBI Gene Expression Omnibus  under accession numbers GSE2592 (mouse Affymetrix data), GSE2593 (overexpression experiment) and GSE2594 (FPD-AML arrays).Mean-rank gene set enrichment tests (MR-GSE)A version of statistical gene set testing was used to investigate associations between the expression profiles obtained from different experiments. Each test uses a set of genes selected as differentially expressed in one data set (the reference dataset) and determines whether the gene set tends to be highly ranked in another dataset (the test dataset). The test statistic is the mean rank of the gene set in the test dataset. This approach, which we call mean-rank gene set enrichment (MR-GSE), is very similar to Tian et al's Tk test [56] and Kim and Volsky's PAGE test [57]. The main difference is that MR-GSE averages the ranks of t-statistics instead of t-statistics themselves, which makes it less influenced by individual genes in the gene set. This has the advantage of giving more weight to gene sets with a larger number of active genes, and it also allows us to use the same testing procedure with a range of ranking procedures other than t-statistics. Where possible, MR-GSE is used with moderated t-statistics rather than ordinary t-statistics, as these are preferable for microarray analysis including gene set testing [56,58]. Unlike earlier Gene Set Enrichment Analysis methods [59], MR-GSE can be used to test individual gene sets in isolation and has good power even for microarray experiments with small to moderate sample sizes.The null hypothesis tested by MR-GSE is that the gene set is randomly chosen. When the reference and test datasets share the same microarray platform, p-values can be computed using Wilcoxon two-sample rank tests [60]. When the reference and test datasets are based on different microarray platforms (cDNA vs Affymetrix), the p-values were instead computed using random permutations of probes on the reference arrays. This was done to avoid any bias arising from probe selection on the cDNA platform or from multiple probe-sets for individual genes on the Affymetrix platform.For the integration of gene expression profiling data and biological processes regulated by RUNX1, genes were ranked in the test datasets by absolute moderated t-statistic. For the correlation with clinical AML samples, the test dataset was the previously published expression profiling data on 285 AML patients and 8 healthy individuals [23]. In this case, the Affymetrix probe-sets were ranked according to their correlation with the 11 RUNX1 probe-sets across the 293 RNA samples. Correlations were computed using Gene Recommender [61], which provides a very robust correlation measure suitable for this purpose. Probe-sets were also ranked by moderated t-statistic on their ability to distinguish the healthy patients from the 22 patients with t(8;21) or from the 18 patients with inv(16).The MR-GSE p-values are computed by permuting genes rather than permuting arrays. This is necessary because the tests are designed for use with small numbers of arrays. The computation necessarily assumes that different genes have statistically independent expression values within experimental groups. When the gene set contains genes which are highly interdependent, and which vary substantially between biological replicates, the test may be anti-conservative. We checked the independence assumption for our data by computing average inter-gene correlations using REML. The inter-gene correlations were found to be generally very small at the expression level (data not shown), suggesting that the MS-GSE results are meaningful on our data.Bioinformatic identification of biological processes and cross-platform comparisonEnrichment of a gene ontology annotation in a dataset of differentially expressed genes compared to the genes present on the array was determined using the GOStat program [62]. For the MR-GSE test, relevant gene sets were taken from published reviews or independent microarray data (see Additional File 1: Table S4)BrdU proliferation assayThe Cell Proliferation ELISA, BrdU kit (Roche) was used to measure proliferation of cell lines derived from two independent families, including the family used for the microarray experiment (Pedigree 2) [9] and an additional family harboring a nonsense mutation Y260X present outside of the Runt domain (Pedigree 3, affected individuals III:7 and IV:4 and one unaffected individual III:8 [9]). Briefly, the cells were split into 96-well plates at an equal density. BrdU was added to the cells for 4 hours and the cells were then treated according to the manufacturer's protocol. The optical density (OD450) was measured on an ELISA plate reader. Technical triplicates and two independent experiments were performed. A two-way ANOVA (analysis of variance) test was performed.Tubulin polymerization assaySoluble (cytosolic) and polymerized (cytoskeletal) fractions of tubulin were separated from the cell lines treated with or without 4 μg/ml of Taxol as described [63]. The same cell lines used for the proliferation assay were assessed. Results were expressed as a percentage of polymerized tubulin by dividing the densitometric value of polymerized tubulin (insoluble) by the total tubulin content (sum of densitometric value of soluble and polymerized tubulin). Three independent experiments were performed and a two-way ANOVA was done.Glycophorin A assayBlood samples were collected in EDTA-tubes, with informed consent, from seven individuals heterozygous (MN phenotype) at the glycophorin A locus. These include: a FPD-AML patient harboring a frameshift mutation (N69fsX94) and her unaffected sister, a second FPD-AML patient harboring a nonsense mutation (Pedigree 3 (Y260X), individual IV:4) [9] and 4 independent unaffected individuals. The assay is described in detail in Additional File 1. A two-way ANOVA test was performed to compare the 5 controls to the 2 affected individuals.Luciferase reporter assayGenomic regions overlapping the conserved binding sites (300–400 bps) were amplified from BACs and cloned into pGL3-Basic vector (Promega #E1751). Each construct was co-transfected into HeLa cells using lipofectamine 2000 (Invitrogen) along with pSCOT plasmids expressing RUNX1 and CBFβ or empty vector to keep the amount of plasmid constant. For normalization, 20 ng of pRL-TK vector (Renilla luciferase Promega #E2241) was also co-transfected. The luciferase activities were measured using the Dual-Luciferase Reporter Assay System (Promega #E1910). The increase or decrease in luciferase activity was determined as a function of the endogenous activity of each construct.cDNA panel productionThe human cDNA panel was generated as described [26]. The relative amount of each cDNA was normalized according to housekeeping gene levels. More details are described in Additional File 1.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsJM designed the experiments and analysis, performed the majority of the experiments and wrote the manuscript. KMS performed the statistical analysis of the Affymetrix data and participated in the Gene Set Enrichment analysis. RE participated in the design of the experiments. KBP participated in the generation of adenovirus particles. TB participated in the design of the bioinformatics analyses. CC participated in the luciferase reporter assay. MER participated in the microarray analyses. FS performed the ANOVA tests and participated in the cross-platform comparison. PC participated in the luciferase reporter assay. ML performed the tubulin polymerization assay. XS performed the GPA assay. YI provided vital Runx1 knockout embryos. WHR provided vital patient samples. MSH provided vital patient samples. MO provided vital Runx1 knockout embryos. DRT participated in the design and analysis of the GPA assay. TPS participated in the design of the bioinformatics analyses. MK participated in the design and analysis of the tubulin polymerization assay. GKS generated the statistical analysis of the cDNA microarray data, the statistical comparison to AML samples and supervised the statistical components of the article. HSS designed the experiments and analysis and participated in the writing of the manuscript. All authors read and approved the final manuscript.Supplementary MaterialAdditional File 1Additional Methods, Figures S1 to S4 and Tables S3 to S6. Additional figures and tables to support the statistical and bioinformatics analyses described in the manuscript.Click here for fileAdditional File 2Table S1. Gene expression profiling results. Summary of the gene expression profiling results, oPOSSUM and corresponding mouse Affymetrix data for each clone Accession number present on the human cDNA array.Click here for fileAdditional File 3Table S2. Gene expression profiling data for E8.5 and E12 Runx1 knockout embryos.Click here for fileAdditional File 4Figure S4. Supporting graphs for the Gene Set Enrichment analysis.Click here for fileAdditional File 5Figure S5. Expression pattern of RUNX1 and a subset of differentially expressed genes.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2529320.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529320",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529320\nAUTHORS: Alfredo S Negri, Bhakti Prinsi, Mara Rossoni, Osvaldo Failla, Attilio Scienza, Maurizio Cocucci, Luca Espen\n\nABSTRACT:\nBackgroundGrape ripening represents the third phase of the double sigmoidal curve of berry development and is characterized by deep changes in the organoleptic characteristics. In this process, the skin plays a central role in the synthesis of many compounds of interest (e.g. anthocyanins and aroma volatiles) and represents a fundamental protective barrier against damage by physical injuries and pathogen attacks. In order to improve the knowledge on the role of this tissue during ripening, changes in the protein expression in the skin of the red cultivar Barbera at five different stages from véraison to full maturation were studied by performing a comparative 2-DE analysis.ResultsThe proteomic analysis revealed that 80 spots were differentially expressed throughout berry ripening. Applying a two-way hierarchical clustering analysis to these variations, a clear difference between the first two samplings (up to 14 days after véraison) and the following three (from 28 to 49 days after véraison) emerged, thus suggesting that the most relevant changes in protein expression occurred in the first weeks of ripening. By means of LC-ESI-MS/MS analysis, 69 proteins were characterized. Many of these variations were related to proteins involved in responses to stress (38%), glycolysis and gluconeogenesis (13%), C-compounds and carbohydrate metabolism (13%) and amino acid metabolism (10%).ConclusionThese results give new insights to the skin proteome evolution during ripening, thus underlining some interesting traits of this tissue. In this view, we observed the ripening-related induction of many enzymes involved in primary metabolism, including those of the last five steps of the glycolytic pathway, which had been described as down-regulated in previous studies performed on whole fruit. Moreover, these data emphasize the relevance of this tissue as a physical barrier exerting an important part in berry protection. In fact, the level of many proteins involved in (a)biotic stress responses remarkably changed through the five stages taken into consideration, thus suggesting that their expression may be developmentally regulated.\n\nBODY:\nBackgroundGrape berry is a typical true fruit originating from the ovary and is formed by skin, flesh, seeds and a complete vascular system. They all have specific properties that are directly linked to their particular physiological roles during berry development and seed dispersal.The growth of this non-climacteric fruit is summarized by the well known double-sigmoidal curve and is divided into an initial and rapid growth, a subsequent lag phase and a second period of growth corresponding to berry ripening [1,2]. During the first phase, embryo formation takes place in the seeds and the berry enlarges through frequent cell divisions, accompanied by the accumulation of many solutes, such as malic acid, tartaric acid and tannins [3,4]. The lag phase is characterized by the lack of any changes in berry weight and volume and its end coincides with the onset of ripening. This stage, which is referred to the French word véraison, is detectable in red cultivars where the change in skin colour takes place due to the start of anthocyanins synthesis. It is important to observe that at this time phloem unloading shifts to an apoplasmic pathway that is accompanied by a parallel change of the role of xylem in the water budgets [5-7]. Furthermore, ripening is characterized by profound changes in berry composition. The concentrations of some metabolites, among which malic acid is the most important, decrease while the levels of other molecules, such as glucose, fructose, volatile aroma compounds and anthocyanins (in red cultivars), greatly increase [4,8-10]. Moreover, berries start to soften at véraison and this event is mainly linked to significant changes in the cell wall composition [11-14].In all growth phases, the very active metabolism of the skin deeply influences the final characteristics of the grape berry. This tissue, which is formed by a single layer of clear epidermal cells and a few hypodermal layers beneath the epidermis, is in fact the site of the synthesis of anthocyanins and aroma compounds [4,8,10,15] and also represents a fundamental protective barrier against damage by physical injuries and pathogen attacks [16]. The composition of this tissue depends on both the particular genetic background of the cultivar and the environmental conditions. These factors play a central role in influencing colour, aroma and other organoleptic properties of wine [4,17-21].The impact of gene and protein expression patterns in determining the specificity of the skin in comparison to the other berry tissues is a crucial aspect that must be considered. In this view, two recent studies of the mRNA expression profiles in isolated skins have been published using oligonucleotide or cDNA microarrays [17,22]. Waters and co-workers provided a first description on the main events characterizing the shift in gene expression in this tissue around véraison [22]. On the other hand, Grimplet and co-workers compared the mRNA expression profiles of the three major tissues of the berry (skin, pulp and seeds) at maturity. The results of this analysis highlighted that the skin transcriptome presented the most distant fingerprint from the global set, since the categories related to housekeeping processes (i.e. protein fate, cell cycle and DNA processing) were under-represented while those related to secondary, amino acid and lipid metabolism were highly expressed, if compared to pulp and seeds [17].The widening of genomic information obtained in the last few years has also paved the way to the study of protein expression. Recently, some proteomic studies have been performed on grape berry. A 2-DE analysis of the mesocarp profile conducted by Sarry and co-workers [23] allowed the identification of 67 proteins using MALDI-MS, thus providing clues to the sugar and organic acid metabolism in ripe berry pulp. More recently, the first analysis of the skin proteome has been performed by comparing, two by two, three different ripening stages in Cabernet Sauvignon berries [24]. This paper mainly reports differences in the expression of pathogenesis-related proteins and of some enzymes involved in anthocyanin biosynthesis. Giribaldi and co-workers [25], on the other hand, focused their attention on the proteome of whole berries of cv. Nebbiolo during a longer period of time, ranging from one month after flowering to complete ripening. These studies provided a first profile of grape proteomes, also describing some dynamic changes taking place in growing berries, although further efforts are still necessary in order to unravel the physiological events that characterize the grape berry ripening and the specific roles of the different tissues at the protein level.A crucial step in a 2-DE analysis is the procedure adopted for protein extraction. As with many other fruits, grape is a recalcitrant plant material because of the high concentration of interfering compounds such as phenolics, terpenes, organic acids, ions, carbohydrates and proteolytic and oxidative enzymes [26-31]. This aspect is particularly onerous for investigations of the skin, where some of these compounds are present at very high concentrations. For this tissue, the phenol extraction method followed by ammonium acetate in methanol precipitation appears to be the most appropriate protocol up to now [24,32].In order to obtain further information on protein expression changes in the skin during berry ripening, a comparative 2-DE analysis was performed on a time-course experimental design made up of five different stages from véraison to full ripening of Barbera, a widely cultivated red variety typical of northern Italy. In order to associate the proteome changes to the events characterizing the ripening process, some biochemical parameters were also measured. In this study, it was reported that 80 spots significantly changed their relative volumes among the different stages. Sixty-nine of them were identified by LC-ESI-MS/MS and the corresponding proteins were classified on the basis of their putative functions. Some of these proteins were associated with glycolysis and other carbohydrate pathways of the primary metabolism and were found to increase in the skin tissue during ripening.Results and discussion2-DE and image analysis2-DE analysis was performed on five consecutive stages of ripening that, as described in the Methods section, were also defined through the determination of some physiological parameters.Proteins were extracted from the berry skin samples of cultivar Barbera previously washed in acetone through a protocol which made use of phenol followed by precipitation in ammonium acetate in methanol, which was previously indicated to be appropriate for this recalcitrant tissue [24]. 2-DE gels are shown in Fig. 1. The average number of detected spots was about 850 for each stage and did not vary significantly among the five different conditions. To ascertain the quantitative changes in the proteomic maps, their relative spot volumes (%Vol) were evaluated by software-assisted analysis. The ANOVA test (p < 0.01), coupled with a threshold of two-fold change in level, revealed 80 spots as being differentially expressed throughout berry ripening.Figure 12-DE maps of five stages through the ripening of Barbera. 2-DE maps of five different ripening stages from véraison until full ripeness of cultivar Barbera berry skins. The véraison stage (0 DAV) was considered as the moment when 50% of the berries started to change colour. Proteins (200 μg) were separated by IEF at pH 3–10, followed by 12.5% SDS PAGE and visualized by cCBB-staining.Hierarchical clustering analysisThe differentially expressed spots were subjected to two-way hierarchical clustering analysis using the PermutMatrix software (Figure 2). Looking at the clustering of columns, which mirrors the distances among the different stages of berry ripening, it is evident that the bunch order reflects the sequential succession of samples, while there is a clear difference between the first two samplings and the following three. These results suggested that the most important changes in protein expression took place between the second and the third stage. Nevertheless, this behaviour appeared different from the data emerging from oligo/microarrays studies in which the most dramatic changes in the transcriptome were found immediately after véraison and appeared well correlated with the start of the ripening process [22,33]. The fact that most of the observed changes in this proteomic analysis did not refer to véraison, but to a period between 14 and 28 days after véraison (DAV), may reflect peculiar features of cv. Barbera that is characterized by a longer period of ripening, compared to the cultivar Shiraz which was used in the works cited above. Anyway, it is important to underline that comparisons among studies using different genotypes need to be evaluated with extreme caution. In addition, when dealing with in-field grown plants, the relevance of environmental factors should not be excluded since they affect the gene and/or protein expression [18].Figure 2Clustering analysis of the spots that resulted to change their relative volumes during ripening. Two-way hierarchical clustering analysis of the 80 spots that showed at least a two-fold change in the relative spot volumes (ANOVA, p < 0.01) in the five different ripening stages of grape berry skins of cultivar Barbera. The véraison stage (0 DAV) was considered as the moment when 50% of the berries started to change colour. The clustering analysis was performed with PermutMatrix graphical interface after Z-score normalization of the averages of relative spot values (n = 6). Pearson's distance and Ward's algorithm were used for the analysis. Each coloured cell represents the average of the relative spot value, according to the colour scale at the bottom of the figure.As for the row clustering, two main trends were observed: one related to proteins whose levels increase during maturation, the other describing those declining as the berry ripens. Most of the spots belonged to the first class (62.5%), agreeing with the observation that the number of genes whose expression is switched on during ripening is far greater than the amount of genes switched off [22]. The clustering analysis also indicated that the trends of expression relative to the last two ripening stages were closely grouped, suggesting that no evident expression changes took place in that period.Protein identification and functional distributionAmong the 80 differentially expressed spots analyzed by LC-ESI-MS/MS, 69 were identified, listed in Table 1 and shown in Figure 3 which is referred to a gel of the fifth stage. The functional distribution of the identified proteins was performed according to MIPS FunCat annotation and is shown in Figure 4.Table 1List of spots identified by LC-ESI-MS/MS and bioinformatic analysis. Proteins were classified according to MIPS FunCat. Additional data about mass spectrometry are reported in the additional file 1.Spot IDAccession numberProtein descriptionName abbreviationMraMrbpIapIba.a. cov.c (%)Glycolysis and gluconeogenesis365Q9ZSQ4Cytoplasmic phosphoglucomutasePGluM68.1763.126.165.499.6397Q429082,3-bisphosphoglycerate-independent phosphoglycerate mutasePGlyM-163.3461.185.835.395.7561P42896EnolaseENO-152.4847.915.895.5635.5596P42896EnolaseENO-252.0447.916.145.5631.5829CAN81988Phosphoglycerate kinase(d)PGK-140.6642.426.156.2927.9863CAN81988Phosphoglycerate kinase(d)PGK-239.0642.426.216.2937.1902ABC75834Glyceraldehyde-3-phosphate dehydrogenaseG3PDH-137.4836.767.486.7225.1937P26518Glyceraldehyde-3-phosphate dehydrogenaseG3PDH-236.7736.987.947.0922.01767Q429082,3-bisphosphoglycerate-independent phosphoglycerate mutasePGlyM-262.1561.185.785.3916.3C-compound and carbohydrate metabolism191AAC26045Aconitase-iron regulated protein 1ACO102.8198.095.965.9514.2325CAN60522Transketolase(d)TK-174.7773.775.976.3615.3327CAN60522Transketolase(d)TK-274.3473.776.036.3610.1378P51615NADP-dependent malic enzymeNADP-ME66.0065.236.106.0920.6412AAB47171Vacuolar invertase 1GIN1-159.0771.554.274.607.9413CAN69570Putative oxalyl-CoA decarboxylase(d)OxD60.1761.065.985.9423.8431AAB47171Vacuolar invertase 1GIN1-259.2071.554.334.608.1851P52904Pyruvate dehydrogenase E1 component subunit β, mitochondrial precursorPDHE139.5938.795.175.887.21109CAN78176Xyloglucan endotransglycosylase(d)XET31.6533.186.195.9812.5Photosynthesis1088CAN61828Manganese-stablising protein/photosystem II polypeptide(d)MnSpPSII31.9133.235.395.8712.2Nucleobase metabolism844AAU14832Adenosine kinase isoform 1SADK40.0437.445.605.0719.7Amino acid metabolism172CAN63089Glycine cleavage system P-protein(d)GCPp109.41112.816.376.994.9270CAN73135Cobalamin-independent methionine synthase(d)MetSy-183.4781.645.976.1911.9273CAN73135Cobalamin-independent methionine synthase(d)MetSy-282.3881.645.986.1914.4572NP_193129Serine hydroxymethyltransferase 4SHM452.8851.727.276.8012.3612AAO92257γ-aminobutyrate transaminase subunit precursor isozyme 3ATpL350.9857.246.656.7220.6654AAG09278Ornithine aminotransferaseOAT48.5651.326.216.449.4815P37833Aspartate aminotransferase cytoplasmicAsAT41.4344.517.317.7517.7Transcription1189BAF46352α chain of nascent polypeptide associated complexPAC28.7821.924.064.3233.71511ABE01085BTF3BTF317.2617.345.526.3211.9Protein synthesis1606AAL13082Putative glycine-rich RNA-binding proteinGlyRp13.5617.335.337.8430.3Protein destination442Q43116Protein disulfide-isomerase precursorPDIpr58.2255.564.924.9529.7490CAN68309Heat shock chaperonin-binding motif(d)HSC56.0441.044.944.9417.11449CAN60868Molecular chaperone(d)MChap-119.6418.236.596.786.91513CAN65631Molecular chaperone(d)MChap-217.2618.155.736.178.81533P2788018.2 kDa class I heat shock proteinHsp18.216.5618.176.855.8112.0Cellular communication/signal transduction1016CAN81470Annexin(d)Annex34.8635.196.927.1329.4Secondary metabolism986CAN60921Kynurenine formamidase(d)KF35.4629.875.545.159.61008CAI56335Isoflavone reductase-like protein 6IFRL635.2333.936.096.0230.81028CAI56334Isoflavone reductase-like protein 5IFRL534.3833.896.215.7625.5Stress362NP_001031620Binding – stress inducible protein(d)BSP68.1763.716.056.0014.9521AAL83720CatalaseCAT54.4456.987.106.7113.0810AAB41022Polyphenol oxidasePPO-141.2667.396.886.398.4819AAB41022Polyphenol oxidasePPO-240.5067.396.646.3915.0826AAB41022Polyphenol oxidasePPO-341.0967.396.816.396.1843AAB41022Polyphenol oxidasePPO-439.9667.396.436.3917.5876AAB41022Polyphenol oxidasePPO-538.9867.395.996.399.6906CAN78553Late embryogenesis abundant protein(d)LEA37.7134.944.434.6722.41071CAB601541,3 β glucanaseGlucβ-132.2613.375.996.1139.31075CAB915541,3 β glucanaseGlucβ-232.6537.466.449.4515.61148AAQ10093Class IV chitinaseChit4-130.1927.534.575.389.11177AAB65776Class IV endochitinaseEnChit428.5027.244.935.3821.11226AAQ10093Class IV chitinaseChit4-227.6627.536.875.3814.41240AAQ10093Class IV chitinaseChit4-326.9927.537.355.3814.41316AAB61590VVTL1TLP24.6223.974.695.099.01318ABC86744Abscisic stress ripening proteinASR-124.3016.695.815.6830.21358ABC86744Abscisic stress ripening proteinASR-223.9416.695.775.6830.21385ABB02395Temperature-induced lipocalinTInLi22.8721.546.426.6313.01408AAQ03092Glutathione peroxidaseGPOX21.5318.536.526.1323.81417ABC86744Abscisic stress ripening proteinASR-321.2216.695.735.6826.21444CAC16165Pathogenesis-related protein 10PR10-119.7617.136.115.9622.81481AAB41022Polyphenol oxidasePPO-618.5067.394.916.395.91482AAB41022Polyphenol oxidasePPO-718.4167.394.996.3910.71508CAN83049Pathogenesis-related protein Bet v I family(d)PRBetv117.2017.105.155.1217.21524ABD78554Pathogenesis-related protein 10.1PR10-216.7517.456.616.0730.21768AAB41022Polyphenol oxidasePPO-818.4567.394.796.396.4Unclassified476CAN67811Dihydrolipoamide dehydrogenase(d)Uncla-156.9449.576.137.189.61181CAN6447914-3-3 protein(d)Uncla-228.2928.784.674.7816.11441ABK64186CBS domain-containing proteinUncla-319.8422.256.959.2425.2Unknown1083NP_001061484Protein of unknown function DUF52 family(d)Unk32.5333.556.226.1116.4a: Experimental molecular weight (kDa) or isoelectric pointb: Theoretical molecular weight (kDa) or isoelectric point.c: amino acid coverage (%)d: The protein was reported as a hypothetical protein. In the features, the similarity and function of the identified genes has been annotated by the authors according to Gene Ontology .Figure 3Protein profiles of identified proteins. Identified proteins are indicated in a 2-DE gel representative of the fifth ripening stage with spot name abbreviation corresponding to those in Table 1, Figure 6 and 7. Spots showing an increased or a decreased expression during ripening are indicated in red and in green, respectively.Figure 4Functional categories distribution of the identified proteins. Functional distribution of the identified proteins (Table 1) according to the annotation in the MIPS FunCat.Most of the observed variations are related to response to biotic or abiotic stresses (38%), glycolysis and gluconeogenesis (13%), C-compound and carbohydrate metabolism (13%) and amino acid metabolism (10%). The proportion of proteins involved in stress responses was quite high if compared to the functional distributions previously observed in the proteome of whole berries and isolated mesocarp, in which these proteins ranged from 8% to 19% of the identified spots [[25] and [23], respectively]. These results paralleled a recent large-scale mRNA expression analysis on the three main berry tissues [17] as well as the skin proteome analysis of cultivar Cabernet Sauvignon where most of the proteins over-expressed at maturity were involved in pathogen response [24]. This massive expression of proteins involved in stress responses may be essential to the protective function of the skin as a physical barrier between the environment and the inner tissues.Although it is known that the biosynthesis of anthocyanins and the transcription of related genes are induced at véraison [9], no proteins related to this pathway were found. This failure could be ascribed to the experimental conditions used in this work. In fact, in a very preliminary analysis conducted on different genotypes using a narrower pH range (4–7) we found some really low expressed spots that were referable to enzymes involved in anthocyanin synthesis (data not shown). Nevertheless, Robinson and Davies reported that enzymes involved in this pathway are present at low levels making their assay difficult [13].Pathogenesis-related proteinsPathogenesis-related (PR) proteins belonging to class IV chitinases (Chit4), β-1,3-glucanases (Glucβ) and thaumatin-like protein (TLP) were found (Table 1 and Figure 3). PR proteins matched to a group of spots, generally low expressed at véraison, whose abundance abruptly rose up to the point of representing about the 20% of the total spot volume in the protein profile of ripe berries (Figure 5). Both chitinase and β-1,3-glucanase are known to have antifungal activity and presumably hydrolyse the cell walls of fungal hyphae [38,39]. In agreement with the previous proteomic studies on grape ripening [24,25], two spots were identified corresponding to β-1,3-glucanase (spots 1071 and 1075) which accumulated after véraison. However, data regarding the behaviour of this enzyme during berry ripening are contradictory. More than one study [38,40] pointed out that, beside the surge of chitinase activity during ripening, no β-1,3-glucanase activity was detected in grape at any stages of berry development while it was reported that the gene is expressed [37]. In spite of this, Deytieux and co-workers [24] associated the assay of the enzyme activity to the proteomic profile and found that, even if weakly correlated, both the expression and the activity of β-1,3-glucanase increased during ripening.Figure 5Changes in the expression of proteins involved in stress response. Changes in the relative spot volumes of the proteins (Table 1) involved in stress responses during five different ripening stages from véraison until full ripening of cultivar Barbera grape berry skins. The véraison stage (0 DAV) was considered as the moment when 50% of the berries started to change colour. Proteins were grouped according to their functions. Values are the mean ± SE of six 2-DE gels derived from two independent biological samples analyzed in triplicate.In addition to their involvement in osmotic stress, a role in defence against fungi in grape berries has also been suggested for thaumatin-like proteins [38]. It is interesting to observe that several studies provide evidence of the fact that chitinases and thaumatin-like proteins accumulate during berry ripening even in the absence of pathogen infections [24,38-40]. According to these results, we observed a sharp increase of these proteins moving from véraison to full maturation, suggesting that their expression may be developmentally regulated (Figure 5).Oxidative stress-related proteinsIt has been proposed that the oxidative stress may play a developmental role in the ripening process [41-43]. As far as it concerns grape, this hypothesis is still a matter of debate. The data regarding the expression and the activities of proteins involved in ROS detoxification still remain unclear [25,33,37,44]. Recently, Pilati and co-workers [18] found that an oxidative burst occurs at véraison and that this event may modulate the expression of a gene set. Nevertheless, among the differentially expressed proteins during ripening, we identified some enzymes that are known to be involved in the oxidative stress response (e.g. PPO, polyphenol oxidase; GPOX, glutathione peroxidase; CAT, catalase; TInLi, temperature-induced lipocalin, Figures 3 and 5).Polyphenol oxidases catalyze the formation of o-quinones, molecules involved in browning reactions as a consequence of pathogen infection, wounding and organ senescence, through the O2-dependent oxidation of monophenols and o-diphenols [45]. In addition to the described defensive role, these ubiquitous enzymes may contribute to the biosynthetic pathways leading to proanthocyanidin [46] and aurone [47]. In our work we identified 8 spots corresponding to PPO, whose expression was high at véraison and dropped during ripening (Table 1, Figures 3 and 5). This trend is in agreement with previous reports on this class of enzymes which are generally highly expressed and active in young developing tissues [48-51]. Dry and Robinson [52] described that the protein is synthesized as a 67 kDa precursor which is imported into the chloroplast and processed to remove a 10.6 kDa chloroplast transit peptide from the N-terminus and a 16.2 kDa peptide of unknown function from the C-terminus, thus resulting in a catalytic unit of 40.5 kDa. On the basis of the matched peptides and the deduced masses, five of the characterized proteins (spots 810, 819, 826, 843 and 876) may represent the catalytic unit. Interestingly, it is possible that spots 1481, 1482 and 1768, identified as PPO and having deduced masses of around 18 kDa, may correspond to the C-terminus of the enzyme. This was supported by the similarity of the molecular weight and by the evidence that the detected tryptic peptides are comprised in the part of the sequence between the hypothesized cleavage site and the C-terminus. This may indicate that the small terminal portion of PPO is maintained in skin cells after the cleavage from the catalytic unit. The role of this fragment is not known but it was recently indicated that its tertiary structure is likely to be similar to that of hemocyanin, an oxygen-binding protein isolated in the blood of molluscs whose main function resides in O2-storage and transport [53].A spot corresponding to catalase (CAT, spot 521) presented a four-fold increase in abundance during ripening. An opposite behaviour was described for this enzyme in some recent reports on whole berries [25,40]. Although the influence of some factors can not be excluded, such as the genetic background and the environmental and seasonal conditions, these results could be explained by considering them as specific traits of the skin. For instance, it was recently discovered that the concentrations of ascorbate and glutathione in apple epidermis were 3- to 7-fold higher than in the underlying mesocarp [54]. In this view, we also observed a clear increase in the expression of a glutathione peroxidase (GPOX, spot 1408, Figure 5).Proteins involved in C-metabolismAmong the characterized proteins, many are involved in primary activities, such as glycolysis, gluconeogenesis, C-compounds and carbohydrate metabolism (Table 1 and Figure 3). A general picture of some traits of carbon metabolism showing the trend of these proteins is depicted in Figure 7.The understanding of grape assimilate partitioning, i.e. the process which determines the way carbohydrates are transported to the berry and how they are allocated, significantly improved in recent years. Sucrose is the preferred sugar for long-distance transfer in this species and is produced through photosynthesis in the mesophyll of mature leaves and conveyed to the berry from the phloem [55]. Until véraison most of the sugar imported into the berry is metabolized and so there is little storage. After véraison, there is an upturn in sugar levels, among which glucose and fructose, that are the most representative carbohydrates, are accumulated in roughly equal amounts in the vacuoles of the mesocarp cells [4]. A number of reports indicates that, during ripening, the localization of sucrose hydrolysis shifts from the vacuole to the apoplast [7,22,56]. This transition is associated to a decrease in the expression and activity of vacuolar invertases and a concomitant upturn of apoplastic acid invertases [7]. In agreement with these reports, we identified two spots (spots 412 and 431) corresponding to a vacuolar invertase, GIN1, showing a strong reduction in their expression after véraison (Figure 6).Figure 6Changes in the expression of proteins involved in C- and N-metabolism or with other functions. Changes in the relative spot volumes of the identified proteins belonging to the indicated functional categories (Table 1), during five different ripening stages of cv. Barbera grape berry skins from véraison until full ripening. The véraison stage (0 DAV) was considered as the moment when 50% of the berries started to change colour. Proteins were grouped according to their functions. Values are the mean ± SE of six 2-DE gels derived from two independent biological samples analyzed in triplicate.Figure 7Schematic overview of the enzymes involved in sugar and organic acid metabolisms and their connection with some intermediary activities that changed in expression in grape berry skins during five different ripe stages from véraison until full ripening. The expression was evaluated by measuring relative spot volumes in the 2-DE analysis. Green or red arrows indicate whether the abundance of the identified proteins decreased or increased during ripening, respectively. IRV1, cell wall invertase, GIN1, vacuolar invertase; Susy, sucrose synthase; UGP, UDP-glucose-pyrophosphorylase; PGluM, phosphogluco-mutase; PGI, phosphogluco-isomerase; PFK, phosphofructokinase; ALD, aldolase; TPI, triosephosphate-isomerase; G3PDH, glyceraldehyde-3-phosphate-dehydrogenase; PGK, phosphoglycerate-kinase; PGlyM, phosphoglycerate-mutase; ENO, enolase; PK, pyruvate kinase; PDC, pyruvate decarboxylase; NADP-ME, NADP-dependent malic enzyme; ADH, alcohol dehydrogenase; PDH, Pyruvate dehydrogenase.The measured drop in titratable acidity is mainly ascribed to the catabolism of malate accumulated in the vacuole during stages I and II of berry development [57]. It has been suggested that this acid is degraded in grape via at least three pathways, mainly by the cytosolic NADP-malic enzyme (NADP-ME), which catalyzes the oxidative decarboxylation of malate into pyruvate and CO2 [58], and, to a lesser extent, by PEP carboxykinase and the cytosolic malate dehydrogenase (cMDH) [59]. In our work we identified a spot corresponding to NADP-ME whose amount gradually increased during ripening (Figure 6). The role of this enzyme during berry development is still a matter of debate: in their tissue-specific transcriptional profile of ripe skins, Grimplet and co-workers [17] recently pointed out that the mRNA levels of several enzymes involved in malate metabolism are higher in the skins than in pulp and seeds.In the past, several papers concerning the whole berry [25] and isolated pulp or seeds [59] reported that glycolysis is down-regulated after véraison. Differently, in some transcriptomic analysis conducted on the whole berry it was found that some enzymes belonging to this pathway were induced during ripening [33,40]. We have been the first, to our knowledge, who found that several glycolytic enzymes strongly increased in the skin during ripening (Figure 6). Most of them, e.g. phosphoglycerate mutase (PGlyM, spots 397 and 1767), enolase (ENO, spots 561 and 596), glyceraldeyde-3-phosphate dehydrogenase (G3PDH, spot 902 and 937) and phosphoglycerate kinase (PGK, spot 863), related to the energy-conserving reactions of glycolysis. These data underline the importance of distinguishing among the different berry tissues in order to understand the ripening process. In other words, the tissues could express different trends for glycolysis during ripening. In this view, we also found the concomitant high expression of NADP-ME as well as of the non-oxidative activities of the pentose phosphate pathway, such as the highly induced transketolase (TK, spots 325 and 327). These enzymes may be required in the skin for satisfying the large demand for carbon skeletons of the biosynthetic pathways operating in this tissue during ripening (e.g. anthocyanin synthesis).Pyruvate may be channelled into the Krebs cycle and is converted to Acetyl-CoA by the pyruvate dehydrogenase. According to an increase in fluxes towards TCA cycle, it has been found that the subunit E1 of this enzyme (PDHE1, spot 851) is more abundantly expressed towards maturity. Aconitase (ACO) is an enzyme of the TCA and glyoxylate cycles catalyzing the reversible conversion of citrate to isocitrate. The importance of this enzyme was emphasized by Carrari and co-workers [60] who studied the Aco-1 tomato mutant which is characterized by a reduced expression of aconitase. Biochemical analysis of the leaves of this genotype suggested that Aco-1 exhibited a restricted flux through the Krebs cycle and reduced levels of Krebs cycle intermediates, with an elevated rate of photosynthesis and sucrose synthesis. The fact that Aco-1 leaves were also characterized by a different amino acid profile, indicates that this activity may have a role in controlling the C/N ratio and amino acid biosynthesis. We observed a spot corresponding to ACO (spot 191) whose expression sharply increased during ripening (Figure 6) as previously reported for cv. Cabernet Sauvignon skins [24] and, at the transcriptomic level, for citrus fruit flesh [61].Oxalyl-CoA decarboxylase (OxD, spot 413) is another protein whose levels increased during ripening. This enzyme catalyses the irreversible decarboxylation of Oxalyl-CoA, derived from glyoxylic acid, to produce formyl-CoA. This activity has already been associated to grape skin during ripening [24], but further analyses are required in order to clarify its role in this process, as far as that of one- and two-carbon compounds.Proteins involved in N-metabolismIt has been observed that the amino acid content of the berry rises significantly during maturation and that the relative amount of different amino acids changes, with proline and arginine generally being predominant [62]. Stines and co-workers [63] suggested that proline accumulation may be achieved via the ornithine pathway under the control of ornithine aminotransferase (OAT), which constitutes a bridge between proline and arginine metabolism. In support of this view, we identified a very low abundance spot (< 0.1 %Vol) corresponding to OAT (spot 654) which sharply increased in expression during ripening (Figure 6).As previously described by Giribaldi and co-workers [25], in our study we found the protein cobalamin-independent methionine synthase (MetSy, spots 270 and 273), which catalyzes the final step of methionine biosynthesis. The exact role of this enzyme, whose expression peaked in the middle of ripening, still remains unclear.Interestingly, we identified a spot corresponding to a subunit precursor of the enzyme γ-aminobutyrate transaminase (ATpL3, spot 612) which is involved in the shunt of the aminoacid γ-aminobutyrate (GABA). To our knowledge there is no evidence of the involvement of this enzyme in the maturation of the grape berry, but it is known that it is involved in the ripening of other non-climacteric fruits, such as citrus [61,64]. According to the hypothesis proposed for citrus fruit, the GABA shunt may be active, among other things, in the regulation of cytoplasmic pH, due to the H+-consuming decarboxylation of glutamate, during the period of late development and ripening following citrate release from the vacuole [61].Other proteinsThe most abundant protein found in the present work belongs to the family of ABA stress responsive elements (ASR, ca. 13% of the total volume at the first stage and ca. 6% thereafter). According to previous results, the spots 1318, 1358 and 1417 (Figure 3), which are referable to ASR, showed a downward trend during ripening (Figure 5). ASR are known to be involved in abiotic stress and fruit ripening, even though their exact role is still elusive [65,66]. The effective function has been questioned because of their very high expression level, the fact that the observed molecular masses were higher than the predicted theoretical values by a range of about 5–10 kDa and because they were found mainly in the cell wall enriched fraction [23,67].Heat shock proteins (HSP) are usually involved in stabilizing protein folding in response to different kinds of stimuli. We identified three spots corresponding to chaperones of a predicted mass of around 18 kDa (MChap and Hsp18.2, spots 1449, 1513 and 1533, Figure 3) and a heat shock chaperonin binding motif protein (HSC, spot 490) whose levels decreased after véraison (Figure 6). This evidence reinforces the conclusions of da Silva and co-workers [47] who supposed that the peak of several HSPs expression at véraison, followed by their sudden drop, could be linked to the intense redirection of metabolism that is necessary to stabilize old and newly synthesized proteins.Finally, some proteins characterized in this study were involved in transcription (spots 1189 and 1511), protein synthesis (spot 1606), signal transduction (spot 1016) and secondary metabolism (spots 986, 1008 and 1028). Further work is necessary to define the effective role of these proteins in the skin during ripening.ConclusionThis work gives new insights to the skin proteome evolution during ripening, focusing on some interesting traits of this tissue. In this view, we observed the ripening-related induction of the enzymes of the last five steps of glycolysis, although they had been described as down-regulated in previous studies performed on whole fruit. These variations were accompanied by the rise of the levels of other important proteins of primary metabolism, such as malic enzyme, aconitase, pyruvate dehydrogenase and transketolase.These results paved the way for investigations on the role of this tissue that has to respond to specific metabolic requests being the site of important biosynthetic pathway (e.g. anthocyanin). Moreover, the data emphasize the relevance of the skin as a physical barrier playing an important role in berry protection. In fact, the levels of many proteins known to take part in (a)biotic stress responses vary during the five analyzed stages. Many of them (i.e. chitinase, thaumatin-like, abscissic stress ripening protein, polyphenol oxidase) are the most expressed proteins found in this work and are characterized by the most abrupt variations in accordance to their possible developmental regulation.MethodsPlant material and experimental designExperimental material was harvested during the 2005 growing season from Vitis vinifera L. cv. Barbera grapevines, grown at the Experimental Station of the Ente Regionale per i Servizi all'Agricoltura e alle Foreste (E.R.S.A.F.) of Regione Lombardia (Pavia, Italy).Samples were collected at five different ripening stages from véraison, until full ripening (corresponding to 58, 72, 86, 100, 107 days after blooming). We considered the véraison stage as the moment when 50% of the berries started to change colour.Two hundred berries were collected at each sampling date. Berries were equally sampled on a single cluster per plant across 20 plants. Immediately after harvest, the skins were collected by squishing the berries in order to remove seeds and the bulk of the mesocarp, then pressing and smearing the inner part of the skin on two layers of cheesecloth to completely take away the residual pulp. Skins samples were split into two technical replicates. The samples were frozen in liquid nitrogen and stored at -80°C until use. Each technical replicates was subjected to independent protein extraction. Three gels were run for each extraction. At all stages, samples of whole fresh berries, obtained as described above, were immediately used to measure total soluble solids, pH and titratable acidity.Determination of physiological parametersIn order to assess the progress of grape berry ripening and to associate the physiological phases to the observed changes in protein expression, total solids, pH, titratable acidity and anthocyanins were evaluated on five stages of ripening, starting from véraison to full maturation.Total soluble solids (°BRIX), pH and titratable acidity were measured in grape juice, obtained by pressing fresh berries with a small hand-crank press, using a hand held refractometer (ATAGO CO., Ltd), a pH meter (Hanna HI 221) and an automatic titrator (Crison Compact Titrator) titrating in the presence of NaOH. Anthocyanins were extracted from the skins as previously described by Fumagalli and co-workers [68]. The anthocyanins concentration was evaluated by measuring the absorbance of the extract at a wavelength of 535 nm and referring the values to a malvidin-3-glucoside calibration curve.Considering the whole period, a sharp increase in the anthocyanin content of the skin, soluble solids and pH in berry juice was measured, while a reduction in titratable acidity occurred at the same time (Figure 8). In detail, we observed a 10-fold surge in the anthocyanin level and a 2-fold upturn of soluble solids, accompanied by a pH shift of 0.4 and a 3-fold decrease in titratable acidity. The rate of sugars and anthocyanins accumulation as well as the changes in pH and titratable acidity were almost constant until the fourth stage, while no significant variations for these parameters were observed between the fourth and the fifth sampling stages.Figure 8Biochemical changes occurring during the ripening of Barbera berries. Changes in the physiological parameters were measured during five different ripening stages of cultivar Barbera grape berries from véraison until full ripening. The véraison stage (58 days after blooming) was considered as the moment when 50% of the berries started to change colour. A, total soluble solids; B, titratable acidity; C, berry juice pH; D, total anthocyanin contents. The data are the means ± SE of three experiments run in triplicate (n = 9).Protein extraction and quantificationFrozen samples (5 g) were finely powdered in liquid nitrogen using a pestle and mortar, homogenized with cold (-20°C) acetone, washed twice on Whatman 41 filter paper (Whatman International Ltd) with cold acetone and finally dried under vacuum. The acetone powder was then resuspended in 20 mL of extraction buffer [0.7 M sucrose, 0.5 M Tris-HCl pH 8, 10 mM disodium EDTA salt, 4 mM ascorbic acid, 1 mM PMSF, 1 μM leupeptin, 0.1 mg mL-1 Pefabloc (Fluka), 0.4% (v/v) β-mercaptoethanol] on ice, incubated in a 4°C cold room under shaking for 30 min and then centrifuged at 13000 g for 30 min. The resultant supernatant was extracted as previously described by Hurkman and Tanaka [32] by the addition of an equal volume of ice-cold Tris-buffered phenol (pH 8). The sample was shaken for 30 min at 4°C, incubated for 2 h at 4°C and finally centrifuged at 5000 g for 20 min at 4°C to separate the phases. The upper phenol phase was collected, while the aqueous phase at the bottom was back-extracted with an equal volume of phenol. Proteins were precipitated by the addition of five volumes of ice-cold 0.1 M ammonium acetate in methanol to the phenol phase, then vortexed briefly and finally incubated at -20°C overnight. Precipitated proteins were recovered by centrifuging at 13000 g for 30 min, then washed again with cold methanolic ammonium acetate and three times with cold 80% (v/v) acetone. The final pellet was dried under vacuum and dissolved in IEF buffer [7 M urea, 2 M thiourea, 3% (w/v) CHAPS, 1% (v/v) NP-40, 50 mg mL-1 DTT and 2% (v/v) IPG Buffer pH 3–10 (GE Healthcare)] by vortexing and incubating for 1 h at room temperature. The sample was centrifuged at 10000 g for 10 min and the supernatant stored at -80°C until further use. The protein concentration was determined by 2-D Quant Kit (GE Healthcare).2-DEThe protein sample (200 μg) was loaded on pH 3–10, 24 cm IPG strips passively rehydrated overnight in 7 M urea, 2 M thiourea, 3% (w/v) CHAPS, 1% (v/v) NP-40, 10 mg mL-1 DTT and 0.5% (v/v) IPG Buffer pH 3–10. IEF was performed at 20°C with current limit of 50 μA/strip for about 90 kVh in an Ettan IPGphor (GE Healthcare) using the following settings: 5 min gradient 200 V, 1 h at 200 V, 5 min gradient 500 V, 1 h at 500 V, 5 min gradient 1000 V, 6 h at 1000 V, 3 h gradient 8000 V and 9 h at 8000 V. After IEF, strips were equilibrated by gentle shaking for 15 min in an equilibration buffer [100 mM Tris-HCl pH 6.8, 7 M urea, 2 M thiourea, 30% (w/v) glycerol, 2% (w/v) SDS] added with 0.5% (w/v) DTT for disulfide bridges reduction and for an additional 15 min in the same equilibration buffer to which was added 0.002% (w/v) bromophenol blue and 4.5% w/v iodoacetamide for cysteine alkylation. Second-dimensional SDS-PAGE [69] was run in 12.5% acrylamide gels using the ETTAN DALT six apparatus (GE Healthcare). Running was first conducted at 5 W/gel for 30 min followed by 15 W/gel until the bromophenol blue line ran off.Protein visualization and image and data analysisProteins were stained using the colloidal Coomassie Brilliant Blue G-250 (cCBB) procedure, as previously described by Neuhoff and co-workers [70]. The gels were scanned in an Epson Expression 1680 Pro Scanner and analyzed with ImageMaster 2-D Platinum Software (GE Healthcare). Automatic matching was complemented by manual matching. Molecular weights of the spots were deduced on the basis of the migration of SigmaMarkers™ wide range (MW 6.500 – 205.000), while pI was determined according to the strip manufacturer's instructions (GE Healthcare).Relative spot volumes of the six replicate gels of the five ripening stages were compared and were analyzed according to the ANOVA test to verify whether the changes were statistically significant (p < 0.01). Only spots showing at least a two-fold change in their relative volumes were considered for successive analysis. Significant differences were analyzed through the two-way hierarchical clustering methodology using the software PermutMatrix [71,72]. For this purpose, the data produced by the analysis of 2-DE gels were converted into a binary matrix replacing the missing values by zero. The row by row normalization of data was performed using the classical zero-mean and unit-standard deviation technique. Pearson's distance and Ward's algorithm were used for the analysis.In-gel digestion, mass spectrometry and protein characterizationSpots were excised from cCBB-stained 2-DE gels and in-gel digested as previously described by Magni and co-workers [73]. The extracted tryptic fragments were resuspended in 0.1% (v/v) formic acid and analysed by LC-ESI-MS/MS. For all the experiments a Finnigan LCQ Deca XP MAX IT mass spectrometer equipped with a Finnigan Surveyor (MS Pump Plus) HPLC system (Thermo Electron Corporation) was used. Chromatography separations were conducted on a BioBasic C18 column (180 μm I.D. × 150 mm length and 5 μm particle size), using a linear gradient from 5% to 80% solvent B [solvent A: 0.05% (v/v) formic acid; solvent B: ACN containing 0.05% (v/v) formic acid] with a flow of 2.5 μL/min. The capillary temperature and the spray voltage were set at 220°C and at 3.0 kV, respectively. For MS/MS scans the normalized collision energy was set at 35%. Acquisitions were performed in data-dependent MS/MS scanning mode and enabling a dynamic exclusion window of 3 min.Protein identifications were conducted by using TurboSEQUEST® incorporated in BioworksBrowser 3.2 (Thermo Electron Corporation) by correlation of uninterpreted spectra to the entries of NCBI NR non-redundant (i), Vitis protein subset (ii) and Vitis EST subset (iii) databases extracted from the NCBI NR non-redundant database (ii) and ESTdb others (iii), downloaded from the National Center for Biotechnology Information (NCBI). The software was set to allow two missed cleavages per peptide and to take into account fixed modification of cysteine carboxyamidomethylation and variable modification of methionine oxidation. The parent ion and fragment ion mass tolerance were set to ± 2 Da and ± 1 Da, respectively. In order to identify proteins, only peptides with Xcorr ≥ 1.5 (+1 charge state), ≥ 2.0 (+2 charge state), ≥ 2.5 (≥ 3 charge state), peptide probability < 1 × 10-3, ΔCn ≥ 0.1 and Sf ≥ 0.70 were considered. Regarding protein identification by sequence similarity search, identified peptides were aligned against the NCBI NR non-redundant database using the FASTS algorithm [74]. Theoretical molecular weight and pI of characterized proteins were calculated by processing sequence entries at . Protein functions were assigned to MIPS FunCat  according to their role described in the literature.AbbreviationsNP-40 octylphenoxy polyethoxy ethanol; cCBB Colloidal Coomassie Brilliant Blue.Authors' contributionsASN carried out protein extraction, 2-DE, gel analysis, clustering and statistical analysis and wrote the initial manuscript draft. BP contributed to the conception of the experimental design, carried out protein characterization by LC-ESI-MS/MS, analyzed the MS data, participated in writing the methods section of the manuscript. MR performed metabolite analyses. OF and AS participated to the manuscript revision. MC contributed to the interpretation of the results and took part to the critical revision of the manuscript. LE conceived the study, coordinated the experiments, wrote and edited the manuscript. All authors read and approved the final manuscript.Supplementary MaterialAdditional file 1Data on protein identification by LC-ESI-MS/MS and bioinformatic analysis. table shows the sequence of all the peptides identified by MS/MS fragmentation and the statistical information related to peptides, proteins and alignment analyses.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2529322.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529322",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529322\nAUTHORS: Wei Dai, Jens M Teodoridis, Janet Graham, Constanze Zeller, Tim HM Huang, Pearlly Yan, J Keith Vass, Robert Brown, Jim Paul\n\nABSTRACT:\nBackgroundHypermethylation of promoter CpG islands is strongly correlated to transcriptional gene silencing and epigenetic maintenance of the silenced state. As well as its role in tumor development, CpG island methylation contributes to the acquisition of resistance to chemotherapy. Differential Methylation Hybridisation (DMH) is one technique used for genome-wide DNA methylation analysis. The study of such microarray data sets should ideally account for the specific biological features of DNA methylation and the non-symmetrical distribution of the ratios of unmethylated and methylated sequences hybridised on the array. We have therefore developed a novel algorithm tailored to this type of data, Methylation Linear Discriminant Analysis (MLDA).ResultsMLDA was programmed in R (version 2.7.0) and the package is available at CRAN [1]. This approach utilizes linear regression models of non-normalised hybridisation data to define methylation status. Log-transformed signal intensities of unmethylated controls on the microarray are used as a reference. The signal intensities of DNA samples digested with methylation sensitive restriction enzymes and mock digested are then transformed to the likelihood of a locus being methylated using this reference. We tested the ability of MLDA to identify loci differentially methylated as analysed by DMH between cisplatin sensitive and resistant ovarian cancer cell lines. MLDA identified 115 differentially methylated loci and 23 out of 26 of these loci have been independently validated by Methylation Specific PCR and/or bisulphite pyrosequencing.ConclusionMLDA has advantages for analyzing methylation data from CpG island microarrays, since there is a clear rational for the definition of methylation status, it uses DMH data without between-group normalisation and is less influenced by cross-hybridisation of loci. The MLDA algorithm successfully identified differentially methylated loci between two classes of samples analysed by DMH using CpG island microarrays.\n\nBODY:\nBackgroundDNA methylation frequently occurs in mammalian DNA at the 5 position of cytosine in CpG dinucleotides. It has been estimated that over 70% of cytosines of CpG dinucleotides are methylated in the human genome. CpG dinucleotides are under-represented in the genome and methylated CpG dinucleotides predominantly occur within repetitive elements [2]. However, there are CpG rich regions of the genome which generally remain unmethylated [3]. These CpG rich regions are known as CpG islands and are frequently located in the promoter or the first exon regions of approximately 60% of all genes [4]. The unmethylated status of CpG islands is thought to be a prerequisite state to maintain the linked gene in an active transcribed and transcriptional permissive state.Differential Methylation Hybridisation (DMH) is one of several techniques for examining CpG island methylation at a genome-wide scale that has been applied to the identification of aberrantly methylated gene promoters in various cancers [5-12]. Nouzova et al[13] modified the original method by using digestion with a methylation-dependent enzyme, McrBC. This enzyme only cleaves methylated CpG DNA sequences. Within-sample comparison is applied after competitive hybridisation with McrBC digested DNA and undigested (mock digested) DNA labelled with Cy3 and Cy5. If a locus is unmethylated the signal intensities of Cy3 and Cy5 are equivalent, while if methylated the Cy5/Cy3 (undigested/digested) ratio is greater than one. However, no common reference is generally used in the modified DMH method, and the unequal representation of methylated and unmethylated sequences due to competitive hybridisation may reduce sensitivity and specificity to detect differential methylation.Currently, Significance Analysis of Microarrays (SAM) [14] and Prediction Analysis for Microarrays (PAM) [15] are commonly applied in DNA methylation analysis. Based on the change in hybridisation relative to the standard deviation of repeated measurements, SAM assigns each gene a score that is an extension of the t-statistic. For significant genes with a score over a certain threshold, SAM uses permutations to estimate the false discovery rate (FDR). It has been implemented in many studies of gene expression data [16-21] as well as DMH data, e.g. Wei et al. [22] applied SAM to find the differential methylation of CpG island loci between ovarian caner patient groups with short and long progression-free survival (PFS). However, SAM assumes that the microarray data conform to approximate normality and symmetry, leading to the loss of power in the analysis of DMH data that are inherently skewed due to the biological features of DNA methylation in cancer and competitive hybridisation on DMH arrays (Figure 1).Figure 1Distribution of log-transformed ratio of gene expression data in breast cancer and DMH data in A2780 cell line. The left histogram shows the distribution of log-transformed ratios (cy3/cy5) in gene expression profiling data from a previous study of breast cancer 36] which is symmetric, while the right histogram shows the log-transformed ratios (undigested/digested) of DMH data from the present study which is skewed.In the modified DMH method, the ratios of raw signal intensities (undigested/digested) greater than 1 reflect the various methylation levels [13]. A ratio cut-off is generally used to identify the hypermethylated loci [7]. However, this is an arbitrary value and does not necessarily accurately reflect the various sources of variation in the experiment. It is therefore desirable to develop an algorithm to more objectively assess the methylation status of loci from DMH data.PAM is a nearest centroid shrinkage method that identifies those genes that discriminate best between classes. This technique shrinks the class gene centroid towards the overall centroid by a \"threshold\" amount after standardizing each gene by its within class standard deviation. The \"threshold\" is identified by cross-validation. This approach was applied in the study by Wei et al. [22] and showed certain power in the identification of differentially methylated loci, but PAM is designed for class prediction rather than class comparison. Although the class predictor used in PAM can reflect the difference between classes, a large number of loci actually differentially methylated between the classes are excluded to improve the accuracy of prediction.Although normalisation has become a standard procedure for the study of microarray data and is necessary for SAM and PAM analysis, unbalanced shifts in methylation status between class samples in DMH limit the use of between-class normalisation which assumes the changes are roughly symmetric. Thus, the differential methylation can be masked by the over-correction of normalisation and it would be preferable to use a method of analysis that does not require normalisation of the data.Since PAM and SAM may have limitations for analysing DMH data, we have developed an alternative approach based on the specific features and known biological properties of the arrays used for DMH analysis. The algorithm is named as Methylation Linear Discriminant Analysis (MLDA) and has been applied to identify a set of loci differentially methylated between ovarian cisplatin sensitive and resistant cancer cell lines.ResultsOutline of MLDAIn this study, we have developed a novel approach, named MLDA, for analysing CpG island microarray hybridisation data that allows the identification of differentially methylated loci. MLDA was programmed in R (version 2.7.0) and the package is available at CRAN [1]. This approach uses three relatively simple linear regression models. The first one is constructed by the log-transformed signal intensities of unmethylated features and used as the reference for unmethylation (Figure 2b). The second one is the intermediate model constructed through the point corresponding to the 97.5-quantiles residual below the first linear regression line (Figure 2c). The features with a standardised residual less than 2 from this intermediate model are used to generate the third model which is used as the reference for methylation (Figure 2d). The log likelihood ratio of a locus being methylated is then proportional to the difference between the squared standardised residual from the methylated line and that from the unmethylated line. The log likelihood threshold of zero then provides a more rational basis for distinguishing between methylated and unmethylated loci than a robust undigested/digested ratio of 1.5, as it takes into account the observed variability in the experiment.Figure 2An illustration of unmethylated and methylated model construction in MLDA in A2780 cell line. a: Three patterns can be observed on the scatter plot of log-transformed Cy3 (undigested) against log-transformed Cy5 (digested) intensities. b: The unmethylated model constructed using 94 mitochondrial sequences as a unmethylation reference. c: The intermediate model constructed through the 97.5 quantile residual. The point X is the 97.5 quantile residual. The microarray probes colored in blue (standardised residual to the intermediate model is less than 2) are selected to construct the methylated model. d: Methylated (in blue) and unmethylated (in red) models in A2780 cell line.In our approach the consistency and inconsistency rates of log likelihood ratios on dye-swapped/duplicate arrays are used to determine methylation and unmethylation cut-offs, which keep the consistency rate (CR) relatively high (about 140%) and the inconsistency rate (IR) low (about 1%). Each loci is assigned a score based on the cut-offs using the weighted methylation scoring scheme. The feature consistently identified as methylated candidates on dye-swapped/duplicate arrays are scored as 1; similarly unmethylated features are scored as -1; the rest of the feature are assigned a weighted score corresponding to their location on the plot of log-likelihood ratios (Figure 3).Figure 3Weighted scoring scheme. The microarray probes consistently identified as methylated candidates on dye-swap arrays were scored 1; similarly unmethylated microarray probes were scored -1. The rest of the microarray probes were assigned a weighted score based on their location on the plot. LRmeth: log likelihood ratio cut-off for methylated loci; LRunmeth: log likelihood ratio cut-off for unmethylated loci. LR: log likelihood ratio on dye-swapped arrays.The averaged score for each locus is calculated in each sample class (e.g. resistant or sensitive) and plotted against each other. A robust regression model is then fitted to these data. The standardised residuals from the robust regression model are assumed to follow a normal distribution N(μ, σ2). The outliers of the standardised residuals are identified as the differentially methylated loci between the class samples.DMH datasetMLDA was applied to identify the CGIs differentially methylated from DMH data derived from sensitive A2780 derivatives (A2780, A2780p3, A2780p5, A2780p6, A2780p13, A2780p14) and isogenically matched, resistant lines [23] derived by multiple exposures to cytotoxic levels of cisplatin and which are 2–5 fold resistant to cisplatin in clonogenic assays (A2780cp70, A2780/MCP1, A2780/MCP2, A2780/MCP3, A2780/MCP4, A2780/MCP5, A2780/MCP6, A2780/MCP7, A2780/MCP8, A2780/MCP9). After background correction, the log-transformed digested and undigested intensities of the 13056 microarray probes show three approximately parallel linear patterns (Figure 2a). The first pattern (digested/undigested is close to 1) represents the unmethylated sequences. The second pattern represents either hemi-methylated sequences or the unmethylated sequences cross-hybridised with the methylated ones on the panel. The third pattern represents the methylated sequences in target DNA. The methylated and unmethylated loci in target DNA can be characterised by a linear regression model for each pattern. As previously mentioned, normalisation may not be appropriate for DMH data, so the log ratios of signal intensities in two classes of samples are not at the same level (Figure 4). Normalisation is not required for MLDA as the determination of the methylation score is based on the data within each experiment.Figure 4Box plot of log ratios of undigested signal intensities against digested signal intensities in 16 cell lines (dye-swapped arrays). The boxes colored in red are the A2780 sensitive cell lines; in blue are the A2780 resistant cell lines. As normalisation is not applied, the center and scale of log ratios for the 16 cell lines are not at the same level.Mitochondrial DNA is unmethylated [24], therefore, the signal intensities of both channels of microarray probes for mitochondrial sequences are expected to be equal. However, a bi-modal distribution is observed in the log-transformed fluorescence ratios (digested/undigested) of 121 mitochondrial sequences. The first peak represents the unmethylated mitochondrial sequences and the second lower peak is assumed to be the mitochondrial sequences cross-hybridised with other methylated sequences on the panel. Thus, we selected 94 of 121 mitochondrial sequences that were consistently unmethylated through all the cell lines and used them as the unmethylation reference in target DNA.The parameters of those two models in all 16 cell lines were estimated (Table 1). The slope of the unmethylated regression line constructed by 94 mitochondrial sequences is indeed close to 1. After computing the log-likelihood ratios, the methylation and unmethylation cut-offs and associated IRs and CRs were determined from the dye-swapped array pairs (details in Method section). As shown in Figure 5, IR tends to rise with the increase of CR slowly, but starts to increase dramatically when the CR goes above 140%, at which point IR is generally about 1%. We have therefore used CR > 140% and IR < 1% as the criteria for determining the methylation and unmethylation cut-offs. Each locus was scored using the weighted scoring scheme based on those cut-offs. The averaged scores in 6 cisplatin-sensitive cell lines and 10 cisplatin-resistant cell lines were used to construct a robust regression model. Figure 6a shows that the standardised residuals (residual/σ) from the robust regression model roughly follow a normal distribution. The positive and negative outliers are determined as described in Method section.Figure 5CR against IR in 16 cell lines. X axis is the consistency rate (CR) and y axis is the inconsistency rate (IR). IR tends to rise with the increase of CR slowly, but starts to increase dramatically when the CR goes above 140%, at which point the inconsistency rate is generally about 1%. Not all cell lines could reach this point e.g. MCP3.Figure 6Outliers identifications. a: Distribution of the observed (histogram) standardised residuals and the theoretical distribution based on the fitted model (dashed smooth line in red). The red and blue solid line are the positive and negative cut-offs, respectively. b: Scatter plot of sensitive scores against resistant scores in A2780 series cell lines. The hypermethylated loci are colored in red and hypomethylated loci are in blue. The robust regression model is Y = 0.9956X + 0.0019.Table 1Parameters of linear models in MLDA for 16 cell lines in DMH dataset IUnmethylation linear regression modelcell lineintercetp (α)slope (β)σR2interceptds (α)slopeds (β)σdsR2dsA2780-0.00031.01220.17270.98290.00051.05740.18970.978A2780p3-0.00031.03430.12120.98970.00181.10650.14250.9882A2780p5-0.00031.01380.16840.984-0.00021.07280.14250.9883A2780p60.00020.99140.16050.9778-0.00011.00120.16380.9747A2780p13-0.00121.0240.16280.9786-0.00051.07440.14360.9852A2780p14-0.00221.04990.15230.9809-0.00091.0340.20690.9691A2780cp700.00130.96040.25320.9524-0.00021.01190.24020.9479MCP10.00020.99460.1450.9803-0.00231.1120.14520.9836MCP20.00000.9980.1370.9727-0.00281.04750.17190.9653MCP30.00040.99320.22530.9183-0.00231.05170.27950.8978MCP40.00060.98380.17390.9718-0.00281.0770.19470.9751MCP50.00090.98570.24640.9639-0.00081.0170.21660.9692MCP6-0.00221.03520.1220.9751-0.00681.12830.1540.9752MCP7-0.00051.00790.13790.9791-0.00451.15290.15880.9764MCP8-0.00281.05780.19030.9431-0.00681.11930.18850.9575MCP9-0.00171.03310.18340.9614-0.00911.15380.16910.9674Methylation linear regression modelcell lineintercetp (α)slope (β)σR2interceptds (α)slopeds (β)σdsR2dsA2780-0.88390.89170.14910.9438-0.80550.90860.17060.926A2780p3-1.16720.97970.15180.9553-0.74140.97740.1790.9438A2780p5-0.89910.89780.15150.9476-0.92460.97660.16730.9523A2780p6-0.94550.95620.18380.9378-1.19950.96410.17880.9324A2780p13-1.89180.98070.25350.8962-1.80490.96520.29360.8512A2780p14-1.56370.91420.25490.8857-1.44680.90660.21280.8837A2780cp70-1.03170.85010.15810.9115-1.30740.89670.15410.9265MCP1-1.1990.97810.16920.9467-1.09351.03840.17750.9525MCP2-0.80370.92920.14860.9557-0.97380.93810.21760.8848MCP3-1.12440.91510.17550.9482-0.93030.92050.25990.8098MCP4-1.43260.91710.14180.966-1.62050.9610.23480.8323MCP5-1.11870.94250.18390.9404-1.20070.95460.17570.9295MCP6-1.2460.9250.19660.9294-1.21820.98260.22480.8989MCP7-1.89720.99090.19770.9442-1.48941.01390.24580.8886MCP8-1.02190.99050.19750.9468-0.47350.94210.2280.8761MCP9-1.33990.98370.20730.9352-1.14971.00780.19670.9115ds: dye swapσ: standard deviationR2: coefficient of determinationFinally, 115 loci were identified as candidates differentially methylated between A2780 sensitive and these resistant cell lines (additional file 1). Noticeably, 113 of 115 loci (p = 8.8 × 10-3, outlier detection test [25]) were hypermethylated, but only 2 loci (p < 0.001, outlier detection test) lost methylation in the resistant cell lines (Figure 6b). This is consistent with the unbalanced shift in DMH data and indicates cisplatin treatment of cells selects preferentially for hypermethylation of loci, rather than hypomethylation in these tumor cells.Validation of differential methylationTo confirm the differential methylation of loci identified in this study, we experimentally tested the methylation of 26 loci by methylation-specific PCR (MSP) and/or pyrosequencing of bisulphite modified DNA [26] in sensitive A2780 derivatives and cisplatin resistant derivatives. Twenty-three out of the 26 loci have been confirmed as differentially methylated (additional file 1). It should be noted that MSP and pyrosequencing only examine methylation at a limited number of CpG sites of the sequence present on the DMH analysis. It is possible that the loci which were not confirmed as differentially methylated are methylated at other CpG sites which are detected by DMH but not targeted by MSP and/or pyrosequencing primers and so 23 out of 26 loci confirmed as differentially methylated is a minimum estimation.To compare the results from MLDA, SAM and PAM, we analysed the DMH dataset by all three methods. MLDA identified 115 loci (113 hypermethylated and 2 hypomethylated loci, misclassification error < 0.001), SAM identified 152 loci (149 hypermethylated and 3 hypomethylated loci, misclassification error = 0.227, FDR = 6.17 × 10-3), and PAM found 24 hypermethylated loci (misclassification error = 0.084, FDR < 0.001) in the resistant cell lines. Twenty-four loci identified by all three methods are listed in Table 2.Table 224 loci identified by MLDA, PAM and SAM as differentially methylated candidates in the comparison between A2780 cisplatin sensitive and cisplatin multiple-selected resistant cell lines.microarray IDstatusvalidationMLDA rank*PAM rankSAM rankCGI***gene symbol**GenBank AccessionChromosome66_G_6hypermethylatedYes111Yes121_D_9hypermethylatedYes256YesCRABP1NM_0043781539_E_1hypermethylatedND322No122_D_9hypermethylatedNo41111YesSOX12NM_00694320123_D_9hypermethylatedNo51010YesSOX12NM_0069432051_H_8hypermethylatedYes61819NoFEZF2NM_018008358_A_1hypermethylatedND72222Yes80_H_5hypermethylatedND81416No21_A_11hypermethylatedYes91718YesNTN4NM_0212291238_D_7hypermethylatedYes112324YesAGBL2NM_0247831140_E_1hypermethylatedND121917No18_A_7hypermethylatedND132020NoEDIL3NM_005711555_F_8hypermethylatedYes1499YesBC127881BC1278817122_B_1hypermethylatedND151614No109_A_6hypermethylatedNo181213Yes41_D_9hypermethylatedYes2288YesWNT1NM_0054301242_D_9hypermethylatedND2344No119_A_6hypermethylatedYes2465YesNR2E1NM_003269663_A_8hypermethylatedND261515No6_D_4hypermethylatedYes3133YesLMX1ANM_177398117_H_9hypermethylatedYes341312YesHRASLS3NM_00629065_D_4hypermethylatedYes3577YesLMX1ANM_177398124_D_3hypermethylatedYes752423YesSP5NM_0010038452122_G_1hypermethylatedND1012121YesND: not done. Yes: validated. No: not validatedMLDA rank*: the rank of standardised residuals to the robust regression line constructed by the averaged sensitive scores against averaged resistant scoresGene symbol**: only the gene of which transcription start site (TSS) is within 5 kb span of the lociCGI***: CpG island defined by Gardiner-Garden and Frommer [31].DiscussionHypermethylation of promoter CpG islands is strongly correlated to transcriptional gene silencing and epigenetic maintenance of the silenced state and is a potential rich source of biomarkers of cancer. Differential Methylation Hybridisation (DMH) is one technique used for genome-wide DNA methylation analysis. The study of such microarray data sets should ideally account for the specific biological features of DNA methylation and the non-symmetrical distribution of the ratios of unmethylated and methylated sequences hybridised on the array. We have therefore developed a novel algorithm tailored to this type of data, Methylation Linear Discriminant Analysis (MLDA). MLDA utilises log likelihood ratios representing the relative probability that loci are methylated instead of log ratios of signal intensities used in previous studies [6-10,27]. Validation of 23/26 identified loci using independent methods of methylation analysis shows that MLDA can robustly identify differential methylated loci between ovarian cancer sensitive and resistant cell lines without requiring the data to be normalised.Although a log likelihood ratio above zero means that the locus tends to be methylated, we did not use zero as the cut-off to determine the number of methylated and unmethylated sequences, as the existence of cross-hybridisation and measurement errors in the DMH assay makes this unreliable. To increase the precision of the methylation classification, we used the inconsistency (IR) and consistency (CR) rates between the dye-swap arrays to determine likelihood ratio cut-offs for methylation and unmethylation and assigned each locus a methylation score based on the position relative to these cut-offs. As shown in Figure 5, not all cell lines can reach the point that CR is around 140% and IR is about 1%. IR and CR need to be carefully selected as the methylation scores of loci are consequently influenced by the change of IR and CR. We also observed a lower CR (about 120%) and a higher IR (about 2%) in another CpG island array using DMH (data not shown), therefore, further examination of what factors influence the achievable CR and IR rates may improve the utility of the MLDA approach.Data on methylation status for 121 mitochondrial derived sequences were available in this study. Mitochondrial sequences would be expected to be unmethylated. We used 94 mitochondrial sequences to construct unmethylated linear model at the beginning of the study, and indeed, 93 of 121 mitochondrial loci were defined as unmethylated and 25 loci being of uncertain methylation status by MLDA. However, three mitochondrial loci were identified as hypermethylated candidates in the resistant ovarian carcinoma cell lines by both MLDA and SAM. One explanation of this discrepancy is that all these three loci have more than one BLAT hit indicating the existence of homology with nuclear DNA sequences, raising the possibility of hybridisation with these nuclear DNA sequences which may be differentially methylated. As shown in Figure 2a, the loci in the middle pattern represent either hemi-methylated sequences or the unmethylated sequences cross-hybridised with the methylated ones on the panel. No specific allowance is made for these intermediate points in analysis by SAM and PAM, whereas MLDA attempts specifically to down-weight these points in the identification of the methylation regression line. By giving a lower weighted score (close to 0) (Figure 3) to those loci, MLDA reduces the influence of cross-hybridisation among this group of sequences. Of course cross-hybridisation may also occur in the loci in the other two patterns (methylated and unmethylated patterns), but it is not possible for any mathematical approach to identify this.The misclassification error of MLDA based on the methylation score is much lower than that for either SAM or PAM based on the log ratios, indicating the potential of MLDA methylation scores to be used as a reliable discriminator between classes of samples.ConclusionWe have developed a novel method, named MLDA, for genome-wide DNA methylation studies. MLDA can transform the signal intensities to log-likelihood ratios through three linear regression models. Using this approach MLDA allows determination of the methylation status of a locus based on dye-swapped/duplicate arrays. The method has been applied to assess the methylation status of each locus and identified 115 loci that exhibit differential methylation between A2780 sensitive and resistant cell lines. A minimum of 23 out of 26 loci have been confirmed by independent methods as differentially methylated.MethodsFirst, all intensity values were log transformed. A multiplicative background correction was applied to correct signal intensities for the background noise in each array. After background correction, the log-transformed digested and undigested intensities show three approximately parallel linear patterns (Figure 2a). The first pattern (digested/undigested is close to 1) represents the unmethylated sequences. The second pattern represents either hemi-methylated sequences or the unmethylated sequences cross-hybridised with the methylated ones on the panel. The third pattern represents the methylated sequences in target DNA. The methylated and unmethylated loci in target DNA can be characterised by a linear regression model for each pattern. The distance of each spot to the methylated and unmethylated lines respectively can then be estimated by standardised residuals. The log likelihood ratio of a locus being methylated is then proportional to the difference between the squared standardised residual from the methylated line and that from the unmethylated one. The algorithm based around this regression approach is named Methylation Linear Discriminant Analysis (MLDA) and was programmed in R version 2.7.0.Log-likelihood ratio transformationa An univariate linear regression model was constructed for the unmethylated probes (e.g. mitochondrial derived features) using formula (1) where α is the intercept, β is the slope of the model, and ξ is the error representing the unpredicted or unexplained variation in the model (Figure 2b). The parameters of regression line were estimated by the method of least squares (formula 2 and 3).(1)Gi = α + βRi + ξi i = 1,2,3.......k(2)β^=∑(Ri−R¯)(Gi−G¯)∑(Ri−R¯)2i=1,2,3.......k(3)α^=β^R¯−G¯k is the number of unmethylated controls on DMH array. Gi and Ri are the logarithmic-transformed digested and undigested intensities of microarray probes for mitochondrial sequences, respectively. G¯ and R¯ are the averaged logarithmic-transformed undigested and digested intensities of the k unmethylated controls.b. The scale estimate σmito associated with the error term in the linear regression model was estimated from the residuals from the observed k points to the fitted line. The most extreme 10% of residuals was omitted from either end of the distribution to minimise the impact of extreme residuals on this estimate.c. The standardised residuals of all the microarray probes to the unmethylation regression line were calculated as formula (4).(4)SRmito=residualsmitoσmitod. The point corresponding to the 97.5-quantiles residual below the unmethylation line is represented as X (R.975, G.975). The intermediate linear model (Figure 2c) was constructed through point X with a slope assumed to be 1 and the intercept estimated as formula (5).(5)α^=G.975−R.975+1.96σmitoe. The standardised residuals of all the microarray probes to the line with slope 1 and intercept estimated from (5) were calculated as formula (6). The variance of the residuals to the intermediate model was assumed to be similar as that in the mitochondrial model.(6)SR.975=residuals.975σmitof. The microarray probes with standardised residuals less than 2 were included for later robust regression analysis. The line estimated from this regression analysis represents the methylation regression line (Figure 2d).g. The scale estimate σmeth of the methylation regression line was estimated using only those microarray probes below the line, with the most extreme 5% removed.h. The standardised residuals of all the microarray probes to the methylated regression line were calculated as formula (7). The log likelihood ratio (LR) of all the microarray probes was estimated by formula (8) for further analysis.(7)SRmeth=residualsmethσmeth(8)LR = SR2mito - SR2methDetermination of log likelihood ratio cut-offsTwo inconsistency rates (IRmeth and IRunmeth) and two consistency rates (CRmeth and CRunmeth) between dye-swap arrays were used to determine the log like likelihood ratio threshold. IRmeth (formula 9) represents the rate of the microarray probes identified as methylated in one array but as unmethylated in the other one, while IRunmeth (formula 10) is the rate of the microarray probes identified as unmethylated in one array but as methylated in the other one. CRmeth (formula 11) and CRunmeth (formula 12) are the rates for the spots identified as methylated (CRmeth) and unmethylated (CRunmeth) in both dye-swap arrays (Figure 7).Figure 7Determination of methylation and unmethylation cut-offs of likelihood ratios on dye-swapped arrays. LRmeth: log likelihood ratio cut-off for methylated spots; LRunmeth: log likelihood ratio cut-off for unmethylated spots.Table 3Classification of lociUnmethylateduncertainmethylatedUnmethylatedabcUncertaindefMethylatedghi(9)IRmeth=cc+f+i+gg+h+i(10)IRun meth=ga+d+g+ca+b+c(11)CRmeth=ic+f+i+ig+h+i(12)CRmeth=aa+b+c+aa+d+gb The log likelihood ratio thresholds (LRmeth and LRunmeth) for methylated and unmethylated microarray probes, which kept the IR rates low (at or close to 1%) and the CR rates high (at or close to 140%), were used as the cut-offs for methylated and unmethylated loci. IR tends to rise with the increase of CR slowly, but starts to increase dramatically when the CR goes above 140%, at which point the inconsistency rate is generally about 1%. We have therefore used CR > 140% and IR < 1% as the criteria for determining the methylation cut-offs.Identification of robust regression outliersEach microarray probe was scored based on the cut-offs of likelihood ratios for methylation and unmethylation on dye-swap arrays using the weighted methylation scoring scheme shown in Figure 3. The microarray probes consistently identified as methylated candidates on dye-swap arrays were scored of 1; similarly unmethylated microarray probes were scored of -1. The rest of the microarray probes were assigned a weighted score based on their location on the plot.A robust regression model [28] was constructed with the averaged scores in one class of samples as the explanatory variable, and the corresponding scores in the other class of samples as the dependent variable. The degree of trimming was determined according to Barnett et al. [29] when estimating the variance of residuals to the robust linear regression model.It was assumed that the standardised residuals (SRs) from the robust regression line followed a normal distribution N(μ, σ2). μ and σ were estimated excluding outliers using the MAD-Median Rule [30]. The p value for each SR cut-off was calculated as described by Simon et al [25]. This p-value reflects the probability of observing a group of more extreme residuals from the fitted normal distribution. Microarray probes were identified as outliers if their SRs were larger than the cut-off for which the p-value was less than 0.01.Estimation of misclassification rateThe misclassification rate was estimated by drawing bootstrap samples 500 times with replacement from the two classes (sensitive and resistant) and carrying out hierarchical clustering based on the loci identified as differentially methylated using weighted scores for MLDA and log ratios without between-group normalisation for SAM and PAM, respectively. Clustering was carried out using Euclidean distance as the distance metric, and clusters were agglomerated using the average linkage criterion. The clustering tree was cut into two groups and the number of misclassified cell lines was counted. The misclassification rate was obtained from the averaged number of misclassified samples in 500 bootstraps divided by the total number of samples.SAM and PAM analysisThe raw signal intensities of each channel were subtracted by the median signal intensities of corresponding channel of controls on HCGI12K array. After this correction, SAM in samr package and PAM in pamr package were applied using log ratios (digested/undigested) in R version 2.7.0. Between-group normalisation was not used in SAM and PAM to avoid over-correction masking the differential methylation.Authors' contributionsWD conducted the statistical analysis and algorithm development supervised by JP and RB. The DMH data was produced by JMT in collaboration with TH and PY. RB, JP and KV conceived the study. JMT, JG and CZ conducted validation by MSP or pyrosequencing in RB's lab. WD, JP and RB prepared the manuscript with review by all authors. Funding was obtained by RB and TH.Supplementary MaterialAdditional file 1115 differential methylated candidates identified by MLDA in A2780 series cell lines.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2529338.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529338",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529338\nAUTHORS: Kai Tao, Min Fang, Joseph Alroy, G Gary Sahagian\n\nABSTRACT:\nBackgroundThe 4T1 mouse mammary tumor cell line is one of only a few breast cancer models with the capacity to metastasize efficiently to sites affected in human breast cancer. Here we describe two 4T1 cell lines modified to facilitate analysis of tumor growth and metastasis and evaluation of gene function in vivo. New information regarding the involvement of innate and acquired immunity in metastasis and other characteristics of the model relevant to its use in the study of late stage breast cancer are reported.MethodsThe lines were engineered for stable expression of firefly luciferase to allow tracking and quantitation of the cells in vivo. Biophotonic imaging was used to characterize growth and metastasis of the lines in vivo and an improved gene expression approach was used to characterize the basis for the metastatic phenotype that was observed.ResultsGrowth of cells at the primary site was biphasic with metastasis detected during the second growth phase 5–6 weeks after introduction of the cells. Regression of growth, which occurred in weeks 3–4, was associated with extensive necrosis and infiltration of leukocytes. Biphasic tumor growth did not occur in BALB/c SCID mice indicating involvement of an acquired immune response in the effect. Hematopoiesis in spleen and liver and elevated levels of circulating leukocytes were observed at week 2 and increased progressively until death at week 6–8. Gene expression analysis revealed an association of several secreted factors including colony stimulatory factors, cytokines and chemokines, acute phase proteins, angiogenesis factors and ECM modifying proteins with the 4T1 metastatic phenotype. Signaling pathways likely to be responsible for production of these factors were also identified.ConclusionThe production of factors that stimulate angiogenesis and ECM modification and induce hematopoiesis, recruitment and activation of leukocytes suggest that 4T1 tumor cells play a more direct role than previously appreciated in orchestrating changes in the tumor environment conducive to tumor cell dissemination and metastasis. The new cell lines will greatly facilitate the study of late stage breast and preclinical assessment of cancer drugs and other therapeutics particularly those targeting immune system effects on tumor metastasis.\n\nBODY:\nBackgroundWhile investigation of the molecular basis of tumor metastasis has in large part focused on proliferation and dissemination of tumor cells from the primary tumor, later events that occur at sites of metastasis are most often responsible for patient mortality and morbidity. From a clinical standpoint, an understanding of the disease at metastatic sites is paramount since the number of breast cancer patients with detectable or occult metastases at the time of diagnosis is substantial and most patients will develop metastatic lesions at some point during the course of the disease. Metastasis is generally treated as a systemic disease with chemotherapy and/or radiation even though factors involved in establishment and growth of metastatic lesions differ from one site to the next and may differ in response to therapeutics. While currently used therapeutic regimens are capable of slowing the progression of metastatic disease, rarely is it possible to stop or reverse the process. Treatments that address the nature of metastatic disease at the site of metastasis could provide more effective therapeutic results for patients afflicted with the later stages of the disease.A major impediment for the study of metastasis has been the availability of suitable models that faithfully represent the metastatic process as it occurs in vivo. Xenograft models in which human tumor cells are introduced into immunocompromised mice have been used extensively for the study of tumor growth and metastasis and to validate specific gene products as drug targets for cancer therapy. While some human xenograft models can approximate primary tumor growth in mice, replication of tumor metastasis is more problematic [1-3]. Human tumor cells generally metastasize poorly in mice and when metastasis does occur, unexpected metastatic characteristics are often observed. In contrast, murine tumor cell models often metastasize more effectively and display metastatic characteristics more similar to those observed in cancer patients [4]. Given the importance of microenvironment and tumor-host interactions in tumor cell behavior, this is not surprising. Syngeneic mouse models such as the 4T1 model described here also have the important advantage of allowing analyses to be carried out in animals with normal immune function. Because the immune system plays an important role in the development and progression of cancer, models that can be used in immunocompetent mice are essential for analysis of cancer progression and evaluation of therapeutics for cancer treatment.The 4T1 mammary carcinoma cell line was originally isolated by Fred Miller and coworkers at the Karmanos Cancer Institute [5,6]. Its use has increased in recent years because of its high propensity to metastasize to bone and other sites [7,8]. When introduced orthotopically, 4T1 is capable of metastasis to several organs affected in breast cancer including lungs, liver and brain, as well as bone [7,9-11]. 4T1 sibling cell lines with different metastatic properties have been isolated and characterized. These lines were isolated from the same spontaneous arising BALB/c mammary tumor [5,6] but appear to have followed divergent pathways for acquisition of their metastatic phenotypes [12].We have modified the 4T1 cell line for optimal use as a model for the study of late stage breast cancer. A modified line (4T1-12B) expressing high levels of firefly luciferase to allow non-invasive longitudinal imaging of in vivo growth and metastasis was isolated. A similar line (4T1-1V) was further modified by insertion of a FLP recombinase target (FRT) site into the 4T1 genome. The FRT site facilitates rapid generation of genetically modified isogenic cell lines for investigation of effector gene function. The extent and kinetics of metastasis to organs affected in human breast cancer indicated extensive colonization of lungs and liver in most animals within a six week period with lower efficiency of metastasis to bone, brain and other sites. Innate and adaptive immune responses were shown to play important roles in growth and metastasis of the lines in BALB/c mice. Analysis of gene expression comparing 4T1 and two of its non-metastatic sibling cell lines suggested prominent roles for several signaling pathways and secreted factors in directing microenvironmental changes within the tumor leading to tumor cell dissemination and metastasis.MethodsMaterialsThe luciferase-containing pGL3-Control vector was obtained from Promega. The pKO-puro vector was from Stratagene. The pSHAG-1 vector was provided by Dr. G. Hannon at Cold Spring Harbor Laboratory. Other vectors including pcDNA5/FRT, pOG44, pFRT/lacZeo, and the Gateway Vector Conversion System Reading Frame Cassette C.1 were obtained from Invitrogen. Dulbecco's modified Eagle's medium (DMEM), Dulbecco's phosphate-buffered saline without calcium and magnesium (PBS), fetal bovine serum (FBS), newborn calf serum (NCS), non-essential amino acids (NEAA), penicillin, streptomycin and lipofectamine PLUS reagent were from Invitrogen. Puromycin and hygromycin were from Sigma. Luciferin was obtained from Caliper Life Sciences.Cell cultureThe 67NR, 168FARN and 4T1 mouse mammary tumor cell lines were obtained from Dr. Fred Miller at Karmanos Cancer Institute. Cells were cultured in high glucose DMEM supplemented 5% FBS, 5% NCS, NEAA and antibiotics (100 units/ml penicillin and 100 μg/ml streptomycin) at 37°C in a humidified atmosphere containing 5% CO2. Except where indicated, analyses were performed on same passage cells within 2 weeks after thawing. All cell lines used in the study were tested and shown to be free of mycoplasma and viral contamination.Expression of luciferase and puromycin resistance in 4T1 cell lines4T1 cells were cotransfected with firefly luciferase-containing pGL-3-Control vector and the puromycin resistance vector, pKO-puro, at a ratio of 10:1 using Lipofectamine PLUS as described by the vendor [Invitrogen]. Transfected cells were selected with puromycin at a final concentration of 10 μg/ml and several colonies were picked and expanded for analysis. Colonies displaying the highest level of luciferase expression were injected into mammary fat pads of female BALB/c mice and imaged 6 weeks later before and after sacrifice and necropsy, as described below. One cell line, designated 4T1-12B, which retained high level expression of luciferase in the absence of puromycin and displayed metastatic properties similar to the parental line, was retained for further analysis and modification. Sublines were obtained from the 4T1-12B line by limiting dilution cloning.Incorporation of FRT site into 4T1 cell lineTo introduce the FLP recombinase target (FRT) site in the 4T1 genome, cells were transfected with pFRT/lacZeo using lipofectamine PLUS and selected with 100 μg/ml zeocin. Colonies were picked and expanded and six clonal lines with a single integration of the vector, as determined by Southern blotting, were identified. The expanded lines were analyzed for efficiency of transfection and targeting and for in vivo tumor growth and metastasis. One line, designated 4T1-1V, displayed growth and metastatic characteristics similar to the parental 4T1 line and efficient transfection and targeting to the FRT site with pcDNA5/FRT (Invitrogen) and was retained for further analysis.Construction of shRNA targeting vectorThe pcDNA5/FRT targeting vector was modified to allow transfer of the shRNA expression cassettes from pSHAG-type shRNA expression vectors [13] to the pcDNA5/FRT vector by Gateway site-specific recombination. Resulting vectors can then be used to target shRNA expression cassettes to FRT sites in 4T1-1V and other FRT-containing cell lines for creation of isogenic cell lines. A Gateway cloning site was inserted into the vector by blunt end ligation of Reading Frame Cassette C.1 (Invitrogen) into the vector's Bgl II site.Knockdown of luciferase with shRNA targeting vectorTo test the efficacy of the construct, an empty expression cassette and a cassette encoding a previously tested luciferase shRNA [13] were transferred to the modified pcDNA5/FRT from corresponding pSHAG-1 vectors. The 4T1-1V cells were then cotransfected with each construct and pOG44 vector at a ratio of 1:10 using Lipofectamine PLUS. Transfected cells were selected with hygromycin B at a final concentration of 200 μg/ml. The pOG44 vector encodes FLP recombinase which directs insertion of the modified pcDNA5/FRT targeting vector into the cell's FRT site; hygromycin B resistance is conferred upon insertion of the vector into the site.Expanded clones of hygromycin B resistant cells as well as resistant cell pools from each transfection were then assayed for luciferase activity using a Turner Designs Model TD-20/20 Luminometer.Biophotonic imaging of animals and organsLuciferase-expressing cell lines were plated at 40% confluency and cultured for 24 h. The cells were then trypsinized, washed, and resuspended in DMEM at 107 cells/ml and kept on ice before injection. Aliquots (100 μl) of the cells were injected into the no. 4 or no. 9 fatpad of 4–6 week old female BALB/c, athymic BALB/c nude or BALB/c SCID mice using a 26-gauge needle. Only cell preparations with viability > 97%, as determined by Trypan Blue exclusion, were used for injection.At various times up to 6 weeks, animals were injected intraperitoneally with 100 μl of D-luciferin (10 mg/ml) in PBS, and after 10 min, imaged under anesthesia with 2.5% isofluorane in a Xenogen IVIS 200 biophotonic imager. At experimental endpoints, luciferin-injected animals were sacrificed and organs and hind limbs were removed and imaged within 15 minutes after injection. Luminescence is expressed as photons/sec/ROI (region of interest) minus background luminescence for a similarly sized region. All experiments with animals were carried out according to guidelines for the care and use of experimental animals and were approved by the Tufts University Institutional Animal Care and Use Committee.Chip hybridizations and analysis of expression dataBiotin-labeled cRNAs were prepared from 250 ng of total RNA using Ambion's TotalPrep RNA Amplification Kit. Chip hybridizations, washing, Cy3-streptavidin (Amersham Biosciences) labeling, and scanning were performed on an Illumina BeadStation 500 platform using reagents and protocols provided by the manufacturer. cRNA samples were hybridized to Illumina MouseRef-8 BeadChips which cover 24,048 RefSeq transcripts. The manufacturing principle of randomly distributing large populations of oligonucleotide-coated beads across the available positions on the chip enables 30 intensity measurements per feature on average, and produces quantitative results closely matching those obtained by Q-PCR [14].Biotin-labeled cRNAs were prepared from 3 biological replicates of cultured 4T1, 67NR and 168FARN cells for hybridization to the chips. Cells were plated at 5 × 105 cells in 10 cm culture dishes and after 3 days, when the cells had reached confluence, the medium was changed and the cells were cultured for an additional 24 hours. Total RNA was isolated using the Absolutely RNA Kit from Stratagene and checked for integrity using an Agilent Lab-on-a-Chip Bioanalyzer.Initial analysis of the data was carried out using Illumina's BeadStudio software. Raw data for each sample were background-subtracted and normalized using the \"cubic spline\" algorithm. Further statistical analysis of the data was carried out using programs associated with BRB Array Tools [15]. Differentially expressed genes for 4T1 samples relative to 67NR and 168FARN samples were determined separately with the Class Comparison program using the random variance model and a p value of 0.0001. Signals less than 10 were set to 10 to eliminate the inaccuracy of analyzing genes expressed at near background levels from being scored as differentially expressed. Genes that differed significantly (p < 0.0001) by > 2-fold for both comparisons (4T1/67NR and 4T1/168FARN) were considered those associated with the metastatic phenotype of the 4T1 cell line. Because of the high level of reproducibility that was achieved, a relatively high level of stringency (p < 0.0001) was chosen for selection of significant differences. As described in RESULTS, this permitted a very low level of false positives with minimal loss of true positives.Histochemistry and hematological analysisStandard H & E staining of paraffin embedded tissue was used for histological examination of primary tumors and metastases. Stained sections were examined and photographed using an Olympus Vanox-T microscope and an Olympus U-PMTVC CCD camera. Blood from control and tumor bearing animals was collected by cardiac puncture and analyzed by the Pathology Department at Tufts University Cummings School of Veterinary Medicine using standard hematological procedures.ResultsCharacteristics of luciferase-expressing 4T1 cell linesThe luciferase-expressing 4T1-12B cell line was cloned from 4T1 cells co-transfected with vectors encoding firefly luciferase and puromycin drug resistance. The level of luciferase expression is high enough to allow imaging of as few as 10 cells in vitro using a Xenogen IVIS 200 biophotonic imager; the cells are fully resistant to inclusion of 10 μg/ml of puromycin in the culture medium. The modified cell line displayed a doubling time of 12 hours in culture and a plating efficiency of 34%. Expression of luciferase persisted when the cells were cultured for extended periods (> 2 months) in the absence of selective pressure, a characteristic critical for reliable quantitation of tumor growth and metastasis in vivo.Imaging of a representative female BALB/c mouse six weeks after mammary fat pad injection of 4T1-12B cells is shown in Figure 1. At six weeks, primary tumors generally reach a size of 1 cm3 or more and metastasis to the thoracic region is apparent in most animals. Imaging of visceral organs and hind limbs revealed metastases in several organs including lungs, liver, bone and brain, sites affected in human breast cancer.Figure 1Imaging of animals and organs at six weeks. 4T1-12B cells (106) were implanted into the mammary fat pad of a normal female BALB/c mouse. After six weeks the whole animal and organs were imaged as described in MATERIALS AND METHODS. The relationship between color and light intensity in arbitrary units (counts) for the whole animal images is given by the color bar at the right side of the figure.A compilation of results for several animals injected with the 4T1-12B line and several clones isolated from the modified line after extended time in culture is shown in Table 1. All animals injected with the 4T1-12B line displayed metastases in lungs at six weeks with substantial numbers displaying metastasis to liver (5/6), spleen (3/6) and bone (2/6). Metastases were occasionally found in lymph nodes, brain, intestine, kidneys and adrenals. Recloned sublines isolated from the 4T1-12B line displayed a similar spectrum of organ metastasis indicating that most if not all of the cells in the preparation are tumorigenic and metastatic and that the tumorigenic and metastatic properties of the cells are stable when cells are expanded for as many as 20 to 30 generations in culture.Table 1Summary of sites of metastasis for 4T1-12B1Cell LinePrimaryLungsSpleenLiverBoneOther4T1-12B line6(6)26(6)3(6)5(6)2(6)Brain 1(6)Intestine 1(6)Kidney 1(6)4T1-12B recloned lines (5)10(10)8(10)2(10)3(10)6(10)Intestine 3(10)Kidney 1(10)1Cells were introduced into fatpads of normal female BALB/c mice and imaged after 5–6 weeks as described in MATERIALS AND METHODS.2Number of animals positive (total number of animals)Generation of the 4T1-1V cell line involved transfection of the 4T1 line with luciferase and puromycin vectors and a vector containing a FLP recombinase targeting (FRT) site contained within a lacZ-zeo fusion protein expression cassette as described in MATERIALS AND METHODS. The 4T1-1V line was shown to contain a single site of integration of the FRT vector, to be readily susceptible to integration by FRT-containing targeting vectors, to stably express luciferase, and to have metastatic characteristics similar to the 4T1-12B line and its sublines. A plasmid containing a Gateway cloning site for insertion of small hairpin siRNA sequences was constructed from the FRT targeting vector, pcDNA5/FRT, and tested for its ability to be incorporated into the genome of the 4T1-1V line in an FRT-dependent manner (Figure 2). A vector carrying a previously tested sequence [13] encoding a small hairpin RNA for firefly luciferase was shown to effectively inhibit expression of luciferase in the 4T1-1V line.Figure 2Targeting shRNAs and cDNAs to FRT site in 4T1-1V. (Top) The diagram shows the results of FLP recombinase-dependent insertion of the FRT targeting vector carrying a cDNA and/or siRNA, into the genome. The promoter driving expression of the lacZ-zeo fusion protein before insertion drives expression of the hygromycin resistance gene after insertion allowing hygromycin selection of cells that had undergone targeted insertion of the vector. (Bottom) Cells were cotransfected with the indicated vector and an expression vector encoding FLP recombinase. Cell pools and clones were isolated from the transfected cells and assayed for luciferase expression as described in MATERIALS AND METHODS. Light emission in arbitrary units per milligram of cell protein is shown for pools and clones transfected with empty targeting vector or targeting vector encoding luciferase shRNA. Error bars represent standard deviations for empty vector (n = 6) and luciferase siRNA vector (n = 7) transfected clones.Progression of tumor growth and metastasis in vivoThe results of a longitudinal study of primary tumor growth and metastasis of the 4T1-1V line are shown in Figure 3 and Table 2. Biophotonic imaging of animals each week over a six week period after implantation of 4T1-1V cells in the abdominal no. 9 (or no. 4) mammary fat pad revealed several previously unidentified characteristics of the 4T1 model. Tumor growth at the site as measured by biophotonic imaging was found to occur in a biphasic fashion with rapid growth during the first two weeks, regression between weeks 2 and 4, and increased growth again in weeks 5 and 6 (Fig. 3, Top Panels). Metastasis became apparent in the thoracic region and lower limbs in weeks 5 and 6 of the second growth phase although metastasizing cells probably seeded these sites earlier [8,9]. Examination of light emission from organs removed from the animals at week 6, revealed a spectrum of organ metastasis similar to that observed for the 4T1-12B line.Figure 3Progression of tumor growth and metastasis. (Left) 4T1-1V cells (106) were introduced into mammary fat pads of normal female BALB/c mice and the animals were imaged on a weekly basis for six weeks. The animals were sacrificed at the end of the sixth week and organs and hind limbs were removed and imaged. Images for two representative animals are shown. (Top Right) Quantitation of light emission from primary tumor over the six week period. (Bottom Right) H&E staining of a section from a primary tumor illustrating a central area of necrosis infiltrated by leukocytes and neoplastic cells at the periphery. The neoplastic cells are poorly differentiated and characterized by the presence of large hyperchromatic nuclei and relatively small amount of cytoplasm. Identifiable neutrophils and mast cells that appear to be located extravascularly were observed in non-necrotic areas of the tissue.Table 2Kinetics and extent of metastasis for 4T1-1V1Time after injection (days)PrimaryLungsSpleenLiverBoneKidney/AdrenalsOther8–143(3)2------15–213(3)1(3)-----22–285(5)5(5)1(5)2(5)0(5)1(5)Intestine 1(5)29–357(7)7(7)2(7)6(7)3(7)2(7)Intestine 1(7)36–427(7)7(7)5(7)6(7)6(7)2(7)Brain 1(7)Heart 1(7)1Cells were introduced into fatpads of normal female BALB/c mice and imaged at the indicated times as described in MATERIALS AND METHODS.2Number of animals positive (total number of animals)Further analysis revealed that biphasic growth at the primary site was related to immune system function. The regression that was observed in weeks 2 through 4 in normal BALB/c mice was associated with necrosis and infiltration of leukocytes (Fig. 3, Bottom Right Panel). Biphasic tumor growth did not occur in athymic nude or SCID BALB/c mice (Fig. 4) suggesting involvement of an acquired immune response in the effect. Antibodies directed against multiple 4T1 cell antigens were found in the sera of mice at week 6 (data not shown) further supporting involvement of an acquired immune system response in the regressive process.Figure 4Involvement of immune system in primary tumor growth. 4T1-12B cells (106) were implanted into the mammary fat pad of two normal (○), athymic nude (□) and SCID (△) BALB/c mice and imaged weekly as described in MATERIALS AND METHODS. Average luminescence +/- sd for each time point is plotted.Imaging of animals and organs at various times after introduction of 4T1-1V cells in the fat pad revealed a clear progression of metastasis first to lungs (beginning around 3 weeks) and later to liver, bone and spleen (weeks 3–6) with occasional metastasis to brain, heart and intestines at the later times (Table 2). Tumor cells were detected in lymph nodes adjacent to primary tumors and elsewhere in the animal consistent with previous studies suggesting that 4T1 cells metastasize via the lymphatic system as well as hematogenously [9,12].The results of histological examination of metastases at selected times and sites are shown in Figure 5. In lungs and kidneys metastases were found within or in close proximity to afferent vessels and in most cases appeared infiltrative. In adrenals and liver metastases were more localized, often appearing spherical in nature. Metastasis to bone was prevalent throughout the skeletal system including skull, ribs, sternum, and limbs. Bone-associated osteoclasts were often observed in areas adjacent to bone metastases indicating increased osteoclastogenesis and elevated degradation of bone in these areas (Fig. 6).Figure 5Metastasis to lungs, kidneys, adrenals and liver. (A) Lung at 3 weeks showing metastases adjacent to blood vessels. (B) Tumor cells in a major vessel of the kidney at week 6. Note infiltration of tumor cells into kidney parenchyma (arrowhead, left panel). (C) Tumor-laden adrenal gland at 6 weeks with multiple spherically-shaped metastases. (D) Large metastasis on the surface of the liver at week 6. Note abnormal appearance of liver parenchyma and high levels of leukocytes in parenchyma (arrowheads, right panel) and sinusoids (arrows, right panel). Specimens were obtained from the experiment described in Figure 3.Figure 6Metastasis to bone. (Top Left) Metastasis near joint between femur and tibia at week 6. Note extensive degradation of bone adjacent to the upper surface of the tumor. (Top Right) Interface between tumor and bone at higher magnification. Note osteoclasts (arrowheads) lining the lower surface of the bone. (Bottom Left and Right) Femoral metastasis within the joint itself at low and high resolution at 6 weeks. Specimens were obtained from the experiment described in Figure 3.A progressive increase in hematopoiesis was observed throughout the 6 week time course as primary tumors progressed and metastases developed at distant sites. This was evidenced by increasing levels of circulating neutrophils and other leukocytes (Table 3) and by enlargement of the spleen and liver resulting from extramedullary hematopoiesis that developed in these organs (Figs. 7 and 8). Extramedullary hematopoiesis was apparent by week 2 when primary tumors began to regress and continued to increase until death ensued between weeks 6 and 8. Immature myelocytic cells (Band N) were found in the circulation at week 4. The histology of spleen and liver and the composition and levels of circulating leukocytes are consistent with expansion of granulocyte lineages with circulating leukocytes reaching leukemia-like levels by the end of the observation period (6 weeks).Table 3Circulating white cell analysis1WBC PopulationControl (n = 1)Week 1 (n = 3)Week 4 (n = 4)WBC5.30024.20056.2104.20066.0002.80015.700134.000Seg N0.424 (8%)31.386 (33%)34.850 (62%)0.546 (13%)48.840 (74%)1.148 (41%)8.164 (52%)99.160 (74%)Band N--3.373 (6%)-3.962 (6%)-0.157 (1%)10.720 (8%)Metamyelocytes--0.562 (1%)-0.660 (1%)--2.680 (2%)Lymphocytes4.823 (91%)2.772 (66%)16.863 (30%)3.654 (87%)10.560 (16%)1.624 (58%)6.594 (42%)20.100 (15%)Monocytes----1.320 (2%)-0.314 (2%)-Eosinophils0.053 (1%)0.042 (1%)0.562 (1%)-0.660 (1%)0.028 (1%)0.471 (3%)1.340 (1%)1Whole blood was collected by cardiac puncture at the indicated time after orthotopic introduction of tumor cells as described in MATERIALS AND METHODS.2Total number of WBC/ml blood in millions for each animal tested.3Number of cells/ml blood in millions for each animal tested (% of total WBC for animal)Figure 7Hematopoiesis in spleen. Spleen at 1 (Top) and 6 (Bottom) weeks. Spleen appears normal at week 1. Extensive extramedullary hematopoiesis is apparent at week 6 as evidenced by the presence of megakaryocytes (arrow heads). Specimens were obtained from the experiment described in Figure 3.Figure 8Hematopoiesis in liver. Liver at 1 (Top), 2 (Middle) and 6 (Bottom) weeks. Liver appears normal at week 1. Islands of extramedullary hematopoiesis are seen at week 2 and extensive hematopoiesis throughout the liver is apparent at week 6. Note increased proportion of nucleated cells in blood vessels at weeks 2 and 6. Specimens were obtained from the experiment described in Figure 3.Genes associated with the 4T1 metastatic phenotypeGene expression analysis was carried out on the 4T1 cell line and two of its sibling lines, 67NR and 168FARN to identify expression differences associated with the 4T1 metastatic phenotype. Both of the sibling lines are non-metastatic when introduced orthotopically into BALB/c mice [9]. The 67NR line displays little if any dissemination from the primary site, whereas the 168FARN line displays dissemination to lymph nodes, but not to blood or distant organs [9]. Using Illumina MouseRef-8 BeadChip arrays, multiple replicates, and carefully controlled culture conditions, highly significant (p < 0.0001) expression data for differences as low as 1.2-fold were achieved. Of the 24,048 genes represented on the arrays, 1.8% or 430 genes (347 annotated) differed by 2-fold or more in the 4T1 line relative to the other two lines (Fig. 9). The median false discovery rate for these genes was less than 1 in 500. The majority of all 2-fold differences were found to be significant (p > 0.0001) for both the 4T1/67NR (66.4%) and 4T1/168FARN (98.7%) comparisons. These results indicate a very high level of confidence in the genelists that were produced from the data.Figure 9Genes with altered expression in 4T1 vs. 67NR and 168FARN. Genes with 2-fold expression differences for 4T1 vs. 67NR and 4T1 vs. 168FARN were determined as described in MATERIALS AND METHODS. Genes in the intersection between the two comparisons are those considered to be associated the metastatic phenotype of the 4T1 cell line. Genes in the intersection represent 1.8% of the total genes analyzed, 52% of genes differentially expressed for 4T1 vs. 67NR and 45% of those differentially expressed for 4T1 vs. 168FARN.Ingenuity Pathway Analysis (IPA) of genes differentially expressed in 4T1 relative to the two non-metastatic lines revealed significant association with cancer and other diseases including hematological and inflammatory disease (Table 4), findings consistent with the high level of inflammation and hematopoiesis observed for the 4T1 lines in vivo. Also consistent with the 4T1 metastatic phenotype was association with cell movement, cell signaling, cell growth, proliferation and death, and cell to cell signaling and interaction (Table 4). Many of the expression differences that characterize the 4T1 phenotype including those known to be involved in metastasis and/or tumorigenesis are listed in Additional File 1.Table 4Ontological analysis1p-value# moleculesDiseases and DisordersCancer5.07E-14-5.72E-04100Hematological Disease1.80E-08-5.72E-0451Connective Tissue Disorders3.41E-08-1.44E-0435Dermatological Diseases and Conditions1.75E-07-5.37E-0441Inflammatory Disease8.81E-07-5.51E-0438Molecular and Cellular FunctionsCell Movement1.18E-15-5.72E-0469Cell Signaling1.30E-13-2.04E-04106Cell Death1.67E-13-5.32E-0496Cellular Growth and Proliferation1.03E-12-5.79E-04111Cell to Cell Signaling and Interaction1.94E-08-5.19E-04631304 annotated gene were included in the analysisGenes differentially expressed in 4T1 were categorized with respect to cellular location and function (Fig. 10). Among the genes are substantial numbers involved in cell adhesion, migration, angiogenesis, and extracellular matrix modification; cytoskeleton function; cell proliferation, apoptosis and survival; cellular metabolism; and inflammation and immune response. Altered expression of several transcription factors and genes involved in chromatin modification that regulate these processes were also observed. Elevated expression of genes associated with tight junctions (Cldn3, Cldn4, Cldn7 and Tjp2), adherins junctions (Cdh1 and Vil1) focal adhesions (Itga3, Itga6 and Lama5), and intermediate filaments (Krt1-18 and Krt2-7) indicate that the 4T1 line has greater epithelial character than the non-metastatic lines. An increased propensity for extracellular matrix (ECM) remodeling is suggested by elevated expression of matrix metalloproteinases (Mmp3, Mmp9 and Mmp13), urokinase-type plasminogen activator (Plau) and secreted protease inhibitors (Serpina3g, Serpin2 and Lcn2).Figure 104T1 genes categorized by cellular location and function. Genes associated with the 4T1 metastatic phenotype that fall into the categories shown are listed in the figure. Genes shown in red are elevated in 4T1 and genes shown in blue reduced in 4T1. The blue line to the right of middle represents the plasma membrane with genes falling to the right of it representing secreted genes. The blue rectangle to the left of middle represents intracellular membranes and genes falling inside the rectangle are genes located within intracellular organelles.Signaling pathways associated with phenotypeSeveral signaling pathways appear to be activated in 4T1 cells (Fig. 11). Most conspicuous is activation of the Jak/Stat pathway as indicated by elevated expression of Jak2 and Stat1, decreased expression of Socs1 and increased expression of several Stat target genes (Myc, Irf1, Igsf3g and Usp20) (Fig. 11A). Also conspicuous is activation of p38 MAPK (Mapk12) as indicated by increased expression of CCAAT/enhancer binding protein beta (Cebpb) and high levels of expression of Cebpb/NFκB target cytokines (Ccl5/RANTES, Csf2, Csf3 and Tslp) and acute phase proteins (Saa3, C3 and Lcn2). Increased expression of TIAM1 (Tiam1) and genes in the IL-1 and TNF-α pathways (Il1a, Tnfrsf19, Traf1, Card10) suggest that these pathways, which are known to activate p38 MAPK, may be involved in the expression of Cebpb/NFκB targets. Targets of p38 MAPK are known to activate the Jak/Stat pathway (Fig. 11) so it is therefore likely that p38 MAPK signaling is responsible for the activation of the Jak/STAT pathway in these cells. Elevated expression of Wnt (Wnt10a) and its receptors (Fzd6 and Fzd7) suggests activation of the Wnt pathway (Fig. 11B). While some Wnt pathway targets (Myc and Plau) displayed elevated expression, other known targets (c-jun, cycD and Fosl1) did not. It may be that the purpose of altered expression of Wnt pathway ligand and receptors is to increase β-catenin levels to support junctional complexes that are more prevalent in these cells. Both the canonical and the non-canonical Wnt pathway are known to play an important role in establishment and maintenance of cellular junctions [16]. Finally, 4T1 displayed significantly reduced levels of CDK2-associated protein 1 (CDK2ap1), p53 and two p53 targets, cyclin-dependent kinase inhibitor p21 (Cdkn1a) and cyclin G (Ccng1) (Fig. 11C). These alterations would be expected to accelerate the early phase of the cell cycle (G1 → S) and attenuate the DNA damage response. Expression of a third p53 target, Gadd45, was elevated. Gadd45 is regulated by hypoxia and glucose deprivation as well as by p53. Elevated expression of several genes known to be sensitive to hypoxia and/or glucose deprivation (Pfkfb3, Vegfc, Flt1 and Trib3), suggest that elevated expression of Gadd45 may be due to these factors and that 4T1 cells exist in a state of stress or pseudo-stress even under optimal culture conditions.Figure 11Pathways involved in metastatic phenotype of 4T1. Pathways shown are based on known Kegg pathways with modifications based on recent literature relating 4T1 phenotype genes to these pathways. Red and blue boxes represent genes that are up-regulated or down-regulated by > 2-fold, respectively. Red arrows represent chemotaxis. Numbers outside the boxes give the higher of the two ratios for expression in 4T1 relative to the non-metastatic 67NR and 168FARN lines.Alterations related to tumor microenvironmentA variety of factors produced at elevated levels by 4T1 cells are secreted cytokines, chemokines, acute phase proteins and proteases that interact locally and systemically with the host to produce, recruit and activate cells of hematopoietic origin capable of remodeling the tumor microenvironment and facilitating tumor cell dissemination. These factors and their expected effects on the tumor microenvironment are depicted in Figure 11D. Two important modulators of endothelial cell function produced by 4T1 cells are vascular endothelial growth factor C (Vegfc) and angiopoietin 2 (Agpt2). VEGF-C interacts with VEGF receptors on endothelial cells to stimulate angiogenesis and lymphangiogenesis when existing vessels are destabilized. Angiopoietin 2 destabilizes vessels by antagonizing the stabilizing effects of angiopoietin 1. Together, these factors would be expected to induce both angiogenesis and lymphangiogenesis, increase tumor vascularization and provide routes of escape of tumor cells. The inhibitory effect of semaphorin 3F (Sema3f) on angiogenesis would be expected to shift vessel development toward lymphangiogenesis. Angiopoietin 2 and VEGF-C also serves as chemotactic factors for recruitment of circulating monocytes and macrophages.Macrophages and other cells recruited to the tumor are produced in the bone marrow and other tissues by hematopoiesis. Colony stimulating factors GM-CSF (Csf2) and G-CSF (Csf3) produced and secreted by 4T1 cells stimulate hematopoiesis along the myeloid lineages and are likely to be responsible for the high levels of hematopoiesis and circulating leukocytes observed when tumors from 4T1 cells are established in vivo [17]. Several factors produced by 4T1 cells are known to play a role in recruitment of hematopoietic cells to tissues. RANTES (Ccl5) is chemotactic for mast cells [18] and fragments generated autocatalytically from complement C3 (C3) are capable of stimulating mast cells to release TNFα, histamine, cytokines and other factors that can act to recruit a wide range of cells including monocytes, dendritic cells, neutrophils, eosinophils and lymphocytes. RANTES (Ccl5) is also known to stimulate secretion of interleukin 8 (Il8) from macrophages [19]. Interleukin 8 is also released by stromal fibroblasts in response to interleukin 1α produced by 4T1 cells. Interleukin 8 along with chemokines (Cxcl6, Cxcl1) released by 4T1 are chemotactic for neutrophils [20,21]. Finally, matrix metalloproteinases produced by macrophages, fibroblasts and neutrophils recruited to the tumor would add to the already high levels of matrix metalloproteinases released by the tumor cells themselves thereby creating a high potential for dissolution of matrix and cell-matrix interactions, a condition likely to facilitate tumor cell invasion and metastasis.Previous studies have indicated that populations of immature myeloid cells called myeloid derived suppressor cells (MDSC) are induced by tumors and that these cells facilitate tumor growth and metastasis by suppressing the immune response [22]. Ectopic expression of interleukin 1β in 4T1 cells has been shown to increase MDSC levels and stimulate growth and metastasis of 4T1 tumors in vivo [23]. Because interleukin 1α rather than interleukin 1β is the predominant form of interleukin 1 produced by 4T1 cells, and because the two cytokines have similar biological activity, it is likely that expression of interleukin 1α by 4T1 is involved in production of MDSC and their effects on growth and metastasis of 4T1 in vivo.DiscussionHere we report on the generation of two clonal 4T1 cell lines (4T1-12B and 4T1-1V), both of which stably express firefly luciferase at a high level in the absence of selective pressure, and one (4T1-1V) which was also modified by addition of an FRT site in its genome. These lines were shown to have metastatic characteristics similar to the parental 4T1 line displaying metastasis to bone, lungs, and liver and brain organs primarily affected in human breast cancer. The ability to image the cells ex vivo with high sensitivity allowed detection and quantitation of metastases in affected organs more effectively than has been possible previously. Luciferase-expressing 4T1 variant cell pools and lines with increased propensity for metastasis to brain, liver, and bone have recently been isolated (to be published elsewhere). These variants will further expand the repertoire of syngeneic models available for the study of late stage breast cancer.An acquired immune response was found to play an important role in regulating 4T1 tumor growth and metastasis. 4T1 tumors established in normal BALB/c mice displayed a substantial loss of tumor cells beginning 2–3 weeks after introduction. This effect was not apparent in BALB/c nude and BALB/c SCID mice, in which 4T1 cells in tumors proliferated rapidly and continuously. Antibodies directed against several 4T1 antigens were detected in sera from normal tumor-bearing BALB/c mice further supporting the involvement of an acquired immune response to the cells. Myeloid derived suppressor cells (MDSC) which are known to be induced in 4T1 tumor-bearing mice are likely to be involved in establishment and maintenance of 4T1 tumors by attenuating the immune response to allow survival of the tumor in weeks 3–4 and re-emergence of tumor growth in weeks 5–6. Further work will be required to determine the actual role that MDSC and other immune system components play in regulating the growth and survival of 4T1 tumors.Metastasis of 4T1 tumors is associated with extensive necrosis and inflammation within the primary tumor and hematopoiesis in several mouse organs including spleen and liver. Elevated hematopoiesis has recently been reported for the 4T1 model [17,24]. Whether or not a causal relationship exists between these processes and metastasis remains to be demonstrated although two observations suggest that there may be such a connection. First, the extent of necrosis is greater in 4T1 tumors than those derived from less metastatic sibling cell lines (67NR, 168FARN) as indicated by the occurrence of large areas of visible necrosis in the 4T1 tumors. Second, a causal relationship between inflammation and metastasis is supported by the inhibitory effect of the COX-2 inhibitor, SC-236, on metastasis of 4T1 after primary tumor excision [25]. Inflammation is known to have a positive effect on metastasis in several systems [26-28] and is likely to have a pro-metastatic effect in this system as well.Inflammation in metastatic tumors is generally thought to result from signals produced by dying cells and ECM fragments in areas of insufficient vascularization [29]. A noteworthy finding of this study is that 4T1 tumor cells, when cultured under optimal growth conditions, produce a wide range of factors capable of inducing production, recruitment and activation of inflammatory cells. These factors include colony stimulating factors GM-CSF (Csf2) and G-CSF (Csf3); cytokines Ccl5, Cxcl1, Cxcl6 and Tslp; angiogenic factors Agpt2 and Vegfc; and acute phase proteins Saa3, C3, and Lcn2. While this does not preclude the involvement of cell death in initiating an inflammatory response in the tumor, it does suggest that the tumor cells themselves may play a more direct and active role in directing pro-metastatic inflammatory processes than previously envisioned.The methodology used in this study for analysis of gene expression yielded highly significant data characterizing the 4T1 metastatic phenotype. The majority of genes that differed by more than 2-fold in 4T1 relative to the two non-metastatic sibling lines examined displayed an exceptionally high level of significance (p < 0.0001) and genelists obtained at this level of significance displayed very low false positive rates. The statistics argue that the results obtained provide a relatively complete and accurate picture of expression differences associated with the 4T1 phenotype. The high level of accuracy and reproducibility that was achieved is attributed to use of the Illumina BeadChip platform and analysis of cells cultured under carefully controlled growth conditions that minimize differences between biological replicates. The data obtained from this study provide detailed information regarding the genes and pathways involved in breast cancer progression for this model and will be particularly useful for further analysis of the pathological processes responsible for progression to a metastatic phenotype.Unlike many cell lines used as xenograft models, subclones of the 4T1-12B cell line that had undergone more than 20 doublings were found to be homogeneous with respect to metastatic properties. These cells also display a high plating efficiency and no visibly apparent differentiation in culture or in vivo. Thus, the cells resemble stem cells found in populations of cell lines such as MCF7 [30] in that they are self renewing, but differ in that they do not appear to differentiate. While more work is need to determine the basis for this property, the characteristic has utility for studies aimed at determining gene function since clonal lines in which a specific genes have been over-expressed or knocked down can be expected to retain the properties of the parental line from which they were derived. In this regard, the FRT site in the 4T1-1V line will be useful for production of isogenic lines for analysis of gene function.ConclusionIn conclusion, this study provides basic information for those interested in using two imagable 4T1 breast cancer models developed in this laboratory. Several characteristics of these models make them particularly attractive for the study of late stage breast cancer. First and foremost, because of their syngeneic nature, they provide a highly physiologic system suitable for analysis of innate and acquired immune system roles in tumor growth and metastasis. The relatively complete gene expression data provided offer numerous avenues for further study of the molecular and pathologic basis for these and other processes related to late stage breast cancer. To our knowledge, the 4T1 model is the only system that has the capacity to metastasize to all organs affected in breast cancer in humans when introduced orthotopically. For this reason, and because of the ease of use and reproducibility that can be achieved, these imagable models provide ideal systems for determining anti-metastatic effects of cancer drugs and therapeutic regimens and is well suited for investigating the molecular, cellular and pathologic basis for metastasis to specific organs and tissues.AbbreviationscRNA: complementary RNA; DMEM: Dulbecco's Minimum Essential Medium; ECM: extracellular matrix; FBS: fetal bovine serum; FLP: flippase; FRT site: FLP recombinase targeting site; GEM: genetically engineered mouse; IPA: Ingenuity Pathway Analysis; MDSC: myeloid derived suppressor cells; NCS: normal calf serum; NEAA: nonessential amino acids; PBS: phosphate buffered saline; Q-PCR: quantitative PCR; ROI: region of interest; siRNA: small inhibitory RNA; shRNA: short hairpin RNA.Competing interestsCell lines described in this study are licensed by Tufts University for commercial use. Royalties are split between Tufts University (including GGS), Wayne State University and the NIH.Authors' contributionsKT acquired and analyzed imaging data. MF acquired and analyzed imaging and gene expression data. JA evaluated histology data. GGS conceived of the study, analyzed imaging data, and wrote the manuscript. KT and MF contributed equally to the study. All authors read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:Supplementary MaterialAdditional File 1Gene table. A list of genes associated with the 4T1 metastatic phenotype.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2529342.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529342",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529342\nAUTHORS: Asquad Sultan, Helen Gaskell, Sheena Derry, R Andrew Moore\n\nABSTRACT:\nBackgroundDuloxetine hydrochloride is a reuptake inhibitor of 5-hydroxytryptamine and norepinephrine used to treat depression, generalized anxiety disorder, neuropathic pain, and stress incontinence in women. We investigated the efficacy of duloxetine in painful diabetic neuropathy and fibromyalgia to allow comparison with other antidepressants.MethodsWe searched PubMed, EMBASE (via Ovid), and Cochrane CENTRAL up to June 2008 for randomised controlled trials using duloxetine to treat neuropathic pain.ResultsWe identified six trials with 1,696 patients: 1,510 were treated with duloxetine and 706 with placebo. All patients had established baseline pain of at least moderate severity. Trial duration was 12 to 13 weeks. Three trials enrolled patients with painful diabetic neuropathy (PDN) and three enrolled patients with fibromyalgia. The number needed to treat (NNT) for at least 50% pain relief at 12 to 13 weeks with duloxetine 60 mg versus placebo (1,211 patients in the total comparison) was 5.8 (95% CI 4.5 to 8.4), and for duloxetine 120 mg (1,410 patients) was 5.7 (4.5 to 5.7). There was no difference in NNTs between PDN and fibromyalgia. With all doses of duloxetine combined (20/60/120 mg) there were fewer withdrawals for lack of efficacy than with placebo (number needed to treat to prevent one withdrawal 20 (13 to 42)), but more withdrawals due to adverse events (number needed to harm (NNH) 15 (11 to 25)). Nausea, somnolence, constipation, and reduced appetite were all more common with duloxetine than placebo (NNH values 6.3, 11, 11, and 18 respectively). The results for duloxetine are compared with published data for other antidepressants in neuropathic pain.ConclusionDuloxetine is equally effective for the treatment of PDN and fibromyalgia, judged by the outcome of at least 50% pain relief over 12 weeks, and is well tolerated. The NNT of 6 for 50% pain relief suggests that this is likely to be a useful drug in these difficult-to-treat conditions, where typically only a minority of patients respond. Comparing duloxetine with antidepressants for pain relief in DPN shows inadequacies in the evidence for efficacy of antidepressants, which are currently recommended in PDN care pathways.\n\nBODY:\nBackgroundNeuropathic pain is the consequence of damage to the central nervous system (e.g. cerebrovascular accident, multiple sclerosis or spinal cord injury) or peripheral nervous system (e.g. painful diabetic neuropathy (PDN), postherpetic neuralgia (PHN), surgery). It has a significant negative impact on quality of life [1]. Some patients with neuropathic pain respond well to treatment and others show no obvious response [2-4]. No pharmacological intervention produces meaningful relief for more than half the patients with neuropathic pain [5].The incidence of PHN and trigeminal neuralgia and PDN together is almost 0.1% per year in the UK [6]. The incidence of neuropathic pain is growing, presumably because of increased numbers of older persons and diabetics, amongst whom about one in five develop painful neuropathy at some stage. Neuropathic pain is quite common in general medical practice with about 1% point prevalence in UK if fibromyalgia, PDN, PHN, and trigeminal neuralgia are included [7].The most common pharmacological approaches to the management of neuropathic pain include antidepressants (tricyclic antidepressants, serotonin and norepinephrine reuptake inhibitors), antiepileptics (valproate, carbamazepine, gabapentin, pregabalin), opioids, other analgesics, topical lidocaine patch, and topical capsaicin. The evidence for these has been reviewed extensively [4,8-16].5-hydroxytryptamine (5HT) and norepinephrine (NE) are involved in the modulation of endogenous analgesic mechanisms via descending inhibitory pain pathways in the brain and spinal cord [17]. Disinhibition and imbalance of 5HT and NE in endogenous pain inhibitory pathways could contribute to persistent pain. An increase in 5HT and NE may increase inhibition of painful signals, improving pain relief.Duloxetine hydrochloride is a serotonin-norepinephrine reuptake inhibitor used to treat depression, generalized anxiety disorder, neuropathic pain, and stress incontinence in women. We investigated the efficacy of duloxetine in the management of PDN and fibromyalgia as duloxetine had not been included in the most recent systematic reviews, including one of antidepressants [13]. Duloxetine in PDN alone has been the subject of a recent post hoc analysis [18].MethodsWe searched PubMed, EMBASE (via Ovid), and Cochrane CENTRAL up to June 2008 for randomised controlled trials using duloxetine to treat neuropathic pain. The detailed search strategy included use of the drug name \"duloxetine\" anywhere in an article, together with \"randomized controlled trial\" as subject heading, publication type or text word; this was modified appropriately for different databases. Reference lists of retrieved articles and reviews were also searched for relevant trials. We contacted Boehringer Ingelheim Limited as a UK distributor for duloxetine in neuropathic pain to enquire about relevant published or unpublished studies, and examined an on-line register [19].Included trials had to be randomised, double blind, placebo controlled, and use duloxetine to treat adult patients with painful neuropathies of any cause. Trials had to have a minimum of 10 patients per treatment arm, and a planned duration of at least four weeks.The abstracts were read, and potentially useful reports retrieved in full. No information was taken from posters or abstracts. Decisions on inclusion or exclusion of trials, assessment of trial quality and validity and all data extraction were made independently by three reviewers, with discrepancies resolved by consensus.Methodological quality of included studies was assessed using a validated 5-point scale [20] utilising reporting of randomisation, blinding, and withdrawals. The maximum score possible was 5 points, and no study could be included with fewer than 2 points (one for randomisation and one for blinding). Study validity was assessed using a validated 16-point scale [21].Data were abstracted into a standard form. Information extracted from the trials included details of the patients (number, age, sex, pain syndrome), duloxetine dose, and permitted rescue analgesia. The primary outcome sought was 50% pain relief. Other measures of pain relief were abstracted where reported. Secondary outcomes were withdrawals (all cause, lack of efficacy and adverse events) and adverse events (patients with at least one adverse event, serious adverse events, and specific adverse events).Guidelines for quality of reporting of meta-analyses were followed where appropriate [22]. The prior intention was to pool data where there was clinical and methodological homogeneity, with similar patients, dose, duration, outcomes, and comparators, but not where numbers of events were small, and random chance might well dominate effects of treatment [23]. Homogeneity tests and funnel plots, though commonly used in meta-analysis, were not used because they have been found to be unreliable [24,25]. Instead, clinical homogeneity was examined graphically [26]. Relative benefit (or risk) and number needed to treat or harm (NNT or NNH) were calculated with 95% confidence intervals. Relative benefit or risk was calculated using a fixed effects model [27] with no statistically significant difference between treatments assumed when the 95% confidence intervals included unity. We added 0.5 to treatment and comparator arms of trials in which at least one arm had no events. NNT or NNH was calculated [28] using the pooled number of observations only when there was a statistically significant difference of relative benefit or risk (where the confidence interval did not include 1). We used the following definitions:• When significantly more beneficial outcomes occurred with duloxetine than placebo, we used the term number needed to treat (NNT).• When significantly fewer adverse events occurred with duloxetine than placebo we used the term the number-needed-to-treat to prevent one adverse event (NNTp).• When significantly more adverse events occurred with duloxetine than placebo we used the term the number-needed-to-harm to cause one adverse event (NNH).Statistical significance of any difference between NNT for different doses was assumed if there was no overlap of the confidence intervals, and additionally tested using the z statistic [29]. RevMan 5.0.12 was used to analyse continuous data. There was a prior intention to carry out sensitivity analyses for high versus low trial quality (<3 vs ≥ 3) and validity (<9 vs ≥ 9), duloxetine dose, and pain syndrome. A minimum of two trials and 250 patients was required in any sensitivity analysis [23].ResultsWe identified six trials satisfying the inclusion criteria [30-35]. Details of the included studies are in Additional File 1. A total of 2,216 patients were included, 1,510 treated with duloxetine and 706 with placebo. Three trials [32-34] enrolled patients with PDN and three [30,31,35] enrolled patients with fibromyalgia, in which 23% to 38% had a diagnosis of major depressive disorder. The trials in PDN excluded patients with any diagnosed psychological disorder. We did not include any trials in which the primary problem was a major psychiatric disorder but with a secondary painful condition [36-41]. All patients had established baseline pain of at least moderate severity, measured using established scales. The mean age in the trials ranged between 49 and 61 years, and the majority of patients were Caucasian. One trial [31] enrolled only women, and the others between 5% and 61% men.Trial duration was 12 to 13 weeks. One trial [33] had a 13-week continuation phase, but results for the first 13 weeks (acute phase) only are analysed here, to make it comparable with the other trials. Duloxetine was used at doses of 20, 60, or 120 mg daily, with titration up to the 120 mg dose, which was given as a divided dose of 60 mg twice daily. Up to 2 g acetaminophen daily was permitted as rescue medication in the fibromyalgia trials, and up to 4 g daily in the PDN trials.Trials were of good methodological quality, with three scoring 5/5, two scoring 4/5, and one scoring 3/5 on the Oxford Quality Score [20]. Two scored 16/16 and four scored 13/16 on the Oxford Pain Validity Score [21]. No sensitivity analyses were therefore carried out for these criteria.Efficacy50% Pain ReliefAll six trials reported the outcome of at least 50% pain relief over baseline in the 24-hour average pain score by the end of the trial, and results are summarised in Figure 1 and Table 1. Trials were consistent, and overall 41% of patients achieved 50% pain relief with any dose of duloxetine compared with 24% with placebo. Combining all doses in both conditions (2,216 patients), the NNT for one patient to achieve at least 50% pain relief with duloxetine compared with placebo was 5.9 (95% CI 4.8 to 7.7).Figure 1Proportion of patients with at least 50% pain relief with duloxetine 60 mg or 120 mg and placebo in individual trials. Pink circles are fibromyalgia trials. Inset scale shows number in comparison.Table 1Summary of efficacy and adverse event outcomes in duloxetine trialsNumber ofPercent withOutcomeDose (daily maximum)TrialsPatientsDuloxetinePlaceboRelative benefit or risk (95% CI)NNT/NNTp/NNH (95% CI)Efficacy50% PR All20/60/120 mg62,21641241.7 (1.4 to 1.9)5.9 (4.8 to 7.7)50% PR PDN60/120 mg31,02447271.7 (1.4 to 2.1)5.1 (3.9 to 7.3)50% PR fibromyalgia60/120 mg399637211.7 (1.4 to 2.1)6.4 (4.7 to 9.9)50% PR60 mg51,21143261.7 (1.4 to 2.0)5.8 (4.4 to 8.4)50% PR120 mg61,41042241.7 (1.5 to 2.0)5.7 (4.5 to 7.8)Adverse events generalWithdrawal – all cause20/60/120 mg62,41830261.2 (1.1 to 1.4)26 (13 to 426)Withdrawal – LoE20/60/120 mg51,872490.5 (0.4 to 0.7)20 (13 to 42)Withdrawal – AE20/60/120 mg62,2201581.8 (1.4 to 2.4)15 (11 to 25)Any AE60/120 mg41,24382671.2 (1.2 to 1.3)6.7 (5.0 to 10)Serious AE60/120 mg31,034230.8 (0.4 to 1.7)not calculatedSpecific adverse eventsNausea20/60/120 mg31,14529103.0 (2.2 to 4.3)5.3 (4.3 to 6.9)Somnolence20/60/120 mg31,1451442.9 (1.7 to 4.9)11 (8.0 to 16)Constipation20/60/120 mg31,1451333.6 (2.0 to 6.5)11 (8.3 to 16)Decreased appetite20/60/120 mg2811714.9 (1.7 to 14)18 (12 to 34)PR – pain relief; PDN – painful diabetic neuropathy; LoE – lack of efficacy; AE – adverse eventIn the right hand column, bold font is used for NNT – number needed to treat; normal font for NNTp – number need to treat to prevent; italic font for NNH – number needed to harmFive of the trials used 60 mg, and all six used 120 mg; only 66 patients (in two treatment arms) received the 20 mg dose. The dose of duloxetine made little difference to the result (Figure 1, Table 1). There was no difference in the proportion of patients achieving at least 50% pain relief with 60 mg and 120 mg (z = 0.13; p = 0.89).There was no significant difference in the proportion of patients achieving at least 50% pain relief with PDN or fibromyalgia (z = 0.95; p = 0.34). The proportion of patients with this outcome was slightly lower for both placebo and duloxetine groups in the fibromyalgia trials (Table 1), with similar NNTs for both conditions.Average pain score (APS)Five of the trials [31-35] recorded daily 24-hour average pain scores (APS) on a 0–10 scale, and reported this as a weekly mean, as well as the change from baseline to final weekly mean. The change in weekly mean APS on treatment was compared with placebo over the 12 or 13 weeks. Figure 2 shows the calculations for different doses of duloxetine in different pain syndromes. The weighted mean difference for duloxetine 60 mg compared with placebo was 1.0 (0.71 to 1.4), and for duloxetine 120 mg compared with placebo was 0.9 (0.49 to 1.3). There was no difference in response between patients with PDN and fibromyalgia.Figure 2Mean change from baseline to endpoint on the 24-hour average pain score (APS) for treatment compared to placebo over 12 to 13 weeks, by duloxetine dose (60 mg and 120 mg) and condition (diabetic neuropathy and fibromyalgia).WithdrawalsWithdrawals for any cause occurred in slightly more patients with duloxetine (30%) than placebo (28%); the NNTp for all cause withdrawal with duloxetine rather than placebo was 26 (13 to 426) (Table 1). Withdrawals due to lack of efficacy occurred in significantly fewer patients (4%) taking duloxetine than placebo (9%); the NNTp for lack of efficacy withdrawal with duloxetine rather than placebo was 17 (12 to 35) (Table 1).Withdrawals for any cause or for lack of efficacy did not differ significantly between the 60 mg and 120 mg doses, although for any cause they were consistently 4% to 5% lower for 60 mg than 120 mg, except for Russell et al [35] where the rates were almost identical.Adverse eventsWithdrawalsWithdrawals due to adverse events occurred significantly more often with duloxetine (15%) than placebo (8%). The NNH was 15 (11 to 25) (Table 1). They were 2% to 8% lower with 60 mg than 120 mg, giving an NNH of 19 (11 to 86) for 120 mg compared to 60 mg.Any adverse eventThe \"at least one adverse event\" criterion was met in significantly more patients taking duloxetine (82%) than placebo (67%) in the four trials that reported this outcome. The NNH was 6.7 (5.0 to 10) (Table 1).Serious adverse eventsSerious adverse events were reported in only three trials; one trial did not report this outcome [30], one did not report it for the 13-week phase [35], and one did not separate rates between groups [32]. In the three trials reporting serious adverse events they were uncommon and not significantly different between duloxetine or placebo, at about 2–3% over the 12 weeks of the trials (Table 1). Russell et al [35] reported that serious adverse events were infrequent over the full 6 months of the trial.Specific adverse eventsOnly three trials [31,32,34] provided numbers of patients experiencing specific treatment emergent adverse events over 12 to 13 weeks. There were statistically significant increases in nausea (29% vs 10%), somnolence (14% vs 4%), constipation (13% vs 3%) and decreased appetite (7% vs 1%) with all doses of duloxetine compared with placebo (Table 1). There were small mean increases in laboratory tests and vital signs, but these were transient and not considered clinically relevant by the trialists.DiscussionThis systematic review differs from the only other that considers duloxetine [18]. That company-sponsored review was able to pool data from the three PDN trials. It calculated NNTs for at least 50% pain relief (with identical results to those calculated here), and also gave NNTs for at least 30% pain relief. It demonstrated the stability of NNTs over two to 12 weeks, an important observation, and no difference in estimate depending on treatment of dropouts. This review differs in demonstrating that the efficacy of duloxetine is similar in PDN and fibromyalgia, and also makes an informed comparison with other evidence on antidepressant treatments for neuropathic pain.For evidence to be credible, it has to fulfil criteria of quality, validity, and size [42]. The evidence here on duloxetine does that. Trials were randomised, and double blind, and quality and validity scores indicated low chance of bias. The trials were of sufficient length (12 or 13 weeks) to make them clinically relevant, and the outcome reported of at least 50% pain relief was a high hurdle. Most older neuropathic pain studies used less stringent measures, including undefined \"improvement\" as an outcome, and only trials of pregabalin have also consistently used at least 50% pain relief. Finally, with information on over 2,200 patients, including over 1,000 patients with PDN, the data set for duloxetine fulfils the requirements of size [23] and is much larger than any previous data set for antidepressants in neuropathic pain [13].Significantly more patients achieved the outcome of at least 50% pain relief with duloxetine (41%) than with placebo (24%) over 12 weeks. The outcome of 50% pain represents substantial clinical pain relief, and an NNT of 6 suggests that this is likely to be a useful drug in these difficult-to-treat conditions, where typically only a minority of patients respond. There was no dose response between 60 mg and 120 mg, nor was there any significant difference in the duloxetine response between PDN or fibromyalgia. There was a similar lack of dose response in this range for use of duloxetine in major depressive disorder [43].Duloxetine was well tolerated in the trials, with fewer withdrawals due to adverse events with 60 mg than with 120 mg. Most adverse events were reported to be mild or moderate, with nausea, somnolence, constipation, decreased appetite and dry mouth frequently mentioned. In stress incontinence duloxetine affects the resting tone and contraction of the urethral striated sphincter muscle. It might be expected to cause symptoms of urinary hesitancy in patients without incontinence, but urinary problems were not reported in any of these trials, or in trials of duloxetine in depression [for example [44,45]].This review has some limitations. Firstly, pain intensity measurements used to calculate our primary outcome of at least 50% pain relief were derived from average pain intensity scores during the previous 24 hours. Secondly, the trials were 12 to 13 weeks in duration, and although they demonstrated a sustained response and good tolerability over this period, they provided no information for longer-term efficacy or safety. Russell et al [35] included a 13-week continuation phase, and reported continuing efficacy and tolerability, as have open-label extension studies in neuropathic pain lasting 26 and 52 weeks [46,47].Duloxetine has been widely trialled in other conditions, in particular depression and stress-induced incontinence in women, and the many trials have been subject to systematic review in those therapeutic areas. An evaluation of cardiovascular safety in 42 placebo controlled trials involving 8,500 patients concluded that duloxetine did not appear to be associated with significant cardiovascular risks [48].Finally, the studies in fibromyalgia included some patients with depression. Although only a minority were depressed (Additional file 1), it could be argued that duloxetine reduced pain intensity by improving depression. We identified a small number of other trials in patients with psychiatric disorders with painful physical symptoms (not neuropathic pain) [36-41]. Although none reported our primary outcome of 50% pain relief, they did use the same scales to record pain intensity, and all reported improvements in pain with duloxetine 60 mg that did not entirely correlate with improvements in depression. Fava et al. estimated that 50% of the total effect of duloxetine on overall pain was independent of changes in depression [37]. A counter view was that duloxetine was ineffective in treating pain in depression [49]. In the trials included in this review, about one third of the patients with fibromyalgia also had major depressive disorder. It would be difficult to attribute all of the analgesic effect of duloxetine in these trials to improvements in depression in a minority, although one could not rule out a contributory effect. One of these trials estimated that around 20% of the overall treatment effect with 120 mg and 30% with 60 mg was due to improvements in depressive symptoms, the remainder being due to duloxetine's direct effect on pain reduction [35].The quantity and quality of randomised trial data in neuropathic pain is limited. Table 2 shows the results for duloxetine 60 mg and 120 mg over 12 weeks compared with other results for antidepressants calculated from a recent Cochrane review [13]. Only patients with PDN are reported in Table 2 for duloxetine, in order to keep to similar inclusion criteria. Even so, the amount of evidence on duloxetine dominates the evidence available, almost doubling the number of patients studied previously with antidepressants. Table 2 shows the curious tendency for smaller amounts of information to be associated with greater benefit, either as higher values for relative risk or lower values for NNT. Size and quality may be linked: only one trial in the Cochrane meta-analysis had over 100 participants, while all the duloxetine trials had over 200, and many older studies used poorly defined outcomes of improvement, probably less stringent than that of at least 50% pain relief. Some small older studies also had a crossover rather than parallel design. This presents problems in determining relative efficacy among antidepressants for treatment of neuropathic pain.Table 2Summary of efficacy in antidepressant meta-analysisNumber ofPercent withAntidepressantOutcomeTrialsPatientsPlaceboActiveRelative benefit (95% CI)NNT (95% CI)Duloxetine 60/120 mgat least 50% pain relief31,02427471.7 (1.4 to 2.1)5.1 (3.9 to 7.3)Amitriptyline all dosesglobal improvement1058832642.0 (1.6 to 2.4)3.2 (2.6 to 4.2)Other antidepressantsglobal improvement321612504.2 (2.5 to 7.0)2.6 (2.0 to 3.7)Venlafaxine all dosesglobal improvement320025572.3 (1.6 to 3.4)3.1 (2.2 to 5.1)Desipramine all dosesglobal improvement27810595.8 (2.2 to 15)2.1 (1.5 to 3.3)Imipramine all dosesglobal improvement25859719 (3.9 to 89)1.1 (1.0 to 1.2)NNT – Number need to treatSeveral evidence-based recommendations for the treatment of neuropathic pain place use of antidepressants early in any care pathway [5,9,50]. Comparing the evidence for different therapies within a class, and between classes, is key to determining the most effective, and most cost-effective pathway. In that circumstance, it is not enough just to calculate an NNT. The quality and credibility of the evidence behind those calculations needs to be evaluated, a function of the utility and validity of outcomes, and to have sufficient numbers of patients or events to avoid random chance. The example of duloxetine provides a firm evidential base for within and between class comparisons.ConclusionDuloxetine is equally effective for the treatment of PDN and fibromyalgia, judged by the outcome of at least 50% pain relief over 12 to 13 weeks, and is well tolerated. The NNT of 6 for this outcome suggests that this is likely to be a useful drug in these difficult-to-treat conditions, where typically only a minority of patients respond. Doses higher than 60 mg do not provide additional pain relief, but do cause slightly more withdrawals due to adverse events. Comparing duloxetine with other antidepressants for pain relief in PDN shows inadequacies in the evidence for efficacy of antidepressants, which are currently recommended in PDN care pathways.Competing interestsRAM, and SD have received research support from charities, government and industry sources at various times, and HG from government, but no such support was received for this work. No author has any direct stock holding in any pharmaceutical company.Authors' contributionsAS, SD and HG carried out searches, selected studies and carried out data extraction. AS, SD and AM were involved with analysis. All authors were involved with writing, and all authors read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:Supplementary MaterialAdditional file 1Included studies. Details of the design and outcomes of the included studies.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2529376.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529376",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529376\nAUTHORS: Malcolm E. Fisher, Allyson K. Clelland, Andrew Bain, Richard A. Baldock, Paula Murphy, Helen Downie, Cheryll Tickle, Duncan R. Davidson, Richard A. Buckland\n\nABSTRACT:\nChick embryos are good models for vertebrate development due to their accessibility and manipulability. Recent large increases in available genomic data from both whole genome sequencing and EST projects provide opportunities for identifying many new developmentally important chicken genes. Traditional methods of documenting when and where specific genes are expressed in embryos using wholemount and section in-situ hybridisation do not readily allow appreciation of 3-dimensional (3D) patterns of expression, but this can be accomplished by the recently developed microscopy technique, Optical Projection Tomography (OPT). Here we show that OPT data on the developing chick wing from different labs can be reliably integrated into a common database, that OPT is efficient in capturing 3D gene expression domains and that such domains can be meaningfully compared. Novel protocols are used to compare 3D expression domains of 7 genes known to be involved in chick wing development. This reveals previously unappreciated relationships and demonstrates the potential, using modern genomic resources, for building a large scale 3D atlas of gene expression. Such an atlas could be extended to include other types of data, such as fate maps, and the approach is also more generally applicable to embryos, organs and tissues.\n\nBODY:\nIntroductionThe ease of access to the embryo and subsequent manipulability has made the chick a reliable and powerful system for developmental biology. The power of this system has been enhanced recently by the availability of genomic data from both whole genome sequencing (2004) and large-scale EST projects (Boardman et al., 2002; Carre et al., 2006; Hubbard et al., 2005; Kim et al., 2006). This opens up new opportunities for identifying all the genes that mediate the development of the embryo and its constituent parts and then using high throughput methods to test their function (Brown et al., 2003). One of the first steps in this process is to document where and when specific genes are expressed in the embryo and a large amount of gene expression data is being generated. A repository of chick embryo gene expression data, GEISHA, has already been pioneered by Antin and colleagues (Bell et al., 2004). This consists of a collection of photographs of embryos in which gene expression has been assayed mainly by using whole mount in-situ hybridisation, in some cases accompanied by sectioned material. The database that we describe here is different in that gene expression is visualised in 3D using OPT and expression patterns are mapped onto 3D digitised images. Here we describe how we have started to establish such a database of 3D gene expression patterns for the developing chick wing and investigated some of the practicalities involved.OPT was developed at the MRC Human Genetics Unit in Edinburgh and is one of a number of new microscopy techniques that have been developed in the last few years that allow capture of 3D image data. OPT has already been used to study the development of human, mouse, fly and plant embryos (DeLaurier et al., 2006; Kerwin et al., 2004; Lee et al., 2006; McGurk et al., 2007; Sharpe et al., 2002) and one of its advantages is that it captures the three dimensional distribution of gene expression in an intact embryo.In order to compare large numbers of gene expression patterns, a number of recent atlas projects have taken the approach of mapping gene expression data to digital reference models. For example, the Edinburgh Mouse Atlas of Gene Expression (EMAGE) deals with 2D data in this way (Baldock et al., 2003), section in-situ derived gene expression data in the mouse brain have been mapped to 3D models generated by Micromagnetic resonance imaging (Lein et al., 2007) and a Zebrafish 3D anatomical Atlas has been produced based on sectioned material for the projection of gene expression data (Verbeek et al., 1999). Projects such as EMAGE (Baldock et al., 2003) and GENEPAINT (Visel et al., 2004) have begun to build large queryable databases containing both whole mount and section in-situ data for mouse embryos. Since the system has already been set up for mouse embryos, the establishment of a parallel database for the chick should allow direct comparisons between gene expression patterns in the two organisms. We have adopted the Bookstein thin plate spine algorithm for mapping our 3D data, which has previously been used extensively in morphometric analysis (Albertson and Kocher, 2001; Bruner et al., 2004; Harmon, 2007; Yeh, 2002). Once gene expression data are assembled in a digital atlas, powerful modern data mining techniques can be used to examine relationships potentially leading to unexpected discoveries.We have focussed on the developing chick wing bud. The wing bud is a good test system for investigating the power of a 3D database because it is a structure with no significant morphological features at early stages. Many insights into the mechanisms that pattern the vertebrate limb have come from studies on chick embryos (Tickle, 2004). In the long term, the ability to compare multiple patterns of gene expression should enable us to identify synexpression domains, complementary patterns and possibly also discrete boundaries of gene expression. We have optimised protocols to maximise consistency of initial whole mount in-situ hybridisation, OPT capture and mapping data to a reference model. In this paper, we have studied a number of previously described genes in terms of their expression patterns. This has allowed us to perform a pairwise analysis of overlap of expression for an initial set of genes and to identify some previously unappreciated features of expression with respect to dorso-ventral distribution. We have also used computational techniques based on microarray analysis to look for specific regions of the limb where genes are co-expressed.Materials and methodsEmbryo preparationWhite Leghorn chick eggs were incubated in a humidified incubator at 38 °C for the appropriate time for the desired developmental stage as determined by the Hamburger and Hamilton stages (Hamburger and Hamilton, 1951). Eggs were then windowed, embryos removed to ice-cold Phosphate Buffered Saline (PBS) (0.02 M phosphate, 0.15 M NaCl) and cleaned of extra-embryonic membranes. Eyes and forebrain were punctured with a tungsten needle to reduce trapping, and embryos were transferred to 4% (w/v) ice-cold paraformaldehyde (PFA) overnight. The embryos were then put through a graded methanol series at 4 °C; ending in 2, 100% methanol washes and stored at − 20 °C until use.Plasmid preparation and probe synthesisThe plasmids used for the different genes were Shh (Echelard et al., 1993), Fgf8 (Crossley et al., 1996), Msx1 (Hill et al., 1989), and Tbx3 (Isaac et al., 1998). All plasmids were linearised and transcribed according to their sources. EST clones acquired from ARK genomics were used as probes for Wnt3a (EST clone 603102629F1), Wnt5a (EST clone 603799237F1), Lmx1 (EST clone 603127966F1) and HoxD13 (EST clone 603499362F1). All EST clones were in pBluescript II KS+, which was linearised with Not1 (NEB) and transcribed with T3 RNA polymerase (Roche) to produce antisense probes. Plasmids were grown up using standard protocols and purified using Qiagen plasmid mini kits and individual clones were sequenced to check their identity.The RNA probes were synthesised accordance with standard protocols (Maniatis et al., 1982; Nieto et al., 1996) and purified using the ProbeQuant G-50 spin column system (Amersham Biosciences). In some cases probe purification was performed using phenol chloroform extraction and Lithium Chloride precipitation as detailed in Nieto et al. (1996).In-situ hybridisationThe in-situ protocol used was a modified version of that of Nieto et al. (1996), full details of the modified protocol are in supplementary materials.Before scanning under UV, embryos require some further washes to remove excess NBT–BCIP. Embryos were washed twice for 10 min in PBS at RT. Embryos were then moved to 10× TBST and allowed to equilibrate at RT, this should take between 10 and 20 min depending on the size of the embryo. Embryos were then washed 3 times for 20 min in 1× TBST and left to wash overnight in fresh 1× TBST at 4 °C. Embryos were washed 3 times for 5 min in PBT at RT and then fixed overnight in 4%PFA–DEPC–PBS at 4 °C. Embryos were washed briefly a further 2 times in PBS and then refixed in formal saline.Section in-situ hybridisationSection in-situs were performed on HH22 embryonic limbs according to the method of Moorman et al. (2001). The Wnt5a probe was the same as above.Tyramide Signal Amplification (TSA)To modify the in-situ protocol for the fluorescent TSA colour reaction, glutaraldehyde was removed in the fixation step to minimise autofluorescence. Secondary detection was performed using a peroxidase (POD) linked anti-DIG antibody. The colour reaction was performed according to the Alexa Fluor 568 kit manufacturer’s instructions (Invitrogen — T-20914) with volumes increased to accommodate whole embryos.OPT sample preparation and scanningStandard reference embryos were fixed in 4% PFA/0.2% glutaraldehyde mix, which produces a stronger autofluorescent signal than PFA alone. The addition of 0.2% glutaraldehyde to the fix was not necessary for embryos that had been in-situ hydbridised, due to the presence of glutaraldehyde in the fixative steps of that protocol. Reference embryos were stored in 100% methanol until scanning, at which point they were taken back through a methanol series to PBS and briefly to water. Embryos having undergone in-situ hybridisation were washed 3 times for 20 min in PBS to remove storage fixative. In order to remove excess salts, embryos were washed twice for 10 min in distilled water and subsequently left overnight in distilled water followed by 1 wash of 10 min in fresh distilled water. OPT scanning was carried out following the protocol set out in Sharpe et al. (2002), for more detailed protocols see supplementary materials. The resulting output was in the Wlz file format which the Edinburgh Mouse atlas Project (EMAP) and EMAGE projects have utilised and which can store 3D grey scale information. The MRC HGU has produced a set of software tools for manipulating data in this format and these tools were used to convert the data into a format that could be imported in to the AMIRA software package.3D mappingThe mapping of the 3D gene expression data to the reference models was performed using the AMIRA 4.1 software from Mercury Computer Systems. The data to be mapped were first roughly aligned with the reference model. Two corresponding sets of landmarks were then set up between an isosurface for the reference embryo and an isosurface for the fluorescent/anatomical data from the scan to be mapped. The landmarks were based on prominent morphological landmarks such as the AER, the region where the limb attaches to the flank and to proportional distances along the main axes of the limb. The fluorescent/anatomical data was warped, using a Bookstein thin plate spline method (Bookstein, 1989) provided by the AMIRA software and based on the previously defined landmark sets. Provided the resulting warped fluorescent/anatomical data seemed consistent with the reference limb’s morphology the same warp was then applied to the brightfield channel data. For full details see methods in supplementary materials.Real-time PCRChick embryos (incubated for 4 days at 38 °C) i.e., approx. stage 22–23, were harvested in ice-cold PBS, the limb buds removed and the distal third cut off with tungsten needles. These pieces were immediately transferred to RNALater (Qiagen). A similar procedure was carried out on the proximal and median thirds. RNA preps were made of pooled limb sections with 20 sections for each region using the Qiagen RNA Easy micro kit, and checked using an Agilent bioanalyzer, using their RNA 6000 nanochip. The integrity values of all RNA samples were between 9.9 and 10 and the 28S and 18S ratio between 1.9 and 2.1. These values indicate little degradation or contamination.Real-time PCR was carried out using an Applied BioSystems HT-7900 machine in a manner similar to Jesmin et al. (2007) using the FastStart Taqman Probe master (Rox) standard reaction mix (Roche). Primers were selected from the Roche Universal Library using their online software. For the Wnt5a reaction, probe 52 was used, and for β-actin, probe 43 was used. Here chick probes are not automatically checked against other possible hybridisation targets, so this was carried out manually by Blasting the candidate sequences against the chick genome in Ensembl. The primers used for the Wnt5a reaction were: forward 5′catgatgaacctacacaatga 3′; reverse 5′ ccacgtcagccaggttgta 3′. And for the β-actin reaction were: F 5′ cacacaagtgcccatttacga 3′; R 5′ caagtccagacgcaggatg 3′. For full protocol see supplementary materials.Computational analysisSimple arithmetical analyses were produced using either the AMIRA software package or Wlz based software tools. AMIRA’s arithmetic module was used to produce averages of multiple datasets and to produce masked datasets for domains of coexpression. Wlz based software developed by the MRC HGU was used to derive medians from multiple datasets and also to derive mean grey level intensity values for both discrete domains and serial sections through the limb. For fuller details of these image manipulations see supplementary materials and methods.For more complex computational analyses each of the experimental gene-expression spatial distributions has been mapped into the standard coordinate framework defined by the model limb. To analyse the gene-expression patterns we first divided the limb into 560 non-overlapping sub-regions each of 10 × 10 × 10 voxels. Each of these was used to sample the experimental gene-expression patterns. For each experimental pattern, the mean gene-expression strength (integrated optical density divided by the volume) within each box was calculated. If the box was partially external to the limb then only the intersecting volume was considered. By this means a 2D matrix of mean expression strengths across the limb was calculated. Each row of the matrix for a given gene is a low-resolution representation of the pattern and each column for a given sample-region is the genetic “signature” for that spatial location.The resulting matrix of gene expression values was analysed using the TMEV4 package from TIGR. The data were analysed using a hierarchical clustering method (Eisen et al., 1998) to produce a nested tree of gene expression pattern similarity based on a Euclidean distance metric. A nested tree was also produced of the similarity of the individual sample regions of the limb making up the 3D data model based on gene expression. The resulting tree was then used to identify clusters of similar expression made of small groups of regions at the terminus of long branches. These regions were used to produce larger 3D domains corresponding to the whole volume occupied by the regions comprising each cluster, which were subsequently visualised using the AMIRA software package.Results and discussionAssessment of efficiency of whole mount in-situ hybridisation and scanning with OPT as a method of detecting gene expression domainsAn initial technical issue we encountered with OPT scanning of WISH (whole mount in-situ hybridisation) specimens was that strong in-situ colour reaction staining can block autofluorescence and prevent the capture of portions of the anatomical data required for subsequent mapping to a reference model. To obviate this problem, we identified a particular depth of staining with the NBT–BCIP substrate, suitable for OPT scanning, which captures an extensive range of the expression pattern and allows the visualisation of dynamic gradients without causing a substantial dropout of the anatomical data necessary for mapping (see supplementary materials Fig. S1). To test our standard in-situ hybridisation protocol, monitor probe penetration of the embryonic limb, and assess the ability of the OPT system to identify graded patterns of expression within deep tissues, we focussed on Wnt5a. Wnt5a has been reported to have a proximo-distal gradient of expression based on both radioactive section in-situ hybridisation and northern blots of distinct portions of the limb (Dealy et al., 1993) and is known to be expressed throughout mesodermal regions of the limb.We first assayed expression by WISH (method modified after Nieto et al., 1996 see supplementary data) Fig. 1A). The Wnt5a whole mount was scanned using OPT and gene expression data mapped onto a reference limb (Fig. 2D), from which virtual sections were derived (Fig. 1B). These virtual sections were then compared with section in-situs (Fig. 1C) performed as in Moorman et al. (2001). This comparison shows that the virtual section captures the extent and range of the Wnt5a expression pattern as accurately as the section in-situ with the exception of some apical ectodermal ridge (AER) expression (Fig. 1C arrowed). For a further illustration of the effective capture of expression patterns using OPT see supplementary data (Fig. S3–5).We then measured the mean grey level signal intensity in all individual sections along the proximo-distal axis. The plot of these data (Fig. 1D) shows low levels of signal in the proximal region (Red, slice 1–28), either very low expression or background. The mean grey level intensity then climbs steeply in the medial region (Orange, slice 29–52) from ∼ 30 up to 100. In the distal region (Green, 53–75) the mean grey level intensity increases less steeply and then levels off at a mean intensity of around 160. In the final 5 slices the mean grey level intensity drops rapidly. This shows the capability of the OPT imaging method to allow a detailed analysis of graded patterns of expression.We also compared Wnt5a expression levels as measured from OPT scans of whole mount in-situs with real-time RT-PCR analysis (Fig. 1F). For both OPT and RT-PCR analyses the limb bud was divided into three regions of equal length along the proximo-distal axis designated proximal, medial and distal (Fig. 1E). For RT-PCR the sample tissue was dissected into the three regions of equal length and samples from 10 embryos were pooled. In the case of the OPT data this division was performed digitally using AMIRA’s segmentation software based upon guidance from the researcher who performed the initial microdissection. The real time RT-PCR data produced relative values for the expression of Wnt5a as follows; expression in the proximal region was taken as the reference expression level, the medial region had a 5 fold increase over the proximal region and the distal region a 19.7 fold increase. The OPT based analysis produced relative values for medial and distal regions of 4.5 fold and 6 fold increases over the value for the proximal region respectively. Therefore, WISH/OPT captures the graded nature of the expression along the proximo-distal axis, indeed the correspondence in the proximal and medial regions is striking, but not across the whole quantitative range captured using RT-PCR as the correspondence falls off dramatically at the higher levels seen in the distal region. The limitations in the captured range of expression may be due to limitations of the WISH detection method, i.e., a nonlinear relationship between signal intensity and RNA quantity. This allows comparisons of the level of expression of a particular gene within a particular scan but not the comparison of absolute levels between different scans, although high and low regions of expression could be compared.The quality of data capture for the gradient of Wnt5A, both from selected domains and from serial virtual sections, suggests that WISH/OPT is suitable for examining complex patterns of graded expression in tissues within developing embryos, but not for quantification of signal with high accuracy and quantitative comparisons of mRNA levels between samples.Reference models for comparative analysisAn initial requirement for meaningful comparison of gene expression patterns is a common spatial reference framework onto which different patterns can be mapped. We have produced a panel of such reference frameworks for several embryonic stages; whole embryos, isolated and fixed according to a standardised protocol, were collected at Hamburger Hamilton (HH) (Hamburger and Hamilton, 1951) stages from HH18 through to HH25. Embryos were then scanned by OPT using autofluorescence stimulated by a UV lamp and reconstructed to produce 3D models. These models were rendered using the AMIRA software package to show morphology and gross anatomy of the embryos (Figs. 2A–H). This resulted in a clear 3D visualisation of the embryo and revealed details such as the AER – the thickened layer of epithelium that rims the distal limb buds – in the models from HH20 onwards (Figs. 2A–F, blue arrows indicate AER). This is best appreciated when the model is rotated (Fig. 2H″, blue arrows indicate AER) (to view models in 3D see movies in supplementary data Fig. S6). Measurements of the length (L; along a line from anterior join of bud and trunk to posterior join) and width (W; a line from distal tip to trunk perpendicular to line L) of wing buds were made on the dorsal plan view (Fig. 2, Table 1) and the L/W ratio calculated in order to stage the reference embryo (Fig. S2). The procedure was the same as one would use to stage a living chick embryo using the staging criteria of Hamburger and Hamilton (Fig. 2, Table 2) (supplementary Fig. S2).Reference models for specific regions, such as the developing limb buds (Figs. 2A′–H′), can be extracted from whole embryo models and used for mapping of gene expression in these regions. Within the AMIRA program, such extracted models remain in register with the model of the whole embryo from which they are derived, thus allowing maintenance of a consistent positional system between all expression patterns mapped. These relative positions are also maintained when files are exported to the Wlz file format, which can be used to store 3 dimensional greyscale image data, using software tools developed by the MRC HGU. Subsequent mappings and analyses of gene expression reported here were performed on an extracted data set for the right wing bud of the late stage HH22 reference embryo (Fig. 2E′). Further reference sets for other stages can be easily generated.Reproducibility of 3D mapping of gene expression within and between labsTo test the reliability of our in-situ protocol and 3D mapping, we focussed on the expression domain of Sonic hedgehog (Shh) in HH22 wing buds. Shh is expressed in the polarizing region at the posterior margin of the wing bud and Shh expression correlates with maps of polarizing activity (Riddle et al., 1993). We compared data generated from wing buds in a single round of whole mount in-situ hybridisation experiments and from wing buds processed in three different labs.A round of in-situ hybridisations using standardised protocols for Shh expression were carried out on 4 embryos and the sense control probe was used on 2 embryos. All 4 embryos assayed using the Shh anti-sense probe were treated identically and detection was carried out in the same tube. Nevertheless there were differences in intensity of staining of Shh transcripts in the polarizing region (Figs. 3A–D) with wing buds of one embryo (Fig. 3C) showing very faint staining. Wing buds of control embryos (data not shown) had no visible background staining. In-situs of embryos from Edinburgh (Fig. 3E) and Dublin (Fig. 3F) showed similar localisation of Shh transcripts in the wing bud and one embryo from each site was then scanned together with the four embryos from the run carried out in Dundee.All six embryos were OPT scanned through different channels to capture a) autofluorescence to represent anatomy and b) the staining pattern under visible light. Having reconstructed 3D representations of each, we then digitally extracted the right wing bud and accompanying flank using the same spatial parameters. Co-visualisation of both anatomy (grey) and expression (orange) with volume rendering (Figs. 3A′–F′) shows the representation of the original in-situ data (Figs. 3A–F) following OPT scanning and reconstruction.The patterns of Shh expression in each of the six wing buds were then mapped in 3D to the HH22 reference wing (Fig. 2E). Figs. 3A″–F″ shows heat maps of intensity of Shh expression in one section, taken across the antero-posterior/dorso-ventral (A-Po/Do-V) axes of the HH22 stage reference wing bud in a plane situated next to the AER (Fig. 3L), for the individual patterns of expression for the six source wing buds. Signal intensity of mapped gene expression data is represented by the heatmap in Fig. 3A″ corresponding to grey scale values between 1 and 255, this measure of intensity is not suitable for precise quantitative comparisons between samples but allows visualisation of the differing levels of expression within a sample within the limits of the WISH methodology. Of the 4 wing buds from the Dundee laboratory one showed a weak signal (Fig. 3C″), with a maximal intensity of only 28. The sense controls were both very clean with no signal (data not shown). Scans for Edinburgh (Fig. 3E″) and Dublin (Fig. 3F″) had a localisation very similar to Fig. 3A″ but an intensity of expression much closer to Fig. 3D″.Both unique domains of expression and intersecting domains of expression can be derived for these data sets. Visualisation of the unique expression domains in 3 dimensions shows no unique domains for the specimens shown in A–D and only small peripheral regions for the higher intensity signal data sets from specimens in E and F (Data not shown). The intersect of the expression domains between all the scans (Fig. 3G) is restricted by the small domain of the outlying data set (Fig. 3C″), if we remove this outlying data from the calculation we have an intersect domain almost twice as large (Fig. 3H).Clearly, there is some variation between individual scans and an occasional extreme outlier, but our data suggest a clear common domain of gene expression is identifiable. To produce a reliable and robust domain which could compensate for the variations we see in individual in-situs, we incorporated the data from our multiple scans into one domain. We produced means of the data from the scans and corrected for background from the controls, using both raw and normalised data (Figs. 3I–J). The mean of the datasets (Fig. 3I) was heavily influenced by high intensity samples (compare Fig. 3E″ and Fig. 3I). To correct for the variation in signal intensity, whether as a result of differences in the in-situ itself or from the scanning steps, data sets were normalised by stretching their entire grey value range to cover the maximal range of 0–255. Such correction had almost no effect (Fig. 3J), although there was a small extension of the domain proximally. As an alternative to normalisation to account for variability and extreme outlying values we derived a median value for each voxel based on the grey levels of all of the scans (Fig. 3K). This median value seemed less dominated by outliers and a better representation of the range seen across the differing samples although it produced a more conservative domain than the simple mean since it removed areas where signal was not apparent in more than half of the samples.These results suggest that best practice for producing a reliable domain of expression for comparison is to perform several in-situs developed to suitable stain intensity and merge the resultant data. A minimum of four scans seems advisable to contribute to the merged data set and these may then be mapped to a common reference and a mean or median expression pattern derived. A mean of the patterns appears to better emphasise the extent of the expression domain while the median produces a more conservative domain less affected by outliers. As our analysis shows, more complicated treatment of the data seems to make little difference to the resulting domains although, for in-situs with persistently low signal, normalisation might help emphasise gradients of expression in some cases.This averaging or otherwise merging together of multiple samples is less necessary in the case of previously well characterised genes where the expected pattern of expression is already known and a representative sample can be confidently identified. This approach should be most valuable in the situation where the gene expression is either poorly characterised or unknown, as is likely to be the case in large scale screens.Comparative analysis of 3D Gene expression patternsOPT is a rapid method of capturing the 3D expression pattern and can allow data on multiple genes to be integrated into a common framework, therefore we used 3D warping of OPT data to our reference models to produce such an integrated data set. Expression patterns for the genes Shh, HoxD13, Fgf8, Msx1, Lmx1, Wnt5a and Tbx3 (for representative in-situs see supplementary Fig. S6) were mapped to our HH22 reference model (Fig. 2E) using AMIRA’s 3D warping capabilities. This stage was chosen as it represents a well-developed limb bud but still consisting mainly of undifferentiated mesenchyme cells. These genes were chosen for particular characteristics of expression such as dorsal restriction, Lmx1, specific expression in the AER, Fgf8, specific expression in the mesoderm, Shh, or particular gradients of expression, such as the proximo-distal gradient of Wnt5a. Particular specimens for scanning were chosen based on in-situ quality in comparison to others processed with them, usually the best example from 5–6 in-situs. The resulting mappings were then visualised in 3D (see supplementary materials Fig. S8), virtual sections were derived along specific planes (Fig. 4A), gene expression patterns and intensity were visualised on the anterior–posterior/dorsal–ventral plane (Fig. 4A.i). The expression domains of several pairs of genes were co-visualised on the antero-posterior/dorso-ventral (A-Po/Do-V) (Fig. 4C, section plane in Fig. 4A.i), antero-posterior/proximo-distal (A-Po/Pr-Di) (Fig. 4D, section plane in Fig. 4A.ii) and dorso-ventral/proximo-distal (Do-V/Pr-Di) (Fig. 4E, section plane in Fig. 4A.iii) planes to allow some specific comparisons to be drawn (Figs. 4C–E; for 3D visualisations see supplementary materials Fig. S11–18). These mappings show both expected features, such as the dorsal expression pattern of Lmx1, and novel features such as an apparent gradient of expression from ventral to dorsal in Wnt5a. Indeed several dorso-ventral asymmetries were by far the most striking features to emerge.One such previously unappreciated asymmetry was in Shh expression (Fig. 4B), which shows that Shh expression extends further anteriorly on the ventral side of the limb. A transverse section through the limb along plane E shows that the asymmetry is more complex and the domain of Shh expression is skewed with the more proximal regions of the limb showing a more dorsal expression of Shh (Fig. 4E.ii). Planes taken at more anterior levels lose this obvious skewing (see supplementary Fig. S9) and elements of this distribution are confirmed by both normal Shh whole mount in-situ and a double in-situ for Shh and Fgf8 (see supplementary materials Fig. S10). This may explain the observations of Yang and Niswander (1995) who reported no apparent dorso-ventral asymmetry in Shh expression at HH24 based on sectioned whole mount in-situ hybridisation data. Alternatively this may be due to the different stages assayed. A more dorsal distribution in the more proximal regions of the limb may suggest that a dorsally localised signal, such as Wnt7a, might be maintaining dorsal expression proximally while more distal expression would be maintained evenly across the dorso-ventral axis by signals from the apical ectodermal ridge. Indeed it is already known that Wnt7a plays a role in maintaining Shh expression in the polarising region (Parr and McMahon, 1995; Yang and Niswander, 1995). Furthermore Kawakami et al. (1999) reported that Frizzled10 (a Wnt receptor) colocalises with Shh dorsally and suggested that Shh expression in this part of the polarising region might be regulated by Wnt7a (Kawakami et al., 2000).The pairwise comparisons similarly produced both expected and unexpected results. HoxD13 shows striking asymmetrical dorso-ventral expression (Figs. 4C.v and E.v) and the ventral margin of HoxD13 expression appears to coincide with that of Lmx1 (Figs. 4C.viii and E.viii; Movie in supplementary Fig. S18), suggesting a possible regulatory relationship. The dorso-ventral asymmetry in Hoxd13 had been previously noted by Duboule et al. from 3D reconstructions based on radioactive section in-situs (Olivo et al., 1993) but at this time Lmx1 was not known. Recent lineage tracing studies in the developing mouse limb have shown the existence of a dorsoventral lineage restriction compartment further suggesting that there is still considerable complexity in the dorsoventral organisation of the limb to be discovered (Arques et al., 2007).Comparison of Shh and Fgf8 expression (Figs. 4D.iii and E.iii; movie in supplementary Fig. S13) shows an unexpected overlap of expression in the mesoderm. This shows that there are important limits of spatial resolution to the current mapping procedure given the well-characterised expression of these genes in mesoderm and apical ectodermal ridge respectively. Since the AER is one of the principle morphological features of the limb, it is heavily used in the landmarking process, which is the first stage of mapping expression data to the reference limb. Strong in-situ staining, as seen with Fgf8, can block autofluorescence and prevent the capture of the anatomical data for the ridge. A comparison of Wnt5a and Fgf8 (Movie in supplementary Fig. S11) also shows considerable overlap except for the most anterior region of Fgf8 expression (Fig. 4C.i) and a persistent ‘leading edge’ of Fgf8 expression in the most distal portion of the limb (Fig. 4E.i). In this case, we would expect to see considerable overlap due to expression of Wnt5a in the ridge, although it is not clear whether the common domain of expression accurately represents just ridge expression.OPT using Tyramide Signal AmplificationTo address the problem of the loss of anatomical landmarks when looking at strongly expressed ectodermal genes, such as Fgf8, we used a Tyramide Signal Amplification (TSA) kit (Invitrogen) to enhance a fluorescent colour reaction and avoid the blocking effect seen with chromagenic substrates such as NBT–BCIP. We compared the expression of Tbx3 and Fgf8 at HH24, mapped to the HH24 reference limb (Fig. 2G′), from a normal NBT–BCIP based colour reaction for Tbx3 and a TSA fluorescent colour reaction for Fgf8 (Figs. 5A, B). This approach provided a much more accurate localisation of Fgf8 expression to the apical ridge compared to that seen in Fig. 4. When we carried out a pairwise comparison between Fgf8 and Tbx3 expression patterns, the overlap was very much reduced compared to that seen between Fgf8 and Shh in Fig. 4D.iii and compared with Fgf8 and Tbx3 from our original data set in which we had used an NBT–BCIP reaction for both genes (Fig. 5C). While these mapped localisations are not sufficient for accurately discriminating expression from expressing tissues close together or very thin tissue layers a database containing such mapped data would be linked to the original 3D scans allowing visualisation of the data.Computational comparisons of mapped gene expression patternsDespite the shortcomings discussed above, our overall mapping strategy provides a valuable tool for analysing spatial and temporal relationships of multiple complex patterns.To begin to apply computational methods to analyse these multiple data sets at once and to look for similarities in patterns of gene expression, we utilised software produced by the MRC HGU to manipulate 3D image data in the Wlz format, a format used by EMAGE. AMIRA files were converted to Wlz format and a set of sub-regions of 10 × 10 × 10 voxels were defined for the early HH22 reference limb producing a coarse sampling of the 3D model. The expression data for the previously analysed genes and an additional gene, Wnt3a, were then used to derive mean expression values from the grey levels of each OPT scan for the newly defined volumes. This produced a matrix of common positional IDs and expression levels for each gene.These data were then analysed using hierarchical clustering for both the genes and the regions defined by the coarse sampling. This analysis was performed and visualised using the TMEV package (http://www.tm4.org/mev.html) (Fig. 6A). The clustering of these genes fits our expectations with Fgf8 and Wnt3a, both known to be expressed in the apical ridge (Barrow et al., 2003; Kengaku et al., 1998; McQueeney et al., 2002), treeing out together. Visualisation of the clustered regions shows that they largely form contiguous spatial domains (Figs. 6B–G; supplementary Fig. S19); some of these domains are associated with specific regions of the limb. Cluster B lies in a plane through the limb along the proximal–distal axis at the level of the anterior margin of Shh expression (Fig. 6B), cluster D is associated with the AER in the anterior of the limb (Fig. 6D) while cluster E is in the dorsal margin of the limb (Fig. 6E). None of the clusters visualised here corresponds simply to the expression pattern of one particular gene.This simple level of clustering computational analysis shows the potential for methods developed to study gene expression data from other sources, such as microarray data, to be applied to the study of 3D gene expression data. Not only can we identify similarly expressed genes using this method but it should also be possible to identify specific regions of the limb that may be important in the regulation of gene expression, such as novel signalling centres.ConclusionWe have developed improved technology for the production of 3D atlases of gene expression (Baldock et al., 2003). Specifically, we have shown that OPT is a reliable and efficient way of visualising 3D patterns of gene expression in the chick limb and have been able to directly compare different patterns on reference models using a 3D warping technique. This technique allows visualisation both of samples too large for confocal microscopy (Welten et al., 2006) and those too small for good resolution with microMRI (see Li et al., 2007 for visualisation of chick wings at later stages). Furthermore, visualisation of gene expression is still rudimentary with microMRI (Liu et al., 2007a).As more 3D patterns of gene expression are mapped, simple pairwise comparisons will no longer be sufficient to analyse more complex relationships between groups of genes and then the computational approaches we have used here will be greatly beneficial, indeed similar methods have been used to study the expression of around 20,000 genes in the mouse brain (Lein et al., 2007; Liu et al., 2007b). Although we have only compared 7 genes known to be expressed in chick limb development we have already revealed some previously unappreciated asymmetries and relationships, particularly with respect to the dorso-ventral axis. These techniques will also be more generally applicable to different developing structures and organisms.\n\nREFERENCES:\n1. AlbertsonR.C.KocherT.D.Assessing morphological differences in an adaptive trait: a landmark-based morphometric approachJ. Exp. Zool.289200138540311351326\n2. ArquesC.G.DoohanR.SharpeJ.TorresM.Cell tracing reveals a dorsoventral lineage restriction plane in the mouse limb bud mesenchymeDevelopment13420073713372217715176\n3. BaldockR.A.BardJ.B.BurgerA.BurtonN.ChristiansenJ.FengG.HillB.HoughtonD.KaufmanM.RaoJ.SharpeJ.RossA.StevensonP.VenkataramanS.WaterhouseA.YangY.DavidsonD.R.EMAP and EMAGE: a framework for understanding spatially organized dataNeuroinformatics1200330932515043218\n4. BarrowJ.R.ThomasK.R.Boussadia-ZahuiO.MooreR.KemlerR.CapecchiM.R.McMahonA.P.Ectodermal Wnt3/beta-catenin signaling is required for the establishment and maintenance of the apical ectodermal ridgeGenes Dev.17200339440912569130\n5. BellG.W.YatskievychT.A.AntinP.B.GEISHA, a whole-mount in situ hybridization gene expression screen in chicken embryosDev. Dyn.229200467768714991723\n6. BoardmanP.E.Sanz-EzquerroJ.OvertonI.M.BurtD.W.BoschE.FongW.T.TickleC.BrownW.R.WilsonS.A.HubbardS.J.A comprehensive collection of chicken cDNAsCurr. Biol.1220021965196912445392\n7. BooksteinF.Principal warps: thin-plate splines and the decomposition of deformationsIEEE Trans. Pattern Anal. Mach. Intell.111989567585\n8. BrownW.R.HubbardS.J.TickleC.WilsonS.A.The chicken as a model for large-scale analysis of vertebrate gene functionNat. Rev., Genet.42003879812560806\n9. BrunerE.SaracinoB.RicciF.TafuriM.PassarelloP.ManziG.Midsagittal cranial shape variation in the genus Homo by geometric morphometricsColl. Antropol.2820049911215636068\n10. CarreW.WangX.PorterT.E.NysY.TangJ.BernbergE.MorganR.BurnsideJ.AggreyS.E.SimonJ.CogburnL.A.Chicken genomics resource: sequencing and annotation of 35,407 ESTs from single and multiple tissue cDNA libraries and CAP3 assembly of a chicken gene indexPhysiol. Genomics25200651452416554550\n11. CrossleyP.H.MinowadaG.MacArthurC.A.MartinG.R.Roles for FGF8 in the induction, initiation, and maintenance of chick limb developmentCell8419961271368548816\n12. DealyC.N.RothA.FerrariD.BrownA.M.KosherR.A.Wnt-5a and Wnt-7a are expressed in the developing chick limb bud in a manner suggesting roles in pattern formation along the proximodistal and dorsoventral axesMech. Dev.4319931751868297789\n13. DeLaurierA.SchweitzerR.LoganM.Pitx1 determines the morphology of muscle, tendon, and bones of the hindlimbDev. Biol.2992006223416989801\n14. EchelardY.EpsteinD.J.St-JacquesB.ShenL.MohlerJ.McMahonJ.A.McMahonA.P.Sonic hedgehog, a member of a family of putative signaling molecules, is implicated in the regulation of CNS polarityCell751993141714307916661\n15. EisenM.B.SpellmanP.T.BrownP.O.BotsteinD.Cluster analysis and display of genome-wide expression patternsProc. Natl. Acad. Sci. U. S. A.95199814863148689843981\n16. HamburgerV.HamiltonH.L.A series of normal stages in the development of the chick embryoJ. Morphol.881951499224539719\n17. HarmonE.H.The shape of the hominoid proximal femur: a geometric morphometric analysisJ. Anat.210200717018517310545\n18. HillR.E.JonesP.F.ReesA.R.SimeC.M.JusticeM.J.CopelandN.G.JenkinsN.A.GrahamE.DavidsonD.R.A new family of mouse homeo box-containing genes: molecular structure, chromosomal location, and developmental expression of Hox-7.1Genes Dev.3198926372565278\n19. HubbardS.J.GrafhamD.V.BeattieK.J.OvertonI.M.McLarenS.R.CroningM.D.BoardmanP.E.BonfieldJ.K.BurnsideJ.DaviesR.M.FarrellE.R.FrancisM.D.Griffiths-JonesS.HumphrayS.J.HylandC.ScottC.E.TangH.TaylorR.G.TickleC.BrownW.R.BirneyE.RogersJ.WilsonS.A.Transcriptome analysis for the chicken based on 19,626 finished cDNA sequences and 485,337 expressed sequence tagsGenome Res.15200517418315590942\n20. International Chicken Genome Sequencing ConsortiumSequence and comparative analysis of the chicken genome provide unique perspectives on vertebrate evolutionNature432200469571615592404\n21. IsaacA.Rodriguez-EstebanC.RyanA.AltabefM.TsukuiT.PatelK.TickleC.Izpisua-BelmonteJ.C.Tbx genes and limb identity in chick embryo developmentDevelopment1251998186718759550719\n22. JesminS.ZaediS.ShimojoN.IemitsuM.MasuzawaK.YamaguchiN.MowaC.N.MaedaS.HattoriY.MiyauchiT.Endothelin antagonism normalizes VEGF signaling and cardiac function in STZ-induced diabetic rat heartsAm. J. Physiol.: Endocrinol. Metab.2922007E1030E104017148754\n23. KawakamiY.WadaN.NishimatsuS.I.IshikawaT.NojiS.NohnoT.Involvement of Wnt-5a in chondrogenic pattern formation in the chick limb budDev. Growth Differ.411999294010445500\n24. KawakamiY.WadaN.NishimatsuS.NohnoT.Involvement of frizzled-10 in Wnt-7a signaling during chick limb developmentDev. Growth Differ.42200056156911142678\n25. KengakuM.CapdevilaJ.Rodriguez-EstebanC.De La PenaJ.JohnsonR.L.BelmonteJ.C.TabinC.J.Distinct WNT pathways regulating AER formation and dorsoventral polarity in the chick limb budScience2801998127412779596583\n26. KerwinJ.ScottM.SharpeJ.PuellesL.RobsonS.C.Martinez-de-la-TorreM.FerranJ.L.FengG.BaldockR.StrachanT.DavidsonD.LindsayS.3 dimensional modelling of early human brain development using optical projection tomographyBMC Neurosci.520042715298700\n27. KimH.LimD.HanB.K.SungS.JeonM.MoonS.KangY.NamJ.HanJ.Y.ChickGCE: a novel germ cell EST database for studying the early developmental stage in chickensGenomics88200625225716714094\n28. LeeK.AvondoJ.MorrisonH.BlotL.StarkM.SharpeJ.BanghamA.CoenE.Visualizing plant development and gene expression in three dimensions using optical projection tomographyPlant Cell1820062145215616905654\n29. LeinE.S.HawrylyczM.J.AoN.AyresM.BensingerA.BernardA.BoeA.F.BoguskiM.S.BrockwayK.S.ByrnesE.J.ChenL.ChenL.ChenT.M.ChinM.C.ChongJ.CrookB.E.CzaplinskaA.DangC.N.DattaS.DeeN.R.DesakiA.L.DestaT.DiepE.DolbeareT.A.DonelanM.J.DongH.W.DoughertyJ.G.DuncanB.J.EbbertA.J.EicheleG.EstinL.K.FaberC.FacerB.A.FieldsR.FischerS.R.FlissT.P.FrensleyC.GatesS.N.GlattfelderK.J.HalversonK.R.HartM.R.HohmannJ.G.HowellM.P.JeungD.P.JohnsonR.A.KarrP.T.KawalR.KidneyJ.M.KnapikR.H.KuanC.L.LakeJ.H.LarameeA.R.LarsenK.D.LauC.LemonT.A.LiangA.J.LiuY.LuongL.T.MichaelsJ.MorganJ.J.MorganR.J.MortrudM.T.MosquedaN.F.NgL.L.NgR.OrtaG.J.OverlyC.C.PakT.H.ParryS.E.PathakS.D.PearsonO.C.PuchalskiR.B.RileyZ.L.RockettH.R.RowlandS.A.RoyallJ.J.RuizM.J.SarnoN.R.SchaffnitK.ShapovalovaN.V.SivisayT.SlaughterbeckC.R.SmithS.C.SmithK.A.SmithB.I.SodtA.J.StewartN.N.StumpfK.R.SunkinS.M.SutramM.TamA.TeemerC.D.ThallerC.ThompsonC.L.VarnamL.R.ViselA.WhitlockR.M.WohnoutkaP.E.WolkeyC.K.WongV.Y.Genome-wide atlas of gene expression in the adult mouse brainNature445200716817617151600\n30. LiX.LiuJ.DaveyM.G.DuceS.JaberiJ.LiuG.DavidsonG.TenentS.MahoodR.BrownP.CunninghamC.BainA.BeattieK.McDonaldL.SchmidtK.TowersM.TickleC.ChudekJ.A.Micro-magnetic resonance imaging of avian embryosJ. Anat.211200779880918045352\n31. LiuC.H.KimY.R.RenJ.Q.EichlerF.RosenB.R.LiuP.K.Imaging cerebral gene transcripts in live animalsJ. Neurosci.27200771372217234603\n32. LiuZ.YanS.F.WalkerJ.R.ZwingmanT.A.JiangT.LiJ.ZhouY.Study of gene function based on spatial co-expression in a high-resolution mouse brain atlasBMC Syst. Biol.120071917437647\n33. ManiatisT.FritschE.SambroookJ.Molecular Cloning: A Laboratory Manual1982Cold Spring Harbor Laboratory PressNew York\n34. McGurkL.MorrisonH.KeeganL.P.SharpeJ.O’ConnellM.A.Three-dimensional imaging of Drosophila melanogasterPLoS ONE22007e83417786206\n35. McQueeneyK.SouferR.DealyC.N.Beta-catenin-dependent Wnt signaling in apical ectodermal ridge induction and FGF8 expression in normal and limbless mutant chick limbsDev. Growth Differ.44200231532512175366\n36. MoormanA.F.HouwelingA.C.de BoerP.A.ChristoffelsV.M.Sensitive nonradioactive detection of mRNA in tissue sections: novel application of the whole-mount in situ hybridization protocolJ. Histochem. Cytochem.4920011811118473\n37. NietoM.A.PatelK.WilkinsonD.G.In situ hybridization analysis of chick embryos in whole mount and tissue sectionsMethods Cell Biol.5119962192358722478\n38. OlivoJ.Izpisúa BelmonteJ.C.TickleC.BoulinC.DubouleD.Reconstruction from serial sections: a tool for developmental biology. Application to Hox genes expression in chicken wing budsBioImaging11993115158\n39. ParrB.A.McMahonA.P.Dorsalizing signal Wnt-7a required for normal polarity of D-V and A-P axes of mouse limbNature37419953503537885472\n40. RiddleR.D.JohnsonR.L.LauferE.TabinC.Sonic hedgehog mediates the polarizing activity of the ZPACell751993140114168269518\n41. SharpeJ.AhlgrenU.PerryP.HillB.RossA.Hecksher-SorensenJ.BaldockR.DavidsonD.Optical projection tomography as a tool for 3D microscopy and gene expression studiesScience296200254154511964482\n42. TickleC.The contribution of chicken embryology to the understanding of vertebrate limb developmentMech. Dev.12120041019102915296968\n43. VerbeekF.J.den BroederM.J.BoonP.J.B.B.DoerryE.van RaaijE.J.ZivkovicD.Standard 3D digital atlas of zebrafish embryonic development for projection of experimental dataProc. SPIE39641999242252\n44. ViselA.ThallerC.EicheleG.GenePaint.org: an atlas of gene expression patterns in the mouse embryoNucleic Acids Res.322004D552D55614681479\n45. WeltenM.C.de HaanS.B.van den BoogertN.NoordermeerJ.N.LamersG.E.SpainkH.P.MeijerA.H.VerbeekF.J.zebraFISH: Fluorescent in situ hybridization protocol and 3D imaging of gene expression patternsZebrafish3200646547618377226\n46. YangY.NiswanderL.Interaction between the signaling molecules WNT7a and SHH during vertebrate limb development: dorsal signals regulate anteropo7sterior patterningCell8019959399477697724\n47. YehJ.The effect of miniaturized body size on skeletal morphology in frogsEvol. Int. J. Org. Evol.562002628641"
+}

batch_0/PMC2530517.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2530517",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2530517\nAUTHORS: Sabika Firasat, S. Amer Riazuddin, J. Fielding Hejtmancik, Sheikh Riazuddin\n\nABSTRACT:\nPurposeTwo consanguineous Pakistani families with autosomal recessive primary congenital glaucoma were recruited to identify the disease locus.MethodsOphthalmic examinations including slit lamp biomicroscopy and applanation tonometry were employed to classify the phenotype. Blood samples were collected and genomic DNA was extracted. A genome wide scan was performed on both families with 382 polymorphic microsatellite markers. Two point LOD scores were calculated, and haplotypes were constructed to define the disease interval.ResultsClinical records and ophthalmic examinations suggest that affected individuals in families PKGL005 and PKGL025 have primary congenital glaucoma. Maximum two-point LOD scores of 5.88 with D14S61 at θ=0 and 6.19 with D14S43 at θ=0 were obtained for families PKGL005 and PKGL025, respectively. Haplotype analysis defined the disease locus as spanning a 6.56 cM (~4.2 Mb) genetic interval flanked by D14S289 proximally and D14S85 distally.ConclusionsLinkage analysis localizes autosomal recessive primary congenital glaucoma to chromosome 14q24.2–24.3 in consanguineous Pakistani families.\n\nBODY:\nIntroductionGlaucoma is the second leading cause of vision loss, and approximately 15% of blindness worldwide result from glaucoma [1]. It is a group of poorly understood neurodegenerative disorders that are usually associated with elevated intraocular pressure [2]. Glaucoma is clinically and genetically heterogeneous with several different forms, each with diverse causes and severities. Clinically, it is characterized by slow but progressive degeneration of retinal ganglion cells and their axons, leading to deterioration of the visual field and to optic nerve atrophy.Although rare, primary congenital glaucoma (PCG) is the most common form of glaucoma in infants with an overall occurrence of 1 in 10,000 births [3]. It is prevalent in countries where consanguinity is common with incidence as high as 1 in 1,250 births in the Slovak population, 1 in 2,500 births in Saudi Arabia, and 1 in 3,300 births in the state of Andhra Pradesh in India [4,5]. PCG is an inherited ocular congenital anomaly of the trabecular meshwork and anterior chamber angle [6-9]. This leads to the obstruction of aqueous outflow and increased intraocular pressure (IOP) resulting in optic nerve damage leading to childhood blindness. The disease manifests in the neonatal or early infantile period with symptoms of photophobia, epiphora, signs of globe enlargement, edema, opacification of the cornea, and breaks in Descemet's membrane. The mode of inheritance is largely autosomal recessive with variable penetrance, but rare cases of pseudo dominance are also seen in families with multiple consanguinity [10-13]. To date, three genetic loci have been reported for autosomal recessive PCG, GLC3A (2p21; OMIM 231300), GLC3B (1p36; OMIM 600975), and GLC3C (14q24.3), with pathogenic mutations only reported in the human cytochrome P450 gene (CYP1B1; OMIM 601771) [14,15]. It is significant to note that CYP1B1 mutations have also been reported in patients with early onset of primary open-angle glaucoma. Additionally, autosomal dominant forms of PCG have been reported, and MYOC, a gene associated with primary open-angle glaucoma, is reported to play a possible role in the pathogenesis [16,17].The current study is aimed to explore the genetic basis of PCG in the Pakistani population. A genome wide linkage analysis was performed, which showed segregation of PCG in two consanguineous Pakistani families. Microsatellite markers on chromosome 14q24.2–24.3 cosegregated with the disease phenotype and defined the disease locus as spanning a 6.56 cM (~4.2 Mb) genetic interval flanked by D14S289 proximally and D14S85 distally.MethodsThirteen consanguineous Pakistani families with PCG were recruited to participate in this study to understand the genetic aspects of glaucoma at the Centre of Excellence in Molecular Biology (Lahore, Pakistan). Institutional Review Board approval was obtained for this study from the Centre of Excellence in Molecular Biology (CEMB). The participating subjects gave informed consent consistent with the tenets of the Declaration of Helsinki. Both families described in this study are from the Punjab province of Pakistan.A detailed medical history was obtained by interviewing family members. All of the ophthalmic examinations including slit lamp biomicroscopy and applanation tonometry were completed at the Layton Rahmatullah Benevolent Trust (LRBT) hospital (Lahore, Pakistan). Diagnosis of PCG was based on established criteria that include measurement of IOP, measurement of corneal diameters, and observation of optic nerve head where possible as well as symptoms of corneal edema including photophobia, buphthalmos, cloudy cornea, and excessive tearing. Patients with elevated IOP associated with other systemic or ocular abnormalities were excluded. Blood samples were collected from affected and unaffected family members. DNA was extracted by a non-organic method as described by Grimberg et al. [18].Genotype analysisA genome wide scan was performed with 382 highly polymorphic fluorescent markers from the ABI PRISM Linkage Mapping Set MD-10 (Applied Biosystems, Foster City, CA) having an average spacing of 10 cM. Multiplex polymerase chain reactions (PCRs) were performed in a 5 μl mixture containing 40 ng genomic DNA, various combinations of 10 μM dye labeled primer pairs, 0.5 μl 10X GeneAmp PCR Buffer II, 0.5 μl 10mM Gene Amp dNTP mix, 2.5 mM MgCl2, and 0.2 U of Taq DNA polymerase (AmpliTaq Gold Enzyme; Applied Biosystems). Amplification was performed in a GeneAmp PCR System 9700 (Applied Biosystems). Initial denaturation was performed for 5 min at 95 °C followed by 10 cycles for 15 s at 94 °C, for 15 s at 55 °C, and for 30 s at 72 °C, and then 20 cycles for 15 s at 89 °C, for 15 s at 55 °C, and for 30 s at 72 °C. The final extension was performed for 10 min at 72 °C followed by a final hold at 4 °C. PCR products from each DNA sample were pooled and mixed with a loading cocktail containing HD-400 size standards (PE Applied Biosystems). The resulting PCR products were separated in an ABI 3100 DNA Analyzer and analyzed by using the GeneMapper software package (Applied Biosystems).Linkage analysisTwo point linkage analysis were performed using the FASTLINK version of MLINK from the LINKAGE Program Package (provided in the public domain by the Human Genome Mapping Project Resources Centre, Cambridge, UK) [19,20]. Maximum LOD scores were calculated using ILINK. Autosomal recessive PCG was analyzed as a fully penetrant trait with an affected allele frequency of 0.001. The marker order and distances between the markers were obtained from the Marshfield database. For the initial genome scan, equal allele frequencies were assumed while for fine mapping, allele frequencies were estimated from 100 unrelated and unaffected individuals from the Punjab province of Pakistan.Mutation screeningIndividual exons were amplified by PCR using primer pairs designed by using the primer3 program (primer sequences and annealing temperatures are available upon request). Amplifications were performed in 25 μl reactions containing 50 ng of genomic DNA, 2.5 μl 10X GeneAmp PCR Buffer II, 8 pmoles of each primer, 2.5 mM dNTP, 2.5 mM MgCl2, and 0.2 U Taq DNA polymerase. Amplification was performed in a GeneAmp PCR System 9700 (Applied Biosystems). PCR amplification consisted of a denaturation step at 96 °C for 5 min followed by 40 cycles, each cycle starting at 96 °C for 45 s followed by 57 °C for 45 s and 72 °C for 1 min. PCR products were analyzed on 2% agarose gel and purified by ethanol precipitation. The PCR primers for each exon were used for bidirectional sequencing using Big Dye Terminator Ready reaction mix (Applied Biosystems) according to manufacturer instructions. Sequencing products were precipitated and resuspended in 10 μl of formamide and denatured at 95 °C for 5 min. Sequencing was performed on an ABI PRISM 3100 Automated sequencer (Applied Biosystems). Sequencing results were assembled by the ABI PRISM sequencing analysis software version 3.7 and analyzed using Chromas software (version 1.45).ResultsThe two families reported here, PKGL005 and PKGL025, are from the Punjab province of Pakistan. Ophthalmic examinations and medical history for both families concluded that a total of 11 affected individuals in both families have primary congenital glaucoma (PCG). The symptoms of PCG in affected individuals of PKGL005 appeared in the first three years of life. Visual acuity was confined to light perception and/or counting fingers. The cup to disc ratios of affected individuals 11 and 41 were 0.8 (OD) and 1.0/0.3 (OD/OS), and the recorded IOPs for individuals 11 and 41 were 32/20 mm Hg (OD/OS) and 38/30 mm Hg (OD/OS), respectively (Table 1). On the other hand, symptoms of PCG in PKGL025 were either present at birth or appeared in the first six weeks of life. Visual acuity was reduced to counting figures and/or light perception with bilateral buphthalmos eyes. The IOPs for affected individuals in PKGL025 were either above the normal range or was controlled by medical treatment (Table 1).Table 1Clinical features of affected individuals of families PKGL005 and PKGL025.Family numberIndividual IDGenderAge of onsetAge at time of studyMaximum IOP (OD/OS)C/D ratio (OD/OS)Visual acuity (OD/OS)Other changesPKGL00511M3 years8 years32/20*0.8/NVCF/CFMegalocorneaPKGL00541M3 years6 years38/301.0/0.3CF/CFMegalocorneaPKGL02515FBy birth4 years20*/24*0.3/0.3CF/CFBuphthalmosPKGL02528MBy birth8 months25/26NACF/CFBuphthalmosPKGL02525FBy birth5 years12*/16*0.9/0.9CF/CFMegalocornea, cornea hazePKGL02524FBy birth15 yearsNA/38NV/1.0NPL/NPLBuphthalmosAn asterisk indicates that IOP is controlled by medical or surgical treatment. IOP, intraocular pressure; OD, right eye; OS, left eye; PL, perception of light; NPL, no perception of light; HM, hand motion; NA, not available; NV, no view because of eyeball atrophy or corneal opacity; CF, counting fingers.Initially, all reported loci for PCG were excluded for linkage using closely spaced microsatellite markers (data not shown). A genome wide scan was completed with the ABI MD10 panel, which consisted of 382 polymorphic microsatellite markers and spaced at an average of 10 cM across the whole genome. During the genome-wide scan, LOD scores above 1.5 were obtained for markers D6S308, D10S59, D10S1652, D11S1314, D14S74, D14S68, D16S404, and D18S53 in PKGL005 and for markers D2S112, D3S1279, D9S1776, D14S74, and D21S263 in PKGL025. Of these markers, D6S308, D10S59, D10S1652, D11S1314, D16S404, and D18S53 have closely flanking markers yielding large negative LOD scores in PKGL005. Similarly, in PKGL025, D2S112, D3S1279, D9S1776, and D21S263 have closely flanking markers yielding large negative LOD scores. Linkage to markers other than chromosome 14q markers that showed LOD scores greater than 1.5 during the genome-wide scan was further excluded by haplotype analysis of closely flanking markers.Two point linkage analysis provided the first evidence of linkage to markers at 14q24.2–24.3 with maximum LOD scores of 5.88 and 6.19 with markers D14S61 and D14S43 at θ=0 for families PKGL005 and PKGL025, respectively. Additional STR markers selected from the NCBI and Marshfield databases were genotyped to define the linkage interval in these families. Two point LOD scores of 4.96, 5.60, 4.01, 4.84, 4.76, 5.88, 3.50, and 3.69 with D14S77, D14S43, D14S284, D14S1036, D14S85, D14S61, D14S59, and D14S1008 at θ=0 were obtained for PKGL005 (Table 2). Similarly, two point LOD scores 4.66, 6.19, 4.44, 5.28, 3.19, and 5.38 with D14S77, D14S43, D14S284, D14S1036, D14S85, and D14S74 at θ=0 were obtained for PKGL025 (Table 3).Table 2Two point LOD scores of PKGL005 with chromosome 14q markers.MarkercMMbTwo-point LOD score values at recombination fraction (θ=)Zmaxθ max0.000.010.030.050.070.090.100.200.30D14S6369.1844.71-5.50-0.380.420.700.830.890.900.720.370.900.10D14S25876.2850.65-5.49-0.380.410.690.830.890.890.720.370.890.09D14S28978.2051.630.971.892.132.132.061.961.901.140.382.130.03D14S7780.8253.634.964.844.614.384.153.913.802.621.514.960.00D14S4384.1654.985.605.495.244.984.734.464.333.021.745.600.00D14S28484.6955.754.013.893.653.423.182.942.821.720.844.010.00D14S7684.6955.821.961.881.711.561.391.251.180.590.991.960.00D14S103684.6955.834.844.734.484.233.983.733.612.361.234.840.00D14S8584.76–4.764.654.404.153.903.653.522.291.174.760.00D14S6186.2956.375.885.755.505.244.984.714.583.231.905.880.00D14S5987.3658.113.503.403.202.992.792.592.491.530.753.500.00D14S7487.3658.70-0.062.953.183.163.062.932.701.961.063.180.03D14S100889.1959.943.693.613.433.243.062.882.791.901.103.690.00D14S60691.62–-0.082.192.462.472.412.312.251.540.812.470.05D14S97493.76–-2.140.250.600.700.720.710.700.460.210.720.07LOD scores were calculated at different θ values for each marker with the FASTLINK version of MLINK from the LINKAGE program package. Maximum LOD scores for each marker were calculated using ILINK.Table 3Two point LOD scores of PKGL025 with chromosome 14q markers.MarkercMMbTwo-point LOD score values at recombination fraction (θ=)Zmaxθ max0.000.010.030.050.070.090.100.200.30D14S6369.1844.71-2.502.002.502.752.752.752.752.001.252.750.03D14S25876.2850.652.504.004.254.254.004.003.752.751.504.250.03D14S28978.2051.631.562.692.942.942.812.752.691.881.002.940.03D14S7780.8253.634.664.594.444.284.093.913.812.781.694.660.00D14S4384.1654.986.196.035.725.445.134.844.693.191.726.190.00D14S28484.6955.754.444.314.063.813.633.383.252.061.064.440.00D14S7684.6955.822.322.272.172.071.971.861.811.290.792.320.00D14S103684.6955.835.285.164.914.634.384.133.992.691.445.280.00D14S8584.69–3.193.133.062.942.812.682.621.811.063.190.00D14S6186.2956.37-4.341.912.222.252.222.162.091.470.782.250.05D14S5987.3658.11-4.345.755.755.755.505.505.254.002.505.750.01D14S7487.3658.705.385.255.004.754.504.254.132.751.505.380.00D14S100889.1959.94-4.342.753.503.753.503.503.502.501.503.750.05D14S60691.62–-4.342.162.843.003.032.972.912.091.093.030.07D14S97493.76–-4.341.502.192.382.442.442.381.750.942.440.07LOD scores were calculated at different θ values for each marker with the FASTLINK version of MLINK from the LINKAGE program package. Maximum LOD scores for each marker were calculated using ILINK.Haplotype analysis supports the results of linkage analysis as shown in Figure 1. There is a proximal recombination in affected individual 19 of PKGL025 at D14S63 and in affected individuals 28 and 41 of PKGL005 at D14S289. Similarly, there is distal recombination in affected individual 28 of PKGL025 at D14S606 and in affected individual 41 of PKGL005 and unaffected individual 23 of PKGL025 at D14S74 as well as in unaffected individual 14 of PKGL025 at D14S85. Taken together, these results suggest the disease locus lies in a 6.56 cM (~4.2 Mb) region flanked by markers D14S289 and D14S85. As marker D14S1036 is uninformative for individual 10 of PKGL025, it is possible that the distal boundary lies proximal to marker D14S1036. Alleles for D14S77, D14S43, D14S284, D14S76, and D14S1036 were homozygous for all affected individuals in families PKGL005 and PKGL025 whereas the normal individuals are either heterozygous carriers of the disease allele or are homozygous for the normal allele.Figure 1Pedigree of families PKGL005 and PKGL025. Squares denote males, circles indicate females, filled symbols represent affected individuals, double lines between individuals indicate consanguinity, and a diagonal line through a symbol signify that the family member is deceased. The haplotypes of 15 adjacent chromosome 14q14.2–24.3 microsatellite markers for families PKGL005 (A) and family PKGL025 (B) are shown with alleles forming the risk haplotype shaded black, alleles cosegregating with primary congenital glaucoma (PCG) but not showing homozygosity shaded gray, and alleles not cosegregating with PCG shown in white.The critical interval on chromosome 14q24.2–24.3 harbors coenzyme Q6 homolog (COQ6), which encodes a flavin-dependent monooxygenase in Saccharomyces cerevisiae. This suggests a functional similarity with CYP1B1. We investigated the COQ6 gene to identify the mutation leading to the disease phenotype in these families by sequencing all coding exons, exon-intron boundaries, and the 5'-untranslated region, but we did not find any pathogenic mutations in this gene. Our sequencing results identified previously reported SNPs rs3213692 and rs2074930 in PKGL005 and rs17552038, rs3213692, and rs7141392 in PKGL025.DiscussionHere, we report autosomal recessive primary congenital glaucoma (PCG) in two large consanguineous Pakistani families, mapped to chromosome 14q24.2–24.3. Maximum LOD scores of 5.88 and 6.19 with markers D14S61 and D14S43 at θ=0 for families PKGL005 and PKGL025, respectively, the lack of LOD scores above 2.0 for any markers other than chromosome 14q in the entire genome scan, and the disease haplotype segregating with the disease phenotype in both families strongly suggest that the PCG locus maps to chromosome 14q24.2–24.3 in these families. Haplotype analysis of these two families refines the disease interval to a 6.56 cM (~4.2 Mb) region flanked by markers D14S289 and D14S85. Localization of the disease interval to 14q24.2–24.3 in two consanguineous Pakistani families strongly suggests genetic heterogeneity of primary congenital glaucoma.To date, three PCG loci have been mapped to chromosomes 2p21 (GLC3A), 1p36 (GLC3B), and 14q24.3 (GLC3C) whereas mutations associated with PCG have only been reported in the CYP1B1 gene [13-15,21]. Previously, GLC3C was localized to chromosome 14q24.3 flanked by markers D14S61 and D14S1000 as shown in Figure 2 [22]. In PKGL025, individual 14 delineates the distal boundary at marker D14S85, strongly suggesting that the disease locus in PKGL025 does not overlap with GLC3C. As both families in this study come from similar geographical and racial backgrounds, haplotype analysis of both families strongly suggests that the region flanked by markers D14S289 and D14S85 harbors the disease causing gene. However, we cannot rule out the possibility that the disease phenotype in these two families is caused by two different mutations, and the pathogenic mutation for PKGL005 may be present in a gene localized in a region overlapping with the GLC3C locus.Figure 2Schematic representation of linkage on chromosome 14q24.2–24.3 in families PKGL005 and PKGL025. Filled circles denote STR markers, and solid vertical lines represent the chromosomal intervals in which markers are homozygous for affected members of each of the two families.The critical interval on chromosome 14q24.2–24.3 harbors 97 genes including coenzyme Q6 homolog (COQ6), WD repeat domain 21A (WDR21A), and ceh-10 homeo domain containing homolog (CHX10). COQ6 is a lipid soluble antioxidant and an obligatory component of the respiratory chain and uncoupling proteins [23,24]. COQ6 in Saccharomyces cerevisiae encodes a flavin-dependent monooxygenase, suggesting a functional similarity with CYP1B1, a mixed-function monooxygenase that belongs to the cytochrome P450 1B subfamily [25]. We sequenced all the coding exons and exon-intron boundaries as well as the 5’ and 3’ regions of affected individuals of families PKGL005 and PKGL025; however we did not identify any pathogenic mutation.WDR21A belongs to the WD repeat protein family. Members of this family are involved in a variety of cellular processes including cell cycle progression, signal transduction, apoptosis, and gene regulation. Mutations in WDR36, also a member of the WD repeat protein family, have been associated with adult-onset primary open-angle glaucoma (POAG) [26]. In contrast, CHX10 is homeobox transcription factor gene that is expressed in progenitor cells of the developing neuroretina and in the inner nuclear layer of the mature retina. In humans, CHX10 mutations are associated with microphthalmia with cataracts and iris abnormalities, isolated microphthalmia with coloboma 3, isolated microphthalmia 2, and isolated microphthalmia with cloudy corneas [27-29]. Similarly, mutations in CHX10 cause microphthalmia, progressive degeneration of the retina, and an absence of the optic nerve in mice [30]. We are currently sequencing these two candidate genes to identify any pathogenic mutations.In summary, we have localized autosomal recessive PCG to chromosome 14q24.2–24.3 in two consanguineous Pakistani families. Identification of the PCG causing gene at this locus will help to unveil the underlying molecular complexity of primary congenital glaucoma and will be a valuable addition to the existing repertoire of glaucoma genetics, particularly of PCG. Finally, it will be helpful in screening for carrier status and genetic counseling of PCG families especially in the Pakistani population to prevent severe visual impairment and blindness.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2530859.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2530859",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2530859\nAUTHORS: Jianping Chen, Han Liang, Ariel Fernández\n\nABSTRACT:\nA proteomic association study between protein three-dimensional structure and transcriptional and post-transcriptional regulation in yeast and human.\n\nBODY:\nBackgroundThe coordination of protein roles to achieve specific biological functions requires the spatial/temporal concurrence of proteins so that they can form complexes [1,2] or, in general, operate within a module [2-4]. In turn, this concurrence is tightly coordinated through the regulation of gene expression, as suggested by established correlations between the transcriptome and the interactome [5,6]. However, structure-encoded factors that may quantitatively control such correlations have not been identified. So far, protein structure has not provided organizing clues for the integration of large-scale descriptions of the molecular phenotype.As reported in this work, by exploiting a structure-based analysis of protein associations [7,8] and their correlated expression patterns, we identify a structural attribute, protein vulnerability, and show that it commits gene expression patterns in a quantifiable manner. More specifically, protein vulnerability is shown to determine the extent of co-expression of genes containing protein-encoding interactive domains in metabolic adaptation phases [9,10] or tissue types [11,12], while extreme vulnerability promotes significant post-transcriptional regulatory control.Soluble proteins maintain the integrity of their functional structures provided the amide and carbonyl groups paired through hydrogen bonds are adequately shielded from water attack, preventing backbone hydration and, generally, the concurrent total or partial denaturation of the soluble structure [13,14]. As shown in this work, this integrity is often ensured through the formation of protein complexes, which become more or less obligatory depending on the extent of structure vulnerability and the level of backbone protection provided by the association [13]. By adopting vulnerability as a structural indicator of dosage imbalance effects, the extent of reliance on binding partnerships is precisely quantified and shown to be an organizing factor for the yeast and human transcriptome.ResultsProtection of a vulnerable protein and co-expression demandsWe start by defining vulnerability ν of a soluble protein structure as the ratio of solvent-exposed backbone hydrogen bonds (SEBHs) to the overall number of such bonds (Figure 1). The SEHBs may be computationally identified from atomic coordinates (Materials and methods). Thus, while backbone hydrogen bonds are determinants of the basic structural motifs [15,16], the SEHBs represent local weaknesses of the structure.Figure 1Hydrogen-bond pattern and structural vulnerabilities (SEBHs) of the yeast SH3 domain and the human prion protein PrPC. (a) Hydrogen-bond pattern and structural vulnerabilities (SEBHs) of the yeast SH3 domain from a S. cerevisiae 40.4 kDa protein (PDB.1SSH) [17]. The ribbon display is included as a visual aid. The protein backbone is shown as virtual bonds (blue) joining consecutive α-carbons in the peptide chain. Light-grey segments represent well protected backbone hydrogen bonds, and green segments represent SEBHs. The extent of solvent-exposure extent of a hydrogen bond was determined from atomic coordinates by calculating the number of nonpolar groups within its microenvironment (Materials and methods). SEBHs are those backbone hydrogen bonds protected by an insufficient number of nonpolar groups as statistically defined in Materials and methods. The level of structure vulnerability ν, defined as the ratio of SEBHs to the overall number of backbone hydrogen bonds, is 19.0% (ν = 4/21). (b) SEBH-pattern for the cellular structure of the human prion protein PrPC (PDB.1QM0) [18]. Its vulnerability parameter is ν = 63.0%, making it the most vulnerable soluble folder of all structures of unbound proteins reported in the PDB.Figure 1a shows the vulnerability pattern of a well protected soluble protein, the yeast SH3 signaling domain [17], with ν = 19.0%. Figure 1b shows the most vulnerable protein structure for an autonomous folder in the Protein Data Bank (PDB) (ν = 63.0%), the cellular form of the 90-230 fragment of the human prion protein PrPC (PDB.1QM0) [18]. This extreme case was detected after exhaustive computation of the ν parameter for all conformations of isolated (those not in a complex) polypeptide chains reported in the PDB (Materials and methods). Figure 2 shows the most vulnerable structure adopted by a protein chain within a yeast complex: subunit 1 from the cytochrome b-c1 complex (COR1/YBL045C). Unlikely to be found in isolation, this structure is found within the mitochondrial respiratory chain complex III [19].Figure 2Ribbon representation and vulnerability (SEBH) pattern of subunit 1 from the cytochrome b-c1 complex. (a) Ribbon representation and (b) vulnerability (SEBH) pattern of subunit 1 from the cytochrome b-c1 complex (PDB.1KB9) [19]. In b, red segments represent virtual protein backbone bonds, light-grey segments represent well protected backbone hydrogen bonds, and those green segments represent SEBHs. In the cytochrome complex, this protein adopts a highly vulnerable (ν = 57.3%) conformation.A vulnerable soluble structure gains extra protection of its backbone hydrogen bonds through forming complexes, as nonpolar groups of a binding partner contribute to expel water molecules from the microenvironment of the preformed bonds [13]. On the other hand, the SEBHs promote their own dehydration as a means to stabilize and strengthen the hydrogen bond [14].To delineate the role of structure vulnerability as an organizing integrative factor in large-scale descriptions of the molecular phenotype, we first examined the Pfam-filtered [7] protein complexes for yeast [8] and human [20]. These associations involve domains whose PDB-reported homologs are involved in complexes.This work quantitatively examines the relationship between the structural vulnerability of a protein and the extent of co-expression of genes encoding its binding partners. Thus, the extent of co-expression, η (i, j), for two genes i, j encoding interacting proteins is measured by the expression correlation of the two genes normalized to the average correlation over the interactome (Materials and methods). In consonance, the expression correlation of a complex, η (complex), may be defined by the maximum expression correlation over its constitutive underlying pairwise interactions (see Additional data files 7-9 for alternative definitions).Thus, the most highly correlated yeast complex (overall η (complex) = 3.61) with full PDB-reported representation is the mitochondrial respiratory chain complex III shown in Figure 3a (PDB.1KB9[19]). The most vulnerable protein within the complex (ν = 57%) is subunit 1 from the cytochrome b-c1 complex (Gene/ORF = COR1/YBL045C, shown in red). Its peptide chain conformation, with the SEBH pattern described in Figure 2, is involved in the most highly correlated interaction (η = 3.61) within the complex (Figure 3b,c). The binding partner in this interaction is subunit 2 of cytochrome b-c1 (Gene/ORF = QCR2/YPR191W, blue chain in Figure 3a). Figure 3c shows the mutual protection of preformed SEBHs in the two subunits along part of their association interface (red, COR1 residues 42-119; blue, QCR2 residues 250-331). This intermolecular mutual 'wrapping' of local weaknesses illustrates the fact that the association contributes to maintain structural integrity (Figure 3c).Figure 3Mutual protection of SEBHs in the two subunits of mitochondrial respiratory chain complex III. (a) Ribbon representation of mitochondrial respiratory chain complex III (PDB.1KB9). The high structure vulnerability of subunit 1 (red; compare Figure 2) renders it highly needy for interaction with other subunits of the complex to maintain its structural integrity. (b) SEBH pattern for subunit 1 (red) and subunit 2 (blue). The interacting pair is characterized by a very high expression correlation η = 3.61. The yellow square highlights the part of the interface shown in detail in (c). (c) Illustration of mutual protections of SEBHs in the two subunits along part of their interface. One side-chain bond (between α and β carbons) is displayed. The thin blue lines, which connect β-carbons in one protein with centers of hydrogen bonds in the other protein, represent mutual protections of hydrogen bonds across the protein-association interface. Thus, a thin line is shown whenever the side chain of one protein is contributing with nonpolar groups to the microenvironment of a preformed hydrogen bond in its binding partner.We examined the role of structure vulnerability as a factor governing the extent of co-expression of binding partners in illustrative yeast complexes (Figure 4a; Additional data file 1). Structure-based protein-protein interactions were curated through the Pfam database, so that two proteins were considered to interact with each other if their respective domains (or homolog domains) were reported in a PDB complex [8,21]. The expression correlation, η, for each interaction pair within a complex was determined at the mRNA level of the genes whose open reading frames (ORFs) contained the interacting domains (Materials and methods). Vulnerabilities were computed either directly from PDB files, when available, as described in Figure 1, or from atomic coordinates generated by homology threading using the Pfam-homolog domain as template (Materials and methods). In the latter case, side-chain equilibration, constrained by a fixed homology-threaded backbone, was obtained from constrained molecular dynamics simulations (Materials and methods). We then determined the maximum ν-value for each interactive pair and, using the comprehensive microarray database for Saccharomyces cerevisiae glucose→ glycerol metabolic adaptation [22], we computed the expression correlation η for each Pfam interaction. A tight (η-ν) correlation (R2 = 0.891) is obtained and shown to hold across the illustrative yeast complexes (Figure 4a) and, furthermore, to hold across all 1,354 pairs of interacting proteins in the yeast interactome with Pfam representation (Figure 4b,c; Additional data file 2). The (η-ν) correlation implies that the protection of a functionally competent protein structure in yeast drives co-expression of its binding partners to an extent that is determined by the structure vulnerability.Figure 4Correlation between maximum structure vulnerability ν and co-expression similarity η for yeast protein interactions. (a) Correlation between maximum structure vulnerability ν and co-expression similarity η for interactions within specific yeast complexes. The ν-parameter of an interaction is defined as the maximum vulnerability between the two interacting partners, and the η-parameter is the ratio of their expression correlation to the (non-zero) expected correlation over all interacting pairs in the proteome. (b) (η-ν) correlation for all Pfam-filtered yeast protein interactions. Red points represent interactions involving extremely vulnerable proteins, including confirmed yeast prions (Additional data file 5). (c) (η-ν) correlation of Pfam-filtered yeast protein interactions involving only PDB-reported proteins. The red data point represents an interaction involving an extremely vulnerable protein, and the green point represents an interaction involving an extremely vulnerable protein reported to be a prion protein (ERF2) [24-26].In selecting the yeast transcriptome [22], particular attention was focused on the 'perturbative' nature of the change triggering the structural remodeling of the proteomic network across different phases. A more extensive remodeling on a vastly larger scale, as in the complete yeast developmental cycle [23], cannot be treated as a perturbation since it clearly alters the modular structure of the proteome network [4] and, consequently, yields a weaker (η-ν) correlation (Additional data file 10).Structure vulnerability is not only an organizing factor for the metabolic-adaptation transcriptome but also steers the organization of tissue-based transcriptomes. This is revealed by a similar comparative analysis of the most comprehensive protein-encoding gene-expression data for human [11] and the structure-represented interactome [20]. Thus, a clear (η-ν) correlation is apparent between the co-expression of 607 gene pairs and the maximum structure vulnerability for each pair of interacting domains encoded in the ORFs of the respective genes (Figure 5; Additional data file 3).Figure 5(η - ν) correlation for human protein interactions. (a) The (η-ν) correlation for all Pfam-filtered human protein interactions. Red points represent interactions involving extremely vulnerable proteins (Additional data file 4). (b) The correlation over Pfam-filtered human protein interactions that involve only PDB-reported proteins. The red point represents an interaction containing an extremely vulnerable protein.Other human transcriptomes based on normal tissue expression were examined (see, for example, [24]), but none provided statistically significant (>>10 genes pairs) representation for the gene pairs for which interactome data also exist [20], as needed for the present study.Post-transcriptional regulation of the expression of highly vulnerable proteinsIn contrast with the tighter yeast correlation, a few but significant outlier pairs (Figure 5, red data points) are found beyond the confidence band defined by a width of two Gaussian dispersions from the linear (η-ν) fit. To rationalize this fact, we identified 115 human genes with ORFs encoding extremely vulnerable proteins (Additional data file 4). Consistent with the definition of structure vulnerability (Figure 1), the latter proteins are identified by large sequences (≥ 30 residues) of amino acids that are poor protectors of backbone hydrogen bonds. In principle, a sizable window of residues unable to protect backbone hydrogen bonds produces a poor folder, yielding a highly vulnerable structure [14,25]. Thus, these sequences are either probably unable to sustain a stable soluble structure, or prone to relinquish the folding information encoded in the amino acid sequence in favor of self-aggregation [25]. The poor protectors (G, A, S, Y, N, Q, P) are amino acids possessing side chains with insufficient nonpolar groups, with polar groups too close to the backbone (thus precluding hydrogen-bond protection through clustering of nonpolar groups) [14] or with amphiphilic aggregation-nucleating character (Y) [26-28]. Charged backbone de-protecting side chains (D, E) are excluded since they would entail negative design relative to protein self-aggregation. All outlier interactions in the human (η-ν) correlation involve genes with extreme vulnerability (Figure 5; Additional data file 4). Significantly, when the same criterion for extreme vulnerability is used to scan the yeast genome (Additional data file 5), 85 genes are identified whose ORFs encode the five confirmed prion proteins for this organism [26-29]: PSI+ (SUP35), NU+ (NEW1), PIN+ (RNQ1), URE3 (URE2) and SWI+ (SWI1). This fact is statistically significant (P < 10-10, hypergeometric test) and supports the presumed relationship between structural vulnerability of the soluble fold and aggregation propensity [25].The (η-ν) correlation reported in Figure 5 for human is weaker than the yeast counterpart likely because, in contrast with yeast, mRNA levels are not a reliable surrogate for protein expression levels in human [30,31]. This observation led us to examine post-transcriptional regulation in human genes, to analyze the microRNA (miRNA) targeting of the predicted 115 extremely vulnerable human genes (Additional data files 4 and 6), and to contrast the miRNA-targeting statistics with the generic values across the human genome [31]. To obtain statistics on miRNA targeting, we identified putative target sites in the 3' UTR (untranslated region) of each gene for 162 conserved miRNA families (Materials and methods) [31]. Thus, 7,927 out of 17,444 genes (45.4%) are predicted to contain at least one miRNA target site (Additional data file 6), while 87 out of 105 (82.9%) extremely vulnerable genes are predicted to be targeted genes. Thus, human genes containing extremely vulnerable regions are more frequently targeted by miRNA (P << 1.31 × 10-5, binomial test). In regards to miRNA regulation complexity, the mean number of miRNA target sites for human genes is 2.66 and the median is 0, while the mean number for extremely vulnerable genes is 6.01 and the median is 5. This significant difference (P < 10-16, Wilcox rank test) strongly suggests that the deviation of extremely vulnerable genes from the (η-ν) correlation (Figure 5), with expression correlation evaluated at the level of mRNA expression, can be explained by post-transcriptional miRNA regulation. This type of regulation influences the final protein expression level. In a broad sense, this analysis highlights the connection between protein structure and gene regulation: extremely vulnerable genes require tight control at the post-transcriptional level.Protein intrinsic disorder and transcriptome organizationThe inability of an isolated protein fold to protect specific intramolecular hydrogen bonds from water attack may lead to structure-competing backbone hydration with concurrent local or global dismantling of the structure [14,25,32]. This view of structural vulnerability suggests a strong correlation between the degree of solvent exposure of intramolecular hydrogen bonds and the local propensity for structural disorder [33-35]: in the absence of binding partners, the inability of a protein domain to exclude water intramolecularly from pre-formed hydrogen bonds may be causative of a loss of structural integrity, and this tendency is marked by the disorder propensity of the domain [32]. These findings led us to regard the predicted extent of disorder in a protein domain as a likely surrogate for its vulnerability and to contrast it with the extent of expression correlation with its interactive partners. The disorder propensity may be determined by a sequence-based score, fd(fd = 1, certainty of disorder; fd = 0, certainty of order), assigned to each residue. In this work, this parameter is generated by the highly accurate predictor of native disorder PONDR-VSL2 [34,35]. The extent of intrinsic disorder of a domain may be defined as the percentage of residues predicted to be disordered relative to a predetermined fd threshold (fd = 0.5).Reexamination of the expression correlations in the yeast and human transcriptomes was carried out, taking into account a proteome-wide sequence-based attribution of the extent of disorder (percentage of residues predicted to be disordered, or 'disorder content') in interacting protein domains. The disorder predictions did not include any structural information on induced fits arising upon forming a complex, and hence, unlike structure vulnerability, the percent predicted disorder is independent of the complex under consideration. This fact introduces deviations in the estimation of vulnerability through disorder content for proteins with extensive disorder content since their conformational plasticity may enable diverse induced-fit conformations with different vulnerabilities (Figure 6a). In yeast, the extent of disorder of the most disordered domain for each pair of interacting domains captures the degree of correlation in the expression patterns required for structure protection (Figure 6a). This is revealed by the correlation between the extent of disorder of the most disordered domain in an interacting pair and the expression correlation η of the two genes encoding the respective interacting domains. While weaker than the η-ν correlation (Figure 4), the η-disorder correlation is still relatively strong for yeast proteins (R2 = 0.752; Figure 6a), implying that disorder content determines the degree of coexpression of binding partners to a significant extent. The large dispersion in disorder extent at high levels of coexpression (approximately 45% dispersion versus approximately 15% for proteins with low disorder/low expression correlation) is indicative that highly disordered chains may adopt structures with very different levels of vulnerability depending on the complex in which they are involved (the η-ν correlation does not widen so significantly for smaller η-values). Thus, the more disordered the chain, the more multi-valued the correspondence between disorder extent and vulnerability, conferring higher dispersion to the η-disorder correlation.Figure 6(η-disorder) correlation for yeast and human protein interactions. Correlation between η-parameter and percent predicted disorder (disorder content) of the most disordered domain for each of (a) the 1,354 Pfam-filtered protein-interaction pairs in yeast and (b) the 607 pairs in human.The η-disorder correlation in human is considerably weaker (R2 = 0.304; Figure 6b) than in yeast. This is partly due to the fact that human proteins have a higher degree of disorder propensity than their yeast orthologs [36] and, hence, they are capable of significantly diversifying their structural adaptation (induced folding) in different complexes. In this context, the extent of disorder becomes a poor surrogate of structural vulnerability, as different ν-values may correspond to a single percent predicted disorder. In addition, post-transcriptional regulation in humans implies that expression correlations at the mRNA level are not reflective of the protein concurrencies modulated by tissue type, as indicated above.To conclude, Figure 6 reveals the role of intrinsic protein disorder in transcriptome organization suggested by exploring the interrelationship between protein vulnerability and disorder propensity.DiscussionSoluble protein structures may be more or less vulnerable to water attack depending on their packing quality. As shown in this work, one way of quantifying the structure vulnerability is by determining the extent of solvent exposure of backbone hydrogen bonds. Within this scheme, local weaknesses in the protein structure may become protected upon forming a complex, as exposed backbone hydrogen bonds become exogenously dehydrated. Vulnerable structures are thus quantitatively reliant on binding partnerships to maintain their integrity, suggesting that vulnerability may be regarded as a structure-based indicator of gene dosage sensitivity [37,38]. This observation is validated by establishing the significance of protein vulnerability or structure protection as an organizing factor in temporal phases (yeast) and tissue-based (human) transcriptomes. Specifically, this role was established by examining the degree of co-expressions of a protein with its binding partners in structure-represented interactions. Thus, for each Pfam-filtered binding partnership, the extent of co-expression across metabolic adaptation phases (yeast) or tissue types (human) was found to depend quantitatively on the structure vulnerability of the proteins involved. Hence, vulnerability may be regarded as an organizing factor encoded in the structure of gene products.Furthermore, as shown in this work, the tight coordination between translation regulation and gene function dictates that extremely vulnerable, and hence 'highly needy', proteins are subject to significant levels of post-transcriptional regulation. In human, this extra regulation is achieved through extensive miRNA targeting of genes coding for extremely vulnerable proteins. In yeast, on the other hand, our results imply that such a regulation is likely achieved through sequestration of the extremely vulnerable proteins into aggregated states. Intriguingly, the 85 yeast genes encoding extremely vulnerable proteins included those for the five confirmed yeast prions [26-29]. This statistically significant result implies that if the extremely vulnerable proteins are themselves translational regulators, this sequestration may directly lead to epigenetic consequences and phenotypic polymorphism [26-28].ConclusionIn this work we adopted a structural biology perspective to reassess the fundamental notion of 'dosage imbalance effect' and examine the implications for gene expression, specifically for transcriptomal organization and post-transcriptional regulation. Thus, vulnerability of protein structures and the concurrent need to maintain structural integrity for functional reasons prove to be quantifiers of dosage imbalance: proteins with a high degree of reliance on binding partnerships to maintain their structural integrity are naturally expected to yield high dosage sensitivity in their respective gene expressions. Hence, structural vulnerability is shown to be a determinant of transcriptome organization across tissues and temporal phases: the need for protein structure protection compels gene co-expression in a quantifiable manner. Extreme vulnerability is shown to require significant additional regulation at the post-transcriptional level, manifested by epigenetic aggregation in yeast and miRNA targeting in human. These latter observations will likely inspire further study of structure-encoded signals that govern post-transcriptional regulation.Materials and methodsExpression data sourcesYeast expression data were obtained from the comprehensive Saccharomyces Genome Database [22]. This complete dataset contains mRNA expression levels during a transition from glucose-fermentative to glycerol-based respiratory growth. Human expression data were taken from the comprehensive Novartis Gene Expression Atlas [11]. This dataset includes 158 array images composed of 79 samples, each of which has two replicates hybridized on the human genome HG-U133A array. We discarded six samples of cancer tissues: ColorectalAdenocarcinoma, leukemialymphoblastic(molt4), lymphomaburkittsRaji, leukemiapromyelocytic, lymphomaburkittsDaudi, and leukemiachronicmyelogenous (k562).Interaction data sourcesProtein interaction curation based on structure provides direct physical interactions [8]. Two proteins were considered to interact with each other when their respective domains or homologs of their respective domains were found in a complex with PDB-reported structure. We obtained curated yeast protein domain interactions from the Structural Interaction Network [8], and filtered them using recently published yeast interaction data [21]. For human, we focused on interactions within complexes. The complex data were obtained from the MIPS/Mammalian Protein Complex Database [20]. We used the protein domain descriptions in the Pfam database [7], and searched for domain-domain interactions using iPfam [39].Expression correlation ηThe expression correlation for a protein-protein interaction is a normalized quantity defined as the Pearson correlation of the expression vectors of the genes encoding for the interacting domains divided by the mean correlation over all gene pairs encoding for interacting domains. The normalization is necessary for comparative analysis across species because different species have different mean expression correlations and, hence, the significance of a correlation is necessarily a relative attribute. Given its statistical nature, the denominator is non-zero for any species since, in a statistical sense, protein pairs that interact are expected to be positively correlated in their expression. We use the Pearson correlation coefficients of expression vectors to determine similarity between expression profiles. For two expression vectors X and Y, the Pearson correlation coefficient Corr(X, Y) is given by:\n                     \n                        \n                           Corr\n                           (\n                           X\n                           ,\n                           Y\n                           )\n                           =\n                           \n                              \n                                 <\n                                 (\n                                 X\n                                 −\n                                 <\n                                 X\n                                 >\n                                 )\n                                 (\n                                 Y\n                                 −\n                                 <\n                                 Y\n                                 >\n                                 )\n                                 >\n                              \n                              \n                                 \n                                    \n                                       <\n                                       \n                                          X\n                                          2\n                                       \n                                       >\n                                       −\n                                       <\n                                       X\n                                       \n                                          >\n                                          2\n                                       \n                                    \n                                 \n                                 \n                                    \n                                       <\n                                       \n                                          Y\n                                          2\n                                       \n                                       >\n                                       −\n                                       <\n                                       Y\n                                       \n                                          >\n                                          2\n                                       \n                                    \n                                 \n                              \n                           \n                        \n                        \n                     \n                  where X, Y are generic coordinates in the vectors X and Y, respectively, and < > indicates mean over the 73 normal tissues (human) [11] or over the 5 metabolic adaptation phases (yeast) [22].Calculation of vulnerability ν and identification of SEBHs for soluble proteinsTo determine the extent of solvent exposure of a backbone hydrogen bond in a soluble protein structure, we determine the extent of bond protection from atomic coordinates. This parameter, denoted ρ, is given by the number of side-chain nonpolar groups contained within a desolvation domain (hydrogen-bond microenvironment) defined as two intersecting balls of fixed radius (the approximate thickness of three water layers) centered at the α-carbons of the residues paired by the hydrogen bond. In structures of PDB-reported soluble proteins, at least two-thirds of the backbone hydrogen bonds are protected on average by ρ = 26.6 ± 7.5 side-chain nonpolar groups for a desolvation ball radius of 6 Å. Thus, SEBHs lie in the tails of the distribution, that is, their microenvironment contains 19 or fewer nonpolar groups, so their ρ-value is below the mean (ρ = 26.6) minus one standard deviation (= 7.5).In cases where the protein structures were unavailable from the PDB, we generated atomic coordinates through homology threading adopting the Pfam homolog as template and using the program Modeller [40-42]. Modeller is a computer program that models three-dimensional structures of proteins subject to spatial constraints [40], and was adopted for homology and comparative protein structure modeling. We thus generate the alignment of the target sequence to be modeled with the Pfam-homolog structure reported in the PDB and the program computes a model with all non-hydrogen atoms. The input for the computation consists of the set of constraints applied to the spatial structure of the amino acid sequence to be modeled and the output is the three-dimensional structure that best satisfies these constraints. The three-dimensional model is obtained by optimization of a molecular probability density function with a variable target function procedure in Cartesian space that employs methods of conjugate gradients and molecular dynamics with simulated annealing.Homolog PDB sourcesYeast PDB homologs were obtained from the Saccharomyces Genome Database [43], and human PDB homologs were from Pfam [44].Micro-RNA targeting analysisFor 17,444 human genes, we identified putative target sites for 162 conserved miRNA families using TargetScanS (version 4.0), a leading target-prediction program [45]. Thus, we obtained the number of target-site types in the 3' UTR of each gene [31]. Among the genes in our analysis: 105 genes were identified as encoding extremely vulnerable proteins; 7,927 out of 17,444 genes (45.4%) are predicted to be miRNA targets (containing at least one type of miRNA target site); and 87 out of 105 genes encoding extremely vulnerable proteins (82.9%) are predicted to be target genes. Thus, genes encoding extremely vulnerable proteins tend to be miRNA target genes (P << 1.31 × 10-5, binomial test).In terms of miRNA regulation complexity, the average number of miRNA target-site types for a human gene is 2.66 and the median number is 0; while the average number for a prion gene is 6.01 and the median is 5. Again, this is highly significant (P < 10-16, Wilcox rank test).Prediction of native disorder of protein domainsThe highly accurate predictor of native disorder PONDR VSL2 [34,35] exploits the length-dependent (heterogenous) amino acid compositions and sequence properties of intrinsically disordered regions to improve prediction performance. Unlike previous PONDR predictors for long disordered regions (>30 residues), it is applicable to disordered regions of any length. The disorder score (0 ≤ fd ≤ 1) is assigned to each residue within a sliding window, representing the predicted propensity of the residue to be in a disordered region (fd = 1, certainty of disorder; fd = 0, certainty of order). The disorder propensity is quantified by a sequence-based score that takes into account residue attributes such as hydrophilicity, aromaticity, and their distribution within the window interrogated.AbbreviationsmiRNA, micro RNA; ORF, open reading frame; PDB, Protein Data Bank; SEBH, solvent-exposed backbone hydrogen bonds; UTR, untranslated region.Authors' contributionsJC provided theoretical insight, designed methodology, generated and collected data, and co-wrote the paper. HL provided theoretical insight, and generated and collected data. AF provided the fundamental concepts and insights, designed methodology and wrote the paper.Additional data filesThe following additional data are available with the online version of this paper. Additional data file 1 provides raw data for Figure 4a. Additional data file 2 provides Raw data for Figure 4b,c. Additional data file 3 provides raw data for Figure 5. Additional data file 4 lists extremely vulnerable proteins in human. Additional data file 5 lists extremely vulnerable yeast proteins. Additional data file 6 lists the predicted number of miRNA targets for human genes. Additional data file 7 outlines the robustness of results with respect to alternative graph-theoretic definitions of co-expression similarity. Additional data file 8 outlines how vulnerability correlates with co-expression similarity in protein complexes. Additional data file 9 provides Raw data: yeast (a) and human (b) complexes examined in Additional data file 8. Additional data file 10 shows the (η-ν) plot obtained for the yeast developmental-phase transcriptome obtained from a comprehensive identification of cell cycle-regulated genes by microarray hybridization [23].Supplementary MaterialAdditional data file 1Data in column A indicate the expression correlation η associated with protein interactions, and data in column B indicate the structure vulnerability ν for interactions within specific complexes. The rest of the columns contain the ORF, domain and structure information (PDB accession code of interacting domain or its Pfam-homologs), respectively, for every pair of interacting proteins.Click here for fileAdditional data file 2Sheet 1 contains all Pfam-filtered yeast protein interactions, while sheet 2 contains only those interactions with both partners having PDB structures. In each sheet, column A lists the expression correlation h of interactions, and columns B and C list the structure vulnerability n of interactions not involving or involving, respectively, extremely vulnerable proteins. The remaining columns contain ORF, domain and structure information (PDB accession code of interacting domain or of its Pfam-homologs) for every pair of interacting proteins.Click here for fileAdditional data file 3Sheet 1 contains all Pfam-filtered human protein interactions, while sheet 2 contains only those interactions with both partners having PDB structures. In each sheet, column A contains the expression correlation h for each interaction, and columns B and C list the structure vulnerability n of interactions not involving or involving, respectively, extremely vulnerable proteins, and the rest of the columns list gene name, protein ID, domain and structure information (PDB accession code of interacting domain or of its Pfam-homologs) of every pair of interacting proteins.Click here for fileAdditional data file 4The extremely vulnerable proteins in human are identified from genome-wide scanning of protein-encoding regions with sequence windows (length ≥ 30) containing mainly amino acids (G, A, S, Y, N, Q, P) that are poor protectors of the protein backbone. An extremely vulnerable protein contains at least one such window with a threshold of three amino acids allowed to be outside the group of poor protectors.Click here for fileAdditional data file 5Extremely vulnerable yeast proteins are determined in the same way as for human (Additional data file 4). The rows marked in green correspond to the five confirmed yeast prions [26-29]: SUP35 (ERF2), URE2, NEW1, RNQ1 and SWI1.Click here for fileAdditional data file 6The number of putative target-site types corresponding to 162 conserved miRNA families determined for 17,444 human genes by interrogation of the 3' UTR using TargetScanS (version 4.0) [45].Click here for fileAdditional data file 7The co-expression similarity for genes i, j, encoding a pair of interacting proteins is alternatively measured as the adjacency aij(β) = (0.5 + 0.5 η (i, j))β, where η (i, j) is the expression correlation for the gene pair i, j and β is a soft threshold [46]. Similarly, the structure vulnerability is alternatively defined as νi, j(β) = ν (i, j)β, where ν (i, j) is the maximum ν-value for the interacting pair. (a, b) (ν (β)-a(β)) correlations for yeast for exponents β = 0.5 (a) and 10 (b). The adjacencies for β = 1 correspond simply to a linear rescaling of η already correlated with ν in Figure 4. (c, d) The same as (a, b) but for human. Notice that high exponents (β > 1) tend to amplify differences in co-expression, yielding lower correlation coefficients (R2 in (ν (β)-a(β)) plots).Click here for fileAdditional data file 8A normalized co-expression similarity γ (β, complex) for all genes encoding proteins that form a complex is obtained from the adjacencies of the pairwise interactions within the complex as: γ (β, complex) = [mediani, j ∈ complexaij(β)]/mediani, j aij(β)], where the median in the denominator extends over all interactive pairs in the interactome. Similarly, the normalized structure vulnerability Λ (β, complex) for complexes is defined as Λ (β, complex) = [mediani, j ∈ complexνij(β)]/mediani, j νij(β)]. (a-c) (Λ (β, complex)-γ (β, complex)) correlation over all 98 yeast complexes with transcriptome representation for exponents β = 0.5 (a), 1 (b) and 10 (c). (d-f) The same as (a-c) but for 53 human complexes.Click here for fileAdditional data file 9(a) Yeast complexes. (b) Human complexes.Click here for fileAdditional data file 10(η-ν) plot obtained for the yeast developmental-phase transcriptome obtained from a comprehensive identification of cell cycle-regulated genes by microarray hybridization [23]Click here for file\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2531081.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2531081",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2531081\nAUTHORS: Enrico M Zardi, Francesco Di Matteo, Daniele Santini, Valentina Uwechie, Pierfilippo Crucitti, Massimiliano Carassiti, Antonio Picardi, Eleonora Perrella, Marco Caricato, Giuseppe Tonini, Roberto Coppola, Antonella Afeltra\n\nABSTRACT:\nDeaths after percutaneous ethanol injection (PEI) into hepatocellular carcinoma (HCC) may occur within a few hours to a few days following the procedure because of hemoperitoneum and haemorrhage from oesophageal varices or hepatic insufficiency. Pancreatitis has been recently reported as a rare lethal complication of intra-arterial PEI, another modality for treating HCCs. In this minireview, we analyze the literature concerning the development of acute pancreatitis after PEI. Pathogenesis of pancreatitis from opioids and ethanol is also addressed. Treatment with opioids to reduce the patient's abdominal pain after PEI in combination with the PEI itself may lead to direct toxic effects, thus favouring the development of pancreatitis.\n\nBODY:\nReviewPercutaneous ethanol injection (PEI) is a widely used procedure for the treatment of hepatocellular carcinoma (HCC), and may be performed via conventional, \"one shot\" or intra-arterial modalities.While conventional PEI is performed under localized anaesthesia and the amount of ethanol injected into the HCC generally does not exceed 10 ml/session, \"one shot\" PEI is performed under general anaesthesia and the amount of administered ethanol is higher, ranging from 20 to 60 ml/session. Intra-arterial PEI is also performed under general anaesthesia, but ethanol (up to 50 ml) is directly injected, through a percutaneous route, into the artery that supplies the HCC after visualizing and puncturing this artery by using colour Doppler and B-mode ultrasound guidance. Interestingly, as demonstrated in a cell culture experimental study on malignant and liver cell lines, the cytotoxic effect of ethanol is dependent upon both its concentration and the exposure time [1].Since at present, the concentration of ethanol is standardized to 95% and the exposure time of the HCC is considered to be practically identical in the two PEI procedures, our opinion is that the development of complications may only depend on the high total dosage of ethanol injected and the patient's clinical conditions. However, according to some authors, no difference in complications (pain and fever excluded) has been reported when using larger doses of ethanol [2].The most frequently reported complication of these three PEI modalities is abdominal pain that may be observed in up to 48% of cases [3].If pain is not tolerated, especially when the doses of injected ethanol are high, the administration of nonopioid or mild opioid analgesics may be required [3]. Since cases of acute pancreatitis after opioid administration have been reported [4-17], we believe that more attention must be given when such drugs are administered. In fact, it is ascertained that there is a close temporal relationship (ranging from 1 to 3 hours) between opioid administration and the development of pancreatitis [7,18].A number of physiopathological studies have elucidated the mechanism through which opioids may induce pancreatitis. These studies most often implicate direct constriction of the sphincter of Oddi [18]; in fact, it has been demonstrated that intravenous morphine increases the intrabiliary pressure by enhancing sphincter of Oddi pressure [14]. It has also been shown that, after biliary sphincterotomy, pancreatitis may occur due to the sphincter spasm [7]. Taking into consideration that sphincter of Oddi dysfunction, a clinical syndrome due to a dyskinesia resulting from a functional alteration of sphincter motility or to stenosis, may occur at any age [19], our opinion is that it should be excluded before giving opioids after PEI in patients with HCC. This caution is very important considering that in patients with idiopathic recurrent pancreatitis, manometric evidence of sphincter of Oddi dysfunction was found to vary between 39 and 90% [20]. Furthermore, most cirrhotic patients with HCC suffer from cholelithiasis, and acute pancreatitis has been reported to occur in association with secondary sphincter of Oddi dysfunction, which is related to biliary calculi in 90% of cases [21]. According to some authors, since cholecistectomy would seem to favour the development of acute pancreatitis after ingestion of therapeutic doses of opioids [7], we believe that pain management with opioids after PEI treatment in cholecistectomized cirrhotics with HCC should be performed with great caution.Furthermore, interesting animal studies have demonstrated that ethanol may have direct effects on the pancreas, such as microcirculatory changes and direct toxic damage to the pancreatic acini [22-24]. Moreover, the mechanism of ethanol-induced pancreatitis has been well-studied in an interesting animal model in which it was demonstrated that sphincter of Oddi dysfunction was implicated in several forms of acute and chronic pancreatitis [25]. In fact, according to the authors, since trans-sphincteric flow, regulated by the sphincter of Oddi which acts as a pump, is a direct measure of sphincter of Oddi function, an alteration of this trans-sphincteric flow after intragastric or i.v. ethanol may indicate Oddi dysfunction; the authors also investigated whether neural mechanisms and gastric mucosal damage might play a role in this process [25], demonstrating that both intragastric and i.v. ethanol administration altered the Oddi trans-sphincteric flow. They also suggested, in accordance with other studies [26-28], that the fall in Oddi trans-sphincteric flow might be due to the direct effects of ethanol, its metabolites (acetaldehyde) and/or other humoural agents (superoxide, endothelin-1) on sphincter of Oddi motility. Furthermore, an effect of ethanol and/or its metabolites on sphincter of Oddi nitrergic innervations was observed [25]. The authors thus concluded that reduced sphincter of Oddi function might contribute to elevated pancreatic duct pressure, which is one of the events required for the onset of acute pancreatitis [25].There are no reports in the literature of acute pancreatitis after treatment of HCC with conventional PEI; in contrast, a case of lethal acute pancreatitis is described as a complication of intra-arterial PEI [29]. This technique can only be performed after the superselective puncture of HCC-supplying arteries, and the extreme technical difficulty of this method provides the major reason for the frequent failures of intra-arterial PEI [30].In an interesting study on large infiltrative HCC treated with intra-arterial PEI, the volume of ethanol intra-arterially injected ranged from 12 to 50 mL [mean, 25 mL ± 13 (63% of total volume injected into tumour)] in a single session and from 0 mL to 50 mL [mean: 15 mL ± 19 (37% of total volume injected)] in the subsequent sessions [29].A higher survival rate compared with that obtained after one-shot PEI [30] was observed with this intra-arterial PEI procedure [29]. However, the authors found that the main specific complication of this procedure, which caused the death of one of their patients, was ethanol reflux into the pancreaticoduodenal artery, a condition that can occur when the arterial branch of the HCC, in which ethanol is injected, originates from a short left hepatic artery close to the origin of the pancreaticoduodenal trunk [31].It is obvious that in this case, the reflux of ethanol in the pancreaticoduodenal trunk was the initial cause of pancreatitis through direct induction of a toxic necrosis of the pancreas. However, we cannot rule out the possibility that also opioids may have contributed to the development of pancreatitis and that the alteration of the Oddi trans-sphincteric flow induced by ethanol may have played a role, although the authors did not mention this possibility [3].Quite recently, we performed a \"one shot\" PEI (a total dose of 50 ml) into two HCC nodules of 4,6 and 3,1 cm respectively, in a patient with Child A cirrhosis. Pain management after the procedure was applied with morphine (10 mg i.v. and 10 mg s.c.), and with paravertebral block (right side) of D3-D5 by means of naropine 0,75 60 mg (total dose). On the next day, the patient developed oedematous head pancreatitis. In order to reduce his abdominal pain, treatment with opioids (morphine 8 mg/i.v. and tramadol 50 mg/i.v.) was maintained until two days after PEI; then, only tramadol 50 mg/i.v b.i.d. was continued until nine days after PEI. Despite an appropriate medical treatment of oedematous head pancreatitis and paralytic ileus (with octreotide, subcutaneous longastatine, hydration infusion and antibiotics), the patient's clinical condition further worsened and free subdiaphragmatic airways, mild abdominal fluid collection and necrosis of the head of the pancreas were observed on a contrast CT. Surgical intervention was mandatory and histological examination of the resected organs showed necrosis of the gallbladder, chronic steatophagic inflammation of the omentum, steatonecrosis of the gastric antrum with microerosive gastritis, haemorrhagic necrosis of the appendix and steatonecrosis of both the pancreatic head and the duodenum. After a few weeks, the patient fell into a hepatic coma and died of multiorgan failure and end-stage hepatic insufficiency.Based on the data available in the literature, our opinion is that acute pancreatitis may develop in cirrhotics with HCC treated with opioids to alleviate their pain after PEI. The mechanism through which ethanol may induce pancreatitis is partially known. After PEI, ethanol cannot easily diffuse into the surrounding non-tumoural tissue, since that tissue is firmer than the tumour structure. Therefore, in this case, the development of pancreatitis may have been favoured by the ensuing treatment with opioids although it cannot be ruled out that ethanol may have played a role; in fact, possible mechanisms of ethanol-induced pancreatitis may be pancreatic duct constriction, Oddi trans-sphincteric flow alteration, metabolic effects, direct cellular toxicity, all of which have been previously discussed [22-25].An experimental animal study on rats with BW7756 hepatoma, performed to compare efficacy and safety of two percutaneous ablation methods [PEI and PAI (percutaneous acetic acid)], showed that PEI had a lower mortality rate for complications than PAI, and that none of the complications from either procedure was due to pancreatitis [32]. In fact, autopsies revealed that the deaths of the rats were due to massive liver necrosis (about 40%) with diaphragma involvement, or to complete inferior vena cava thrombosis with extension to the right atrium.In this experiment, PEI was performed under general anaesthesia and opioid analgesics were not administered: this might be the reason why no evidence of pancreatitis was observed [32].It is true that pancreatitis after treatment with PEI of cirrhotics with HCC is a very rare complication, but these data, taken together, show that both opioids and ethanol may induce acute pancreatitis.It is well established that opioids can favour the development of pancreatitis through a constriction of the Oddi's sphincter. The fact that i.v. ethanol may alter the function of the Oddi's sphincter [25] suggests that both in intra-arterial PEI and in \"one shot\" PEI, the pathogenesis of pancreatitis may have also been due to mechanisms of motility dysfunction of the Oddi's sphincter.Therefore, the combined administration of ethanol and opioids may greatly favour the development of pancreatitis in both procedures.According to Beger et al., mortality after acute pancreatitis is 7.6% when less than 30% of the pancreas is necrotic and 24% when up to 50% of the pancreas is necrotic. However, mortality is 34.3% when there are additional extrapancreatic fluid effusions [33]. According to Rau et al. and Hartwig et al., the mortality rate after acute pancreatitis varies from 20 to 30% [34,35].In animal models of severe necrotizing pancreatitis, mortality is promoted by sepsis and by the development of a systemic inflammatory response syndrome, which, in turn, causes lethal multiorgan failure [36,37].Therefore, given the elevated mortality rate of pancreatitis, more attention is necessary when pain is treated with opioids in cirrhotics with HCC after PEI.Competing interestsThe authors declare that they have no competing interests.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2531091.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2531091",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2531091\nAUTHORS: Pierfrancesco Grima, Roberto Chiavaroli, Paola Calabrese, Paolo Tundo, Piero Grima\n\nABSTRACT:\nHHV-6 has been identified as the aetiologic agent of exanthem subitum in infants and an acute febrile illness in young children. HHV-6 probably remains latent in the body after the primary infection and it reactivates upon host immunosuppression in a manner similar to other human herpes viruses. Primary HHV-6 infection in adults is very rare and it is not clear whether disease manifestations are similar to those observed in children.We report the case of acute hepatitis in a 18-year-old immunocompetent woman presenting with sever jaundice and liver dysfunction. Serum immunoglobulin levels were elevated (3.8 gr/dl) with a titre of anti nucleus antibody of 1:640. Serological data demonstrated the presence of IgM antibodies against human herpesvirus-6 in the serum and of viral DNA on liver biopsy by real time quantitative polymerase chain reaction, with a viral load of 280 genomes/106 of cellular genomes. No other etiologic agents were found to induce hepatitis and the patient was diagnosed as having HHV-6 triggered autoimmune acute hepatitis.\n\nBODY:\nIntroductionHuman herpesvirus 6 (HHV-6) was first isolated from patients with the acquired immunodeficiency syndrome or lymphoproliferative diseases and was named human B lymphotropic virus [1]. HHV-6 has been identified as the etiologic agent of exanthema subitum in infants [2] and an acute febrile illness in young children [3]. Most people are seropositive for HHV-6 by the age of three years [4]. HHV-6 also produces latent or chronic infections [5] and is occasionally reactivated in immunocompromised hosts [1,6]. Furthermore, HHV-6 has been implicated in several diseases in immunocompetent adults, including Kikuchi's lymphadenitis [7] and an infectious mononucleosis-like syndrome that is negative for Epstein-Barr virus and cytomegalovirus [8]. We describe the immunopathological and clinical features of a severe acute hepatitis in a 18-year-old woman that was probably caused by a primary infection with HHV-6.Case presentationA 18-year-old woman was admitted to S.Caterina Novella Hospital on October 10, 2006, with a fifteenday history of flu-like syndrome. She had been healthy and had a history of self-limiting viral infections including measles and rubella in childhood. Physical examination revealed left cervical lymphadenopathy, splenomegaly and sever jaundice. Abnormal laboratory findings included a white blood cell count of 4.9 × 109/L (3% atypical lymphocytes) with large granular cells and anisocytosy in peripheral smear.Liver dysfunction was seen, with an increase in the levels of aspartate aminotransferase (1515 IU/l), alanine aminotransferase (1658 IU/l), lactate dehydrogenase (1080 IU/l) and total bilirubin (18.6 mg/dl). Prothrombin time was 28%. Serum immunoglobulin levels were elevated (3.8 gr/dl) with a titre of anti nucleus antibody (ANA) of 1:640. No antibodies against human immunodeficiency virus (HIV), hepatitis C virus (HCV), Hepatitis B virus (HBV), Cytomegalovirus (CMV), Epstein Barr Virus (EBV) were detected. However anti-HHV-6 antibody (IgG and IgM) were detected with IgM index of 3.2 (cut off for positive control > 1.1). A diagnosis of hepatic failure was made, and liver biopsy was performed during the acute stage. Histologic examination showed moderate infiltration of atypical lymphoid cells and diffuse focal vacuolar degeneration of hepatocytes. The infiltrating lymphocytes were positive for CD3, CD4, and CD8, but negative for CD20. The presence of HHV-6 DNA was shown in liver tissue by polymerase chain reaction (PCR) with a viral load of 280 genomes/106 of cellular genomes, suggesting active viral replication in the hepatocytes. Methylprednisolone was administered for three weeks beginning on the seventh day of hospitalization with dosage of 25 mg every twelve hours. The jaundice, lymphadenopathy and splenomegaly gradually disappeared and patient was sent home on the 35th hospital day with a normal hepatic function and no clinical sequelae. At 2 months HHV6 IgM antibodies decreased and disappeared after 3 months.DiscussionOur data indicate that HHV-6 was the cause of our patient's acute illness. Serologic studies excluded the possibility of active infection by HCV, HBV or other human herpesviruses such CMV and EBV. The presence of HHV-6 IgM antibodies shortly after the onset of liver disease and positive ANA titres suggest that HHV6 or an autoimmune disease may also be involved in the pathogenesis. HHV-6 is a CD4 lymphotropic virus isolated from T-cells cultures derived from the blood of subjects HIV+ [1]. Infection by HHV6 is rapidly controlled by the host immune response, and the virus established a state of latency. Primary infection occurs mostly in early childhood and only rarely in adults, in whom the prevalence of anti-HHV6 IgG is more than 90% [3]. Symptomatic infection is characterized by fever, skin rash (exanthema subitum), sometimes associated with mild respiratory illness, leukopenia and atypical lymphocytosis [3,10]. Recovery is usually rapid and benign, although a more severe course with meningitis, encephalitis, myocarditis or hepatitis, variable from mild hepatitis to fulminant liver failure, has been described [3]. HHV6 has also been associated with interstitial pneumonia and encephalitis in immunocompromised patients. [11] We assume that HHV-6 caused the initial clinical manifestations but an humoral virus-triggered autoimmune reaction, indicated by the positive ANA titre, responding to immunosuppression induced hepatic damage. Manifestations of autoimmune hepatitis have been described repeatedly after infection with hepatitis A, B and C as well as with herpes viruses, namely HSV1, EBV and HHV6 [12,13].ConclusionAutoantibodies may be triggered by a virus specific mechanism to evade immune responses called 'molecular mimicry', when domains on viral proteins closely resembling human self-epitopes are generated [14,15]. Thus, we believe that in addition to causing exanthem subitum in infants and a febrile illness in children, HHV-6 type B can cause an acute and potentially fulminant hepatitis in adults with an autoimmune pathogenetic mechanism.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsPGjr acquisited data and made analysis and interpretation of data, PC assessed ultrasonographic examination, RC assessed liver biopsy, PG analized tissue molecular test and helped to draft the manuscript. All authors read and approved the final manuscript. Informed written consent was received from the patient for publication of the manuscript.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2531166.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2531166",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2531166\nAUTHORS: Anders Skinningsrud, Vidar Stenset, Astrid S Gundersen, Tormod Fladby\n\nABSTRACT:\nBackgroundThe objective was to assess the utility of total tau protein (tTau), the ratio of (tTau)/181 phosphorylated tau protein (P-Tau) and 14-3-3 protein, as diagnostic markers in cerebrospinal fluid (CSF) for Creutzfeldt-Jakob disease (CJD).MethodsCSF samples received from Norwegian hospitals between August 2005 and August 2007 were retrospectively selected from consecutive patients with tTau values > 1200 ng/L (n = 38). The samples from patients clinically diagnosed with CJD (n = 12) were compared to those from patients with other degenerative neurological diseases: Alzheimer's/vascular dementia (AD/VaD, n = 21), other neurological diseases (OND, n = 5). Total Tau, P-Tau, and β-Amyloid (Aβ42) were measured with commercial kits. Additionally, 14-3-3 protein was measured semi-quantitatively by immunoblot.ResultsThe minimum cut-off limits for diagnosis of CJD were chosen from the test results. For tTau the lower limit was fixed at 3000 ng/L, for the tTau/P-Tau ratio it was 60, and for 14-3-3 protein it was 0.75 arbitrary units. For tTau and tTau/P-Tau ratio, all but three CJD patients had levels above the minimum, whereas almost all of the other patients were below. For the 14-3-3 protein, two CJD patients were below the minimum and five were above. Only one of the other patients was higher than the limit. The sensitivities, specificities and diagnostic efficiencies were: tTau 75%, 92%, and 87%; tTau/P-Tau 75%, 96%, and 89%; and 14-3-3 protein 80%, 96%, and 91%.ConclusionThe results suggest that 14-3-3 protein may be the better marker for CJD, tTau/P-Tau ratio and tTau are also efficient markers, but showed slightly inferior diagnostic properties in this study, with tTau/P-Tau marginally better than tTau.\n\nBODY:\nBackgroundThe prion protein disorder Creutzfeldt-Jakob disease (CJD) is a human transmissible spongiform encephalopathy (TSE) that leads to rapid decay of brain tissue. Human prion diseases are either idiopathic such as sporadic or spontaneous CJD (sCJD), genetic or familial (fCJD), acquired such as iatrogenic CJD (iCJD) or variant CJD (vCJD). Variant CJD is related to bovine spongiform encephalopathy (BSE), has clinical and pathological characteristics different from sCJD and has not been described in Norway. The most common form is sCJD, which accounts for about 85% of all known cases. Spontaneous CJD is a rare fatal disorder with rapid progression, an incidence of approximately 1/million/year [1] and a mortality of more than 90% within one year. The etiology of sCJD is not known, but the pathogenesis is related to conversion of the normal membrane prion protein PrPc to the pathological form PrPSc [2]. Immunohistochemical demonstration of PrPSc provides a definite diagnosis of CJD at autopsy or by brain biopsy.In the Norwegian population of about 4.5 million, one would expect approximately 5 cases of sCJD per year. Due to the few specific pre-mortal diagnostic signs, it is difficult to separate sCJD clinically from rapidly progressing Alzheimer's disease (AD) and other rapidly progressing neurological diseases. As TSEs can cross species barriers, there is a public health concern about the ability of TSE to spread from other species to humans. Classical scrapie in sheep is endemic in Norway, and an atypical variant, Nor 98, has appeared [3], but there is no evidence that scrapie has spread from animals to humans.The total concentration of tau protein (tTau) in CSF has been found to separate patients with CJD from those with AD [4]. As an increased concentration of tTau is regarded to be a marker for degradation of neurons, the success of tTau as a marker for CJD depends on whether other neurological diseases have the same high amount and rate of neuronal degradation as CJD. A low concentration of the amyloid precursor protein (APP)-derived 42 amino acid peptide in CSF (Aβ42), has been correlated with brain amyloid deposition in AD [5]. Elevation of 181 phosphorylated tau protein (P-Tau) is also related to brain pathophysiology in AD, possibly as a result of interaction between amyloid and tau metabolism [6]. As low CSF-Aβ42 and elevated P-tau are more specific for AD, one would expect the diagnostic specificity for CJD to rise when the markers are combined. In line with this, the ratio of tTau/P-tau has been described to separate CJD from other neurodegenerative diseases without overlap [7,8]. A separation of sCJD in two clinically different groups according to P-Tau level and a negative prognostic value of elevated P-Tau have been described for CJD [9]. Spontaneous CJD patients with high P-Tau had a shorter disease duration (i.e. they died earlier), had earlier onset of akinetic mutism and a higher incidence of typical EEGs. The occurrence of elevated P-Tau values in CJD will decrease the value of combining tTau and P-Tau in the diagnosis of sCJD because the difference in tTau/P-Tau ratio between the groups would be reduced. The 14-3-3 proteins are evolutionarily conserved proteins present in the cytoplasm of brain neurons at a concentration of about 1% of the total protein content. It has been suggested that 14-3-3 protein as a pre-mortem immunoassay marker, may obviate the need for a brain biopsy in the diagnosis of CJD. Although 14-3-3 protein is known to be non-specific, it is included in the WHO criteria for CJD [10].We wished to investigate the utility of tTau, tTau/P-Tau ratio and 14-3-3 protein measurements in CSF as markers for sCJD. Other diagnostic characteristics for CJD, such as hyper intense magnetic resonance (MR) signals from the basal ganglia, sharp wave complexes in EEG and the examination of the CSF proteins, neuron specific enolase (NSE) and S100 have been evaluated elsewhere [11]. Routine laboratory analyses of the three biological variables, tTau, P-Tau and Aβ42, have been offered at the Department of Clinical Chemistry, Akershus University Hospital, since 2003. Routine analysis for 14-3-3 protein has been performed since November 2007. Neurological, psychiatric and outpatient departments in Norway have also had the opportunity to send CSF-samples for analysis.MethodsPatient selection and CSF analysisCSF samples (n = 691) were received by the Department of Clinical Chemistry, Akershus University Hospital between August 2005 and August 2007 from patients demonstrating symptoms of cognitive decline and possible neurodegenerative disease. This work was supported and approved by the Eastern Norway Regional Health Authority, RHA East, and approved by the Regional Committee for Medical Research Ethics, which also approved an exception from collecting informed patient consent.Total Tau, P-Tau and Aβ42 were analysed in CSF-samples with commercially available enzyme linked immunosorbent assay (ELISA) kits from Innogenetics (Gent, Belgium) adapted to a Tecan Robotic Microplate 150 Processor (Tecan AG, Switzerland). The analyses were performed approximately twice a month. The 14-3-3 protein was analysed by immunoblot with equipment from Invitrogen Corporation Ltd. (Paisley, U.K.) using an antibody against the γ-isoform, anti-14-3-3 gamma, clone CG31-2B (mouse monoclonal IgG1, Upstate biotechnology, Lake Placid, NY, USA). The 14-3-3 protein results were assessed semi-quantitatively using arbitrary units. As standard we used dilutions of a homogenate from normal brain. For semi-quantification we performed image analysis using Fujifilm Multi Gauge software (Fujifilm Corporation, Tokyo, Japan). Thirty-nine patients studied retrospectively had tTau values > 1200 ng/L (the highest standard of the kit). One patient was excluded because no additional sample was available to analyse tTau in dilution. Total Tau, P-Tau and Aβ42 were measured at the Akershus University Hospital between August 2005 and October 2007. 14-3-3 protein was analysed in seven of the 12 CJD patients and 25 of the 26 AD/VaD/OND patients in October and November 2007 in Akershus University Hospital. In one AD patient and in five CJD patients there was no sample available for analysis. Three of these CJD-patients had qualitative 14-3-3 protein results from laboratories outside Norway. Two CJD patients had no 14-3-3 protein determination. The maximum value for normal clinical values for tTau was 300 ng/L for patients < 50 years, 450 ng/L for patients 50–70 years and 500 ng/L for patients > 70 years [12]. For Aβ42, we have previously used values above 450 ng/L for normal levels [13], but after comparison with the Neurochemistry laboratory at Sahlgrenska University Hospital, Gothenburg, Sweden (unpublished data), this has recently been revised to 550 ng/L. The maximum for normal clinical levels for P-Tau was 80 ng/L from the Neurochemistry laboratory in Gothenburg.Sample handlingThe samples for tTau, P-Tau and Aβ42 were initially frozen at -20°C. In February 2006 all old samples were transferred to -80°C and all new samples were kept at -80°C. The samples for 14-3-3 protein were stored for up to 2 years at -80°C before analysis in batches of six. The ELISA standards were run in duplicate and the samples were run singly. The analytical results were read from the corresponding standard curve for each run. No result exceeded the highest standard for P-Tau and Aβ42, but for tTau all the patients in this study exceeded the highest standard point of 1200 ng/L. These samples were diluted and rerun.Patient diagnosisThe patients were clinically diagnosed according to ICD-10 (International Classification of Diseases -10). The criteria were clinical suspicion, MR findings highly suggestive of CJD and three-phase sharp wave complexes on EEG also considered to be characteristic of CJD. Twelve patients had definite or probable CJD (Table 1); 21 patients had Alzheimer's disease (AD), vascular dementia (VaD), mixed dementia (AD/VaD), frontotemporal dementia (FTD) or unspecified dementia and five patients had other neurological diagnoses (OND, Table 2). Seven of the 12 CJD patients had short disease duration, mean 3.7 months (n = 7, SD = 1.5) and a further two patients had unspecified very rapid progression. Three patients had CJD with slower progression, disease duration 15, 22 and >23 months.Table 1Characteristics and results for patients with Creutzfeldt-Jakob diseaseSex/age (years)Diseaseduration1(months)EEGMRAutopsytTau (ng/L)P-Tau4(ng/L)tTau/P-Tau (ng/L)Aβ425 (ng/L)14-3-3 proteinFemale/842TSWC2NC8Denied7,09049173664Not doneMale/756TSWCNCDenied13,830211*65385*Positive7Male/624GD6NCNot done12,910149*83372*Not doneFemale/5822GDNCDenied1,88092*26623Positive7Female/773TSWCHSC9Positive331204569704Positive7Female/674TSWCHSCNo result853266129526*Positive7 0.96Female/64RapidTSWCHSCPositive biopsy 314,06084*167562Positive7 3.68Male/742GDHSCNot done22,46082*2749462.42Female/65>23TSWCHSCNot done1,34372193M0.30Female/7015TSWCHSCNot done2,34893*2510510.35Female/61RapidTSWCHSCNo result11,038482307551.75Male/545TSWCHSCNo result71,900172*971042Positive7 9.701The time from disease recognition to death, 2Three-phase sharp wave complexes, 3Positive for immunochemical examination of pathological prion protein, PrPSc, 4Values above the normal limit used in clinical practice marked with *, 5Values below the normal limit used in clinical practice marked with*, 6Generalized dysrythmia, 714-3-3 protein found positive in different laboratories outside Norway because some hospitals sent CSF samples for 14-3-3 protein analysis to laboratories outside Norway in parallel to sending samples to us. (Three patients both have semi quantitative results from us and a qualitative, positive result, from a foreign laboratory). 8NC, non-specific changes. 9HSC, highly suggestive changes of CJD, M = missing value.Table 2Characteristics, results and diagnosis for patients with Alzheimer's disease, vascular dementia, mixed dementia and other neurological diseasesSex/age (years)tTau1 (ng/L)P-tau2 (ng/L)tTau/P-TauAβ423(ng/L)14-3-3 protein1(arbitrary units)DiagnosisFemale/592120244*8401*MADFemale/642335240*10347*0.00ADFemale/761400220*6505*0.00ADFemale/792443287*9517*0.00ADFemale/85135049277460.59MCIFemale/731380217*67430.15ADFemale/671410199*7413*0.00ADFemale/6216706625348*0.41AD/WernickeFemale/51141027526970.00Possible FTDFemale/661410238*66220.00MCI/Possible ADFemale/64175049369260.55Possible VaDMale/651910792412970.34Unspecified dementiaMale/651280572212440.53Possible AD/VaDFemale/623443242*14312*0.52FTD/Possibly ADFemale/671540238*6369*0.00Probably ADFemale/88272090*30464*0.00AD/VaDMale49135081*179730.00VaDFemale/721780240*7533*0.00ADFemale/801280239*56990.00Probably ADFemale/761480227*7516*0.00AD/VaDFemale/761360207*75730.00ADMale/628,530*7511410430.00Cerebral lymphomaMale/711,2025920487*0.00Cerebral infarctionFemale/691,99283*246740.29Cerebral infarctionFemale/491,26727478070.95*Cerebral lymphomaFemale/321,43349299600.00Cerebellitis1Values above lower limit for CJD marked with*, 2Values above maximum for normal in clinical practice marked with*, 3Values below minimum for normal used in clinical practice marked with*, M = missing value.Data analysisReceiver operating characteristic (ROC) curve analysis was performed for tTau and tTau/PTau ratio with the statistical package Analyse-it (Analyse-it Software, Ltd. Leeds, U.K.). ROC-curve analysis was not performed for 14-3-3 protein because some of the results from other laboratories were qualitative only. The AD, VaD, AD/VaD and other dementia patients were treated as one group, AD/VaD. The lower limit chosen for the diagnosis of CJD was 3000 ng/L for tTau, for the ratio of tTau to P-Tau it was 60 and for 14-3-3 protein it was 0.75 arbitrary units. The limits, based on our test results, were set to obtain the best diagnostic performance for each marker. It could be argued that the cut-off for tTau and 14-3-3 protein could have been set slightly higher. In that case for tTau the sensitivity would decrease and the specificity would increase, and for 14-3-3 protein the sensitivity would also decrease and the specificity would increase to 100%. The sensitivity, specificity, positive and negative predictive values, and diagnostic efficiency for the three markers were calculated manually.ResultsTable 1 presents the characteristics for 12 patients who were diagnosed with CJD, two definite CJD and 10 probable CJD. Eight had MR findings suggestive of CJD. Generalized dysrhythmia and sharp wave complexes on EEG were found in nine patients and generalized dysrhythmia indicating diffuse cerebral pathology in the remaining three. Table 2 presents the characteristics for patients with diagnoses other than CJD. Two patients had VaD or possible VaD, 11 had AD or possible AD, three had AD/VaD or possible AD/VaD, one had AD and Wernicke's encephalopathy, two had possible FTD and two had unspecified dementia. Five patients had OND.Total TauThe results for tTau by diagnostic group and lower limits for CJD are presented in Figure 1A. Eleven patients had tTau values above cut-off (3000 ng/L) and 27 below. Nine of the 12 patients with tTau-values above cut-off had CJD, one had OND (cerebral B-cell lymphoma) and one had possible FTD. Three patients with CJD of long duration had values (1343, 1880 and 2350 ng/L) which are below the chosen lower limit, 3000 ng/L. All but one of the OND patients and all but one of the AD/VaD patients had values below 3000 ng/L.Figure 1Data plots of Tau protein (tTau) (A) and tTau/P-Tau ratio (B) by diagnostic group Creutzfeldt-Jakob disease (CJD), other neurological diseases (OND) and Alzheimer's disease or vascular dementia (AD/VaD). Horizontal dashed lines show chosen lower limit for CJD: tTau: 3000 ng/L, and tTau/P-Tau ratio: 60.Ratio of tTau to P-TauThe results for the ratio tTau/P-Tau by diagnostic group and chosen lower limit for CJD (60) are presented in Figure 1B. Ten patients had tTau/P-Tau ratios above the limit. Nine of these had CJD and one had cerebral B-cell lymphoma. The three patients with CJD of long duration, had values below cut-off (19, 25 and 26). All but one of the OND patients and all AD/VaD patients had values below the limit for CJD.P-TauP-Tau results by diagnostic group and maximum for normal (80 ng/L) are presented in Figure 2A. Six of the 12 CJD patients had P-Tau above normal. In the AD/VaD-group (n = 21) 15 patients (71%) had P-Tau above and six below normal. The P-Tau results in the AD/VaD patients were distributed in two groups, range 27–92 and 200–287 ng/L. One of the OND patients (cerebral infarction) had P-Tau slightly above normal, 81 ng/L.Figure 2Data plots of 181-phosphorylated tau protein (P-Tau) (A) and beta-amyloid (Aβ42) (B) by diagnostic group CJD, OND and AD/VaD with upper (P-Tau) and lower (Aβ42) normal limits used in clinical practice (dashed lines): P-Tau: 80 ng/L, Aβ42: 550 ng/L.Beta amyloid (1–42)The results for Aβ42 by diagnostic group and lower normal limit are presented in Figure 2B. Three of 11 patients (38%) in the CJD group (one missing value) and 11 of the 21 patients (52%) in the AD/VaD group had Aβ42 values below normal. The two CDJ patients with the lowest Aβ42-values, 372 and 385, also had elevated P-Tau results, 149 and 211 ng/L.14-3-3 proteinThe lower limit for CJD diagnosis was set at 0.75 arbitrary units. Figure 3 shows results from the home laboratory. Samples from 32 patients, seven from the CJD group, all five OND patients and 20 from the AD/VaD group, were available for analysis. Five of the seven CJD patients tested positive. Three of the 14-3-3 positive patients had been tested before and had positive results from laboratories outside Norway. Three of the five CJD patients not tested by us, had been tested before and had positive results. These were included in the estimations of diagnostic parameters (Table 3). Ten of the 12 patients with CJD had been examined for 14.3.3-protein by us, other laboratories or both. Eight of these were positive. One of the five patients with OND (cerebral B-cell lymphoma) tested weakly positive, and all the tested patients in the AD/VaD group were negative. Of the three CJD patients with long disease duration, two were negative.Figure 3Data plots of 14-3-3 protein by diagnostic group CJD, OND and AD/VaD with lower limit for CJD (dashed line): 0.75 arbitrary units (data from Akershus University Hospital only).Table 3A comparison between markers of the diagnostic performance for the diagnosis of CJD.tTautTau/P-Tau14-3-3 proteinSensitivity757580Specificity929696Positive predictive value829089Negative predictive value898992Diagnostic efficiency878991Diagnostic performanceCalculations of sensitivity, specificity, positive predictive value, negative predictive value and diagnostic efficiency are presented in Table 3. The ROC curve analysis for the diagnosis of CJD showed that tTau was not significantly different from tTau/P-Tau for the diagnosis of CJD (area under curve 0.897 and 0.918, n.s.).DiscussionAlthough the ROC-curve analysis showed no significant difference between tTau and tTau/P-Tau, the tTau/P-Tau ratio did separate results around the cut-off value more clearly than t-Tau and the specificity and predictive values of a positive test for the CJD diagnosis were slightly better. Thus, our findings seem to agree with those of Riemenschneider et al [7], who found that tTau/P-Tau ratio was a better marker for CJD than tTau. Buerger et al [14] came to the opposite conclusion, but they measured P-Tau phosphorylated in the 232- position and not in the 181-position as measured here and by Riemenschneider et al. Our results suggest that 14-3-3 protein may be a slightly better marker than tTau and tTau/P-Tau.In contrast to P-Tau, Aβ42 did not contribute to providing a more definite diagnosis of CJD. Increased P-Tau was a more specific measure of AD/VaD than low Aβ42, and contributed slightly to the diagnostic separation of CJD from the AD/VaD group. Similar values for Aβ42 in the CJD and AD/VaD groups suggest that comparable amyloid pathology was present in both groups. This indicates that some CJD patients may have acquired CJD in addition to an increased and possibly AD-related, amyloid pathology. The fact that the two CJD patients with the lowest Aβ42 values, also had elevated P-Tau is consistent with an interaction between prion protein and the AD pathological processes [15]. The P-Tau values separated into two groups for AD/VaD patients (Figure 2a). The group with results above 200 ng/L was homogenous clinically because 11 of the 13 patients had the clinical diagnosis of AD, one had FTD/or possibly AD and one had AD/VaD. The group with results below 100 ng/L (n = 8) was clinically more heterogeneous (Table 2). Our results suggest that tTau, tTau/P-Tau and possibly 14-3-3 protein may only be good markers for sCJD of short duration and may not separate the CJD cases with longer duration from the other dementias and OND. This fact may be of considerable importance for the diagnosis of CJD and suggests that brain biopsy should be used more often in dementia cases. In addition, the use of autopsy should be encouraged.None of the patients in this study were tested for hereditary forms of CJD or AD, and the clinical information did not raise any suspicion about hereditary conditions. Seven patients in the AD/VaD group were <65 years. None of them had a family history of dementia. Our results show that in addition to AD and VaD, other rapidly progressing neurological diseases may have the same tTau and P-Tau biomarker pattern as CJD. One patient with cerebral lymphoma had tTau and tTau/P-Tau above the cut-off values for CJD. Another patient with the same diagnosis was slightly positive for 14-3-3 protein. In spite of using both tTau, P-Tau and 14-3-3 protein for the diagnosis of CJD, our data suggest that there will still be a few patients with AD/VaD and OND that cannot be distinguished from CJD using the biomarkers tTau, P-Tau and 14-3-3 protein. Other investigations will usually establish the diagnosis in these cases by imaging or CSF-cytology.It can be argued that this study should have been performed prospectively, i.e. patients clinically suspected to have CJD should have been enrolled consecutively. This would have required a more active cooperation between the clinical centres. We were not able to undertake this task at that time, but it should ideally be done as a follow-up of the present study. Starting with high tTau patients was, however, a cost-effective way to find cases. The reason for not analysing 14-3-3 protein at the same time as the other markers was that the analysis was not set up by us until October 2007.Another criticism that could be raised against our study is the low frequency of histological verification. The reason for this is the right to refuse autopsy and the low autopsy frequency in Norway in general. The importance of obtaining histological verification should be stressed in future prospective CJD studies. Considering the low number of histological verifications, it is possible that some of the CJD patients were wrongly classified. We would argue that this is less likely as eight of the 12 patients (two proven histologically) had MR findings suggestive of CJD. Seven of these also had three-phase sharp wave complexes. Of the remaining four patients, all had non-specific MR changes and two had three-phase sharp wave complexes. The two remaining patients both had non-specific changes on EEG and the disease duration in one of them was four months, highly suggestive of CJD. The other patient had disease duration of 22 months and although 14-3-3 protein was positive, another type of degenerative brain disease cannot be totally excluded.This study also shows a higher prevalence of CJD in Norway than might be expected. We identified 12 cases during a period of two years. With a prevalence of one case per million inhabitants, approximately 10 cases would be expected, which is quite close to the observed number of 12. However, we do not expect that all the Norwegian cases in this period were known to us. The prevalence of CJD may therefore be higher than anticipated. This indicates a need for closer surveillance of human prion diseases in Norway.ConclusionTotal Tau, tTau/P-Tau ratio and 14-3-3 protein are useful, but not entirely sensitive and specific markers for the diagnosis of CJD in CSF. The results indicate that the diagnostic performance of 14-3-3 protein may be slightly better than tTau/P-Tau, which may be slightly better than tTau alone. There is clearly a need for a specific test for CJD, preferably for the misfolded prion protein (PrPSc) itself in readily obtainable biological samples such as blood and CSF. It remains to be seen whether a specific test for PrPSc would be as sensitive and specific as the markers used in this study.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsAS designed the study, drafted the manuscript, and performed the statistical analysis. VS and TF participated in its design and coordination and helped to draft the manuscript. ASG performed ELISA tests, set up the immunoblot for 14-3-3 protein and contributed to the manuscript. All authors read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2531170.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2531170",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2531170\nAUTHORS: Minwook Yoo, Dion F Graybeal\n\nABSTRACT:\nA myxoma is the most common primary tumor of the heart. It has been reported as the source of a cardiogenic embolism. Therefore, it is important for clinicians to detect the myxoma early via echocardiography to prevent complications, such as syncope, sudden death, and cerebral embolic ischemic stroke. This report presents the case of a 54-year-old female whose clinical manifestation of atrial myxoma was an ischemic stroke. Atrial myxoma was later confirmed as the cause of her symptoms via transesophageal echocardiography.\n\nBODY:\nBackgroundCerebral infarction induced by cardiogenic embolism is observed in about 20% of stroke patients. Of those patients, atrial fibrillation is responsible for over 50% of the cardiogenic emboli, while myxomas are observed in only 0.5% of emboli [1]. Atrial myxomas are a very rare source of cardiogenic embolism. Although they are usually asymptomatic, myxomas can develop lethal complications without warning because of their ability to embolize. This report describes a patient who presented with a left-sided hemiparesis. The cause of the patient's right cerebral infarction was a left atrial myxoma which was detected by transesophageal echocardiography (TEE).Case presentationA 54-year-old Caucasian female presented to the emergency room with a 4 day history of left-sided weakness. The patient stated that she was at home when she suddenly felt a sharp pain in her left hand that radiated to her neck. She then lost consciousness and collapsed to the floor. It was not until 4 days later that her friend convinced her to go to the hospital. The patient had a medical history of longstanding hypertension, obstructive sleep apnea, hypothyroidism, and depression. She had been a smoker for 25 years. Her mother also had hypertension and her father had a myocardial infarction at the age of 56. Her height was 161.5 cm and her weight was 127.3 kg. The patient's vital signs were as follows: blood pressure, 153/104 mm Hg; heart rate, 101 beats/minute; respiratory rate, 18/minute; and body temperature, 36.6°C (98.0°F). She was alert and oriented, had left facial paralysis, slight dysarthria and right-sided tongue deviation, but no dysphasia. On cardiac examination, the carotid impulse was normal without a bruit. Her heart had a regular rate and rhythm, and there were normal S1 and S2 heart sounds without murmurs. An EKG showed a normal sinus rhythm. Range of motion was limited to 30° for the left upper and lower extremities. She had 1/5 motor strength on the left side, but 5/5 motor strength on the right side. The deep tendon reflexes were 2+ bilaterally. Her sensation was intact bilaterally. The Babinski and Hoffman signs were both negative. All her laboratory results were normal. A chest X-ray showed a normal cardiac silhouette with no signs of pulmonary edema. A non-contrast computed tomography (CT) scan of the brain revealed multiple low density areas in the right frontal and parietal lobes. Our stroke team started her on intravenous heparin. Metoprolol (Toprol-XL) and furosemide (Lasix) were administered to stabilize her blood pressure. The following day, magnetic resonance imaging (MRI) of the brain demonstrated an acute infarction in the distribution of the right middle cerebral artery (MCA; Figure 1). On the third day of hospitalization, the patient underwent a TEE. A TEE was chosen since the less invasive transthoracic echocardiography (TTE) showed negative imaging for a cardiogenic embolic source. In addition, the patient was obese and the TTE did not provide a comprehensive image. The TEE identified a 4.3 cm × 1.3 cm mass in the left atrium. A cardiac catheterization showed no significant coronary artery disease. The patient was thus diagnosed with a right MCA ischemic infarction and a left atrial myxoma. On the 13thhospital day, the patient underwent successful surgical excision of the myxoma (Figure 2). The biopsy confirmed the diagnosis of myxoma. The patient recovered without any complications and was discharged on the 20th day of hospitalization.Figure 1Axial T1-weighted MRI shows acute infarcts in the right caudate body, and the frontal and parietal lobes. (A), The diffusion weighted image (DWI) presents an abnormal signal corresponding to the restricted diffusion in the right hemisphere. (B), MRA shows no aneurysm, stenosis, or abnormal flow in the visualized vessels of the Circle of Willis, the carotid arteries, and the vertebral arteries. (C).Figure 2Transesophageal echocardiography shows a mobile mass in the left atrium, which does not obstruct the mitral valve. (A), After performing cardiopulmonary bypass, the retractor allows visualization of the left atrium, and the atrium is then opened by a blade. The myxoma is attached from the atrial septum. (B), LA myxoma: tan and jelly-like tissue with an aggregate measurement of 6 cm × 1.5 cm. (C).DiscussionA myxoma is the most common primary tumor of the heart. Primary cardiac neoplasms are rare, with incidences ranging from 0.001–0.3% in autopsy series. Benign tumors account for 75% of primary neoplasms and malignant tumors account for 25%. Myxomas compromise 30–50% of primary cardiac tumors [1]. The majority of myxomas are sporadic, but 7% of patients have a genetic mutation that is inherited in an autosomal dominant manner. Familial myxoma has been well-described as the Carney complex, characterized by hyperpigmentation, cutaneous myxomas, and endocrine adenomas. This tumor is three times more common in females than in males and generally occurs between the third and sixth decades, with an average age of presentation at 43 years [2].Myxomas originate from the mesenchymal cells of the septal endocardium. They are gelatinous with a smooth or lobulated surface and are usually white, yellow, or brown in color. They can present as villous, papillary, sessile, or pedunculated-type growths. Approximately one-half of the cases of myxomas are pedunculated tumors, and these are irregular and more likely to result in emboli because of the mobility of this type of tumor [3]. Sixty to 75% of cardiac myxomas develop in the left atrium, most of which are from the atrial septum near the fossa ovalis. Most other myxomas develop in the right atrium. Fewer than 20 cases of myxomas arising from the right or left ventricle have been reported [4]. Myxomas produce a vascular endothelial growth factor that stimulates angiogenesis and tumor growth and an increased expression of interleukin-6 [5].A myxoma may be completely asymptomatic until it grows large enough to obstruct the mitral or tricuspid valve or fragments that give rise to emboli. Because they are intravascular and friable, myxomas account for most cases of tumor emboli [1]. Embolism occurs in about 30–40% of patients with myxomas. The site of embolism is dependent upon the location of the myxoma (left or right atrium) and the presence of an intracardiac shunt. This is not surprising, given the degree of motion that can be seen on echocardiography and angiography, as the myxoma swings on a small pedicle with each cardiac contraction [6]. Intermittent acute obstruction of the mitral orifice has been reported to produce syncope and even sudden death. Some myxomas produce generalized symptoms resembling an autoimmune disorder, including fever, weight loss, digital clubbing, myalgias, and arthralgias. These patients may have an immune reaction to the neoplasm, as elevated levels of interleukin-6 and elevated levels of antimyocardial antibodies have been described [5].The emboli that occur are either a tumor fragment that is released from the myxoma or a blood clot that is formed on the surface of the myxoma. These resulting emboli can result in infarction, as occurred in our patient. More precisely, it has been reported that 45% of patients with myxomas have neurologic manifestations resulting from embolization [7]. This embolization includes pulmonary embolism, myocardial infarction, mesenteric infarction, retinal artery occlusion, spinal cord ischemia, and stroke [2-7]. Right-sided myxomas cause pulmonary embolization, but left-sided myxomas usually cause systemic embolization. Ischemic infarction of the brain is responsible for the majority of cases of systemic embolization. The MCA is frequently affected by this type of infarction because of the MCA's dominant blood flow [8]. In cases in which frontal or parietal infarction is suspected in a patient with myxoma, the MCA territory should be thoroughly investigated.Usually, the diagnosis is readily established by two-dimensional echocardiography, which is considered the gold standard. TEE may be useful when transthoracic findings are equivocal or confusing. MRI has been of value in diagnosis, providing excellent cardiac definition. Cardiac catheterization is not necessary in the majority of cases, but may be necessary when other cardiac disease is suspected or if other diagnostic studies are equivocal. TTE has a sensitivity of 95% and the sensitivity is nearly 100% [9]. Whether performing TTE or TEE, echocardiography is able to evaluate the location, size, shape, and movement of myxomas. TTE or TEE may also show other cardioembolic sources, such as a patent foramen ovale, mitral valve calcification, or aortic atherosclerosis. Prompt resection is required after the diagnosis, even in asymptomatic patients. It is important that myxomas should be excised with negative margins because any remnant can aggravate an infarction. The recurrence rate is 1~3% after surgery [10]. Therefore, all patients with myxomas are recommended to undergo long-term follow-up with echocardiography. This patient described herein, who was morbidly obese with a BMI of 48.8 kg/m2, is representative of a growing medical problem in the United States. With stroke patients, physicians use TTE routinely when they search for cardiogenic embolic sources. But, in using TTE exclusively, myxomas in the obese will frequently be missed.ConclusionThis case demonstrates the importance of investigating the possibility of cardiogenic source in stroke, as our patient developed cerebral infarction that was caused by an atrial myxoma. It is important that clinicians consider using echocardiography in stroke patients. Treating the atrial myxoma can prevent a cardioembolic stroke and its complications. In conclusion, TEE, as compared to TTE, has many more advantages when physicians search for a cardiogenic embolic source in obese stroke patients. In addition, because obesity has sharply increased in the United States, the importance and use of TEE will increase over time as physicians encounter obese patients with cardiogenic emboli.AbbreviationsMCA: Middle Cerebral ArteryP; CT: Computed Tomography; MRI: Magnetic Resonance Imaging; TEE: Transesophageal Echocardiography; TTE: Transthoracic Echocardiography.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsMY and DG were involved in the clinical assessment and writing the case report. All authors read and approved the final manuscript.ConsentFull written consent was received for the manuscript to be published.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2531176.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2531176",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2531176\nAUTHORS: Andrew R Davies, Matthew J Forshaw, Aadil A Khan, Alia S Noorani, Vanash M Patel, Dirk C Strauss, Robert C Mason\n\nABSTRACT:\nBackgroundThe optimal operative approach for carcinoma at the lower esophagus and esophagogastric junction remains controversial. The aim of this study was to assess a single unit experience of transhiatal esophagectomy in an era when the use of systemic oncological therapies has increased dramatically.Study DesignBetween January 2000 and November 2006, 215 consecutive patients (182 males, 33 females, median age = 65 years) underwent transhiatal esophagectomy; invasive malignancy was detected preoperatively in 188 patients. 90 patients (42%) received neoadjuvant chemotherapy. Prospective data was obtained for these patients and cross-referenced with cancer registry survival data.ResultsThere were 2 in-hospital deaths (0.9%). Major complications included: respiratory complications in 65 patients (30%), cardiovascular complications in 31 patients (14%) and clinically apparent anastomotic leak in 12 patients (6%). Median length of hospital stay was 14 days. The radicality of resection was inversely related to T stage: an R0 resection was achieved in 98–100% of T0/1 tumors and only 14% of T4 tumors. With a median follow up of 26 months, one and five year survival rates were estimated at 81% and 48% respectively.ConclusionTranshiatal esophagectomy is an effective operative approach for tumors of the infracarinal esophagus and the esophagogastric junction. It is associated with low mortality and morbidity and a five survival rate of nearly 50% when combined with neoadjuvant chemotherapy.\n\nBODY:\nIntroductionDuring the last thirty years, there has been a marked increase in the incidence of adenocarcinoma close to the esophagogastric junction whilst the incidence of squamous cell carcinoma of the esophagus has remained relatively unchanged [1]. Surgical resection of tumors in the esophagus and esophagogastric junction has been based upon the concept that, if all neoplastic tissue can be removed, a worthwhile period of survival and possibly cure can be achieved. Despite oncological advances, surgical resection is the only treatment that has repeatedly been shown to prolong survival, albeit in only 30% of patients [2].Transhiatal esophagectomy is often advocated as the preferred surgical approach in patients with benign disease or early tumors or those patients with more advanced disease who would not tolerate a thoracotomy. This approach has been criticized because of the lack of a formal two field lymphadenectomy and the failure to completely resect the tumor under direct vision [2]. Transhiatal esophagectomy has been the favoured operative approach in our institution for managing both carcinoma of the oesophagus below the level of the carina and type I and II tumours of the esophagogastric junction. It has also been utilised for benign lower oesophageal disease including high grade dysplasia. This study evaluates our experience and outcomes with transhiatal esophagectomy in an era in which the use of neoadjuvant chemotherapy became more prevalent.MethodsStudy populationBetween January 2000 and January 2007, 215 patients with benign or malignant disease of the intrathoracic esophagus and type I and II tumours of the esophagogastric junction underwent transhiatal esophagectomy at our institution. Prospective data on these 215 consecutive patients was collected from consultant databases supplemented by cancer registry data and case note review. A further 152 patients underwent transthoracic esophagectomy during the same time period and were excluded from analysis. Ethical committee approval was obtained for this study and the need for individual patient consent was waived.Preoperative evaluation and treatmentRoutine preoperative evaluation involved upper gastrointestinal endoscopy with biopsy, endoscopic ultrasound and computed tomography of the neck, chest and abdomen. Staging laparoscopy and PET scanning were performed on a selective basis. Operative risk analysis included standard blood examination, electrocardiography, echocardiography, pulmonary function tests and cardiopulmonary exercise tests (in higher risk patients). Surgery was offered to medically fit patients following discussion at a multidisciplinary meeting.90 patients in the study group (42%) received preoperative chemotherapy based upon the presence of T3 disease or positive lymph nodes on preoperative staging. The preferred chemotherapy at our institution consisted of three cycles of combination epirubicin, cisplatin and 5-fluorouracil each given over three weeks, following the MAGIC trial protocol [3].Operative techniqueAll patients underwent subtotal esophagectomy and proximal gastrectomy by the transhiatal technique as described in detail by Orringer. [4-6] An initial laparotomy was performed through a rooftop incision to confirm tumour resectability. After abdominal exploration and gastric mobilisation had been performed, the esophageal hiatus was enlarged by splitting the diaphragm anteriorly and retractors were positioned to facilitate exposure of the intrathoracic esophagus up to the level of the carina. This enabled en bloc resection of the esophagus and paraesophageal tissue including the crura and pleura (if indicated) under direct visualisation. Standard lymph node dissection involved lymph nodes in the lower mediastinum, around the esophagogastric junction and along the lesser curvature of the stomach. A radical lymph node dissection was performed at the origins of the left gastric and common hepatic arteries; lymph nodes at the celiac axis were included when enlarged and resectable. A less radical resection was performed for patients with benign disease. Gastrointestinal continuity was re-established with a narrow gastric tube vascularized by the right gastroepiploic artery in all cases, positioned within the posterior mediastinum. An end to side hand sewn single layer esophagogastric anastomosis was fashioned in the neck through a left sided cervical incision. Transmediastinal chest drains and placement of a feeding jejunostomy were performed in all patients.Pathological examinationPathology specimens were processed by three dedicated esophagogastric pathologists according to Royal College of Pathologists' guidelines. [7] Tumors of the esophagogastric junction were categorized according to Siewert's classification based upon macroscopic tumor location, irrespective of the presence of Barrett mucosa. [8] Type I adenocarcinoma of the esophagogastric junction was staged according to esophageal pTNM classification whilst type II adenocarcinoma of the esophagogastric junction were staged according to gastric pTNM classification. [9] To ensure standardized histopathology results, all early specimens were re-categorized according to the latest guidelines.Follow upDuring the immediate postoperative period, patients were kept intubated and ventilated until the following morning. Following extubation, patients were monitored on a surgical High Dependency Unit until well enough to be managed on a surgical ward. Oral nutrition was recommenced if a water soluble contrast swallow examination failed to demonstrate an anastomotic leak on the seventh day.After discharge, patients were routinely followed up at 3–6 monthly intervals. Patients were offered either adjuvant chemotherapy (up to a maximum of 6 cycles) or chemoradiotherapy (if any margins were positive) based upon analysis of the pathological specimen and the histologically determined response to any preoperative treatment. Additional diagnostic procedures were only performed if indicated by the development of any new symptoms suggestive of recurrent disease. In the presence of recurrent disease, further oncological or palliative options were considered. The median duration of postoperative follow up was 26 months (range = 1–82 months) for all patients and 36 months (range = 2–82 months) for those alive at final follow up.StatisticsOverall survival was defined as the time interval from the date of operation until the date of death or most recent follow up. Disease free survival was defined as the time interval from the date of operation until the date of disease recurrence or most recent follow up. Survival curves were calculated according to the Kaplan-Meier method. Univariate group comparisons were calculated using the log rank test. Categorical variables were assessed using Fisher's exact test and continuous variables were assessed by student's t test [10]. A p value < 0.05 was regarded as statistically significant. Statistical analysis was performed with Graphpad Prism v3.0 and Instat v2.0 (GraphPad Software, San Diego California USA).ResultsPreoperative featuresThe demographic details of the 215 patients undergoing transhiatal esophagectomy are shown in Table 1. Dysphagia and weight loss were present in 73% and 48% of patients respectively with preoperatively confirmed malignant tumours. Twenty two patients (10%) had an asymptomatic cancer or high grade dysplasia detected during endoscopic surveillance of Barrretts oesophagus. Three patients (1%) underwent urgent transhiatal esophagectomy following endoscopic tumor perforation. According to the American Society of Anesthesiologists (ASA) classification [11], operative risk was scored as ASA-I (n = 15), ASA-II (n = 125), ASA-III (n = 72) or ASA-IV (n = 3).Table 1Demographic data on 215 patients undergoing transhiatal esophagectomy.DemographicsnSex (M:F)182:33Age (range)65 years (29–83 years)Preoperative indicationAdenocarcinoma162 (75%)Squamous cell carcinoma23 (11%)Other malignant tumours3 (1%)Benign tumours1 (0.5%)High grade dysplasia23 (11%)Benign strictures3 (1%)Preoperative staging (in 188 patients with preoperatively confirmed malignant tumours)T128 (15%)T248 (26%)T3108 (57%)T44 (2%)N0113 (60%)N+75 (40%)Intraoperative surgical findingsOnly one patient required intraoperative conversion to a right posterolateral thoracotomy due to tumor adherence at the carina and difficulties in achieving macroscopic tumor clearance through the esophageal hiatus. Macroscopic tumor clearance could not be achieved in one patient due to the presence of extensive left gastric and celiac axis lymphadenopathy. The median operative time was 151 minutes (range = 93–276 minutes).Postoperative courseThere were two in-hospital deaths during this study (<1%). One patient, a 74 year old man, with a past medical history including pneumonectomy for lung cancer and a previous myocardial infarction, developed respiratory failure requiring prolonged ITU admission and respiratory support; he died from myocardial infarction on day 44. The second patient, a 70 year old man, died from a pulmonary embolus on day 13 in ITU following admission with multiorgan failure secondary to chest sepsis.Major postoperative complications are listed in Table 2. All 12 patients with clinically apparent anastomotic leaks were managed conservatively with opening the cervical wound to allow adequate wound drainage and reduction of oral intake combimed with jejunostomy tube feeding. None of these patients required re-operation for their anastomotic leaks. 10 patients (5%) required re-operation in the early post-operative stage for: bleeding (n = 4), bowel obstruction (n = 3), chyle leak (n = 2) and wound dehiscence (n = 1). Unplanned ITU admission was required in 29 patients (14%), most commonly for respiratory failure. The median ITU stay in this group was 7 days (range 2–44 days). Overall median length of hospital stay was 14 days (range 8–95 days). All patients were discharged directly home and the in-patient stay reflects the need for sufficient mobility and tolerance of an adequate oral diet prior to discharge.Table 2Major postoperative complicationsComplicationn (%)Clinical anastomotic leak12 (5.6)Respiratorya65 (30)Cardiovascular31 (14)Recurrent laryngeal nerve neuropraxia6 (3)Wound infection22 (10)Renal failure6 (3)Chyle leak5 (2)Deep vein thrombosis/pulmonary embolism3 (1)aRespiratory complications are defined as respiratory failure, lower respiratory tract infection and symptomatic pleural effusion requiring drainage.Oncological outcomesHistopathological analysis of the operative specimens in the 215 patients revealed the following tumor types: adenocarcinoma (n = 169), squamous cell carcinoma (n = 22), high grade dysplasia (n = 17), adenosquamous carcinoma (n = 3), benign strictures only (n = 3) and spindle cell tumor (n = 1). In 3 patients, all initially diagnosed with adenocarcinoma, there was a complete pathological response to neoadjuvant chemotherapy whilst, in a further 2 patients, there was residual adenocarcinoma in lymph nodes only. The type of esophagogastric junctional tumour in 169 patients with adenocarcinoma was classified as follows: type I (n = 93), type II (n = 70) or type III (n = 6). All 6 patients with type 3 tumors had been preoperatively staged as type 2 tumours.Macroscopic tumour clearance was achieved in 193 out of 194 patients with pathological evidence of invasive malignancy. Residual microscopic disease was found at the proximal or distal resection margins in 11 patients (5%), all in association with positive circumferential resection margins and involved lymph nodes. Eighty eight patients (46%) were subsequently found to have tumor cells at or within 1 mm of the esophageal adventitia or the gastric serosal surface.The radicality of resection in relation to tumour infiltration and involved lymph nodes is shown in Table 3. The median lymph node yield in all patients was 12 (range 1–52). Both tumour stage and radicality of resection were independent predictors of overall survival on univariate analysis (Figures 1 &2).Table 3Pathology results from 194 patients undergoing transhiatal esophagectomy for invasive malignancy.N0N+R0R1R2% R0 resectionsT0325100%T135741198%T22338412068%T325521957125%T4151617%Figure 1Survival curves comparing overall survival for p (and yp) T0–2 tumours versus p (and yp) T3–4 tumours.Figure 2Survival curves comparing overall survival for R0 and R1–2 resections. There was only one R2 resection.Recurrence and survivalAll patients undergoing transhiatal esophagectomy for benign disease remain alive on follow up. Excluding the two in-hospital deaths, 79 patients (40%) who underwent esophagectomy for invasive malignancy have died on follow up. The causes of death are as follows: locoregional recurrence (n = 14), systemic metastases (n = 27), combination of locoregional recurrence and systemic metastases (n = 29), medical causes (n = 5), ongoing surgical complications (n = 1) and cause unable to be identified (n = 3). In total, 39% of patients developed recurrent disease during the period of study. The median survival for all patients undergoing transhiatal esophagectomy for invasive malignancy was 43 months and the one year and five year survival rates were estimated at 81% and 48% respectively (Figure 3). There was no difference in overall or disease free survival between patients with type I and II adenocarcinoma of the oesophagogastric junction.Figure 3Kaplan Meier survival curves for overall survival of 21 patients with benign disease and 194 patients with invasive malignancy undergoing transhiatal esophagectomy.DiscussionThis study has demonstrated that transhiatal esophagectomy can be associated with a low morbidity and a mortality of less than 1%. Although other units have reported similar results for transhiatal esophagectomy, several multicentre studies and national audits have shown that the mortality for all types of esophagectomy may exceed 10% [12-16]. It is recognised that high volume centres with a concentration of surgical, critical care and interventional radiological expertise achieve better outcomes. [17-19] The rationale for a transhiatal esophagectomy is the avoidance of a thoracotomy, thereby reducing the incidence of pulmonary complications, and the fashioning of a cervical anastomosis so that the clinical consequences of any anastomotic leak are minimized [12,13]. Critics of the transhiatal approach argue that there is a risk of blind intrathoracic injuries such as massive bleeding from the azygous vein, tracheal injury and episodes of cardiac instability resulting from retraction and surgical manipulation within the mediastinum. Case selection for transhiatal esophagectomy is crucial to prevent these problems and also to ensure adequate macroscopic tumor clearance for more proximally located esophageal tumors. It is the authors' policy that only patients with subcarinal tumors identified on preoperative imaging and confirmed by transhiatal dissection to above the proximal macroscopic extent of the tumor are suitable for the transhiatal approach. In the current series, only one patient required intraoperative conversion to a thoracotomy to obtain tumor clearance and 2 patients (1%) required reoperation for bleeding (both of these patients had active intrathoracic bleeding although none were associated with an azygous vein injury). Clinically apparent anastomotic leaks occurred in 6% of patients and all were managed successfully with conservative treatment. The data from this study supports the concept that a transhiatal esophagectomy in appropriately selected patients is safe and feasible.Surgeons who advocate a transthoracic approach argue that neglecting to perform a mediastinal lymphadenectomy risks leaving behind residual tumour, resulting in higher rates of locoregional recurrence and worse overall survival. [20-22] However, the additional value of formal mediastinal lymph node dissection remains controversial in Western patients, especially with the concept that lymph node involvement may reflect systemic micrometastatic disease and that extended resections will not alter the natural history of this disease. Reported differences in recurrence and survival may merely represent a stage migration effect due to an increased accuracy of histological staging. [2,23,24] Portale et al recently suggested that extended en bloc transthoracic resections were significantly associated with better survival rates of up to 50% compared to transhiatal resections and that this could not be ascribed to a stage migration effect. [21] R0 status (defined in this study as clear circumferential and longitudinal margins) is a recognized independent prognostic factor for survival. Advocates of a transthoracic esophagectomy have suggested that the transhiatal approach limits the ability to achieve an R0 resection [20-22]. Macroscopic tumour clearance was achieved in all but one patient in the current study. Longitudinal margin involvement, especially at the proximal margin, has been shown to independently impact on survival via increased loco-regional recurrence. The rate of positive longitudinal margins in this study was 5% which is in keeping with other published series [25]. The problem of a positive gastric resection margin at transhiatal esophagectomy has recently been addressed by DiMusto and Orringer [26]. They achieved a negative gastric margin in 98% of over 1000 patients treated. In the few patients who had a positive gastric margin, they found that 80% die with distant metastases, which would not be influenced by more extensive gastric resection, and, in about 20%, local tumor recurrence in the intrathoracic stomach was usually asymptomatic. They also demonstrated that adjuvant therapy for a positive gastric margin was usually unhelpful. A similar picture was seen in the current study with all five patients with involved distal resection margins developing systemic metastases.The role of circumferential resection margin (CRM) involvement is more controversial. Khan et al concluded that a positive CRM did not influence outcome. [27], but this has been disputed by other studies which suggested that it may independently predict survival [28]. One of these was performed by Maynard and colleagues who recently studied 242 patients undergoing esophagectomy and reported higher rates of local recurrence in patients with a positive CRM. Interestingly, there was no difference in CRM positivity when comparing different operative approaches [29].In our population, CRM involvement was encountered in 46% of patients with malignant disease, predominantly affecting those with T3 tumours, and this was the main limiting factor in achieving an R0 resection. R0 resection rates varied from 97–100% with T0/1 tumours to 0–17% for T3–4 tumours. In keeping with previous studies, R0 resections were significantly associated with improved overall survival and hence the group benefiting most from this operative approach would appear to be those patients with early (T1–2) tumours. [20-22]Advocates of more radical en-bloc transthoracic strategies argue that their approach may reduce rates of CRM involvement although this is yet to be proven [28]. Regardless of the operative technique, it is often difficult to obtain circumferential clearance due to the proximity of vital structures and the lack of any fascial boundaries. [13,28] The local recurrence rates in this study compare favourably to previous studies of both transhiatal and transthoracic esophagectomy [20,21,30,31]. Furthermore, the predominant pattern of recurrence was haematogenous metastatic disease (present in 70% of patients with disease relapse), mirroring the patterns seen with more radical en-bloc strategies [32]. These patterns of early systemic relapse were also noted by Orringer in his analysis of 2000 esophagectomy patients [33].To date, there has been only one randomised controlled trial comparing transthoracic and transhiatal approaches and this failed to show any significant differences in radicality of surgery or survival at the cost of increased postoperative morbidity in the transthoracic group. [34] Recent five year survival data from this trial have again failed to demonstrate a survival benefit for the transthoracic approach although a sub-group of patients with oesophageal cancer and 1–8 involved lymph nodes appear to have improved disease-free survival. This study did not include chemotherapy and overall five year survival rates were 34% (Transhiatal) and 36% (Transthoracic) with in-hopsital mortality of 2% and 7% respectively [35]. Other meta-analyses have attempted to compare the two approaches and have favoured the transhiatal approach in terms of early morbidity and mortality with no long term survival disadvantage [22,36]. Despite this evidence, it remains difficult preoperatively to select the appropriate operative approach for individual patients.Over the last few decades, the survival rates following esophagectomy have significantly improved, largely as a result of improvements in postoperative mortality. The one year survival rate of 81% in the current study for patients with invasive malignancy compares very favorably with the Western standard from the 1990s of 61%. [37] Furthermore, quality of life data suggests patients undergoing a transhiatal approach have fewer physical symptoms and better activity levels in the short term compared to the transthoracic approach although these differences become less evident by 1 year. [38] Several authors have emphasized the central role of surgery in achieving five year survival rates of approximately 50%. [21,30] It is increasingly recognized that there is an important role for oncological treatments in the perioperative management of esophageal and esophagogastric junctional cancer. The survival advantages associated with chemotherapy in both the MRC OEO2 and MRC MAGIC trials have significantly influenced surgical decision making in the UK. [3,39,40] The current series, which combined transhiatal esophagectomy with neoadjuvant chemotherapy in 42% of patients, has achieved equivalent five year survival results to Portale et al but with a greater preponderance of AJCC stage II and III disease. A complete pathological response was seen in 4% of patients receiving neoadjuvant chemotherapy and for many patients, there was little or no histological evidence of response. This emphasizes the need to identify potential responders prior to treatment, and also for the development of new chemotherapeutic agents. [21]The development of high volume centres within the UK and the increasing use of (neo)adjuvant therapies have undoubtedly improved both the short term surgical results as well as the long term oncological outcomes of these patients. In summary, we have shown that transhiatal esophagectomy is a safe approach in appropriately selected patients. Radical resections, postoperative complication rates and survival results were in line with data reported for traditional transthoracic approaches. Some units restrict transhiatal esophagectomy to patients deemed unfit for thoracotomy or to patients with very early tumours or, conversely, locally advanced tumours where the benefits of more radical resections may be limited. However, the authors suggest that transhiatal esophagectomy is at least a viable alternative with certain advantages in terms of post-operative recovery, and ever improving oncological outcomes especially when combined with chemotherapy.Authors' contributionsAD was primary author of the manuscript. MF performed some of the surgery, set up the database and assisted in data collection as well as drafting of the paper. AK, VP and AN were the primary data collectors and also performed the statistical analysis. DS helped conceive the study, performed some of the surgery and assisted in data collection. RM was the consultant in charge, performed the majority of the surgery and made alterations to the final draft prior to submission. All authors read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2531181.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2531181",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2531181\nAUTHORS: Katrin Sebastian, Andreas Borowski, Michael Kuepper, Karlheinz Friedrich\n\nABSTRACT:\nThymic stromal lymphopoietin (TSLP), a novel interleukin-7-like cytokine, triggers dendritic cell-mediated inflammatory responses ultimately executed by T helper cells of the Th2 subtype. TSLP emerged as a central player in the development of allergic symptoms, especially in the airways, and is a prime regulatory cytokine at the interface of virus- or antigen-exposed epithelial cells and dendritic cells (DCs). DCs activated by epithelium-derived TSLP can promote naïve CD4+ T cells to adopt a Th2 phenotype, which in turn recruite eosinophilic and basophilic granulocytes as well as mast cells into the airway mucosa. These different cells secrete inflammatory cytokines and chemokines operative in inducing an allergic inflammation and atopic asthma. TSLP is, thus, involved in the control of both an innate and an adaptive immune response. Since TSLP links contact of allergen with the airway epithelium to the onset and maintainance of the asthmatic syndrome, defining the signal transduction underlying TSLP expression and function is of profound interest for a better understandimg of the disease and for the development of new therapeutics.\n\nBODY:\nBackgroundAtopic asthma is a common inflammatory disorder of the airway epithelium characterized by tissue obstruction and remodeling, bronchial smooth muscle cell hyperreactivity to allergens and chronic bronchial inflammation. It classically involves allergen-driven T helper 2 (Th2) lymphocyte polarisation with coordinate production of interleukin (IL)-4, IL-5, IL-13 and granulocyte-macrophage colony-stimulating factor (GM-CSF), which are encoded in one gene cluster on chromosome 5q31-34 [1]. IL-4 and IL-13 are critically involved in the pathogenesis of allergic asthma by regulating IgE-production by B cells, inducing airway hyperreactivity and triggering key features of airway remodeling, whereas IL-5 is a key factor for eosinophilia [2,3]. Activation of IgE receptors on mast cells triggers the release of preformed vasoactive mediators such as histamine, the synthesis of prostaglandins and leukotrienes, and, via a positive feedback loop, expression IL-4 and IL-13 [2].Its apparent association with airway diseases has recently focussed interest on the novel IL-7-like cytokine thymic stromal lymphopoietin (TSLP). TSLP expression is increased in asthmatic airways and correlates with both the expression of Th2-attracting chemokines and with disease severity [4-6], indicating a link between TSLP and human asthma. Furthermore it was shown that experimental lung-specific expression of TSLP leads to transgene-induced allergic airway inflammation characterized by a massive infiltration of leukocytes, goblet cell hyperplasia, and subepithelial fibrosis, as well as by increased serum IgE levels [7].TSLP is a typical four-helix-bundle cytokine 140 amino acid residues in length and was first cloned in humans in 2001 [8-10]. The human TSLP gene is localized on chromosome 5q22, interestingly close to the gene cluster encoding several Th2-related cytokines such as IL-4, IL-5, IL-9, and IL-13 [7,11]. Human TSLP is produced by different cell types in atopic asthma, mainly by epithelial and smooth muscle cells and induces an inflammatory Th2 response. The TSLP receptor (TSLPR) is a heterodimeric cytokine receptor consisting of the IL-7 receptor alpha chain (IL-7Rα) and a TSLP-specific receptor chain with similarity to the common gamma receptor chain (γc). The TSLPR, also known as CRLF2, is expressed in heart, skeletal muscle, kidney and liver, but also on asthma-relevant dentritic cells [9,12]. In this review, the signal transduction around human TSLP in the cascade of events in the development of atopic asthma is discussed. We first describe the regulation of TSLP production in airway epithelial and other cells, then cover the TSLPR-mediated effects on TSLP target cells such as DCs and mast cells, and finally treat the DC-triggered onset of a specific Th2 response.Regulation of TSLP expressionIn the human airway system, fibroblasts, smooth muscle cells, epithelial cells and mast cells all have the potential to produce TSLP [14-18]. Airway epithelial cells (AECs) were found to have increased TSLP mRNA levels in human asthmatics [4]. Importantly, overexpression of TSLP in AECs induces experimental asthma in mice [7].TSLP expression is enhanced by different stimuli with relevance in asthma. Primary small airway epithelial cells (SAECs) produce biologically active TSLP in response to bacterial peptidoglycan, and lipoteichoic acid as well as to poly I:C (mimicking viral double-stranded RNA) [16]. IL-1β and TNF-α, two cytokines associated with pulmonary inflammation and strongly upregulated in the asthmatic lung [19,20] can, under appropriate conditions, induce human TSLP expression in normal human bronchial epithelial cells (NHBECs) [15,17], SAECs [16] and human airway smooth muscle cells (HASMCs) [18]. Similarly, TGF-β, IFN-β, IL-4, IL-13, and, in particular, a combination of TNF-α and IL-4 or IL-13 upregulate TSLP expression in NHBEs [17].It is established that rhinovirus and respiratory syncytial virus (RSV) can trigger exacerbations of asthma [21]. TSLP expression in human bronchial epithelial cells is stimulated by both viruses and an involvement of signal transduction through p38 and Jun kinase (JNK) has been demonstrated [22]. Stimulation of TSLP expression evoked by rhinoviral dsRNA and RSV proteins via toll-like receptors (TLRs) is synergistically enhanced by IL-4, indicating a contribution of JAK/STAT signalling [17]. The notion of cooperative signalling to TSLP gene transcription by cytokine and toll-like receptors is supported by the observation that tumor necrosis factor- (TNF-) α and IL-4 or IL-13 jointly drive TSLP expression in NHBECs, but none of the factors has a respective effect on its own. The induction of TSLP by combination of TNF-α and Th2 cytokines but not by the individual cytokines suggests that NFκB and STATs cooperate in transcriptional regulation of the TSLP gene [17].HASMCs which also act as effector cells in initiating or perpetuating airway inflammation [23,24], respond by TSLP release to stimulation with TNF-α and IL-1β both in vitro and in vivo [18]. Pharmacological inhibitors of ERK1/2 and p38, but not blockers of phosphatidylinositid-3 kinase (PI-3K) specifically suppress TSLP secretion induced by both factors individually or combination, suggesting that TSLP expression in HASMC is controlled via MAPK pathways [18]. Crosstalk of NFκB- and MAPK pathways is suggested by experiments in cells with mutated mediators NFκB and Ras which show a strong decrease in transcriptional activity of the human TSLP promoter [25].Studies employing deletion constructs of the TSLP gene promoter indicated that a DNA fragment extending from 3.74 to 3.86 kb upstream of the transcriptional start site contains a cis element required for transcriptional induction by IL-1β. Inspection of this ~120-bp sequence revealed consensus cognate elements for NFκB and IRF-1 as well as a putative AP-1 binding site [15]. Mutations in these motifs indicated that the induction of TSLP gene expression seen in cells stimulated with IL-1β is likely to be mediated through NFκB, whose subunits p65/p50 bind to the NFκB cognate motif in the human TSLP promoter. One of the major pathways for NFκB activation involves the phosphorylation of the inhibitor IκBα, which is followed by IκBα degradation and the subsequent migration of NFκB dimers (each monomer consisting of a p50 and a p65 subunit) from cytoplasm to the nucleus [26]. A dominant-negative mutant of IKKβ (IκB) inhibits IL-1β-mediated transcription of the TSLP gene [15].Since TLSP induction in the airway epithelium of asthmatics appears to be a associated with allergen contact, it is important to note that engagement of TLRs by allergene provocation activates NFκB [27]. TLR2, TLR3, TLR8, and TLR9 can all induce human TSLP expression in airway epithelial cells [15,17], suggesting that TSLP may become upregulated in the lung epithelium upon allergen challenge. In line with this hypothesis, we have recently observed that direct stimulation of lung epithelial cells with different allergens induces the expression of TSLP mRNA (Borowski et al., manuscript in preparation).Viral dsRNA is sensed, apart from TLR3, also by the recently identified cytosolic RNA helicases RIG-I and MDA5 [28,29]. Activation of TLR3, RIG-I, and MDA5 by dsRNA is transmitted to transcription factors NFκB and IRF-3, leading to transcriptional upregulation of pro-inflammatory genes and expression of type I interferons including IFN-β. siRNA experiments suggested that TSLP is directly induced by dsRNA in airway epithelial cells, and that the response is mediated by a pathway involving TLR3, NFκB and and IRF-3, but is independent of interferon signalling. Enhancement of dsRNA-dependent TSLP expression by IL-4 is significantly inhibited by siRNA targeting STAT6, supporting the notion of STAT6 as an important transcription factor in the control of TSLP expression [17].Very recent work showed that TSLP expression in the murine lung is influenced by peptidyl-propyl isomerase (PIN1), an important regulator of survival-promoting and proinflammatory cytokines in T-cells. Active PIN1 inactivates adenosine-uridine binding factor 1 (AUF1), whose function is to destabilize mRNA by interaction with adenosine-uridine rich elements. Since TSLP expression is blocked by a PIN1 inhibitor after challenging lung with allerges and the 3'-untranslated region of TSLP mRNA contain an AUF1 binding site, PIN1 is likely to be a modulator of TSLP expression in asthma at the posttranscriptional level [30].Activation of the TSLP receptor and intracellular signal transductionThe specific, low affinity TSLP receptor α chain (TSLPRα) is a member of the hematopoietic (type 1) cytokine receptor family. In combination with the IL-7Rα chain it forms the heterodimeric TSLP receptor (TSLPR) which, upon TSLP binding, transmits signals towards STAT activation and proliferation into the cell interior [8,9,31,32]. The TSLPRα chain has some atypical features for a type 1 cytokine receptor, both in its extracellular and intracellular region. The exodomain, for instance, lacks one of the four cysteine residues conserved within the receptor family, perhaps indicating a unique tertiary structure. Intracellularly, the TSLPRα lacks one of the two conserved sequence boxes present in other cytokine receptors that govern the interaction with Janus kinases (JAKs).Signal transduction emanating from the dimerized TSLPR is similar to signalling from the IL-7R, reflecting the overlapping utilization of the IL-7Rα chain by the two systems. The IL-7 receptor utilizes the γc chain as a dimerization partner for IL-7Rα, which recruits JAK3 via its box1 element. Ligand-induced crosslinking of both the TSLP and IL-7 receptors results in the activation of STAT5 and STAT3 and concomitant specific gene regulation [8,33,34]. However, unlike the IL-7R, the TSLPR appears to drive STAT activity via an uncommon mechanism without an involvement of Janus kinases. Evidence for this interpretation comes from experiments showing that no JAK phosphorylation was evoked by the activated TSLPR and that dominant-negative forms of JAK1 and JAK2 did not block TSLP-mediated STAT5 activation [35]. It was puzzling, however, that fusions of TSLPR cytoplasmic domain with the exodomains of the erythropoietin receptor or CD8 did activate JAK2 upon ligand-dependent homo-dimerization [36-38]. From results obtained with dominant-negative versions of Tec, a role of this cytoplasmic Src-related tyrosine kinase in the TSLPR-mediated activation of STAT5 was inferred [33]. Src-type mediators are also involved in proliferative signaling, TSLP-induced cell proliferation is blocked by the Src family inhibitor PP1 [34]. While both STAT5 activation and cell proliferation require the box1 domain of TSLPRα and IL-7Rα, the single intracellular tyrosine residue of TSLPRα receptor is critical only for proliferative signaling, but not for TSLP-dependent STAT5 activation [34]. Apart from STAT5, TSLP initiates STAT3 phosphorylation in murine DCs without the induction of any of the four Janus kinases (JAKs) [9,31,32] and activates STAT1 in murine pro B cells (A. Wohlmann and K. Friedrich, unpublished results). TSLP does not stimulate the activation of ERK1/2 and p70S6K [8]. Thus, neither the MAPK pathway nor the p70S6K pathway appear to be involved in the signal transduction pathway elicited by TSLP. Details of TSLPR signaling are far from being settled, but the present view is that Src type kinases are mediating the proliferative response and unknown (non-JAK) kinases are critical for STAT activation and, ultimately, regulation of target genes (Figure 1).Figure 1Structure and signal transduction of the heterodimeric TSLP receptor complex. For details see text.Activation of effector cells of the innate and adaptive immune system through TSLPCompelling evidence has been acumulated for a determinative role of TSLP in the initiation and maintenance of the allergic response in the context of atopic asthma [5,6]. Human TSLP is able to directly activate effector cells of the innate immune system like mast cells (MCs), which are known to play an important role in the pathogenesis of atopic diseases [39,40]. Functional receptors for TSLP are expressed in vivo on MCs infiltrating the bronchial mucosa of asthmatic patients as revealed by immunostaining of biopsy specimen [16]. In inflammatory conditions mimicked by the presence of IL-1 and TNF α, TSLP is a potent activator of MCs leading to the production of high levels of proinflammatory Th2 cytokines and chemokines such as IL-5, IL-13, IL-6, GM-CSF, CXCL8, and CCL1 [16] (Figure 2). Signaling pathways underlying this complex gene regulation have not been characterized yet.Figure 2Central role of TSLP in the orchestration of an asthmatic response upon contact of the airway epithelium with allergens or other challenging agents. Intercellular communication among different cell types via cytokines evokes activity of the native (bottom part) as well as the adaptive (upper part) of the immune system. For details see text.Apart from mast cells, the second main sentinel of the innate immune system is represented by DCs localized at the epithelial surface. DCs are also operative in the creation of a microenvironment that directs T cells towards Th1 or Th2 differentiation. A strong Th2 type response is typical in the context of the allergic syndrome. Human TSLP strongly activates immature CD11c+ DCs while it does not appear to have any direct biological effects on B cells, T cells, NK-cells or neutrophils [9,14,41]. TSLP induces DCs to up-regulate the expression of major histocompatibility class I and II and costimulatory molecules, including CD40, CD80, CD86. Importantly, it also strongly upregulates expression of the mRNA for OX40L, a member of the TNF superfamily that has been implicated in the initiation of Th2 cell responses [42-44]. TSLP also stimulates DCs to produce the Th2-attracting chemokines TARC (thymus and activation regulated chemokine, CCL17) and MDC (macrophage derived chemokine, CCL22) [14], as well as IL-8, IL-15 and eotaxin-2, clearly suggesting that TSLP-activated DCs may represent an initial key step in the development of allergic inflammation [34,45]. In the asthmatic bronchial mucosa, elevated expression of TSLP was also accompanied by and correlated with elevated expression of the CCR4 ligands TARC and MDC at the mRNA level [4]. As revealed by a comparative global transcriptome analysis of naive and TSLP treated DCs, TSLP does not stimulate DCs to produce the Th2-polarizing cytokine IL-4 and, at the same time, suppresses the anti-inflammatory cytokine IL-10 as well as interferon- (IFN-) γ [45].When TSLP treated DCs are stimulated with CD40 ligand, they induce the differentiation of CD8+ T cells into cytolytic effector cells which produce IFN-γ as well IL-4 and IL-13. Interestingly, expression of these ctokines has as before been considered mutually exclusive [46] (Figure 2).Indirect effects of TSLP on Th2 differentiation via OX40LOX40L, a member of the TNF superfamily has been identified as crucial mediator of Th2 cell responses driven by DCs [43,47]. In mice, blocking of OX40L by inhibitory antibodies inhibited the immune response induced by TSLP, indicated by reduction of cytokine secretion, Th2 inflammatory cell infiltration and IgE production [48]. It appears that OX40L and IL-4 act synergistically and sequentially in driving Th2 cell responses in cocultures of T cell and DCs activated by TSLP [10,45]. Interestingly, in the presence of IL-12, OX40L is unable to induce a Th2 cell response but rather directs T cell differentiation towards the Th1 phenotype, indicating a functional dominance of IL-12 over TSLP [45].Historically, CD4+ Th2 cells are defined as effector T cells with the capacity to produce IL-4, -5, -10, and -13 [49,50]. IL-4 and IL-13 are typical pro-inflammatory cytokines, but IL-10 does not appear to contribute to allergic inflammation in either humans or mice [51,52], but even suppresses allergic inflammation [53-55]. Importantly, dendritic cells activated by TSLP prime naive CD4+ T cells to differentiate into a particular subtype of Th2 cells that produce the classical Th2 cytokines IL-4, IL-5 and IL-13, but no IL-10. It is also remarkable that these special Th2 cells secrete very high levels of TNF-α [14]. TNF-α is prominent in asthmatic airways and genotypes that correlate with increased TNF-α secretion are associated with an increased risk of asthma [56]. Because of their unique profile of cytokine production, Th2 cells induced by TSLP-activated DCs have been designated inflammatory Th2 cells to discriminate them from the classical regulatory Th2 cells [10] (Figure 2).Th1 and Th2 cell differentiation is regulated by the key transcription factors T-bet for Th1 and GATA-3 and c-Maf for Th2. Th1 cells express high levels of T-bet but low GATA-3 and c-Maf, while Th2 cells show a reverse expression pattern, hence these transcription factors can be used as molecular markers for Th1 or Th2 cells [57]. CD4+ T cells primed by TSLP-activated DCs show the typical Th2 pattern high GATA-3 and c-Maf and low T-bet. However, IL-12 can override this Th2-specific gene regulation by inhibiting GATA-3 and c-Maf and strongly up-regulating T-bet [Ito et al. 2005]. Regulation processes behind this phenomenon may involve temporal IL-12-induced upregulation of the IL-12R signaling subunit (IL-12Rβ2) and concomitant signal transduction via STAT4, which has been observed in CD4+ cells upon activation of OX40 [58].TSLP has also been discussed as a player in the maintenance and regulation of Th2 memory cells. This interpretation comes from the finding that DCs activated by TSLP can induce an expansion of a CD4+ T cell subset expressing the prostaglandin D2 receptor (CRTH2), a property of human Th2 central memory T cells [59]. Interestingly, TSLP-activated DCs enhance the allergy-inducing properties of Th2 memory cells by up-regulating their expression of pro-allergic genes, particularly IL-17RB, the receptor for IL-25. IL-25, in turn, was shown to trigger the proliferation of Th2 memory cells and increase to the production of IL-4, IL-5, and IL-13, but not TNF-α or IFN-γ. These results suggest that IL-25 may costimulate the proliferation and further polarization of Th2 memory cells induced by TSLP-activated DCs [60].Importantly for the development of asthmatic symptoms, the activation of inflammatory Th2 cells through TSLP and their production of the inflammatory cytokines IL-4, IL-5, IL-13 and TNF-α triggers IgE production, eosinophilia, mucus production and fibroblast proliferation [61,62]. The effector mechanisms of atopic asthma ultimately involve IgE-coated mast cells that undergo degranulation upon contact with the allergen and induce an immediate response, leading to the symptoms of local inflammation and bronchospasm [63].ConclusionIn atopic asthma, many different agents such as viruses, bacteria and allergens can induce a TSLP-dependent inflammatory response, leading to an inappropriate activation of both the innate and the adaptive immune system. With regard to the innate branch of the response, TSLP acts on mast cells and dendritic cells as well as, according to recent results, natural killer cells. Mast cells play a prominent role in the development of asthmatic symptoms, because they secrete inflammatory cytokines in response to TSLP. The fact that mast cells also produce TSLP indicates a potential amplification loop by the action of mast cell-derived TSLP on epithelial DCs. A critical switch governed by DCs is the TSLP-induced expression of OX40L, a Th2 cell polarizing signalling molecule. A further emerging role of TSLP is the generation of Th2 memory T cells. The IL-25-mediated collaborative interactions between eosinophils/basophils and Th2 memory cells perhaps propagate a positive feedback loop between innate effector and adaptive immunity, leading to the amplification of allergic inflammation (Figure 2).List of abbreviations usedAEC: Airway Epithelial Cells; DCs: Dendritic Cells; GM-CSF: Granulocyte-Macrophage Colony Stimulating Factor; HASMCs: Human Airway Smooth Muscle Cells; IgE: Immunoglobulin E; IL: Interleukin; IFN: Interferon; JAK: Janus Kinase; JNK: Jun Kinase; MAPK: Mitogen Activated Protein Kinase; MCs: Mast Cells; NFκB: Nuclear Factor κB; NHBECs: Normal Human Bronchial Epithelial Cells; NK cells: Natural Killer Cells; PI-3K: Phosphatidylinositid 3-Kinase; SAECs: Small Airway Epithelial Cells; TLR: Toll-like Receptor; TNF: Tumor Necrosis Factor; TSLP: Thymic Stromal Lymphopoietin; TSLPR: Thymic Stromal Lymphopoietin Receptor; STAT: Signal Transducer and Activator of Transcription.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsKS and AB collected information from the literature and prepared iniatial versions of large parts of the manuscript. MK provided additional information from an immunological and clinical point of view and edited major parts of the manuscript. KHF conceived the overall organization of the manuscript, added extended sections of text and did the final editing. All authors read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2532686.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2532686",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2532686\nAUTHORS: Seung-Jae Lee, Joong-Seok Kim, Kwang-Soo Lee, Jae-Young An, Woojun Kim, Yeong-In Kim, Bum-Soo Kim, So-Lyung Jung\n\nABSTRACT:\nBackgroundSeveral studies have suggested that the specific stroke subtype may influence the presence of leukoaraiosis in patients with ischemic stroke. We investigated the association between stroke subtype and leukoaraiosis in Korean patients with ischemic stroke by MRI.MethodsThere were 594 patients included in this study that were classified as large artery disease, lacune and cardioembolic stroke. For large-artery disease, the analysis focused on the intracranial or extracranial location of the stenosis, and the multiplicity of the stenotic lesions. Leukoaraiosis grading was performed according to the Atherosclerosis Risk in Communities Study.ResultsThere was a significant association between leukoaraiosis and the stroke subtypes; the large-artery-disease group had a higher prevalence of leukoaraiosis than did the other groups (55.4% in the large-artery-disease group, 30.3% in the lacunar group and 14.3% in the cardioembolic group, P = 0.016 by chi-square test). On the multivariate linear regression analysis, age, the presence of hypertension, previous stroke and stroke subtype were independently associated with the presence of leukoaraiosis. In the sub analysis of the large-artery-disease group, the leukoaraiosis had a tendency to be more prevalent in the mixed and intracranial stenosis group than did the extracranial stenosis group (45.5% in the mixed group, 40.3% in the intracranial group and 26.9% in the extracranial group, P = 0.08 by chi-square test).ConclusionThe association of leukoaraiosis with large-artery disease in this study might be due to the relatively high prevalence of intracranial occlusive lesions in Korean stroke patients compared to other ethnic groups.\n\nBODY:\nBackgroundThe term leukoaraiosis (LA) refers to lesions of altered signal intensity on computed tomography (CT) and magnetic resonance imaging (MRI) in the periventricular and subcortical white matter. LA is found during the normal aging process, and in the patients with cerebrovascular disease. It also constitutes the core pathology of Binswanger's disease, a type of vascular dementia. The association of LA with lacunar infarcts rather than territorial infarcts is well documented [1-7]. However, most prior studies have been based on CT findings, not MRI, and reported from western countries.In the present study, we analyzed the association between stroke subtype and LA in Korean stroke patients using MRI.Methods1. PatientsWe initially included 963 consecutive acute ischemic stroke patients admitted to the neurology department from July 2003 to June 2007. All patients underwent detailed clinical evaluation including laboratory tests, chest radiography, transcranial Doppler study, electrocardiography and 24 hour Holter monitoring. In addition, transthoracic echocardiography and brain magnetic resonance imaging (MRI), contrast-enhanced MR angiography (MRA) and/or cerebral angiography were obtained. All results from the evaluations were analyzed according to the diagnostic criteria for stroke mechanisms and etiology based on the TOAST subtype classification system [8]. Among the initial patients, 369 were categorized into stroke of undetermined etiology (mean age ± SD, 68.1 ± 10.2; age range, 31–87 years) and were excluded from the study: of those patients, 245 patients were classified as stroke of two or more potential etiology (164 with lacune plus large-artery disease, 54 with large-artery disease plus cardioemobolism and 27 with lacune plus cardioembolism), 21, 76 and 27 patients were classified as groups of negative evaluation, incomplete evaluation and other determined etiology, respectively.Finally, 594 cases with large artery disease (297 patients), lacune (193 patients) and cardioembolic stroke (104 patients) were enrolled in this study. The ethics committee at our institution approved the study protocol, and all subjects provided written informed consent.2. Risk factor evaluationThe clinical information included age, gender, history of hypertension (defined by the use of an antihypertensive agent before admission or a systolic pressure > 140 mmHg or diastolic pressure > 90 mmHg demonstrated on repeated examinations at least one month after presentation with a stroke), diabetes mellitus (defined as a fasting blood glucose level > 126 mg/dl or a history of being treated for diabetes mellitus) and hyperlipidemia (defined as a total cholesterol level > 200 mg/dl or a low-density lipoprotein cholesterol > 130 mg/dl at the time of presentation or a history of treatment). In addition, regular cigarette smoking, a previous history of ischemic stroke and heart disease (defined as a known history or clinical demonstration of any heart disease, including myocardial infarction, angina pectoris, congestive heart failure, or arrhythmia) were noted3. MR imaging and LA gradingAll patients enrolled underwent conventional MRI on a 1.5-T system (Signa 1.5-T TwinSpeed, General Electric Medical Systems and Archieva 1.5-T, Philips Electronics) within 7 days of the stroke onset. The conventional MRI consisted of transverse T2/T1-weighted, fluid-attenuated inversion recovery (FLAIR) sequences and sagittal T1 with 5-mm-thick slices. Diffusion-weighted imaging was obtained in the transeverse plane using a single-shot echoplanar, spin-echo pulse sequence. A three-dimensional time of flight MRA of the intracranial arteries and contrast-enhanced MRA of the head and neck were also performed on the same system using a neurovascular coil.LA was defined as a periventricular white matter lesion with hyperintensity on T2- weighted and FLAIR images and without prominent hypointensity on T1-weighted images. The LA grading was according to the Atherosclerosis Risk in Communities (ARIC) study [9,10]. Three trained neurologists and two neuroradiologists, blinded to patient data and stroke subtype, graded the LA by consensus. When evaluating the WMH, new (high signal on diffusion-weighted image) and old (definitely low signal on T1-weighted image) infarcts were excluded. If one or both sides of the brain were focally abnormal, estimates were based on the uninvolved side with the principle of symmetry assumed.4. Patterns of stenotic lesions in large artery diseaseThe evaluation of the arterial stenosis was performed by the investigators and > 50% of signal loss on the MRA was considered to be a \"significant stenosis\" for the classification of stroke subtype and the categorization of stenosis pattern. The locations of significant stenosis were categorized as located in the intracranial or extracranial arteries. For the internal carotid artery, an intracranial location was defined when the stenotic lesion was distal to the ophthalmic artery. For the vertebral artery, the differentiation was made at the point where the artery pierced the dura at the level of the foramen magnum. The intracranial extent of the stenosis included up to the M2 of the middle cerebral artery (insular segment which terminates at the circular sulcus of the insula before the operculum) and the A2 segment of the anterior cerebral artery (ascending segment with inferior forward convexity) in the anterior circulation, and the P2 segment of the posterior cerebral artery (ambient segment which extends from the junction between the posterior communicating artery and the posterior cerebral artery to the posterior aspect of midbrain). The area of the stenotic lesion was divided into intracranial or extracranial and anterior or posterior circulation. The stenoses were described as single or multiple according to the number of the areas involved with the arterial stenosis.According to the distribution and pattern of the stenotic lesions, the patients with large-artery disease were categorized as intracranial, extracranial and a mixed (intracranial plus extracranial) group, and as multiple- or single-stenosis groups.5. Statistical analysisWe classified persons with leukoaraiosis of grade 3 or higher as having \"LA\" and of grade 2 or lower as having \"little or no LA\" [10]. The presence or absence of LA was compared between the patients with risk factors and the patients without, the stroke subtypes, the multiple stenosis and single stenosis groups, and among intracranial, mixed and extracranial stenosis groups by chi-square test or independent t-test. In addition, we compared the severity of LA between the groups of different stroke subtypes by one-way ANOVA test and LSD multiple comparison test. Multiple linear regression analysis was used to determine the factors considered independently associated with leukoaraiosis. Values of P < 0.05 were considered statistically significant.ResultsOf the 594 patients, 342 (57.6%) were men and 252 (42.4%) were women (mean age ± SD, 66.8 ± 12.1; age range, 27–97 years). In study population, distinct white matter changes were present in 307 patients (51.7%). The association of the presence of LA to age was statistically significant; the LA group had higher age at onset of stroke than the \"little or no LA\" group. The LA was more frequently observed in female gender, the patients with hypertension and a history of previous ischemic stroke than in the patients without this history (P < 0.05). There was no significant association between the LA and diabetes mellitus or hyperlipidemia; a negative correlation was found with the smoking status and the presence of heart disease (table 1).Table 1Characteristics of stroke patients by leukoaraiosis.LeukoaraiosisCharacteristicsPresent (n = 307)Absent (n = 287)P valueAge (yr)72.1 ± 9.861.1 ± 11.9< 0.001Gender0.001 Male (n = 342)157 (45.9)185 (54.1) Female (n = 252)150 (59.5)102 (40.5)Hypertension (n = 383)229 (59.8)154 (40.2)< 0.001Diabetes mellitus (n = 236)133 (56.4)103 (43.6)0.390Hyperlipidemia (n = 193)95 (49.2)98 (50.8)0.228Heart disease (n = 105)44 (41.9)61 (58.1)0.018  Atrial fibrillation (n = 78)34 (43.6)44(56.4)0.125Regular cigarette smoking (n = 153)57 (37.3)96 (62.7)< 0.001Previous ischemic stroke (n = 87)58 (66.7)29 (33.3)0.002Stroke subtype0.016 Large artery disease(n = 297)170 (57.2)127 (42.8)0.031a0.006b Lacunar(n = 193)93 (48.2)100 (51.8)0.198b Cardioembolic (n = 104)44 (42.3)60 (57.7)Values represent mean ± SD and number of patients with percentages in parenthesisThe two groups were compared by two-sample t-test for continuous variables and chi-square test for nominal variables.a Comparison is with lacunar group, based on the Chi-square test.b Comparison is with cardioembolic groupAmong the 307 patients with LA, 170 patients (55.4%) were in the large-artery-disease group, 93 patients (30.3%) in the lacunar group and 44 patients (14.3%) in the cardioembolic group (P = 0.016 by chi-square test, table 1). In addition, there was overall a significant association between the LA grade and the stroke subtype: the large-artery-disease group had more severe LA disease than did the other groups. Although the lacunar group tended to have a higher LA grade than did the cardioembolic group, there was no significant difference in the LA grade in comparisons between the two groups (table 2). On the multivariate linear regression analysis, using the variables (age, gender, the presence of hypertension, diabetes mellitus, hyperlipidemia, smoking, previous stroke, and stroke subtype), age, the presence of hypertension, previous stroke and stroke subtype were independently associated with the presence of LA (Table 3).Table 2Leukoaraiosis grade by stroke subtype.Stroke subtypeLarge artery diseaseLacunarCardioembolicP value(n = 297)(n = 193)(n = 104)LA grade(mean ± SD)3.37 ± 1.982.86 ± 1.952.54 ± 1.71< 0.001LSD, P value0.004a, < 0.001b0.170bP value refers to the overall association between leukoaraiosis grade and stroke subtype, computed from one-way ANOVA test.a Comparison is with the lacunar group, based on LSD multiple comparison tests.b Comparison is with the cardioembolic groupTable 3Multiple linear regression analysis for the relationship between leukoaraiosis and potential confounding variables.tpAge11.130< 0.001Gender1.0560.292Hypertension4.132< 0.001Diabetes mellitus0.6110.541Hyperlipidemia- 0.1170.907Smoking0.5500.583Previous stroke2.4400.015Stroke subtype2.4490.013Data are t (p value) of the correlation.aR2 = 0.503In the large-artery-disease group, there was a borderline association between the LA and the stenotic areas; the LA was most prevalent in the mixed stenotic group, next was the intracranial stenosis group and the extracranial stenosis group was the least affected. In addition, a more LA was observed in patients with multiple arterial stenoses than in patients with single arterial stenosis (table 4).Table 4Association between the pattern of stenosis and leukoaraiosis in large-artery-disease group.LeukoaraiosisPresent (n = 170)Absent (n = 127)P valueDistribution of stenosis0.080 Mixed (n = 101)46 (45.5)55 (54.5)0.245a0.023b Intracranial (n = 144)58 (40.3)86 (59.7)0.070b Extracranial (n = 52)14 (26.9)38 (73.1)Number of stenotic lesions0.019 Multiple (n = 155)71 (45.8)84 (54.2) Single (n = 142)47 (33.1)95 (66.9)Values represent number of patients with percentages in parenthesisThe two groups were compared by chi-square test.a Comparison is with intracranial groupb Comparison is with extracranial groupDiscussionAlthough the pathophysiology of LA remains speculative, there is evidence to suggest that LA may be linked to cerebral ischemia [11-24]. Selective injury to the cerebral white matter has been noted in a limited number of human conditions characterized by hypoxia/ischemia of the brain such as carbon monoxide poisoning and therapeutic occlusion of the internal carotid artery [11-14]. It has been assumed that the ischemic insult, responsible for LA, results from the vulnerable nature of the long penetrating end-arteries that feed the deep white matter [15,16]. LA has been associated with increasing age, arterial hypertension and other cerebrovascular risk factors [17-22]. In addition, white matter lesions similar to LA can be induced in rats or gerbils by ligating the bilateral common carotid arteries [23,24].Although the previous studies of risk factors for LA have shown different results, advanced age and hypertension have been consistently reported to be highly associated with LA [1]. One recent study indicated that diabetes mellitus seems to be a risk factor for progression rather than new LA development [7]. In our study, similar to previous studies in other ethnic groups, LA was associated with age, hypertension and previous history of ischemic stroke. However, it had no relation with hyperlipidemia and diabetes mellitus. In addition, there was a negative correlation between regular cigarette smoking and LA, as previously reported in one study [21]. However, the relationship between smoking and brain disorder is controversial, and the negative association with LA shown in this study should be discussed in further future study. These differences between the results of previous researches may be due to study variations in patients and/or the definitions of LA used by different investigators.The LA was more frequently observed in the large-artery-disease group than the other subtypes. In addition, the large-artery-disease group had more severe LA than did the other group. The cardioembolic group had the lowest prevalence of LA although there was no significant difference when compared to the lacunar group. This finding suggests that the hypoperfusion that results from large-artery occlusion might be more important to the progression and aggravation of LA. This is supported by the fact that the periventricular white matter, vulnerable to LA, was the distal irrigation field or border zone; these areas are prone to ischemia under conditions of moderate blood flow deficits.However, the results of our study are in general not consistent with prior reports. In most previously reported studies, LA was strongly associated with lacunar strokes rather than non-lacunar, territorial strokes [1-7]. In addition, an inverse correlation between high-grade (> 50%) stenoses of the extracranial arteries and white matter changes has been reported [1,21,25-27]. These observations have been explained by the hypothesis that reduced blood perfusion, in the patients with high-grade stenoses of the extracranial carotid arteries, might alleviate damage to intracerebral arteries by decreasing tensile stress on the arterial walls [21,28]. However, these studies did not divide the territorial infarcts into large artery disease or cardioembolic strokes. Moreover, most prior studies were from western countries, and the results were based on CT findings not MRI, which is better for assessing white matter lesions.In the present study, most of the patients in the large-artery-disease group had intracranial stenotic lesions (82.5%); extracranial lesions were uncommon. The LA tended to be more frequent in patients with intracranial stenoses than in the patients with extracranial stenoses alone. The intracranial location for cerebrovascular atherosclerosis is characteristic of strokes in the Asian population [29-31]. Therefore, the more prevalent LA in patients with large-artery-disease compared to the other subtypes in this study might be explained by the high prevalence of intracranial stenoses in Korean stroke patients.Unlike extracranial stenoses, more prevalent in Caucasians, intracranial stenoses may have a different pathogenesis contributing to the development and progression of LA. First, atherosclerotic stenoses of intracranial arteries can directly occlude the orifice of numerous small perforators penetrating into the deep brain parenchyma. The occlusion of the small perforators promotes extensive hypoperfusion of the periventricular region that is vulnerable to ischemia. Second, the periventricular border zone, under ischemic conditions induced by the stenoses of intracranial arteries, might have less opportunity to be compensated by blood flow via major collateral channels such as the anterior or posterior communicating artery, which can be recruited without difficulty in stroke patients with stenosis of extracranial vessels. Third, compared with the emboli from the stenotic lesions of extracranial arteries, those from an intracranial location might not easily be cleared away by travel along with the blood flow, probably because of the close proximity to the perfused brain tissues.The limitations of this study include the following. One methodological problem is that our results are based on a cross-sectional sample. The longitudinal effect of large-artery atherosclerosis on periventricular white matter changes could not be accurately assessed in this study. In addition, a selection bias might have been present because 369 (38.3%) of the 963 patients were excluded due to the diagnosis of a stroke of undetermined etiology or other determined etiology; we used very strict criteria for the classification of subjects into specific stoke subtypes. For example, suspected large artery disease with < 50% signal loss of the proximal artery on MRA, or supratentorial subcortical infarcts of < 1.5 cm with > 50% stenosis of the parent artery were classified as strokes of undetermined origin. The application of this criterion would increase the specificity and lessen the likelihood of misclassification of patients in the other categories [8]. Moreover, the quantitative analysis of LA by a 0-to-9 grading system might not reflect an accurate estimation of the LA severity. Additional grading systems including volumetric methods for LA are needed.ConclusionThe results of this study showed that LA is significantly associated with large-artery-disease rather than other stroke subtypes including small vessel disease in Korean stroke patients. The differences between our study and previous reports might be due to the high prevalence of intracranial occlusive lesions in patients with cerebrovascular atherosclerosis in the Asian population. Different from the high-grade stenoses of extracranial arteries, the stenoses of intracranial arteries might induce extensive white matter changes by interrupting blood flow to the periventricular border zone vulnerable to ischemia.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsSJL and JSK conceived and coordinated the study, analysed the data and drafted the initial manuscript. All authors were involved in initial literature search and collection of data. Review of initial manuscript for major intellectual content was done by SJL and JSK. All authors read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2532690.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2532690",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2532690\nAUTHORS: Noboru Minakawa, Gabriel O Dida, Gorge O Sonye, Kyoko Futami, Satoshi Kaneko\n\nABSTRACT:\nBackgroundTo combat malaria, the Kenya Ministry of Health and nongovernmental organizations (NGOs) have distributed insecticide-treated nets (ITNs) for use over beds, with coverage for children under five years of age increasing rapidly. Nevertheless, residents of fishing villages have started to use these bed nets for drying fish and fishing in Lake Victoria. This study investigated the extent of bed net misuse in fishing villages.MethodsSeven fishing villages along the lake were surveyed to estimate how widely bed nets were being used for fishing and drying fish. Villagers were asked why they used the bed nets for such purposes.ResultsIn total, 283 bed nets were being used for drying fish. Of these, 239 were long-lasting insecticidal bed nets (LLIN) and 44 were non-long-lasting insecticidal bed nets (NLLIN). Further, 72 of the 283 bed nets were also being used for fishing. The most popular reasons were because the bed nets were inexpensive or free and because fish dried faster on the nets. LLINs were preferred to NLLINs for fishing and drying fish.ConclusionThere is considerable misuse of bed nets for drying fish and fishing. Many villagers are not yet fully convinced of the effectiveness of LLINs for malaria prevention. Such misuses may hamper the efforts of NGOs and governmental health organizations.\n\nBODY:\nBackgroundThe World Health Organization (WHO) announced the Roll Back Malaria (RBM) movement in 1998, with the goal of decreasing malaria deaths by half by 2010 [1]. Several field trials demonstrated that insecticide-treated nets (ITNs) are effective in reducing malaria-related mortality in sub-Saharan Africa [2]; thus, ITNs have become a major tool in RBM. In Kenya, ITNs have been mainly distributed to pregnant women and children under five years of age, either free of charge or at subsidized prices, through programmes of the Kenya Ministry of Health and nongovernmental organizations (NGOs) [3,4]. Consequently, ITN coverage for children under five years of age has increased rapidly from 7% in 2004 to 67% in 2006; this increase has been associated with a 44% reduction in malaria deaths [5].Nevertheless, a study in western Kenya found that 30% of bed net recipients did not adhere to net use [6,7]. Net use tends to decrease during hot weather. Further, ITNs are sometimes used for other purposes such as wedding dresses or fishing in Zambia [8]. Bed nets have also been observed being used for drying a small zooplanktivorous Dcyprinid (Rastrineobola argentea, called \"omena\" in the local language) in fishing villages in the Kenyan part of the Lake Victoria basin (Figure 1), where malaria is endemic. Traditionally, these fish have been dried on papyrus sheets. However, the extent of bed net misuse for this purpose is unknown. The widespread misuse of the nets might hinder the RBM goal. Thus, this study investigated how widely bed nets were used for fishing and drying fish in villages along Lake Victoria in western Kenya.Figure 1Omena fish spread on bed nets on a beach by Lake Victoria.MethodsStudy areaThe study was conducted in seven major fishing villages in the Gambe West sub-district of Suba District, western Kenya. The area is approximately 76.6 km2. Most residents in the sub-district depend on fishing and traditional small-scale farming. The primary targets of the local fishery are Nile perch (Lates niloticus), Nile tilapia (Orechromis niloticus), and omena. Omena and Nile perch each accounted for 43% of the catch in Lake Victoria during the period of 1980 to 2005 [9]. Although Nile perch accounted for > 90% in volume of Kenya's total fish exports during the period of 1985 to 2005 [9], omena is an important protein source for locals [10].Two rainy seasons occur annually from approximately March to June and October to November, but the periods vary by year. Malaria is the leading cause of morbidity and mortality of children in the region [11]. Three species of vectors are known: Anopheles arabiensis, Anopheles gambiae and Anopheles funestus [12].Bed net surveyEach village had its own fish-landing beach, and nearly all captured omena fish were spread out and dried on the beach. The beaches were visited three times in early morning during \"young\" moon periods in February and March 2008 (during the rainy season). Omena fishing is not active during full moon periods. Fishermen use lamps to attract omena towards the boats during the night; this method is not effective under a full moon.When a sheet for drying fish was found, its material was categorized as papyrus, fishing net, bed net, or other. Bed nets were further categorized as long-lasting insecticidal bed net (LLIN) or non-long-lasting insecticidal bed net (NLLIN). NLLINs include nets with and without periodical insecticidal treatments. The size of each sheet in square metres was measured using a tape measure with the permission of the owner. When a sheet had been created from multiple bet nets or fishing nets, the number of nets was counted. Bed nets and fishing nets were measured separately when a sheet consisted of both materials.To determine whether villagers preferred LLINs or NLLNs for drying fish, information on the availability of both types of bed net in the area was necessary. This background information was obtained from a previous survey that was designed to estimate bed net coverage in the study area in August 2007 (Dida, unpublished data). In total, 111 houses were visited, and the numbers of LLINs and NLLNs were counted. The sources of bed nets (i.e., stores, NGOs, or health facilities) and the number of residents in each house were also recorded.Owners of fish-drying sheets that consisted of a bed net were asked the following questions: where and when the bed net was acquired, whether any bed nets were currently used in the house, why bed nets were used for drying fish, whether bed nets had ever been used for fishing in the lake, and when they started using the bed nets for drying fish. The investigated sheets were numbered to avoid duplicating data collection during a subsequent visit.In June 2006, an NGO distributed LLINs mainly to children in the villages. This NGO provided information on the number of LLINs distributed to each village. The interviewees were also asked whether the NGO provided the LLINs in use on the beaches.Statistical analysisThe total sizes of bed nets, fishing nets, and papyrus sheets were estimated for each beach. The mean total sizes were compared using repeated-measures analysis of variance (ANOVA). The Tukey-Kramer honestly significant difference (HSD) test was used for post-hoc multiple comparisons.A paired t-test was used to test the difference in the number of LLINs and NLLINs used for drying fish. The values were log-transformed because of heteroscedasticity. The proportion of LLINs to NLLNs was calculated for each beach and also obtained for houses. These values were arcsine-transformed and then compared between beaches and houses using a chi-square test to determine whether either type of bed net was used preferentially for drying fish.The difference in number between the types of bed nets used for fishing was compared using a Wilcoxon test because transformation did not stabilize the variances. To determine whether either type of bed net was preferably used for fishing, the proportion of LLINs to NLLNs used for fishing was compared with that for drying fish using a paired test.A paired t-test was also used to compare the number of bed nets obtained from NGOs and health facilities with that of bed nets acquired from stores. The proportion of nets from NGOs or health facilities to those from stores was compared between the beaches and the houses using a chi-square test. The significance level was 0.05 for all tests.ResultsThe total number of sheets used for drying fish was 166 at the seven fishing beaches, and the total area was 8295.8 m2 (Table 1). Nearly half of the sheet area was either bed nets or fishing nets; papyrus sheets made up only 9.5% of the sheet area. Bed nets accounted for 15.0 to 83.8% of the total sheet area among the beaches. The repeated-measures ANOVA indicated that the mean sheet area varied significantly among the materials (F = 4.55; df = 2, 20; P = 0.034). The post-hoc multiple comparisons indicated that the area of fishing nets was significantly greater than that of papyrus sheets, but the differences between bed nets and the other materials were insignificant.Table 1Total and mean areas (square metres per village) of bed nets, fishing nets, and papyrus sheets used for drying fish (n = 7).MaterialArea%Mean (SE)Bed nets3686.844.4515.8 (82.7)Fishing nets3770.845.5550.5 (116.5)Papyrus sheets788.19.5112.6 (75.1)Other500.67.1 (4.9)Total area8295.8100.0-Of 166 sheets, 87 consisted of 283 bed nets, of which 238 were LLINs and 44 were NLLINs (Table 2). The paired t-test revealed that significantly more LLINs were found on the beaches compared with NLLINs (t = 7.11, df = 12, P < 0.001). In total, 220 bed nets were found in 111 houses. Of these, 145 bed nets were LLINs, and 75 were NLLINs (Table 3). The proportion of LLINs to NLLINs used for drying fish was significantly greater than that of LLINs to NLLINs in the houses (chi-square test: χ2 = 23.50, P < 0.001). Of 220 bed nets, 87 were obtained from stores and 130 were from NGOs or health facilities (the information on three nets was not available). The mean numbers of total residents and children under five years of age were 3.6 and 0.6 per house, respectively.Table 2Numbers, percentages, and means (per village) of bed nets used for drying fish and fishing (n = 7) and their sources.Number%Mean (SE)Bed nets used for drying fish LLIN23984.534.1 (8.2) NLLIN4415.56.3 (1.8) Total283100.0-Bed nets used for fishing LLIN6894.49.7 (2.4) NLLIN45.60.6 (0.3) Total72100.0-Source of bed nets NGOs or health facilities23984.534.1 (7.8) Stores4415.56.3 (2.2) Total283100.0-Year when bed nets were acquired 200820.7- 200719067.1- 20067426.1- 2003, 2004 and 2005176.0- Total283100.0-Table 3Numbers, percentages, and means (per house) of residents and bed nets in houses (n = 111) and their sources.Number%Mean (SE)Bed nets in houses LLIN14565.91.3 (0.1) NLLIN7534.10.7 (0.1) Total220100.02.0 (0.1)Source of bed nets NGOs or health facilities8740.1- Stores13059.9- Total217*100.0Residents in houses Children under 5 years of age7017.40.6 (0.1) Persons above 5 years of age33482.63.0 (0.1) Total404100.03.6 (0.2)*Information on three bed nets was not available.Among the bed nets found on the beaches, 72 (24.5% of the total nets) nets had been used for fishing. Of these, 68 were LLINs and four were NLLINs. Significantly more LLINs than NLLINs were used for fishing (Wilcoxon test: χ2 = 6.52, P = 0.013). The proportion of LLINs that were used for fishing was significantly greater than that of LLINs that were used for drying fish (t = 2.76, df = 12, P = 0.04).In total, 239 (84.5%) bed nets were obtained either free of charge or at subsidized prices from NGOs and local health facilities, and only 44 (15.5%) nets were purchased at stores. The mean number of bed nets obtained from NGOs or health facilities was significantly greater than that of the bed nets obtained from stores (t = 8.84, df = 12, P < 0.001). The proportion of nets obtained from NGOs or health facilities to those obtained from stores was significantly greater for the beaches than for the houses (chi-square test: χ2 = 38.33, P < 0.001). Of 283 bed nets found on the beaches, 74 (26.1%) and 190 (67.1%) nets were acquired in 2006 and 2007, respectively.All 87 owners of bed nets found on the beaches were interviewed as to why they used the bed nets for drying fish. Of these, 85 owners answered the interview questions. The most popular reasons were because fish dried faster and bed nets were cheap or free (Table 4). Of the 85 owners, only seven were not using bed nets in their houses. Most owners started using bed nets for drying fish in 2006 and 2007.Table 4Reasons for using bed nets for drying fish and the year that this practice was started.Number%Reasons for using bed nets for drying fish Fish dry faster on bed nets6475.3 Inexpensive3844.7 Fish do not stick to bed nets2529.4 Fish dry straight on bed nets1720.0 No other materials for drying fish1618.8 Easy to obtain from NGOs1517.6 Have enough bed nets1416.5 Original colour of fish is retained on bed nets89.4 Strong67.1 Other78.2 Number of interviewees85-Year when started to use bed nets for fish drying 200833.7 20073239.0 20064352.4 200544.9 Number of interviewees82100.0The single NGO distributed 1040 LLINs in six villages (the number of bed nets distributed in one village was not available), of which 170 (16.3%) were being used for drying fish. Among the villages, the percentages of bed net used for drying fish ranged from 5.9 to 43.3%. Of 239 LLINs found on the beaches, 71.1% were from that particular NGO.DiscussionA considerably large number of bed nets were used for drying fish and fishing in the study area adjacent to Lake Victoria. Although the misuse of bed nets for fishing has been reported from Zambia without details [8], their use for drying fish was previously unknown. The traditional method of using papyrus sheets for drying fish was no longer popular in the study area and had been replaced with the method using bed nets and fishing nets.The interviews with bed net owners suggested that bed nets have clear advantages over papyrus sheets for drying fish. Whereas the price of a papyrus sheet ranged between 150 and 200 Kenya shillings (Ksh), a bed net could be obtained from an NGO free of charge or from local health facilities at subsidized prices (usually 50 Ksh). Bed nets were readily available from these organizations, but papyrus sheets were only available in the weekly market in the major local town. In fact, nearly 85% of the bed nets found on the beaches were from NGOs and local health facilities. The villagers also indicated that fish dried faster on the bed nets, which provided greater aeration when laid on grass than did papyrus sheets. They also noted that the fish dried straighter on bed nets, which increased the commercial value of the fish.A larger proportion of LLINs than NLLINs was used for drying fish than shown by the background information from the houses. This suggests that villagers preferred LLINs because the materials used for LLINs are stronger than those used for NLLINs and are more suitable for use outdoors. Moreover, approximately one-fourth of the bed nets found on the beaches were also being used for fishing in the lake, and the proportion of LLINs used for fishing was greater than that for drying fish. Only four NLLINs were being used for fishing. This is reasonable because fishing requires stronger materials than does fish drying. Although LLINs are weaker than real fishing nets, they are much cheaper or free and are at least strong enough to catch small fish such as omena. After fishing, the nets can be used for drying fish while the nets are also being dried on the beaches. For villagers who buy nets at subsidized prices, the use of LLINs as a disposable fishing net must be cost effective, although LLINs used for fishing purposes wear out much faster than those used inside the home. Consequently, as NGOs and health facilities distribute more LLINs, more LLINs may be used for fishing and fish drying.Over 70% of LLINs found on the beaches were from the single NGO; > 15% of the nets distributed by that NGO were used on the beaches, even though the nets were mainly provided to children. The interviews clearly indicate that misuse of the nets started in the period when the Kenya Ministry of Health and NGOs began distributing LLINs. Although data were unavailable, it seems that LLINs were not popular in this area before they began distributing the nets.The proportion of bed nets obtained from NGOs or health facilities to those from stores was greater for the beaches than for the houses. This suggests that villagers preferentially use free or inexpensive bed nets for fishing purposes because the practice does not cost them. For some villagers, fishing might be more important than protection from mosquitoes. Alternately, villagers concerned for their health might have bought bed nets for house use before the NGO's distribution; therefore, the nets provided by the NGO could have been extra nets.Nearly 15% of the interviewees answered that they used bed nets on the beaches because the nets were extra, and > 80% reported that they had bed nets in their houses. However, it is difficult from this survey to conclude whether there were enough bed nets to cover all residents in the houses because information on the number of residents and bed nets in individual houses was not available. However, a previous survey of houses found means of 2.0 bed nets and 3.6 residents per house. Although the house survey was conducted in 2007 (six to seven months before the beach survey), the villagers had started to use most of the bed nets found on the beaches in 2006 and 2007. This suggests that there were not enough bed nets to cover all residents when they started to use the nets for fishing.Bed nets may be reused on beaches after being used in houses considerably. However, the results from this study deny this possibility. Over 90% of bed nets found on the beaches were not older than two years approximately, and nearly 70% them were likely newer than one year old. LLINs were designed to last more than two years. Considering these and the timing that misuse of the nets started in the period when the Kenya Ministry of Health and NGOs began distributing LLINs, it is suspected that the bed nets had been little used in houses. Worn-out bed nets with holes are not suitable for fishing, at least.Because omena fishing is important in villages along the lake [10], misuses of bed nets must be common throughout the lake region. Breeding habitat for malaria vectors is closely associated with lakeshores [12,13], and malaria transmission is high near the lake. The misuses of bed nets must be a substantial drawback for malaria-control programmes involving LLINs in the region.ConclusionThe misuse of bed nets for drying fish and fishing is considerable in the study area. Many villagers are not yet fully convinced of the effectiveness of LLINs for malaria prevention. Misuses of bed nets may hamper the efforts of NGOs and governmental health organizations for malaria prevention.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsNM initiated the study and drafted the manuscript. GD and GS initially identified the misuses of bed nets on the beaches and led the field survey. KF and SK organized and analysed the data. All authors have read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2532694.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2532694",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2532694\nAUTHORS: Nicole L Quinn, Natasha Levenkova, William Chow, Pascal Bouffard, Keith A Boroevich, James R Knight, Thomas P Jarvie, Krzysztof P Lubieniecki, Brian A Desany, Ben F Koop, Timothy T Harkins, William S Davidson\n\nABSTRACT:\nBackgroundWith a whole genome duplication event and wealth of biological data, salmonids are excellent model organisms for studying evolutionary processes, fates of duplicated genes and genetic and physiological processes associated with complex behavioral phenotypes. It is surprising therefore, that no salmonid genome has been sequenced. Atlantic salmon (Salmo salar) is a good representative salmonid for sequencing given its importance in aquaculture and the genomic resources available. However, the size and complexity of the genome combined with the lack of a sequenced reference genome from a closely related fish makes assembly challenging. Given the cost and time limitations of Sanger sequencing as well as recent improvements to next generation sequencing technologies, we examined the feasibility of using the Genome Sequencer (GS) FLX pyrosequencing system to obtain the sequence of a salmonid genome. Eight pooled BACs belonging to a minimum tiling path covering ~1 Mb of the Atlantic salmon genome were sequenced by GS FLX shotgun and Long Paired End sequencing and compared with a ninth BAC sequenced by Sanger sequencing of a shotgun library.ResultsAn initial assembly using only GS FLX shotgun sequences (average read length 248.5 bp) with ~30× coverage allowed gene identification, but was incomplete even when 126 Sanger-generated BAC-end sequences (~0.09× coverage) were incorporated. The addition of paired end sequencing reads (additional ~26× coverage) produced a final assembly comprising 175 contigs assembled into four scaffolds with 171 gaps. Sanger sequencing of the ninth BAC (~10.5× coverage) produced nine contigs and two scaffolds. The number of scaffolds produced by the GS FLX assembly was comparable to Sanger-generated sequencing; however, the number of gaps was much higher in the GS FLX assembly.ConclusionThese results represent the first use of GS FLX paired end reads for de novo sequence assembly. Our data demonstrated that this improved the GS FLX assemblies; however, with respect to de novo sequencing of complex genomes, the GS FLX technology is limited to gene mining and establishing a set of ordered sequence contigs. Currently, for a salmonid reference sequence, it appears that a substantial portion of sequencing should be done using Sanger technology.\n\nBODY:\nBackgroundThe salmonids (salmon, trout and charr) are of considerable environmental, economic and social importance. They contribute to ecosystem health by providing food sources for predators such as bears, eagles, sea lions and whales. As an increasingly popular food choice for humans, salmonid species contribute to local and global economies through fisheries, aquaculture and sport fishing. In addition, they have distinct social importance as they are a traditional food source for indigenous peoples, and play a significant role in their culture and spirituality. Salmonids are also of great scientific interest. The common ancestor of salmonids underwent a whole genome duplication event between 20 and 120 million years ago [1,2]. Thus, the extant salmonid species are considered pseudo-tetraploids whose genomes are in the process of reverting to a stable diploid state. More is known about the biology of salmonids than any other fish group, and in the past 20 years, more than 20,000 reports have been published on their ecology, physiology and genetics. Salmonids, with their genome duplication and wealth of biological data, are excellent model organisms for studying evolutionary processes, fates of duplicated genes and the genetic and physiological processes associated with complex behavioral phenotypes [3]. It is surprising therefore, that no salmonid genome has been sequenced to date.The Atlantic salmon (Salmo salar) is an ideal representative salmonid for genome sequencing given the popularity of this species for aquaculture as well as the extensive genomic resources that are available. The current genomic resources include: a BAC library [4], restriction enzyme fingerprint physical map comprising 223,781 BACs in ~4,300 contigs [5], 207,869 BAC-end sequences that cover ~3.5% of the genome sequence, a linkage map with ~1,600 markers, ~600 of which are integrated with the physical map [6], and > 432,000 ESTs [7,8]. The haploid C-value for Atlantic salmon is estimated to be 3.27 pg [9], or a genome size of approximately 3 × 109 bp, which is very comparable to the sizes of mammalian genomes. The Atlantic salmon genome is highly repetitive, and at least 14 different DNA transposon families whose members are ~1.5 kb have been described [10]. Although five fish genomes have been sequenced (medaka, Oryzias latipes; tiger pufferfish, Takifugu rubripes; green spotted pufferfish, Tetraodon nigriviridis; zebrafish, Danio rerio and stickleback, Gasterosteus aculeatus), they represent euteleostei lineages, and often very derived species that have been separated from salmonids for at least 200 million years [11]. The complexity of the Atlantic salmon genome combined with the lack of a closely related guide sequence means that sequencing and assembly will be extremely challenging.Conventional Sanger sequencing of paired end templates (2–4 kb plasmids, 40 kb fosmids, or ~150 kb BACs) using fluorescent di-deoxy chain terminators and capillary electrophoresis revolutionized the field of genomics (reviewed in [12]). Although this approach remains the gold standard for sequence and assembly quality, limitations with respect to cost, labor-intensiveness and speed, which are largely due to the necessity of generating and arraying cloned shotgun libraries and isolating template DNA for sequencing, have fueled the demand for new approaches to DNA sequencing. In recent years, several novel high-throughput sequencing platforms have entered the market including the SOLiD system by Applied Biosystems [13], the Solexa technology [14], now owned by Illumina, the recently released true Single Molecule Sequencing (tSMS) platform by Helicos [15] and the 454 platform [16], now owned by Roche. Most of these are targeted to the goal of re-sequencing an entire human genome for < $1,000 [17]. This next generation of genome sequencing stands to have major scientific, economic and cultural implications with respect to applications such as personalized medicine, metagenomics and large-scale polymorphism studies on organisms of commercial value whose genomes have already been sequenced. However, the ability of these technologies to sequence the genomes of complex organisms de novo remains unknown.A common feature among the new generation of sequencing procedures is the elimination of the need to clone DNA fragments and the subsequent amplification and purification of DNA templates prior to capillary sequencing. Rather, sequence templates are handled in bulk, and massively parallel sequencing by synthesis or ligation allows the generation of hundreds of thousands to millions of sequences simultaneously.With respect to de novo whole genome sequencing, perhaps the most promising new technology uses a pyrosequencing protocol [18] optimized for solid support and picolitre scale volumes (i.e., pyrosequencing using the 454 system [16]). The 454 pyrosequencing technology [both the Genome Sequencer (GS) 20 and FLX generation systems] has proven very successful for a number of applications such as complete microbial genome sequencing [19] metagenomic and microbial diversity analyses [20,21] ChIP sequencing and epigenetic studies [22,23], genome surveys [24], gene expression profiling [25] and even for sample sequencing fragments of Neanderthal DNA that were extracted from ancient remains [26,27]. Recent accomplishments include its contribution to a high quality draft sequence of the grape genome [28] as well as complete re-sequencing of an individual human genome, for which the assembly was accomplished by mapping 454 reads back to a reference genome [29].Although several studies comparing 454 pyrosequencing with Sanger sequencing have shown that the per base error rates of the two technologies are similar [27,30], 454 pyrosequencing has limitations. The major concerns have been relatively short read lengths (i.e., as of 2007 an average of 100–200 nt compared to 800–1,000 nt for Sanger sequencing), a lack of a paired end protocol and the accuracy of individual reads for repetitive DNA, particularly in the case of monopolymer repeats [12]. Combined, these factors often make it impossible to span repetitive regions, which therefore collapse into single consensus contigs during sequence assemblies and leave unresolved sequence gaps. These issues have recently been addressed with the release of the GS FLX system as well as the Long Paired End sequencing platform. The GS FLX system provides longer read lengths and lower per-base error rates than the previous systems. In addition, the 454 technology offers the longest read length of any of the next generation sequencing systems currently available. Thus, we chose to evaluate the ability of the 454 technology, as it stands, to sequence a complex genome without the aid of high-coverage Sanger-generated reads.With respect to de novo assembly of a complex genome, the most relevant test to date of the capability of the 454 pyrosequencing technology (GS 20 system) involved sequencing four BACs containing inserts of the barley genome, two of which had previously been sequenced using the traditional Sanger approach [30]. The barley genome is relatively large (5.5 × 109 bp) and is comprised of more than 80% repetitive DNA, posing a significant challenge for sequencing. Whereas each BAC contained approximately 100 Kb of genomic DNA, the cumulative size of all consensus sequence contigs per BAC did not reach the actual size of the BAC clones for any of the 454-based assemblies. This was largely due to the pooling of repetitive sequences into single contigs. Thus, while the 454 technology proved useful for identifying genes, it was of limited value for producing long contiguous sequence assemblies [30].Given the significant and ongoing improvements in the 454 technology since the barley BAC analysis, which include longer read lengths and higher sequence accuracy attributable to the release of the GS FLX system, as well as the availability of a paired end protocol, we set out to assess the feasibility of using this technology to sequence the Atlantic salmon genome. Here we report the results of using the GS FLX pyrosequencing system to sequence de novo a 1 Mb region of Atlantic salmon DNA covered by a minimum tiling path comprising eight BACs. We discuss the integration of Atlantic salmon genomic resources such as BAC-end sequences as well as assembly techniques and annotation tools given the lack of a closely related guide sequence. We also address the ability of the GS FLX Long Paired End technology to establish the order of sequence contigs and assemble them into large scaffolds. Finally, we compare the GS FLX assemblies with and without the addition of paired end reads to a Sanger-generated assembly of a ninth BAC from the same region of the genome. This is the first application of the GS FLX Long Paired End system for de novo assembly of a large region from a complex genome. This study represents the most difficult challenge for 454 pyrosequencing thus far, and the results we present can be used to assess the feasibility of this technology for sequencing the Atlantic salmon genome de novo.MethodsEstablishment of minimum tiling path and DNA preparationWe initially chose contig 570 of the Atlantic salmon physical map for analysis due to the presence of the microsatellite marker SsaF43NUIG, which is linked to upper temperature tolerance in rainbow trout [31,32] and Arctic charr [33]. Contigs 2469 and 483 were joined to contig 570 using 'chromosome walking'. Specifically, 40-mer oligonucleotide probes were designed from the BAC-end sequences of the outer-most BACs in the contigs, as determined by the contig order predicted by the physical map, beginning with contig 570. The probes were labeled with γ32P-ATP using T4 polynucleotide kinase (Invitrogen, Burlington, Ont. Canada) and hybridized to filters containing the Atlantic salmon BAC library [4] (CHORI-214; CHORI, BAC-PAC Resources, Oakland, CA, USA.). Filters were exposed to phosphor screens that were scanned and visualized using ImageQuant™ software, giving an image of the 32P-labeled hybridization-positive BACs containing the probe sequence. All hybridization-positive BACs were verified using PCR with the SsaF43NUIG primers [34]. The minimum tiling path across Atlantic salmon contig 483 was established by designing primer sets for sequence tag sites (STSs) in both the SP6 and T7 ends of selected BACs. Using these primers, we screened the BACs that were predicted to overlap with the STS source BAC given the predicted assembly from the Atlantic salmon physical map using PCR, thereby establishing relative BAC orientation and overlap. The minimum tiling path was then established by selecting the minimum number of overlapping BACs required to span the entire contig. We isolated approximately 5 μg of cloned Atlantic salmon BAC DNA from the minimum tiling path BACs using Qiagen's Large Construct kit as per the manufacturer's directions (Qiagen, Mississauga, Ont. Canada). The kit includes an exonuclease digestion step to eliminate E. coli genomic DNA.454 shotgun pyrosequencingThe shotgun sequencing protocol using the 454 sequencing system has been described previously [16]. The salmon BAC results presented here were generated on the GS FLX (454 Life Sciences, Branford, CT) whereas the results presented previously [16] were generated on the GS 20 sequencer, the previous generation instrument. The GS FLX instrument is capable of generating 100 million bp of sequence in approximately 250 bp reads in a 7.5 hour run. Additionally, the GS FLX system has a significantly lower error profile than the GS 20 system.Briefly, to generate the GS FLX shotgun library, the isolated Atlantic salmon BAC DNA was mechanically sheared into fragments, to which process specific A and B adaptors were blunt end ligated. The adaptors contain the amplification and sequencing primers necessary to the GS FLX sequencing process. After adaptor ligation, the fragments were denatured and clonally amplified via emulsion PCR, thereby generating millions of copies of template per bead. The DNA beads were then distributed into picolitre-sized wells on a fibre-optic slide (PicoTiterPlate™), along with a mixture of smaller beads coated with the enzymes required for the pyrosequencing reaction, including the firefly enzyme luciferase. The four DNA nucleotides were then flushed sequentially over the plate. Light signals released upon base incorporation were captured by a CCD camera, and the sequence of bases incorporated per well was stored as a read. DNA extractions were performed at Simon Fraser University (Burnaby, BC, Canada), and library generation and sequencing were performed at 454 Life Sciences (Branford, CT, USA).GS FLX Long Paired End DNA library generation and sequencingGS FLX Long Paired End library generation for 454 sequencing has been described previously [23]. Briefly, DNA was sheared into ~3 kb fragments, EcoRI restriction sites were protected via methylation, and biotinlylated hairpin adaptors (containing an EcoRI site) were ligated to the fragment ends. The fragments were subjected to EcoRI digestion and circularized by ligation of the compatible ends, and subsequently randomly sheared. Biotinlyated linker containing fragments were isolated by streptavidin-affinity purification. These fragments were then subjected to the standard 454 sequencing on the GS FLX system. The paired end reads are recognizable as the known linker (originating from the two hairpin adaptors) surrounded by BAC sequence. When sequenced on the GS FLX, this protocol generates two, ~100 bp tags known to be ~3 kb apart. These paired end reads were used to build the original contigs and to assemble the contigs into scaffolds.GS FLX assembliesA previous version of the Newbler assembler used in performing the assemblies has been described previously [16], and the overall structure and phases of the assembler used here follows the structure described in that paper; however, the algorithms used for the specific phases of assembly have been upgraded. The upgraded Newbler assembler identifies pairwise overlaps between reads, and then uses them to construct multiple alignments of contiguous regions of the dataset. Boundaries where the read-by-read alignments diverge or converge (such as at the boundaries of repeat regions) define breaks in the contig multiple alignments (also called branch points). The resulting data structure consists of a graph, where each node is a contiguous multiple alignment, undirected edges exist between the 5' and 3' ends of the contig nodes, and reads form alignments along paths of the graph. The assembler builds this multiple alignment graph using an adjustable greedy algorithm of taking a 'query' read, finding the pairwise overlaps to it, constructing a multiple alignment of those overlaps, then choosing a subsequent 'query' read from the overlapped reads that are only partially aligned so far (thereby extending the multiple alignment). If any pairwise overlap alignments conflict with the current multiple alignment graph, corrective algorithms use the conflicting alignments to either ignore the new pairwise overlap (if the graph is more consistent) or to correct the constructed multiple alignment (if the new pairwise overlap identifies a misalignment in the graph). These overlaps and multiple alignment algorithms use a combination of nucleotide-space (i.e., the bases of the reads) and flow-space (i.e., the 454 flowgram signal intensities of the reads), where available, to perform the multiple alignment construction.Following the construction of the multiple alignment graph, a series of 'detangling' algorithms are used to simplify the complex regions of the graph, such as overly collapsed regions shorter than the length of the reads (i.e., parts of reads that happened to be near-identical to each other by chance, and so produced overlaps that collapsed into a single multiple alignment region). The nodes in the resulting graph after detangling are considered to be the 'contigs' by the assembler, and those longer than 500 bp are output as the 'large contigs' of the assembly (those longer than 100 bp are output in the set of 'all contigs').If paired end reads are included in the data set (either 454 or Sanger paired ends), then an additional scaffolding step is performed after detangling, to create chains of contig nodes using the paired end information. The pairs from each library where both halves of the pair occur in the same contig are used to calculate expected pair distances for the library. The scaffolding algorithm then performs a greedy algorithm of identifying pairs of nodes where at least two paired end reads have their halves aligned at the ends of the pair of nodes, with the correct alignment direction and expected distance from each other. In addition, the set of paired end reads aligned at those two contig ends must support the unambiguous chaining of the two nodes as immediate neighbors in a scaffold, with fewer than 10% of the paired end reads aligning to other contig nodes in the assembly. The chains of contig nodes found by this greedy algorithm are output as the scaffolds of the assembly.Gene mining of 454 GS FLX assemblies using syntenic regionsSequence contigs > 1,000 bp were analyzed using a variety of sequence similarity searches and gene prediction algorithms that have been incorporated into an in-house computational pipeline and database [35]. Sequences entering this pipeline were screened (masked) for repetitive elements using RepeatMasker 3.1.8 [36] and were searched against the NCBI nr (non-redundant) and Atlantic salmon EST [8] databases using BLAST [37]. A GENSCAN gene model prediction algorithm [38] was used to predict introns and exons, and the resulting predictions were searched against the Uniref50 (clustered sets of sequences from UniProt Knowledgebase) database [39]. Finally, a rps-BLAST against the NCBI CDD (Conserved Domain Database; [40]) was conducted to provide additional information with respect to the predicted genes [see additional File 1].Use of BAC-end sequences to confirm GS FLX scaffold builds and orderThe final scaffold assembly incorporating all data (GS FLX shotgun, paired end and BAC-end reads) was verified by conducting BLAST searches of the 126 BAC-end sequences against the four scaffolds > 10,000 bp and comparing the alignment positions with those predicted by the Atlantic salmon physical map. This method was also used to establish relative scaffold order and to confirm the gene order predicted by the BLAST searches of the 454 shotgun and BAC-end sequence contigs against four published fish genomes.Sanger shotgun sequencing, assembly and annotationThe ninth BAC (S0022P24) of the minimum tiling path was sequenced using standard Sanger sequencing of a shotgun library. Briefly, the purified BAC DNA was sheared by sonication and blunt-end repaired. The sonicated DNA was size fractioned by agarose gel electrophoresis and 2–5 kb fragments were purified using the QIAquick Gel Extraction Kit (Qiagen, Mississauga, Ont. Canada). DNA fragments were ligated into pUC19 plasmid that had been digested with SmaI and treated with shrimp-alkaline phosphatase to produce de-phosphorylated blunt ends. The ligation mixture was used to transform supercompetent E. coli cells (XL1-Blue; Stratagene, La Jolla, CA. USA). Transformed cells were cultured overnight at 37°C on LB/agar plates supplemented with ampicillin (200 mg/L) and 1,920 (5 × 384 well plates) clones were sent to the Michael Smith Genome Sciences Centre for sequencing. The sequences were analyzed for quality using PHRED [41], assembled using PHRAP [42], and viewed using Consed version 15.0 [43]. The S0022P24 assembly was annotated using the same protocol as the GS FLX assemblies (see above).Results and discussionSelection of BACs for GS FLX pyrosequencingUsing chromosome walking, we joined contigs 2469 and 483 to contig 570, and by convention, the new contig was named after the lowest numbered contig within it (i.e., contig 483). Contig 483 contains 195 BACs and includes 126 BAC-end sequences with an average read length of 660 bp. A contig summary can be found in the Atlantic salmon database [6]. Nine BACs were required to span the contig in a minimum tiling path (Fig. 1); eight tiled BACs were selected for GS FLX pyrosequencing and the final (ninth) BAC was sequenced using standard Sanger sequencing of a shotgun library. The estimated length of the minimum tiling path, based on HindIII banding patterns and accounting for overlap between BACs was 1,119,000 bp, with the eight BACs sequenced by GS FLX pyrosequencing accounting for ~950,000 bp. This is probably an underestimate of the true length as doublet and triplet bands may be counted only once.Figure 1Nine BACs within the minimum tiling path (MTP) of Atlantic salmon contig 483. Using the BAC-end sequences, primers were developed to amplify sequence tag sites (STSs – vertical lines), which were used to design and verify a minimum tiling path across the contig. BAC S0022P24 (green line) was sequenced using traditional Sanger sequencing of a shotgun library and the remaining eight BACs (black lines) were sequenced using the GS FLX platform.GS FLX shotgun assemblies with and without BAC-end sequencesWe created a GS FLX shotgun library using eight pooled BACs belonging to a minimum tiling path that spanned approximately 1 Mb of the Atlantic salmon genome. The shotgun run produced 141,746 high quality reads with an average read length of 248.5 bp (Fig. 2a). After filtering for vector and E. coli sequences, 101,705 reads with a total of 30,549,147 bases were assembled into 803 contigs, 149 of which were > 500 bp and therefore defined as large contigs. Note that this definition of a large contig would include all Sanger-generated reads, which typically range from 500–800 bp. The average contig size was 6,381 bp and the largest contig comprised 34,471 bp. The N50 contig size, defined as the largest contig size at which half of the total size of the contigs is represented by contigs larger than the N50 value, was 11,497 bp (Table 1). The second assembly incorporated an additional 89,095 bp in the form of 126 Sanger-generated BAC-end sequences with an average read length of ~660 bp. This effectively added 126 large contigs to the 149 generated by GS FLX shotgun sequencing. Assembling the GS FLX shotgun data with the BAC-end sequences enabled contig joins, thereby decreasing the number of large contigs to 138 and increasing the N50 contig size to 13,455 bp. The average contig size for the second assembly was 6,827 bp and the largest contig size was 38,211 bp. Both assemblies produced an estimated total length of ~1,080,000 bp not including sequence gaps, which is in agreement with the estimate derived from HindIII fragments (Fig. 3). The GS FLX shotgun sequencing produced ~30× coverage of the region and the BAC-end sequences provided an additional ~0.09× coverage.Table 1Summary of GS FLX shotgun assembliesSGSG+BEReads assembled101705102953Singleton reads27952870Large contigsa (> 500 bp)149138Total number of contigs803811Bases in large contigs950826942244Total bases covering region10881031081281Average contig size (bp)63816827N50 contig sizeb (bp)1149713455Largest contig (bp)3447138211> Q40 bases (bp)947699939244GS FLX shotgun assembly alone (SG) and when combined with 126 BAC-end sequences (SG+BE). aContigs are defined as more than one read joined by overlapping sequence. Large contigs defined as greater than 500 bp. bThe N50 contig size is defined as the largest contig size at which half of the total size of the contigs is represented by contigs larger than the N50 value.Figure 2a. Distribution of the read lengths for the GS FLX shotgun sequencing (average 248.5 bp). b. Distribution of read lengths of the GS FLX Long Paired End sequencing. The yellow curve represents the raw reads (average read length 210 bp). These were separated into those containing the linker sequence and those without. The reads containing the linker sequence were separated into two paired end reads, one to the left of the linker (green curve; average read length 93 bp) and those to the right of the linker (red curve; average read length 96 bp). Reads without the linker sequence (blue curve, average read length 191 bp) were added to the assembly as additional shotgun reads.Figure 3HindIII banding patterns of the nine BACs that comprise the minimum tiling path of contig 483 of the Atlantic salmon physical map. Adjacent lanes share some common bands indicating overlap, whereas lanes separated by more than one lane do not share common bands except when HindIII fragments are of the same size by chance. Scale indicates migration distance. The nine tiled BACs were estimated to span 1,119,000 bp with the eight BACs sequenced by the GS FLX system accounting for approximately 950,000 bp as determined by summing the unique bands in each lane.Annotation of GS FLX shotgun contigs > 1,000 bpBLAST results for four fish genomes (medaka, Oryzias latipes; tiger pufferfish, Takifugu rubripes; zebrafish, Danio rerio and stickleback, Gasterosteus aculeatus) against the large contigs from the GS FLX shotgun and BAC-end sequence assembly revealed hits to seven well annotated genes and one hypothetical gene (Fig. 4a). BLAST results against the Tetraodon nigriviridis genome were inconclusive, as most sequence contigs matched to \"un_random\" sequences (sequence contigs and scaffolds that have not been mapped to any Tetraodon chromosome) that collectively spanned over 130 Mb. No genes were identified in any of the fish genomes that were not found in the Atlantic salmon sequence contigs and vice versa, indicating conservation of synteny for this genomic region for these four species. Gene order was conserved across three of the four fish species (medaka, zebrafish and the tiger pufferfish), whereas there were two apparent inversions in the stickleback genome relative to the other genomes (Fig. 4b), which may be an artifact of the preliminary, incomplete assembly of the stickleback genome. Using these results and assuming conservation of gene order among teleosts, we could predict the order of 12 gene-containing sequence contigs relative to one another; however, their order with respect to the remaining 126 large contigs could not be established. This confirmed the utility of GS FLX shotgun sequencing for gene discovery and highlighted the difficulty of using this approach alone to assemble the sequence of a complex genome de novo.Figure 4a. Genes identified in the nine BACs using our in-house annotation pipeline . b. Order of the genes within the minimum tiling path. Comparative synteny analysis against the four published fish genomes (medaka, Oryzias latipes; tiger pufferfish, Takifugu rubripes; green spotted pufferfish, Tetraodon nigriviridis; zebrafish, Danio rerio and stickleback, Gasterosteus aculeatus) enabled the ordering of the gene-containing contigs in the GS FLX assembly of shotgun reads only. This order was confirmed when contigs were assembled into scaffolds with the addition of GS FLX Long Paired End reads. Numbers correspond to contig identity in the Atlantic salmon assemblies; colors coordinate with genes listed in Figure 4a. The grey boxes that correspond to sequence contigs 5 and 685 indicate matches to hypothetical genes. The genes for gonadotropin releasing hormone receptor and the novel protein similar to vertebrate perilipin were found within the Sanger-sequenced BAC and the remaining genes were within the eight BACs sequenced by GS FLX pyrosequencing.Assemblies incorporating GS FLX Long Paired End dataWe constructed a GS FLX Paired End library using DNA from the eight tiled BACs to test its ability to improve the shotgun assembly. After trimming for E. coli and vector sequences, the GS FLX Long Paired End sequencing produced 149,035 high-quality reads with an average read length of 210 bp (Fig. 2b). Of these, 66,739 contained the linker sequence used to construct the paired end library; therefore, they represented the two paired ends of DNA separated by linker. The average read lengths of the paired ends were 93 and 96 bp for left and right sides of the linker, respectively (Fig. 2b). The remaining reads (i.e., those not containing linker) had an average read length of 191 bp (Fig. 2b) and were used in the assembly as additional shotgun reads. After splitting each linker-containing read into two paired ends and adding the remaining reads, 213,118 usable reads were obtained. When assembled, these produced 310 contigs, 203 of which were assembled into six large scaffolds (i.e., > 10,000 bp) with an N50 scaffold size of 197,327 bp and the largest scaffold was 227,111 bp (Table 2). When combined with the GS FLX shotgun reads, the assembly yielded 289 large contigs, 106 of which were assembled into three large scaffolds with an N50 scaffold size of 361,606 bp and the largest scaffold size was 501,016 bp. Finally, when the 126 BAC-end sequences were incorporated, 286 contigs were produced, 175 of which were assembled into four large scaffolds [GenBank: EU481821] with an N50 and largest scaffold value of 538,994 bp. The GS FLX Long Paired End sequencing provided an additional ~26× coverage of the eight tiled BACs, which, when combined with the GS FLX shotgun data resulted in ~56× coverage of the region. So far, the only published use of the GS FLX Long Paired End technology has been for revealing structural variations in the human genome [23]. The results presented here represent the first use of this technology for de novo genome sequence assembly.Table 2Summary of GS FLX Long Paired End assembliesPE onlyPE+SGPE+SG+BES0022P24Large contigsa (> 500 bp)31028928614Average contig size (bp)2686305831498885N50 contig sizeb (bp)41604728563532866Contigs assembed into scaffoldsc2031861759hTotal scaffolds9342Large scaffoldsd (> 10 Kb)6342Average large scaffold size (bp)96257299378226679112155Largest scaffold size (bp)227111501016538994137857N50 scaffold sizee (bp)197327361606538994137857Total gapsf1941831718Maximum gap size (bp)1,8812,1002,131unknownMinimum gap size (bp)448unknownPair distance averageg (bp)268027762782N/APair distance deviation (bp)670694696N/ATotal bases covering region95850710028401000926231017Depth of coverage~26×~56×~56×~10.5×Results for GS FLX Long Range Paired End (PE) assembly alone and when combined with the GS FLX shotgun (SG) data and BAC-end (BE) sequences. aContigs are defined as more than one read joined by overlapping sequence. Large contigs are greater than 500 bp. bThe N50 contig size is defined as the largest contig size at which half of the total size of the contigs is represented by contigs larger than the N50 value. cA scaffold is defined as two or more contigs associated by paired ends. dLarge scaffolds are those consisting of more than 10,000 bp among all contigs therein. eThe N50 scaffold size is defined as the largest scaffold size at which half of the total size of the scaffolds is represented by scaffolds larger than the N50 value. fGaps represent unsequenced regions between two contigs known to be adjacent due to associated paired ends. gAverage pair distance is the average distance between two sections of BAC DNA separated by linker sequence. hAssembly based on large contigs (> 500 bp) consisting of ≥3 reads each.The combination of GS FLX shotgun and Long Paired End reads provided approximately 56× coverage of the 1 Mb region of the salmon genome. We speculate that this represents extensive over-coverage and that similar results could be obtained using fewer reads and less coverage of the region. However, further studies that examine various combinations of coverage from shotgun and paired end libraries are necessary to test this hypothesis and to determine the optimal combination of the two GS FLX read types for genome assembly.Use of BAC-end sequences and minimum tiling path to confirm assembly and order of scaffoldsThe accuracy of the final scaffold assembly was verified by conducting a BLAST search of the 126 BAC-end sequences against the scaffold builds. This also established the order of the four scaffolds relative to one another and confirmed that the aligned sequences followed the order predicted by the minimum tiling path of the eight BACs. These results provided further support for conservation of synteny and gene order of the seven genes in the genomes of Atlantic salmon, medaka, zebrafish and tiger pufferfish. Fig. 5 provides a visual summary of the data, including the minimum tiling path, sequence contigs, scaffolds, predicted genes and BAC-end sequences in the 1 Mb region.Figure 5Summary of the 1 Mb sequenced region for the final assembly incorporating the GS FLX shotgun and paired end data with the 126 BAC-end sequences. This figure summarizes all genes identified within the 1 Mb region and their position, the arrangement of the large scaffolds (order and orientation) as confirmed by the BAC-end sequences, the sequence contigs aligned against the scaffolds, the eight BACs of the minimum tiling path (MTP) including established overlap, and the BAC-end sequences within the region in the order predicted by the Atlantic salmon physical map.Assembly and annotation of the ninth BACSanger sequencing of the shotgun library of the ninth BAC (S0022P24) in the minimum tiling path produced 3,524 confirmed reads and an average confirmed read length of 693.3 bp. PHRAP defines a confirmed read as verification of a read by another read with different chemistry or by an opposite-strand read [44]. This produced a ~10.5× depth of coverage given the estimated BAC size of 231,979 bp. The confirmed reads were assembled into 20 contigs with an average contig size of 8,885 bp and an N50 contig size of 32,866 bp; 14 contigs were defined as large contigs (i.e., > 500 bp). Nine large contigs consisting of three or more reads were assembled into two large scaffolds based on corresponding paired end reads from cloned inserts [GenBank: EU873552]. The average and N50 scaffold sizes were 112,155 and 137,857 bp, respectively. The two scaffolds were oriented relative to one another based on the locations of the T7 and SP6 BAC-end sequences.The Sanger assembly produced a much larger average contig size and N50 contig size than any of the GS FLX assemblies (i.e., with and without paired end and BAC-end sequence reads), which corresponds to fewer contigs produced. This is likely because of the larger average read length of the Sanger sequences. The Sanger assembly produced two scaffolds with eight gaps for a ~230,000 bp region, whereas the final GS FLX assembly produced four scaffolds with 171 gaps for a ~1 MB region. Thus, with respect to the ability to establish the order and orientation of sequence contigs relative to one another, the GS FLX assembly was comparable to a Sanger-based assembly. This, however, was offset by the numerous gaps between contigs within the GS FLX assembly.Sequence annotation using our in-house pipeline (described above) revealed hits to two genes: gonadotropin-releasing hormone receptor type I and a novel protein similar to vertebrate perilipin (Fig. 4a), with the latter located next to the final gene in the BACs sequenced by GS FLX. When the region was compared with regions that were previously identified as being syntenic with other sequenced fish genomes, only that of the zebrafish (Danio rerio) contained both genes. The remaining genomes (medaka, Oryzias latipes; tiger pufferfish, Takifugu rubripes; and stickleback, Gasterosteus aculeatus) only contained the gonadotropin-releasing hormone receptor type I gene with no evidence of the novel protein similar to perilipin or any other genes (Fig. 4b).Nature of gaps in GS FLX assemblyA major concern is that 171 gaps remain between the GS FLX-sequenced contigs within the four final scaffolds. Given that GS 20, and by extension GS FLX, pyrosequencing is known to provide good coverage of genic regions [24], these gaps likely represent repeat regions rather than missed genes. This was supported by synteny analysis, which indicated that the initial assembly covered all genes present within this region in sequenced fish genomes, and by conducting a BLAST search of gap ends, which revealed that many of the gaps bordered known salmonid repetitive elements [10]. A comparison of the overlapping region between the BAC sequenced by the Sanger method and the corresponding region sequenced by GS FLX pyrosequencing (i.e., the region between the BAC-ends S0070O23-T7 and S0022P24-SP6 in Fig. 6), identified two gaps of 893 and 151 bp in the GS FLX assembly. These regions of the Sanger assembly were completely masked by the salmonid-specific repeat masker [45], thus verifying that the GS FLX technology has difficulty with repetitive regions.Figure 6Summary of the Sanger-sequenced BAC (S0022P24). The two genes within the ~200,000 bp region are indicated as well as the nine sequence contigs and two scaffolds (indicated by red and green contigs). The relative orientation of these scaffolds was determined knowing the SP6 and T7 BAC-end sequences. The BAC-end sequences within the region are indicated in the order predicted by the Atlantic salmon physical map. Note that this BAC overlaps with the remainder of the MTP (i.e., that sequenced by GS FLX) at the 70O23-T7 BAC-end.ConclusionWith 30–40% repetitive content and its pseudo-tetraploid nature due to a whole genome duplication event [2], the Atlantic salmon genome poses a significant challenge for sequencing. To date, the strategies to sequence complex vertebrate genomes have been Sanger sequencing of whole genome shotgun libraries (e.g., dog genome [46]), the generation of a library of cloned inserts such as BACs, followed by a 'map-first, sequence second' approach (e.g., pig genome [47]), or a combination of whole genome shotgun sequencing and pooled BAC sequencing [48]. These strategies are dependent on the minimal ability to sequence and assemble a full BAC insert. However, to date, this has proved unsuccessful with respect to complex genomes with any technique other than Sanger sequencing of a subcloned shotgun library [30].The purpose of this study was to assess the feasibility of GS FLX pyrosequencing for de novo assembly of the Atlantic salmon genome given recent advances in read length and the availability of GS FLX Long Paired End technology. We demonstrated that without the inclusion of GS FLX Paired End reads, the GS FLX shotgun technology alone was substantially inferior to Sanger sequencing given the size and number of contigs produced and the inability to establish the relative order and orientation of the contigs. However, the addition of GS FLX Paired End reads vastly improved the capability of 454 pyrosequencing by enabling the assembly of contigs into large scaffolds. Indeed, in terms of the number of scaffolds produced, the GS FLX assembly that included the combined shotgun and paired end reads was comparable to the Sanger assembly. Moreover, the order of the GS FLX scaffolds could be established from information from BAC-end sequences and the Atlantic salmon physical map. However, numerous gaps remained within the scaffolds, which is undesirable when a complete or reference genome sequence is one of the goals. Currently, if the Atlantic salmon genome is to provide a reference sequence for all salmonids, then a substantial proportion of the sequencing will have to be carried out using Sanger technology.Authors' contributionsNLQ, PB, TPJ, BD, JK, TTH, BFK and WSD conceived the project. NLQ established the minimum tiling path and prepared the DNA. PB was responsible for GS FLX pyrosequencing. NL, WC, KAB, JK, KPL and BD performed bioinformatics. NLQ, NL, WC, PB, JK, KAB, KPL and WSD analyzed and interpreted the data. NLQ, TTH and WSD prepared the manuscript.Supplementary MaterialAdditional file 1Summary of information used for sequence annotation. Species, Ensembl names, assembly release date, Genebuild and database versions for all genome sequences used for comparative synteny analyses of the GS FLX shotgun + BAC-end sequence-generated contigs.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2532746.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2532746",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2532746\nAUTHORS: Natasha Webb, Geoff Connolly, Judy Tellam, Alpha S. Yap, Rajiv Khanna\n\nABSTRACT:\nPrevious studies have indicated that Epstein-Barr virus (EBV) can modulate the Wnt pathway in virus-infected cells and this effect is mediated by EBV-encoded oncogene latent membrane protein 1 (LMP1). Here we have reassessed the role of LMP1 in regulating the expression of various mediators of the canonical Wnt cascade. Contradicting the previous finding, we found that the levels of E-cadherin, β-catenin, Glycogen Synthase Kinase 3ß (GSK3β), axin and α-catenin were not affected by the expression of LMP1 sequences from normal B cells or nasopharyngeal carcinoma. Moreover, we also show that LMP1 expression had no detectable effect on the E-cadherin and β-catenin interaction and did not induce transcriptional activation of β-catenin. Taken together these studies demonstrate that EBV-mediated activation of Wnt pathway is not dependent on the expression of LMP1.\n\nBODY:\nIntroductionThe oncogenic potential of Epstein-Barr virus (EBV) is well recognized, and the virus is associated with a number of human malignancies, including Burkitt's lymphoma (BL) and nasopharyngeal carcinoma (NPC) [1]. Each of the EBV-associated malignancies is characterised by a unique viral and cellular phenotype. In most of the EBV-associated malignancies the viral gene expression is often restricted to a limited number of proteins. This limited gene expression is often considered as one of the most important factors in the pathogenesis and escape of these malignancies from immune control (see reviews [2], [3]. EBV-encoded oncogene latent membrane protein 1 (LMP1), has been recognised as one of most crucial latent proteins for EBV-mediated transformation of normal B cells and is uniquely able to induce malignant outgrowth and hyperplasia in transgenic mice [4]. Furthermore, LMP1 is also known to exhibit pleiotropic effects on the cellular phenotype of B cells which include induction of activation antigens, the expression of inhibitors of programmed cell death and NF-κB activation through the TRAF signalling pathway [5]–[7]. Previous studies have shown that LMP1 acts as a constitutively active receptor like molecule independent of the binding of a ligand [1], [8]. The transmembrane domains mediate oligomerization of LMP1 molecules in the plasma membrane, a prerequisite for LMP1 function [1], [9].Over the last few years, there has been increasing evidence to suggest EBV is capable of modulating the Wnt pathway [10]–[13]. In particular, it has been suggested LMP1 expression can repress the expression of E-cadherin [14]–[16]. The current experiments reported here were undertaken to reassess the role of LMP1 in regulating the expression of E-cadherin and to further explore the mechanism by which LMP1 modulates the function of various mediators of the canonical Wnt cascade. Here we show that transient or stable expression of LMP1 sequences from normal B cells and NPC does not impair the expression of E-Cadherin and other mediators of the Wnt pathway. Furthermore, we also demonstrate that LMP1 expression in human cells had minimal effect on the interaction of E-cadherin and β-catenin thus no evidence of β-catenin-mediated transcriptional activation was observed.Results and DiscussionExpression of Wnt pathway mediators in LMP1-positive cellsTo explore the effect of LMP1 on other mediators of the Wnt pathway, we transiently transfected HaCaT and MDCK cells with expression vectors encoding LMP1-GFP or the control EGFP vector. These LMP1 sequences were either derived from the prototype B95.8 isolate, spontaneous LCLs (HS6, QC and PM) or NPC (NPC9 and CAO). After transfection, these cells were examined using confocal microscopy for the expression of E-cadherin, β-catenin or actin. Representative data from a series of experiments is presented in Figure 1 (panel A). In contrast to the previous studies, we observed very little difference in the expression of E-cadherin or β-catenin in LMP1 or EGFP-positive cells. Both HaCaT and MDCK cells showed minimal effect of LMP1 on the expression of E-cadherin, β-catenin. Interestingly, LMP1 sequences from both normal B cells or from NPC showed no effect on the expression of E-cadherin and β-catenin. On the other hand, we did noticed alteration in the organization of actin filaments in LMP1 expressing cells which is consistent with the previous studies published by Dawson and colleagues who also showed actin filament remodelling following LMP1 expression in 3T3 fibroblasts [17]. To ensure that the results described above were not influenced by the covalent linking of LMP1 with EGFP, we also expressed LMP1 protein in HaCaT cells without EGFP and then assessed the expression of β-catenin. Consistent with the data presented above, we observed no significant difference in the pattern of β-catenin expression in HaCaT cells transfected with either pcDNA3.1 (control) or pcDNA3.1 encoding B95.8-LMP1 (Fig. 1, Panel A).10.1371/journal.pone.0003254.g001Figure 1Panel A: Effect of LMP1 on the expression of E-cadherin, β-catenin and actin.HaCaT or MDCK cells were transiently transfected with expression vectors encoding LMP1 protein fused to EGFP. LMP1 sequences were derived from either the prototype B95.8 isolate, spontaneous LCLs (HS6, QC and PM) or NPC biopsies (CAO and NPC9). Following transfection, these cells were cultured for 36–48 h and then assessed for the expression of E-cadherin, β-catenin and actin using confocal microscopy. HaCaT cells transfected with pcDNA3.1 vector with or without B95.8-LMP1 were also assessed for β-catenin expression (bottom panels). Panel B: HEK293 cells transfected with various LMP1 sequences were also processed for SDS-PAGE and immunoblot analysis. Antibodies specific for β-catenin, E-cadherin, α-catenin, GSK3β were used to assess the level of expression each of these components in LMP1 or EGFP expressing cells. In addition, the level of LMP1 and GAPDH expression was also assessed in these cells. Representative data from one of the five different experiments is presented in this figure.To further confirm these observations, we resolved the protein samples (normalised for GFP expression using FACS analysis) from these transfected cells on SDS-PAGE followed by immunoblotting. Data presented in Figure 1, panel B clearly demonstrate that the levels of E-cadherin and β-catenin were largely unaffected by the expression of LMP1. In addition, expression levels of other Wnt pathway mediators and potential modulators (α-catenin and GSK3β) was indistinguishable between LMP1 and EGFP-positive cells. It is important to point out that the lack of any modulation of Wnt pathway mediators by LMP1 in these experiments was not due to either low levels of LMP1 expression or the loss of LMP1-mediated signalling due to covalent linking of EGFP. All LMP1 expression vectors showed normal to high levels of LMP1 expression which was quite comparable to the levels seen in EBV-infected B cells. Furthermore, data presented in figure 2 clearly shows that expression vectors encoding LMP1 protein fused to GFP at the C-terminus are fully capable of activating NF-κB and STAT3 which is comparable to that seen with LMP1 protein without GFP [18]. It is interesting to note that LMP1 sequences displayed some differences with respect to their ability to activate NF-κB and STAT3, although this variation not particularly associated with any specific disease setting from which the LMP1 sequence was isolated.10.1371/journal.pone.0003254.g002Figure 2Effects of the LMP1 expression constructs on the activation of cellular signaling pathways.NFκB (panel A) and STAT (panel B) activation induced by B95-8-LMP1, HS6-LMP1, NPC9-LMP1, QC-LMP1 or CAO-LMP1 was assessed by quantitation of the luciferase produced from a co-transfected reporter plasmids, 3Enh.κB-ConALuc or GRR(5)-Luc. The data were normalized for transfection efficiency by measuring GFP-positive cells and then expressed relative to the activity obtained with the B95-8-LMP1 (100%) without subtracting the basal activity in control pEGFP-N1-transfected cells. Results are the mean and standard deviation of at least four separate experiments.To ensure that these observations were not influenced by the transient expression of LMP1, we established stable transfectants of the HaCaT cell line expressing B95.8-LMP1 and EGFP. These cells were processed for E-cadherin, β-catenin and actin expression using confocal microscopy. Similar to data obtained with transient transfection, HaCaT cells stably expressing LMP1 showed minimal change in the expression of E-cadherin and β-catenin when compared to the cells expressing EGFP alone (Figure 3, panel A). Immunoblot analysis of protein samples from these cells also showed comparable levels of E-cadherin, β-catenin α-catenin, GSK3β, and axin in LMP1 and EGFP expressing cells (Figure 3, Panel B). Furthermore, analysis of β-catenin expression in the cytoplasmic and nuclear fraction of cells expressing LMP1 or EGFP revealed no significant difference (data not shown). Taken together, these data clearly demonstrate that LMP1 does not influence the expression of various mediators and potential modulators of the Wnt cascade and there is no evidence of accumulation of β-catenin either in the cytoplasm or nucleus following expression of LMP1.10.1371/journal.pone.0003254.g003Figure 3Effect of stable LMP1 expression on the canonical Wnt pathway mediators.HaCaT cells were transfected with expression vectors encoding LMP1 protein from the prototype B95.8 isolate. Following transfection these cells were cultured in growth medium supplemented with Geneticin for 3 weeks and stable transfectants isolated by FACSort. These stable transfectants were assessed for the expression of E-cadherin, β-catenin and actin using confocal microscopy (Panel A) and as well for α-catenin, GSK3β and Axin by SDS-PAGE and immunoblotting.Effect of LMP1 on the interaction of E-Cadherin, β-catenin and α-catenin with the surface membraneAlthough the data presented above clearly indicated that LMP1 does not influence the expression of individual components of the Wnt pathway, it is possible that LMP1 may disrupt the interaction of E-cadherin and β-catenin. A series of experiments were designed to immunoprecipitate E-cadherin and β-catenin complexes from LMP1 and EGFP expressing cells. In the first set of experiments, E-Cadherin was immunoprecipitated from the whole cell extract and then resolved on SDS-PAGE followed by immunoblotting. The immunoblots were probed with anti-E-cadherin, anti-β-catenin and anti- α-catenin antibodies. Data presented in Figure 4, panel A shows that LMP1 expression in HaCaT cells had very little effect on the interaction of E-cadherin, β-catenin and α-catenin. To further confirm these observations, we surface labelled LMP1 and EGFP-positive HaCaT cells with NHS-biotin and immunoprecipitated the biotin-labelled proteins with either anti-E-cadherin or anti-β-catenin antibodies. These immunoprecipitates were resolved on SDS-PAGE followed by immunoblotting with streptavidin, anti-E-cadherin or anti-β-catenin antibodies. Similar to the data presented in the panel A, we noticed very little effect of LMP1 expression on the interaction of E-Cadherin and β-catenin on the cell surface (Fig. 4, panel B).10.1371/journal.pone.0003254.g004Figure 4Effect of LMP1 expression on the interaction of E-cadherin and β-catenin.Two different methods were used to assess the effect of LMP1 on this interaction. Firstly, E-cadherin was immunoprecipated from the cells stably expressing either LMP1 or EGFP. These immunoprecipitates were then resolved on SDS-PAGE gel followed by immunoblotting (Panel A). These immunoblots were probed with antibodies specific for E-cadherin, β-catenin and α-catenin. In the second strategy, live HaCaT cells stably expressing either LMP1 or EGFP were initially surface labelled with biotin and then immunoprecipitated with anti- E-cadherin or anti-β-catenin antibodies. These immunoprecipitates were then resolved on SDS-PAGE gel followed by immunoblotting (Panel A). These immunoblots were probed with either streptavidin or antibodies specific for E-cadherin and β-catenin.Previous studies have suggested that EBV-mediated activation of β-catenin involves stabilization of this protein which leads to β-catenin-mediated increased transcriptional activity. To explore the possibility that this effect may be mediated by LMP1, stable transfectants (LMP1 and EGFP-positive) were pretreated with cycloheximide to block fresh protein synthesis and protein samples collected at different time intervals. These samples were then resolved on SDS-PAGE followed by immunoblotting with the β-catenin-specific antibody. Data presented in Figure 5, panel A, shows no evidence of increased stabilization of β-catenin in LMP1 expressing cells. These experiments were repeated at least five times and we were unable to see any firm evidence of LMP1-medaited stabilization of β-catenin.10.1371/journal.pone.0003254.g005Figure 5Panel A: Effect of LMP1 on the half-life of β-catenin.HaCaT cells expressing LMP1-GFP or EGFP pre-treated with cyclohexamide and then equal aliquots of cells were removed at time points 0 min, 45 min, 90 min, 135 min and 180 min; and lysed in RIPA buffer, resolved on SDS-polyacrylamide gel followed by immunoblotting with antibodies specific β-catenin. Protein bands densitometricaly analysed using Imagequant software. Panel B: Assessment of β-catenin transcriptional activity in LMP1 expressing cells. LMP1 or EGFP expressing cells (HaCaT, Hek293 or SW480) were co-transfected with TOPFlash or FOPFlash plasmids. These cells were used in luciferase reporter assays as described in the “Material and Methods” section. The TOPFlash luciferase activity produced by each sample is shown relative to the matching FOPFlash activity produced.Another possible approach to test the stabilization of β-catenin is to assess downstream transcriptional activity mediated by this protein. It is now well established that stabilized β-catenin forms a complex with Tcf/lymphoid enhancer factor transcriptional factors and that this complex transactivates various cellular oncogenes (e.g. c-myc and cyclin D1) which play crucial role in cell transformation and tumour development [19], [20]. To investigate whether LMP1 expression results in β-catenin-mediated transcriptional activation of Tcf, we transfected HaCaT cells with EGFP or LMP1-GFP expression plasmids (B95.8, CAO, HS6, NPC9 or QC) in combination with Tcf reporter plasmids containing three copies of WT Tcf-binding site (TOPFLASH) and three copies of mutated site as a negative control (FOPFLASH) and used these cells in luciferase reporter assays. Representative data one of these experiments is presented in Figure 5, Panel B. Consistent with data presented in panel A, we observed no evidence of β-catenin-mediated transcriptional activity in LMP1-expressing cells.These experiments were undertaken to reassess the role of LMP1 in modulating the canonical Wnt pathway. We have used human epithelial and keratinocyte cell lines to express different sequence variants of LMP1 and studied its effect on the regulation of E-cadherin and β-catenin interaction/function. The effect of LMP1 expression was assessed using both a transient and stable expression system. Confocal microscopic studies showed none of the LMP1 sequences (derived form either normal B cells or NPC) had any dramatic effect on the expression of E-Cadherin and β-catenin. Furthermore, we also found no effect of LMP1 on the interaction of E-Cadherin and β-catenin and the downstream β-catenin-mediated transcriptional activity. Based on this extensive and in depth analysis, we propose that it is unlikely that LMP1 plays any significant role in the modulation of the Wnt pathway. It is very difficult to precisely identify the reason for difference in the results described here and those described previously by other groups [14]–[16]. One possible reason might be that different cell lines respond differentially to LMP1 signalling. For example our studies were primarily based on human epithelial cell line HaCaT, while other groups have used canine epithelial cell line (MDCK). Furthermore, it is also possible that minor differences within the LMP1 sequences used by different groups may differentially impact on the expression of E-Cadherin and other mediators of Wnt pathway.It is important to stress here that our studies do not refute a potential role of EBV in activating β-catenin and its transcriptional activity. It is possible that another EBV protein(s) play a crucial role in regulating β-catenin activity in virus-infected normal and malignant cells. Indeed, recent studies by Morrison and colleagues have shown that EBV-encoded Latent membrane protein 2A (LMP2A) activates β-catenin in epithelial cells through the PI3/AKt pathway [12], [21]. In this context, it is important to point out that LMP2A is consistently expressed in type II malignancies such as NPC where dysregulation of E-cadherin or β-catenin expression has been reported [21]–[23].Materials and MethodsExpression Plasmids and TransfectionLMP1 sequences were amplified from prototype B95.8 isolate (referred to as B95.8-LMP1), NPC (referred to as NPC9-LMP1, CAO-LMP1) or spontaneous lymphoblastoid cell lines (LCL; HS6-LMP1, QC-LMP1, PM-LMP1) using sequence-specific primers and PCR and cloned in frame into the pEGFP-N1 vector (Clontech, Palo Alto, California). Amplified LMP1 sequences were ligated into the EcoR1 and BamHI sites of pEGFP-N1 in frame so as to express a fusion protein with the Green Fluorescent Protein (GFP) at the C-terminus. Human epithelial or keratinocyte cell lines (HaCaT, HEK293, SVMR6) or Madin-Darby canine kidney cell line (MDCK; ATCC no. CCL-34) were transfected with either pEGFP-N1 or LMP1-GFP expression vectors using lipofectamine 2000 (Invitrogen, Calsbad, California). In some experiments, an expression vector (pcDNA3.1) encoding full-length LMP1 without GFP fusion protein was also used These cells were cultured for 36–48 h in RPMI-1640 supplemented with 10% FCS (growth medium) and the transfection efficiency was assessed by the expression of EGFP using FACScalibur (Becton Dickinson, San Diego, California). In some cases transfected cells were cultured in growth medium supplemented with Geneticin (Invitrogen, Calsbad, California, 500 µg/ml) for 3 weeks and stable transfectants isolated by FACSort (Mo-Flo Fluoroscence Activated Cell Sorter, Dako Cytomation, Forte Collins, CO).ImmunofluorescenceFor immunofluorescence studies, LMP1 or EGFP expressing cells were seeded onto coverslips and cultured overnight in growth medium. After incubation, these cells were washed in PBS plus Calcium and Magnesium (PBSCM) and then fixed in 3% paraformaldehyde and permeabilised with 0.1% Saponin (Sigma Aldrich, St Louis, MO) in 5% FCS/PBSCM. After fixation, cells were incubated with monoclonal or polyclonal antibodies specific for LMP1 (Dako, Carpinteria, CA), β-Catenin (BD Transduction Laboratories, San Jose, CA) and E-Cadherin (BD Transduction Laboratories, San Jose, CA), and TRITC conjugated anti-Phalloidin antibody (Sigma Aldrich, St Louis, MO) in 5% FCS/PBSCM and incubated on the cells for 1 hour at room temperature. After washing in 0.1% Saponin/PBSCM, cells were incubated with anti-mouse Cy3 (Jackson ImmunoResearch Laboratories, Baltimore Pike, PA). Finally the cells were washed and examined by confocal microscopy (Leica TCS SP2; Leica, Mannheim, Germany).Surface Labeling, Immunoprecipitation and ImmunoblottingMonolayers of HaCaT cells stably expressing LMP1 or EGFP were washed with cold PBS (pH 8.0) and labelled with 2 mM EZ-Link Sulfo-NHS-Biotin Reagent (Pierce, Rockford, IL). The reaction was quenched with PBS and 100 mM Glycine prior to solubilisation with modified radioimmunoprecipitation (RIPA) buffer (Upstate Inc. Waltham, MA). These cell extracts were initially incubated with E-Cadherin-specific antibody for 2.5 hours followed by an overnight incubation with Protein G Sepharose beads (Roche Diagnostics, Mannheim, Germany). After incubation, beads were extensively washed with RIPA buffer and protein samples resolved using standard SDS-PAGE, transferred to nitrocellulose membrane and incubated with HRP-conjugated streptavidin (Chemicon, Temecula, CA), and antibodies specific for E-Cadherin, β-catenin or α-catenin (BD transduction laboratories, San Jose, CA),. Protein bands were detected using Chemiluminescence Reagent Plus (PerkinElmer, Life Sciences, Boston, MA) and their intensity compared by densitometric analysis using Imagequant software (Molecular Dynamics, Sunnyvale, CA). In some experiments protein samples from EGFP and LMP1-GFP expressing cells were directly resolved on the SDS-PAGE, transferred to nitrocellulose membrane and incubated with HRP-conjugated streptavidin, and antibodies specific for LMP1, E-Cadherin, β-catenin, α-catenin or GSK3β (Becton Dickson Transduction Laboratories, San Jose, CA), GAPDH (Ambion, Austin, TX) and Axin (Zymed Laboratories Inc. San Francisco, CA).Intracellular stability analysis of β-cateninThe human keratinocyte cell line HaCaT was transiently transfected with expression constructs encoding LMP1-GFP or EGFP as described above. At 30 h posttransfection, cyclohexamide (50 µg/ml) was added to 1×106 cells and equal aliquots of cells were removed at time points 0 min, 45 min, 90 min, 135 min and 180 min; and lysed in RIPA buffer, resolved on SDS-polyacrylamide gel, transferred to nitrocellulose membrane and incubated with antibodies specific β-catenin. Protein bands were detected using Chemiluminescence Reagent Plus densitometricaly analysis using Imagequant software.Analysis of β-catenin transcription activityApproximately, 106 HaCaT cells were co-transfected with TOPFlash or FOPFlash (Upstate Biotechnology, Lake Placid, NY). in combination with EGFP or LMP1-GFP expression plasmids (B95.8, CAO, HS6, NPC9 or QC) at a ratio of 1∶2. After incubation for 24 h, a small aliquot of these cells was analysed for GFP expression by FACSCalibur and cell numbers were accordingly standardised based on GFP positive cells. These cells were used in luciferase reporter assays according to the manufacturer's instructions (Promega, Madison, WI). The TOPFlash luciferase activity produced by each sample is shown relative to the matching FOPFlash activity produced.Gene transfection and luciferase reporter assayHaCaT cells were grown to late-log phase of cell growth prior to co-transfection. On the day of transfection, luciferase reporter plasmids for either NF-κB (3.Enh-Luc) or STAT (pGRR5-Luc) (Fielding et al., 2001) were combined with LMP1-GFP expression plasmids (B95–8, CAO, HS6, NPC9 or QC) and pEGFP-N1 expression vector in a 1∶2 ratio. Electroporation was used for the STAT co-transfections where 9 µg of total DNA was used for 5×106 cells, and 0.8 µg total DNA was added to 2×105 cells when using Effectene™ (Qiagen) for the NF-κB co-transfections. After growth for 24 hours in RPMI medium supplemented with antibiotics and 10% FCS, cells were harvested and resuspended in PBS supplemented with 2% FCS. To measure successful transfection a small aliquot was taken for FACScan (Becton Dickson) analysis of GFP expression, and the volume remaining was subsequently standardized to the sample of lowest efficiency. The transfection efficiencies were generally between 40–50%. To report the transcription factor activity, the cells were pelleted and lysed with 120 µl of Cell Culture Lysis Reagent (Promega) and the luciferase activity measured on 20 µl triplicates by addition of Luciferase Assay Reagent (Promega). The luciferase activity produced by each sample was calculated relative to the activity produced by the prototype B95-8-LMP1.\n\nREFERENCES:\n1. KieffEFieldsBNKnipeDMHowleyPM\n1996\nEpstein-Barr virus and its replication.\nVirology\nPhiladelphia\nRaven Press\n2343\n2396\n2. KhannaRBurrowsSR\n2000\nRole of cytotoxic T lymphocytes in Epstein-Barr virus-associated diseases.\nAnnu Rev Microbiol\n54\n19\n48\n11018123\n3. KhannaRMossDJGandhiM\n2005\nApplications of emerging immunotherapeutic strategies for Epstein-Barr virus-associated malignancies.\nNature Clinical Practice Oncolcogy\n2\n138\n149\n4. KulwichitWEdwardsRHDavenportEMBaskarJFGodfreyV\n1998\nExpression of the Epstein-Barr virus latent membrane protein 1 induces B cell lymphoma in transgenic mice.\nProcNatlAcadSciUSA\n95\n11963\n11968\n5. HendersonSRoweMGregoryCCroom-CarterDWangF\n1991\nInduction of bcl-2 expression by Epstein-Barr virus latent membrane protein 1 protects infected B cells from programmed cell death.\nCell\n65\n1107\n1115\n1648447\n6. MillerWECheshireJLBaldwinASJrRaab-TraubN\n1998\nThe NPC derived C15 LMP1 protein confers enhanced activation of NF-kappa B and induction of the EGFR in epithelial cells.\nOncogene\n16\n1869\n1877\n9583684\n7. MillerWEMosialosGKieffERaab-TraubN\n1997\nEpstein-Barr virus LMP1 induction of the epidermal growth factor receptor is mediated through a TRAF signaling pathway distinct from NF-kappaB activation.\nJournal of Virology\n71\n586\n594\n8985387\n8. LamNSugdenB\n2003 Jun 16\nLMP1, a viral relative of the TNF receptor family, signals principally from intracellular compartments.\nEMBO J\n22(12)\n3027\n3038\n12805217\n9. IzumiKMKayeKMKieffED\n1997\nThe Epstein-Barr virus LMP1 amino acid sequence that engages tumor necrosis factor receptor associated factors is critical for primary B lymphocyte growth transformation.\nProc Natl Acad Sci U S A\n94\n1447\n1452\n9037073\n10. ShackelfordJMaierCPaganoJS\n2003\nEpstein-Barr virus activates beta-catenin in type III latently infected B lymphocyte lines: association with deubiquitinating enzymes.\nProc Natl Acad Sci U S A\n100\n15572\n15576\n14663138\n11. MorrisonJAKlingelhutzAJRaab-TraubN\n2003\nEpstein-Barr virus latent membrane protein 2A activates beta-catenin signaling in epithelial cells.\nJournal of Virology\n77\n12276\n12284\n14581564\n12. MorrisonJARaab-TraubN\n2005\nRoles of the ITAM and PY motifs of Epstein-Barr virus latent membrane protein 2A in the inhibition of epithelial cell differentiation and activation of {beta}-catenin signaling.\nJournal of Virology\n79\n2375\n2382\n15681438\n13. EverlyDNJrKusanoSRaab-TraubN\n2004\nAccumulation of cytoplasmic beta-catenin and nuclear glycogen synthase kinase 3beta in Epstein-Barr virus-infected cells.\nJournal of Virology\n78\n11648\n11655\n15479806\n14. TsaiCNTsaiCLTseKPChangHYChangYS\n2002\nThe Epstein-Barr virus oncogene product, latent membrane protein 1, induces the downregulation of E-cadherin gene expression via activation of DNA methyltransferases.\nProc Natl Acad Sci U S A\n99\n10084\n10089\n12110730\n15. LoAKHuangDPLoKWChuiYLLiHM\n2004\nPhenotypic alterations induced by the Hong Kong-prevalent Epstein-Barr virus-encoded LMP1 variant (2117-LMP1) in nasopharyngeal epithelial cells.\nInternational Journal of Cancer\n109\n919\n925\n15027126\n16. TanELSamCK\n2007\nBiological properties of TW01 cells expressing latent membrane protein-1 gene of EBV-derived from nasopharyngeal carcinoma cells at different stages of malignancy.\nExp Oncol\n29\n166\n174\n18004239\n17. DawsonCWTramountanisGEliopoulosAGYoungLS\n2003\nEpstein-Barr virus latent membrane protein 1 (LMP1) activates the phosphatidylinositol 3-kinase/Akt pathway to promote cell survival and induce actin filament remodeling.\nJ Biol Chem\n278\n3694\n3704\n12446712\n18. TellamJConnollyGWebbNDuraiswamyJKhannaR\n2003\nProteasomal targeting of a viral oncogene abrogates oncogenic phenotype and enhances immunogenicity.\nBlood\n102\n4535\n4540\n12920032\n19. HeTCSparksABRagoCHermekingHZawelL\n1998\nIdentification of c-MYC as a target of the APC pathway.\nScience\n281\n1509\n1512\n9727977\n20. TetsuOMcCormickF\n1999\nBeta-catenin regulates expression of cyclin D1 in colon carcinoma cells.\nNature\n398\n422\n426\n10201372\n21. MorrisonJAGulleyMLPathmanathanRRaab-TraubN\n2004\nDifferential signaling pathways are activated in the Epstein-Barr virus-associated malignancies nasopharyngeal carcinoma and Hodgkin lymphoma.\nCancer Research\n64\n5251\n5260\n15289331\n22. KaoRHHuangLCHsuYH\n2002\nMapping the methylation pattern by bisulfite genomic sequencing of the E-cadherin promoter CpG island in nasopharyngeal carcinoma.\nAnticancer Research\n22\n4109\n4113\n12553040\n23. TsaoSWLiuYWangXYuenPWLeungSY\n2003\nThe association of E-cadherin expression and the methylation status of the E-cadherin gene in nasopharyngeal carcinoma cells.\nEuropean Journal of Cancer\n39\n524\n531\n12751385"
+}

batch_0/PMC2532992.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2532992",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2532992\nAUTHORS: Bee Wee, Gina Hadley, Sheena Derry\n\nABSTRACT:\nBackgroundIn contemporary medical research, randomised controlled trials are seen as the gold standard for establishing treatment effects where it is ethical and practical to conduct them. In palliative care such trials are often impractical, unethical, or extremely difficult, with multiple methodological problems. We review the utility of Cochrane reviews in informing palliative care practice.MethodsPublished reviews in palliative care registered with the Cochrane Pain, Palliative and Supportive Care Group as of December 2007 were obtained from the Cochrane Database of Systematic Reviews, issue 1, 2008. We reviewed the quality and quantity of primary studies available for each review, assessed the quality of the review process, and judged the strength of the evidence presented. There was no prior intention to perform any statistical analyses.Results25 published systematic reviews were identified. Numbers of included trials ranged from none to 54. Within each review, included trials were heterogeneous with respect to patients, interventions, and outcomes, and the number of patients contributing to any single analysis was generally much lower than the total included in the review. A variety of tools were used to assess trial quality; seven reviews did not use this information to exclude low quality studies, weight analyses, or perform sensitivity analysis for effect of low quality. Authors indicated that there were frequently major problems with the primary studies, individually or in aggregate. Our judgment was that the reviewing process was generally good in these reviews, and that conclusions were limited by the number, size, quality and validity of the primary studies.We judged the evidence about 23 of the 25 interventions to be weak. Two reviews had stronger evidence, but with limitations due to methodological heterogeneity or definition of outcomes. No review provided strong evidence of no effect.ConclusionCochrane reviews in palliative care are well performed, but fail to provide good evidence for clinical practice because the primary studies are few in number, small, clinically heterogeneous, and of poor quality and external validity. They are useful in highlighting the weakness of the evidence base and problems in performing trials in palliative care.\n\nBODY:\nBackgroundIn contemporary medical research, randomised controlled trials are seen as the gold standard for establishing treatment effects where it is ethical and practical to conduct them. In palliative care, randomised controlled trials may be impractical, unethical, or extremely difficult, with multiple methodological problems. The fact and nature of these issues with palliative care trials has been frequently commented upon [1-3]. Frequently encountered problems include recruitment and attrition, insufficient numbers of patients for any comparison, clinical heterogeneity between patients (condition palliated, comorbidity), heterogeneity in treatments (intervention, dose, duration), different outcomes reported, and use of non-standard scales. A palliative care Outcomes Working Group has recently made recommendations on outcomes they consider to be important in this context and how they might be sought in clinical trials [4].Trials that have been done in palliative care are often small, diverse in nature and outcomes, and with high attrition rates, making meta-analysis, and even qualitative systematic review, impractical, unsatisfactory, or both. Moreover, some aspects of palliative care are difficult to capture, given the nature of palliative care as a person-centred approach, in which individual packages of care are often the norm [5].With this background, the value of systematic reviews of randomized trials in palliative care might be questioned. One side of the argument would be that without a sufficiency of trials satisfying criteria of quality, validity, and size [6] systematic reviews are worthless. Another would see systematic reviews as a necessary first step to obtaining more evidence; despite their limitations, they at least tell us what we don't know, and may indicate how to improve.This review set out to examine a subset of Cochrane reviews published under the auspices of the Pain, Palliative, and Supportive Care Review Group, to ascertain the number of successfully completed palliative care systematic reviews from this source over the last nine years, to assess their quality and the strength of the evidence presented to guide clinical practice.MethodsA list of published reviews relating to palliative care and registered with the Cochrane Pain, Palliative and Supportive Care Group was obtained from the Review Group Coordinator as of December 2007. Copies of each review were obtained from the Cochrane Database of Systematic Reviews using the most recent upload, issue 1, 2008.The following information was extracted from each review:• Number of studies included�� Number of patients included• Condition palliated• Intervention• Trial design (randomized, observational)• Measures of quality and/or validity used• Whether exclusions due to poor quality were made, or a sensitivity analysis presented• Whether a pooled analysis was done• Original authors' conclusion on efficacy• Original authors' conclusion on strength of evidence• Original authors' implications for future research.Two reviewers (GH, SD) independently carried out data extraction, using a standard form, and assessed the quality of each review using the Oxman & Guyatt Index of Scientific Quality [7]. To determine the strength of the evidence presented, for each review we assessed the quality of the included studies, based on randomization and blinding since these characteristics are known to affect potential bias [8], and the number of patients available for any analysis, because small numbers are prone to random error [9,10]. Any discrepancies were resolved by consensus.There was no prior intention to perform any statistical analyses. What was intended was an evaluation of this set of systematic reviews in palliative care based on the quantity and quality of primary studies available, and the quality of the review process itself, in order to determine their utility for informing clinical practice.ResultsDetails of the 25 published systematic reviews [11-35] are in Additional file 1, together with the conclusions of the original authors. The first of these Cochrane reviews was published in 1999, and the most recent in 2007. Sixteen of the reviews concerned drug interventions for pain or other reasons, three involved radiotherapy, three complementary therapy, and one each for a mineral supplement, supportive care, and pleurodesis. Only five of the studies were published before 2003, and the rate of publication was five per year since 2004.Primary studiesNumber of trials and patientsThe numbers of included trials ranged from none to 54. Thirteen had fewer than five controlled trials, and 16 had fewer than 10 trials. Three reviews had between 11 and 20 trials, and six more than 20 trials. Six reviews had information on fewer than 100 patients in total in controlled trials, fourteen had fewer than 500, while eight had between 1000 and 5000, and one more than 6,000 patients (see Additional file 1: Included reviews). Within each review, included trials were frequently heterogeneous, with differing interventions (drug, dose, route, technique) and reported outcomes, so that the number of patients contributing to any single analysis was nearly always much lower than the total number of patients included in the review.All the reviews sought randomised controlled trials for inclusion. Five reviews [22,25,26,30,34] sought uncontrolled studies, but only two analysed these in the absence of randomised trials [22,26]. Two reviews [28,31] found no studies that met their inclusion criteria.Types of patientsEighteen reviews included trials involving only cancer patients. In most cases the type of cancer or site of the primary cancer was not restricted. One review included only AIDS patients [27], two included mixed diagnoses of cancer, lung disease, cardiac failure, cystic fibrosis, and elderly patients [20,33], and one included patients with cancer or unspecified \"terminal illness\" [14].Original authors' assessment of quality of included studiesAll reviews with included studies assessed their quality, with the exception of Ballantyne [22] and Quigley [26], who found no randomised trials and included mainly retrospective studies, audits, or case reports, and uncontrolled prospective cohort studies. A number of scales were used. Most (18/23) used the Oxford Quality Score [8], and of these, three [16,29,30] additionally used the Oxford Pain Validity Scale [36], two [14,21] used Rinck [1], one [17] used Detsky [37], and another [15] used both Juni [38] and Delphi [39]. Shaw [24] graded trials according to criteria in the Cochrane Handbook [40], Feuer [34] according to Mann [41], and Ezzo et al [18] used their own set of five questions. Seventeen reviews also assessed allocation concealment using Cochrane criteria [42] in at least some of the included trials. Eight of the reviews that assessed trial quality did not use the information to exclude low quality studies, weight analyses, or perform sensitivity analysis for effect of low quality [11-15,23-25]. For details of quality scores of included studies see Additional file 2 (Adequacy of included studies), and of the quality scoring tools used see Additional file 3 (Quality and validity tools).The original authors themselves indicated that there were frequently major problems with the primary studies, individually or in aggregate. These included low numbers (either in total or available for pooled analysis) in 18 cases, the lack of useful outcomes in 10, methodological heterogeneity in eight, design problems in five, and clinical heterogeneity in two. For example, one review stated that we \" ... need more larger studies with standardised outcomes of clinical relevance and clearer definitions of best supportive care\" [21], while another stated that \"Trials were too ...... short term for results to be meaningful\" and that \"Clinically relevant questions to address include which compounds are most beneficial, optimal dose and administration route, when prophylactic therapy ... should be started ...\" [29].Reviewers' assessment of quality of reviewsThe methods used in these 25 reviews appeared to be sound. We attempted to use the Oxman & Guyatt Index of Scientific Quality [7], which asks questions about review methods. All the reviews had effective search strategies, and all looked at methodological quality in some way. However, deficiencies in the primary studies made judgment about assessment of validity and combining data close to impossible, as it was for the original authors. For instance, many reviews made no attempt to combine studies in a pooled analysis because of clinical heterogeneity and diverse interventions and outcomes, a decision that we felt to be correct.We also felt that an overall Oxman & Guyatt score for these reviews was inappropriate because it attempts to measure flaws in the reviewing process. Our judgment was that the reviewing process was generally good in these reviews. Limited amounts and quality of data limited conclusions about efficacy or harm, most importantly lack of patient numbers, poor/inconsistent reporting, frequent use of non-standard outcome measures, and excluding outcomes which lack clinical relevance, for example patient satisfaction and long-term morbidity.In our assessment of the strength of the evidence presented, we found that of the 25 reviews:• 2 had no data – there were no trials found [28,31];• 2 included uncontrolled trials [22,26], known to be the subject of significant bias [8];• 12 included randomized trials, but with open or non-blinded designs [12,14-17,19-21,23-25,35], again known to be the subject of bias, especially in pain [43];• 4 included randomized trials, with a mix of blind and open designs. Of these:○ Wong [32] included mostly double blind studies, with 3600 patients, but using different drugs, doses, and routes of administration;○ Nicholson [11] had 460 patients and 6/9 trials were double blind, but with different doses, and routes of methadone administration, and different comparators;○ Dewey [13] had 60 patients and 4/5 trials were double blind, but they were insufficiently rigorous to be confident of any effect;○ Ezzo [18] had 1250 patients in acupuncture trials, with a mix of techniques and controls. The trials and review have been criticized elsewhere [44];• 5 included randomised trials with only double blind design. Of these:○ Three had fewer than 100 patients [27,30,34];○ Roque [29] had only 325 patients in 4 trials using different drugs, and doses, in single or multiple dose schedules, and for different duration;○ Jennings [33] had 292 patients in 3 trials, but with different drugs and doses.Two reviews [20,35] were considered to have the strongest evidence, although even for these reviews there were limitations with methodological heterogeneity or definition of outcomes. No review provided strong evidence of no effect. Even reviews with relatively large numbers of trials and patients could not provide strong evidence because of inappropriate comparison or trial design [23] or methodological heterogeneity [17].DiscussionThis systematic review of systematic reviews in palliative care was to question the utility of systematic reviews for informing clinical practice in this area of medicine. It found that 25 reviews were published in the Cochrane Database of Systematic Reviews over nine years, a rate of about 2.7 per year overall, though almost double that rate occurred in the three years to 2007. Despite a respectable level of productivity from this prestigious source, 22/25 reviews could produce only weak evidence of the benefits of any intervention, and even of the two where the evidence was considered to be strong there were caveats.The review processes themselves appeared adequate. Deficiencies lay in the primary studies, which were either missing or scant, or were characterized by heterogeneity in the methods, interventions, patients, and outcomes, which made an overall assessment of benefit or harm impossible. These deficiencies are similar to those identified previously [1-3,5]. The authors of the reviews commonly commented on these deficiencies, and others. The biggest single issue was that of inadequate trials or inadequate patient numbers in high quality trials. In making even this point, the reviews and the reviewers make an important contribution.It is likely that these observations are general to systematic reviews in palliative care. We limited our investigation to reviews from the Cochrane Database published through the auspices of the Palliative Care group, but we would expect such reviews to be no worse, and perhaps better, than non-Cochrane reviews [45,46]. The restriction to Cochrane reviews should not limit any generalisability of these findings, especially as this reasonably sized body of reviews consistently makes the same, or very similar, points.These findings are not a surprise. The dearth of good quality primary studies in the field of palliative care is widely accepted, and those trials that have been done are often known to have weaknesses [1-3,5]. Together, these factors underline the limitations of the knowledge base upon which palliative care has to draw. Whether new guidance about outcomes to be measured in palliative care trials would make a difference [4] remains to be seen, but given the difficulties in design and conduct of palliative care trials, rapid change in the corpus of evidence is unlikely.The challenge for palliative care is the lack of evidence that is available to support it and the inordinate difficulties in obtaining evidence, for example difficulties with recruitment and attrition in an ill and vulnerable population. This has led to calls for a different framework for examining evidence [47]. Part of the problem is that nearly all randomised controlled trials examine single interventions, while in clinical practice that intervention will often form a small part of a much larger overall package of care [5]. Randomised trials of overall packages of care with small or incremental differences between them are unlikely to be able to measure small improvements, and an evaluation of systematic reviews of palliative care services [48] highlighted similar problems to those of palliative care interventions. High patient losses also make interpretation of randomised trials difficult. It is important that palliative care research moves away from dependence on randomized trials, and explores alternative study designs to identify the most effective treatments and packages of care for its patients.There may be alternatives. Nearly all the Cochrane reviews included only randomised trials, and the small number of reviews that did consider non-randomised studies found them to have many of the same problems as randomised trials, with an additional increased risk of bias. We know, from other areas of medicine, that high quality, well-formulated, and impeccably conducted large observational studies, can provide equivalent results to those obtained from randomised controlled trials [6,49,50]. To overcome the play of chance these good quality studies need to be large, and to minimise bias they need to be both prospective and inclusive (i.e. a whole population, or all patients attending a clinic in a defined time). Registry studies are studies based on information from registers that systematically record information from all individuals in a defined population. They can be entire populations, as in the death register in the UK, or all patients with a specific characteristic (eg twins) or condition (eg breast cancer) within a defined population. At least one large registry-based programme for continuous quality improvement aimed at cancer pain is ongoing in Italy [51]. An extensive search for observational studies in palliative care has been undertaken, with the aim of identifying good quality observational studies and aspects of their design that make them reliable and useful (Hadley et al., manuscript in preparation). The proven limitation of controlled trials in palliative care may make registry studies a more acceptable option in future.ConclusionCochrane reviews in palliative care are well performed, but fail to provide good evidence to guide clinical practice because the primary studies are few in number, small, clinically heterogeneous, and of poor quality and external validity. These reviews do, however, tell us how limited the evidence base is, and highlight common deficiencies in primary studies. There are well-documented problems with conducting valid randomised trials in this area, and it may be that for some questions more, and more clinically relevant, information can be obtained from other types of primary study, such as large registry studies.Competing interestsBW and SD have received research support from charities, government and industry sources at various times, but no such support was received for this work. No author has any direct stock holding in any pharmaceutical company.Authors' contributionsGH and SD were involved with planning the study, data extraction, analysis, and preparation of the manuscript. BW was involved with planning the study and preparation of the manuscript. All authors read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:Supplementary MaterialAdditional file 1Included reviews. Details of included reviews.Click here for fileAdditional file 2Adequacy of included studies. Details of design, size, and quality of primary studies in each review.Click here for fileAdditional file 3Quality and validity tools. Details of tools used to assess quality and validity in primary studies.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2532998.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2532998",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2532998\nAUTHORS: Sheryl L Szeinbach, Spencer E Harpe, P Brock Williams, Hanaa Elhefni\n\nABSTRACT:\nBackgroundTest results for allergic disease are especially valuable to allergists and family physicians for clinical evaluation, decisions to treat, and to determine needs for referral.MethodsThis study used a repeated measures design (conjoint analysis) to examine trade offs among clinical parameters that influence the decision of family physicians to use specific IgE blood testing as a diagnostic aid for patients suspected of having allergic rhinitis. Data were extracted from a random sample of 50 family physicians in the Southeastern United States. Physicians evaluated 11 patient profiles containing four clinical parameters: symptom severity (low, medium, high), symptom length (5, 10, 20 years), family history (both parents, mother, neither), and medication use (prescribed antihistamines, nasal spray, over-the-counter medications). Decision to recommend specific IgE testing was elicited as a \"yes\" or \"no\" response. Perceived value of specific IgE blood testing was evaluated according to usefulness as a diagnostic tool compared to skin testing, and not testing.ResultsThe highest odds ratios (OR) associated with decisions to test for allergic rhinitis were obtained for symptom severity (OR, 12.11; 95%CI, 7.1–20.7) and length of symptoms (OR, 1.46; 95%CI, 0.96–2.2) with family history having significant influence in the decision. A moderately positive association between testing issues and testing value was revealed (β = 0.624, t = 5.296, p ≤ 0.001) with 39% of the variance explained by the regression model.ConclusionThe most important parameters considered when testing for allergic rhinitis relate to symptom severity, length of symptoms, and family history. Family physicians recognize that specific IgE blood testing is valuable to their practice.\n\nBODY:\nBackgroundWith the prevalence of allergic rhinitis estimated at 21% – 23% for the European population and 20% – 40% for the western population, appropriate diagnosis and treatment of allergic rhinitis is of global importance [1,2]. Family physicians are usually first approached by patients experiencing symptoms; however, little information exists regarding the rationale to perform specific IgE blood testing, which parameters are most important, and the value of such testing. Given the need to determine if symptoms are truly attributed to allergic mechanisms, it is important that family physicians consider diagnostic testing in conjunction with a careful examination of patient history, clinical evidence, and environmental exposure factors to optimize patient care. The consequences of untreated symptoms can lead to multiple future complications while the consequences of misdiagnosis can lead to inappropriate treatments [3].Chronic rhinitis has detrimental effects on quality of life and work productivity [4,5]. Although medications may control symptoms in some patients, it is difficult to distinguish between allergic rhinitis and non-allergic rhinitis using clinical evaluation and medication trials. Two commonly applied methods are used to uncover an allergic etiology and identify possible causes. These include skin prick tests (SPT), and specific IgE tests that are thought to produce concordant measures on a dichotomous basis for specificity and sensitivity, as well as a propensity toward appropriate diagnoses in relation to the presence of specific IgE antibody levels [6,7]. Decisions to utilize these tests are influenced by experience, patient history, diagnostic accuracy and efficacy of the test, and how well test results relate to symptoms [8,9].When presented with patient complaints and bothersome symptoms that may or may not be related to allergic rhinitis, physicians rely on numerous strategies to make an appropriate diagnosis. How family physicians weigh the importance of these patient-related parameters when recommending specific IgE testing is largely unknown, yet instrumental to determine appropriate treatment and follow-up therapy. To address this research question, we used a trade off approach (conjoint analysis) to evaluate family physicians' preference to recommend specific IgE blood testing with respect to patient symptoms, family history, and medication use. A second approach using visual analog scaling (VAS) was added to validate and compare findings obtained from the conjoint analysis. Visual analog scales have been used extensively in clinical assessment to quantify patient perceptions of disease severity and the impact of symptoms on health [10,11]. Further evaluation was performed to determine if family physicians perceive that testing, as part of the care process was valuable to patient care. As healthcare gatekeepers, family physicians have the best opportunity to construct a baseline assessment of these patients to determine if current treatment strategies are effective, or if patients would benefit from a referral to an allergist or other specialist.MethodsStudy samplePrimary care (family) physicians in a southeastern state in the United States were identified through already established medical societies and physician mailing lists that were complied at the Recruitment and Retention Shared Facility at the University of Alabama, Birmingham. Mailing addresses, telephone numbers, fax numbers, specialty area, and practice affiliation was verified for 424 physicians in Alabama. From the list of 424 physicians, a sample of 150 physicians was randomly selected to participate in the study. Three separate mailings containing 50 questionnaires were sent by priority mail, one week apart to these physicians at their respective practice sites. The questionnaire package contained a letter of invitation to participate in the study, along with a self-addressed stamped envelope for the returned questionnaire. Thirty-two questionnaires were completed and returned during the first month, follow-up reminder calls were conducted three weeks after the initial mailing (77 calls were answered), and 14 surveys were faxed per request by providers. A total of fifty completed surveys were returned within two months. The estimated sample size of 50 was determined for this study following examples for studies with repeated-measures designs [12]. As an incentive, a gift certificate for a local department store was mailed to physicians who completed the questionnaire.Instrument DevelopmentTechniques to evaluate preference include standard gamble, alternate rating (e.g., visual analog) scale, and time trade-off [13]. Besides these techniques, conjoint analysis (a trade off approach among attributes) is another technique that is used to evaluate the importance of preference measurements [14]. Choices are usually presented in the form of profiles that are ranked or rated (e.g., recommend specific IgE testing – yes or no). The part-worth values (coefficients) for each attribute are obtained from the random effects logistic regression model analysis with repeated measures (50 responses × 9 profiles = 450 observations), which follows stated choice experiments based on choice theory [15,16]. Although developed in marketing research, the use of conjoint analysis in health care is becoming a valuable tool [17-20].In an attempt to simplify the conjoint exercise for this study, each attribute was assigned three levels (see Additional file 1). For example, symptom severity was assigned \"high,\" \"medium,\" or \"low,\" and symptom length appeared as symptoms for \"less than 5 years,\" \"symptoms for 10 years,\" and \"symptoms for more than 20 years.\" Family history included \"neither parent has allergic rhinitis,\" \"mother has allergic rhinitis,\" and \"both parents have allergic rhinitis.\" Medication use included the use of \"over-the-counter medications to control allergy symptoms,\" \"prescribed antihistamines,\" and the \"prescribed nasal spray to control allergy symptoms.\" Effects coding was used to construct the numerical values of the profile attributes. The \"low\" level was the reference value and was denoted as -1. The \"high\" level was denoted as +1.A one-third fractional factorial design using repeated measures was chosen to minimize the number of profiles to 9 thereby attempting to avoid respondent fatigue. Two additional profiles were produced manually as holdout profiles for use in validation [21]. Each profile portrayed an individual with a pre-determined set of allergic symptoms and clinical indicators. For the dependent variable, family physicians were asked to provide a \"yes\" or \"no\" response to whether they would recommend specific IgE blood testing for this patient. For the purposes of this study, which specific IgE blood test was used by family physicians was not important, or what type of test (food or inhalant) was performed.The hypothesis for this study was that the estimated part-worth values or coefficients, exponentiated to odds ratios in this study, for each of the four profile attributes were simultaneously equal to zero. In the next section, family physicians were asked to indicate if recommendations to test were influenced by managed care guidelines, value of testing, referral activities, familiarity with specific IgE testing, relation of test results to symptoms, and value of test to practice, the likelihood to use specific IgE testing using a ten-point scale \"less likely to test\" to \"more likely to test.\" Specific IgE blood testing was rated using a scale ('1' = not valuable to '6' highly valuable) for overall value, value compared to skin testing, and value compared to not testing at all. Demographic characteristics of participating family physicians, such as age, gender, years in practice, and practice site information, were elicited in the last section of the questionnaire. To assess questionnaire validity, participants provided an estimate of a patient's overall health status given the impact of various symptoms and clinical indicators with 0 being the \"Worst Possible State\" and 100 being the \"Best Possible State.\" The means for these items were compared to those rankings obtained from the conjoint exercise to offer additional information regarding testing.Data analysisDescriptive statistics were provided for demographic variables. The conjoint exercise data were analyzed using a random effects logistic regression model. This type of model was chosen since it produces standard errors that account for the intra-individual correlation. Assumptions of normality, linearity, and equal variances among the items were evaluated to ensure appropriate interpretation of statistical analyses. All statistical analyses were performed using Stata/SE version 9 (College Station, Texas, USA). An Institutional Review Board from the University of Alabama, Birmingham, granted approval for the study.ResultsDemographicsParticipating physicians (33% response rate) were more likely to be male, between 40 and 60 years of age and with about 20 years of clinical experience in a private practice setting (Table 1). Independent t-test revealed that among older physicians (> 50 years), those with 10 or more years in practice placed a greater value on specific IgE testing than not testing (n = 32; mean = 4.6) compared to those with less than 10 years in practice (n = 18; mean = 3.8; t = 2.2; P = 0.03).Table 1Demographic Characteristics of Family PhysiciansCharacteristicValue (n = 50)Age, y Mean (SD)49 (12.0) Range29 – 79Years in Practice Mean (SD)18.6 (12.4) Range2 – 52Gender, no. (%) Male35 (71.4) Female14 (28.6)Practice type, no. (%) Private/Independent43 (87.8) Managed care setting6 (12.2)Conjoint modelResults from the random effects logistic regression model are presented in Table 2. The interaction between study attributes and demographic characteristics was not significant. Attributes, that are more likely to influence decision to request specific IgE blood testing, were symptom severity, length of time having symptoms, and history for allergic rhinitis reported for both parents. Results reveal the log likelihood = -196.983, Wald χ2 = 94.03, P < 0.0001, with 448 observations for 50 physicians each physician evaluating 9 profiles, thus supporting the hypothesis that the impact of parameters on specific IgE blood testing are not perceived equally. Symptom severity had the greatest impact on physician decisions to test patients for allergic rhinitis (OR, 12.11; 95%CI, 7.1–20.7). Thus, one would expect that physicians would be 12 times more likely to consider the specific IgE blood test for patients with high symptom severity compared to patients with low symptom severity. Although not significant, other attributes such as length of symptoms and both parents having a history of allergic rhinitis influenced physician decisions to test (OR, 1.46; 95%CI, 0.96–2.2: OR, 1.44; CI, 0.95–2.2, respectively). However, some physicians may not be willing to trade among the alternatives when the decision involved a potentially dominant attribute, where symptom severity may be the only reason to recommend specific IgE blood testing. To assess the potentially dominant effect of symptom severity [22], two versions of the model were run – one containing profiles where symptom severity was present and one containing only those where symptom severity was absent (results not shown). In both situations, other parameter estimates were significant indicating the hypothesis that coefficients were simultaneously equal to zero was rejected regardless of the presence of the symptom severity.Table 2Results from the random effects logistic regression modelAttributeLevelOdds RatioP-value95% CISymptom severityHigh12.11<0.001a[7.1, 20.7]Medium1.460.281[0.84, 1.9]Low*0.06Length of symptoms>20 years1.460.073b[0.96, 2.2]5 years to 20 years1.390.074b[0.96, 2.2]<5 years*0.47History of allergic rhinitisBoth parents1.440.089b[0.95, 2.2]Mother only1.200.37[0.80, 1.8]Neither parent*0.58Medication useIntranasal corticosteroids1.120.586[0.75, 1.7]Prescribed antihistamines1.330.171[0.89, 2.0]OTC allergy medications*0.67*Baseline attribute levela p < 0.05; b p < 0.10Log likelihood = -196.983; Wald Chi square = 94.03; p < 0.0001448 observations for 50 individualsValidationTwo methods were used to validate the results of the conjoint exercise – the use of a holdout profile and an alternate rating method. First, was to estimate predictive validity for the holdout profile using the regression model developed from the 9 orthogonal profiles. The relation between the observed response for the holdout profiles and the predicted responses was then examined. The predicted values for the holdout profiles were quite similar to the observed value. The predicted mean probabilities were 82.7% and 78.4% compared to the observed values 70% and 78%, respectively. The differences were not significantly different (t-test; p = 0.162, p = 0.996, respectively) suggesting that the conjoint model exhibits acceptable internal predictive validity. Second, was the use of a VAS where participants responded to each item from the conjoint study presented separately. Lower mean scores obtained for each domain indicated that the particular domain represented choices that were less desirable to the respondent. Symptom severity (mean = 36.7; SD = 16.4) and symptom length (mean – 36.0; SD = 16.6) were ranked the worst followed by medication use (52.6; SD = 21.5), and family history (mean = 61.0; SD = 24.0), revealing consistent response patterns between the conjoint study and the VAS.Impact of testing issues on valueKaiser-Meyer-Olkin measure of sampling adequacy for the final principal components analysis was 0.82 and the significant (p ≤ 0.001) Bartlett test of sphericity supported the use of factor analysis for the items used to assess testing issues [23]. One factor was retained for testing issues accounting for 54.6% of the variance. Two items, difficulty in interpreting test results and insurance coverage were dropped from the analysis. The factor structure was further verified by reanalyzing the reliability of the dimension. Descriptive statistics (item's mean and standard deviation) and Cronbach's alpha for study items are presented in Table 3. Most noteworthy, was that physicians perceived that \"how well the test correlated with symptoms\" was given the highest score (mean = 7.6; SD = 1.9) with respect to \"more likely to use specific IgE testing.\" In addition, physicians perceived that specific IgE testing had significant (p ≤ 0.007) value overall, perceived value compared to not testing, and perceived value was comparable to skin testing. Cronbach's alpha for the remaining nine items for testing issues was 0.90 and 0.86 for the three items consisting of testing value, indicating a high degree of internal consistency or a high signal-to-noise ratio (i.e., error variance minimized) across individuals [24].Table 3Descriptive Statistics and Scale Evaluation for Issues and Test ValueMean (SD)Testing issues * Managed care practice guidelines5.4 (2.3) Patient's perceived value of the test6.5 (1.9) Reduced need to refer patients to allergists6.8 (2.0) Difficulty in interpreting test results4.4 (2.6) Familiarity with test use6.9 (2.3) Patient demand to have the test done7.0 (1.9) Type of allergic rhinitis (intermittent vs. persistent)6.6 (1.9) How well test results relate to symptoms7.6 (1.9) Value of testing to my practice6.9 (2.3)Testing value ** Overall value of specific IgE as a diagnostic tool3.9 (1.1) Compared to skin testing, usefulness of specific IgE blood testing3.9 (1.4) Compared to not testing at all, usefulness of specific IgE blood testing4.3 (1.2)* Issues – scale = 1 less likely to test, 10 = more likely to test; (measure of internal consistency of items – α = 0.90)** Value – scale = 1 not valuable; 6 = highly valuable; (measure of internal consistency of items – α = 0.86)Linear regression analysis was used to assess the relationship between testing issues and testing value. As hypothesized, results using composite scores for testing issues and testing value revealed a moderately positive association between these two dimensions (β = 0.624, t = 5.296, p ≤ 0.001) with (R2 = 0.39) 39% of the variance explained by the model.DiscussionAccording to our results, family physicians consider symptom severity to be the significant determinant, followed by symptom length and family history when recommending the use of specific IgE blood testing for patients suspected of having allergic rhinitis. Physicians in practice for 10 years or more placed greater value on specific IgE testing compared to those in practice for less than ten years. Moreover, results from VAS were consistent with findings from the conjoint study. Our findings were also corroborated in another recent study where VAS for symptom severity compared favorably with standard quality of life measures [25].Professional organizations such as the American Academy of Allergy, Asthma, and Immunology and the European Academy of Allergology and Clinical Immunology recognize that allergic disease is a major health concern often requiring specific allergen avoidance and treatment strategies that are based on positive findings from history and diagnostic testing [26,27]. Results from this study support the positions elicited from the Joint Task Force on Practice Parameters for Rhinitis and Allergic Rhinitis and its Impact on Asthma (ARIA) in that family physicians are capable of recommending specific IgE testing, using the test to confirm allergic disease and identifying possible allergens [28-30]. Also consistent with recommendations from the Joint Task Force, results from the VAS closely approximated the findings of the conjoint study, thus revealing the usefulness of VAS in clinical practice to assess symptom severity for patients suspected of having allergic rhinitis.Values for each item relating to patient perceptions of the test, patient demand to have testing performed, other clinical indicators, and the type of allergic rhinitis were summated to create a composite score. This composite score for testing issues yielded a moderately positive correlation with testing value, thus providing initial evidence that issues associated with testing and the process of care were linked to outcomes such as testing value. Moreover, positively framing the information describing the benefits of testing and the value of testing to patients is also known to influence their expectations of benefits [31].Limitations include a low response rate and a cross-sectional study representing one geographical region. In addition, family physicians may consider attributes that were not evaluated in this study when deciding to request specific IgE blood testing for patients suspected of having allergic rhinitis. Hypothetical profiles were developed for this study and may not include all aspects of information provided by patients to family physicians, reflect what happens in actual clinical practice, and represent the opinions of physicians in other geographical areas.Given the economic burden of allergic rhinitis on society and the research evidence that supports an inverse relationship between health status and specific IgE antibody levels [32-34], current guidelines should be repositioned and possibly modified to allow family physicians to have a more active role in specific IgE blood testing. Although ARIA suggests the SPT as a first line choice when further evaluation of patients is needed, interpretation of test results requires extensive training and experience. Thus, specific IgE testing was examined in this study as a practical choice for primary care physicians. As suggested from this study and supported in the literature, with proper training family physicians would become more adept at quantifying the results from specific IgE blood testing and recognizing when to refer patients (e.g., continued treatment failure, complications, and beyond scope of expertise) to allergists or other specialists [35-38]. Another important aspect of training is the need to consider specific IgE blood test and SPT results in the context of patient history, especially when discrepancy exists between test results and symptoms. Diagnostic testing, per se, is no substitute for a thorough examination of patient symptoms, health status, and medical history. In summary, allergists and family physicians understand that test results coupled with the findings of a careful clinical examination serve as the foundation to establish a strategy for treatment, from which future health outcomes can be evaluated to determine the success of treatment.ConclusionFamily physicians rely on symptom severity, and to some extent on length of time that symptoms are present and family history to determine whether patients should be tested to determine the presence of allergic disease. Physicians with more practice experience placed greater value on specific IgE testing. Findings also revealed a moderately positive association between the issues influencing the use of specific IgE blood testing and test value. Overall, family physicians valued specific IgE blood testing, especially compared to not testing.From the study findings, family physicians can use symptom severity as a gauge in clinical practice to determine if patients should undergo detection and testing for allergic rhinitis or related conditions perhaps much earlier during the process of clinical evaluation, especially in the presence of severe symptoms and a positive family history. Baseline evaluation will also increase the likelihood of determining the correct diagnosis and appropriate treatment, and to ascertain the need for referral. Future research is needed to address the impact of patient expectations and treatment experience on value and other outcome measures.Competing interestsThis study was supported by an unrestricted research grant from Phadia US Inc., Portage, Michigan.Authors' contributionsSLS conceptualized the study, examined the study design, performed the statistical analysis, and drafted the manuscript. SEH setup the study design, performed the statistical analysis, and drafted the manuscript. PBW conceptualized the study, examined the study design, and provided a critical assessment of the manuscript. HE coordinated and managed the collection of data for the study and reviewed the manuscript. All authors approved the manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:Supplementary MaterialAdditional file 1Testing for allergic disease: Parameters considered and test value. Sample profile used for data collection in the conjoint exerciseClick here for file\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2533018.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533018",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533018\nAUTHORS: Abuzer Dirican, Bulent Unal, Cuneyt Kayaalp, Vedat Kirimlioglu\n\nABSTRACT:\nIntroductionHydatid cyst disease is common in some regions of the world and is usually located in the liver and lungs. This report presents two cases of primary hydatid cysts located subcutaneously: one in the medial thigh and one in the left palm between the index and middle fingers.Case presentationsA 64-year-old male farmer visited our hospital because a swelling on the right medial thigh had grown during the last year. Superficial ultrasound and computed tomography revealed a lesion resembling a hydatid cyst. A germinative membrane was encountered during surgical excision. Pathological examination was compatible with a hydatid cyst. The second case involved a 67-year-old male farmer who complained of a swelling that had grown in his left palm in the last year. The preliminary diagnosis was a lipoma. However, a hydatid cyst was diagnosed during surgical excision and after the pathological examination. The patient did not have a history of hydatid cyst disease and hydatid cysts were not detected in other organs. There has been no disease recurrence after following both patients for 3 years.ConclusionA hydatid cyst should be considered in the differential diagnosis of subcutaneous cystic lesions in regions where hydatid cysts are endemic, and should be excised totally, with an intact wall, to avoid recurrence.\n\nBODY:\nIntroductionA hydatid cyst is a parasitosis caused by the larval form of Echinococcus granulosus or rarely Echinococcus alveolaris. The main hosts for E. granulosus are predators such as dogs, wolves, and foxes, while intermediate hosts include sheep, goats, and cattle. Humans are a coincidental intermediate host. The disease is more frequent in the Middle East, Central Europe, Australia, and South America, where the intermediate hosts are common. The organs affected most often are the liver (70%) and lungs (10–15%). Other locations are extremely rare [1]. Primary subcutaneous hydatid cyst is very rare and the incidence is unknown. In this report, we present two cases of primary hydatid cysts located subcutaneously: one in the medial thigh and one in the hand.Case presentationsA 64-year-old male farmer visited our clinic because of a swelling on the medial thigh that had grown during the last year. On physical examination, a mobile, painless, fluctuant, 8 × 9 cm mass was palpated. The overlying skin was normal. The only abnormality in the pre-operative laboratory examination was an increased erythrocyte sedimentation rate (ESR 60 mm/hour). The patient had no history of surgery for a hydatid cyst in another organ. Ultrasound (US) and computed tomography (CT) showed a lesion resembling a hydatid cyst (Fig. 1). During surgical exploration under spinal anesthesia, the skin and subcutaneous layers were incised and the cyst was reached. Hypertonic saline (3% NaCl) was injected into the cyst and after waiting for 10 min, the cyst was completely excised. A germinative membrane was seen during excision (Fig. 2). We thought that the cyst was fertile as it contained daughter cysts. The surgical site was irrigated with 40% povidone iodine (Betadine®) and hypertonic saline. The subcutaneous layers and skin were closed in the standard manner.Figure 1Subcutaneous hydatid cyst in the right medial thigh, displacing the muscles laterally.Figure 2Subcutaneous hydatid cyst in the right medial thigh.Histopathological examination revealed a hydatid cyst, but no additional hydatid cysts were observed on US or CT of the abdomen and thorax; the indirect hemagglutination test for hydatid cysts was negative. The patient was started on albendazole for 3 months (15 mg/kg/day). No findings associated with local or systemic hydatid cysts were detected during a 3-year follow-up period.The other case involved a 67-year-old male farmer who complained of a subcutaneous swelling inside the left palm between the index and middle fingers. Physical examination revealed a subcutaneous immobile 2 × 3 cm mass on the palmar side of the left hand between the thumb and index fingers. Surgical excision was planned with a pre-operative diagnosis of lipoma. A hydatid cyst was considered when a germinative membrane was seen during excision under local anesthesia (Fig. 3). We also thought that the cyst was fertile as it contained daughter cysts as in the previous patient. The cyst space was irrigated with 40% povidone iodine (Betadine®) and hypertonic saline. Total cyst excision and primary closure were performed, and histopathological examination revealed a hydatid cyst. The only abnormality in the pre-operative laboratory examination was an increased ESR (60 mm/hour). The patient had no history of surgery for a hydatid cyst in another organ, and no additional cysts were observed on US and CT of the abdomen and thorax. The indirect hemagglutination test for hydatid cysts was negative, and the patient was placed on albendazole for 3 months (15 mg/kg/day). No findings associated with local or systemic hydatid cysts were detected during a 3-year follow-up period.Figure 3Germinative membrane of cyst localized in the palmar site of the hand.DiscussionHere we report two cases of primary subcutaneous hydatid cysts both treated surgically. In a large series, the distribution of hydatid cysts outside the liver and lungs was reported as 9% of cases [2]. Chevalier et al. reported that the incidence of subcutaneous hydatid cysts was 2%, but some of the patients had hydatid cysts in other organs too [3]. Subcutaneous hydatid cyst may be secondary or primary. In secondary cysts, there is a primary location of hydatid disease like liver, lung, or spleen that is operated or not operated. Reports of primary subcutaneous hydatid cysts are very rare [4-6], and we were unable to find a case of a palmar hydatid cyst in a literature review. In our cases, the hydatid cysts were located subcutaneously, the patients had not undergone previous surgery for hydatid cysts, and no hydatid cysts were found in other organs. Therefore, our patients were diagnosed as having primary subcutaneous hydatid cysts.The mechanism of primary subcutaneous localization is unclear. After being ingested orally, under the action of gastric and intestinal enzymes, the oncosphere is released; it penetrates the intestinal wall, joins the portal system and reaches the liver. If the eggs attach to the liver, an hepatic hydatid cyst takes shape. Parasite eggs can pass to the systemic circulation and cause disease in other end organs. Larvae must pass through two filters (liver and lung) to form a solitary hydatid cyst, but that is very difficult. It is very possible that systemic dissemination via the lymphatic route accounts for cases with solitary cysts in uncommon sites [4]. Direct spread from adjacent sites may be another mechanism of infection provided a microrupture has occurred [7].Diagnosing hydatid cysts is very difficult in patients living outside the endemic regions. Because exposure to the contents of the cyst can cause problems such as anaphylactic reaction and local recurrence, making the diagnosis pre-operatively is important. The diagnosis of a palmar hydatid cyst was not considered in our second patient pre-operatively since the mass was very small and this localization is very rare. When the cyst contents were seen during excision, the possibility of a hydatid cyst was then considered. No anaphylactic reaction developed in either patient.The radiological findings of a thick cyst wall, calcification, daughter cysts, and a germinative membrane separate from the cyst wall are findings specific to hydatid cysts [8]. Our first case was diagnosed according to the appearance of the mass on superficial US and CT.Serology is a useful tool for the diagnosis. The indirect hemagglutination (IHA) test is positive in more than 80% of liver hydatid cysts. However, false negative IHA results can be higher in other located hydatid cyst. In those cases, more specific serologic tests are mandatory. A positive indirect hemagglutination test for hydatid cysts is significant, although negative test results do not indicate the absence of the disease, as in our patients. Therefore, the most important diagnostic tool is the awareness of the physician, particularly for the unusual presentation of the disease.The best treatment option is total surgical excision without opening the cyst. If the cyst cannot be excised without opening, the fluid contents should be removed, the laminated membrane should be totally excised, and the cyst pouch should be irrigated with protoscolicidal solutions [9]. Subcutaneous located cysts are more prone to rupture since they have not been diagnosed pre-operatively. We performed total cyst excision in both cases and irrigated the surgical areas with protoscolicidal agents. Identifying postoperative recurrence of the cyst in endemic regions is very difficult because the probability of formation of a new cyst is high. However, since our patients were still free of disease in the third postoperative year, any subsequent hydatid cyst formation may be considered to be a new infestation.ConclusionHydatid cyst should be considered in the differential diagnosis of subcutaneous cysts in regions where hydatid cysts are endemic. Total excision of the cyst with an intact wall is the best treatment.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsAD is the consultant surgeon who drafted the article and performed the operations. BU assisted in performing the surgery, took the pictures and helped revise the article. CK helped in acquisition of data and technical support. VK performed the literature search and helped in revision. All authors read, appraised and approved the final manuscript.ConsentWritten informed consent was obtained from the patients before publication of this case series and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2533021.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533021",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533021\nAUTHORS: Hao Wang, Jin-Song Liu\n\nABSTRACT:\nBackgroundLong terminal repeat retrotransposons (LTR elements) are ubiquitous Eukaryotic TEs that transpose through RNA intermediates. Accounting for significant proportion of many plant genomes, LTR elements have been well established as one of the major forces underlying the evolution of plant genome size, structure and function. The accessibility of more than 40% of genomic sequences of the model legume Medicago truncatula (Mt) has made the comprehensive study of its LTR elements possible.ResultsWe use a newly developed tool LTR_FINDER to identify LTR retrotransposons in the Mt genome and detect 526 full-length elements as well as a great number of copies related to them. These elements constitute about 9.6% of currently available genomic sequences. They are classified into 85 families of which 64 are reported for the first time. The majority of the LTR retrotransposons belong to either Copia or Gypsy superfamily and the others are categorized as TRIMs or LARDs by their length. We find that the copy-number of Copia-like families is 3 times more than that of Gypsy-like ones but the latter contribute more to the genome. The analysis of PBS and protein-coding domain structure of the LTR families reveals that they tend to use only 4–5 types of tRNAs and many families have quite conservative ORFs besides known TE domains. For several important families, we describe in detail their abundance, conservation, insertion time and structure. We investigate the amplification-deletion pattern of the elements and find that the detectable full-length elements are relatively young and most of them were inserted within the last 0.52 MY. We also estimate that more than ten million bp of the Mt genomic sequences have been removed by the deletion of LTR elements and the removal of the full-length structures in Mt has been more rapid than in rice.ConclusionThis report is the first comprehensive description and analysis of LTR retrotransposons in the Mt genome. Many important novel LTR families were discovered and their evolution is elucidated. Our results may outline the LTR retrotransposon landscape of the model legume.\n\nBODY:\nBackgroundTransposable elements (TEs) are mobile repetitive DNA that have been found in virtually all eukaryotic genomes investigated so far [1-3]. LTR retrotransposons are class I TEs that transpose in a \"copy and paste\" mode via RNA intermediates. Typical structural characters of a LTR retrotransposon include: 1) two highly similar LTR sequences from several hundred to several thousand bp; 2) 4–6 bp target site duplication (TSD) at its 5' and 3' ends; 3) primer binding site (PBS) downstream of 5' LTR and polypurine tract (PPT) upstream of 3' LTR; 4) protein-coding domains of enzymes important to retrotransposition, e.g. Capsid protein (GAG), Aspartic Proteinase (AP), Reverse Transcriptase (RT), Integrase (IN), and RNase H (RH). Sometimes Envelope protein (ENV) may occur as well [4]. In the plant kingdom, LTR elements present a significant fraction of many genomes and even make predominant components of large genomes [5-7]. The amplification and deletion of these elements is considered to be an important mechanism underlying the remarkable genome size variation in plants [8-11]. Moreover, LTR retrotransposons affect genome organization, gene regulation [12,13], novel gene origination [14,15] and other genetic functions. In summary, the dynamics of LTR retrotransposons are thought to be an important source of genome evolution.Medicago truncatula is a model plant of the Fabaceae, the third largest angiosperm family. Because of their vital role in agriculture and environment [16,17], legumes have provoked great interests. The identification and study of LTR elements is one of the basic and indispensable step to understand biology and evolution of this family. The sequencing of Mt opens an unprecedented opportunity to carry out a thorough study of it at the molecular level. Genomic data so far released have made it possible to explore many important facts of the Mt genome, specifically, the characteristics of LTR elements and their interactions with the host organism.In comparison with the Gramineae, the knowledge of LTR retrotransposons in the Fabaceae is relatively limited [18,19]. To date, a few Mt LTR families, e.g. MEGY and Ogre have been well documented [20-22] and some families have been deposited in Repbase [23] and TIGR Plant Repeat Databases [24]. However, little research has been focused on the comprehensive identification and description of LTR retrotransposons based on high-throughput Mt genomic sequences.Here we report the result of the computer-based analysis of LTR retrotransposons in 233 Mb Mt BAC sequences. At least 85 LTR families were found. We analyzed their phylogenetic relationship and structural patterns, with emphasis on several important families. We investigated the amplification-deletion pattern of these LTR elements and found that the removal of LTR elements in Mt has been more rapid than in rice, and more than 10 Mb of LTR retrotransposon sequences have been lost. The present work in about 41% of the whole-genome provides the LTR retrotransposon landscape in Medicago truncatula.Results and DiscussionEstimation of copy-number of LTR elementsThe development of ab initio algorithms [4,25,26] has greatly promoted the identification and analysis of LTR elements in large-scale genomic data. With predicted full-length elements at hand, a widely adopted method to find their related copies is to perform homology search against the host genome. However, subregions of full-length elements may be generated through insertion of other TEs, e.g. nested elements are caused by the insertion of LTR elements one into another [8,9]. In this report, we call these subregions unrelated sequences. Based only on the recognition of structural characters, ab initio prediction is not capable to provide information on such sequences. If an element contains unrelated sequences derived from highly abundant TEs, taking the direct matches as its copies will greatly overestimate its copy-number and exaggerate its contribution to the host genome. Therefore, we developed an algorithm (see Methods) to discriminate unrelated sequences and discard the matches generated by them (pseudo-copies). Using this method we obtained a more accurate estimation of the number of Mt LTR copies.Overview of LTR retrotransposonsThis survey identified 526 full-length LTR elements in 232996 Kb Mt BAC sequences [see Additional file 1]. The following validation process detected more than 16000 copies, corresponding to 22470 Kb sequences, about 9.6% of all the sequences scanned. If this value is kept in the unreleased genomic sequences, the percentage of the LTR retrotransposons in Mt is lower than that in rice (17%–22%) [7,27], yet still remarkably higher than that in Arabidopsis thaliana (1–2%) [28]. The length of the full-length elements is within the ranges of 364 bp to 18.7 Kb and that of the LTRs is 126 bp to 3.5 Kb [see Additional file 2]. Most of the elements showed canonical TG-CA boxes and 4–6 bp TSDs.In this report, We define the LTR family by DNA sequence similarity, following the suggestion of Wicker et al. [3]: two elements belong to the same family if they share 80% (or more) sequence identity in at least 80% of their coding region or internal domain, or within their LTR, or in both. A novel family is discovered when the following standards are met: 1) None of the members in the family belong to the same family with known legume LTR retrotransposons. 2) Besides a full-length member, the family has at least one strong hit (also called strong-hit copies. See Methods).The 526 full-length elements and their related copies thus were classified into 85 families, of which 64 were identified for the first time. The information of these families is listed in Table 1. LTR families are denoted as MtrXX (XX are digits) and the last two columns of Table 1 list the number of full-length and strong-hit copies because these two values provide multi-copy supports for a family. The following sections sometimes mention the total copy-number of a family and the length of its sequences. Such values, however, are estimated by all copies of that family, including full-length, strong-hit and other truncated ones.Table 1Summary of the 85 Medicago truncatula LTR familiesFamilyPre-existing nameSuper-familyBACLocation(bp)LTR size(bp)Element size(bp)FLStrong hitMtr1-CopiaAC150246.134464–4642413091196169184Mtr2COPIA4CopiaAC137553.4599838–1125352045126981129Mtr3-CopiaCR931729.271151–7720712876057522Mtr4-CopiaAC157648.1822659–284466275788915Mtr5-CopiaAC146866.6121051–1277486026698814Mtr6-CopiaAC151725.264572–1022679756551014Mtr7-CopiaCT967304.459158–648996395742713Mtr8-CopiaAC145027.1730490–371886766699812Mtr9SHACOP3CopiaAC159223.127730–327601895031611Mtr10-CopiaAC144482.11104300–10931543650161011Mtr11-CopiaAC165219.2137152–1418892734738710Mtr12-CopiaAC149637.117608–1841813441081157Mtr13-CopiaAC151956.565260–71649203639056Mtr14-CopiaAC144538.23111496–12198513941049056Mtr15COP20CopiaAC121235.2019414–24294346488156Mtr16-CopiaCR931743.132761–38060574530046Mtr17SHACOP12CopiaAC148291.2218569–23612213504455Mtr18-CopiaAC124217.2165448–738501419840324Mtr19-CopiaAC149492.1460016–7218418221216924Mtr20-CopiaAC145330.1954872–60067190519634Mtr21-CopiaAC144760.2799316–104667597535244Mtr22COP10CopiaAC144618.769511–74634324512444Mtr23COP6CopiaAC125476.3052976–57511212453624Mtr24-CopiaAC148470.1469233–73561250432933Mtr25SHACOP4CopiaAC152964.1348071–52925276485523Mtr26-CopiaCT573053.195022–10627516081125423Mtr27-CopiaAC140916.1754292–6546814341117713Mtr28-CopiaAC174349.1054241–58572126433233Mtr29COP12CopiaAC165446.1677640–82389351475023Mtr30MTCOPIA2CopiaAC149471.184193–89268186507623Mtr31-CopiaCT573028.1132334–34027204305913Mtr32COP3CopiaAC138465.2217893–22580181468833Mtr33MTCOPIA1CopiaCT963078.362740–67719269498023Mtr34SHACOP11CopiaAC154867.130673–35442189477013Mtr35SHACOP20CopiaAC149038.247760–52995263523623Mtr36-CopiaAC174295.736459–41224171476622Mtr37-CopiaAC157979.1169871–74887249501712Mtr38-CopiaAC171778.1579794–84281222448822Mtr39-CopiaAC148470.1454043–59356861531422Mtr40SHACOP9CopiaAC175047.2122584–127374190479122Mtr41-CopiaAC174336.998223–102259296470412Mtr42-CopiaAC157506.345694–50555205486212Mtr43-CopiaAC175047.2115144–119648256450512Mtr44-CopiaAC183304.1099523–104053167453122Mtr45-CopiaCR955009.1105072–109675306460422Mtr46COP14CopiaAC170583.612146–16446171430122Mtr47COP21CopiaAC182817.543694–48680290498722Mtr48SHACOP21CopiaAC161106.1330630–35152232452322Mtr49-CopiaAC158377.175346–80903131555822Mtr50-CopiaAC175047.283093–87466242437412Mtr51-CopiaAC123573.4160243–65229213498712Mtr52-CopiaAC135798.3110206–15241330503612Mtr53-CopiaAC137831.2731862–36604218478212Mtr54-CopiaCT963132.516390–21157239503012Mtr55-CopiaAC149634.814602–17988206338712Mtr56SHACOP2CopiaAC146909.2366018–71021274500422Mtr57Ogre1A,B,C,DGypsyAC144405.3085771–104542352918772114137Mtr58-GypsyAC162162.2396896–105667212787723469Mtr59Ogre2,3,4GypsyAC138465.2243789–603382435165503137Mtr60-GypsyAC123573.4134307–46826133512520226Mtr61-GypsyAC147430.969500–7813120728632110Mtr62-GypsyAC160097.2742464–4551720768661110Mtr63-GypsyCU024896.374551–822512222770136Mtr64-GypsyAC146759.2974288–8543512181114845Mtr65-GypsyCT963073.380019–85667313564945Mtr66-GypsyAC150705.1677136–842991943716424Mtr67-GypsyAC140773.2023404–376057211420234Mtr68-GypsyAC125481.2324806–33987330918223Mtr69-GypsyAC157375.285556–930732125751833Mtr70-GypsyAC144591.1013089–2966123701657313Mtr71-GypsyCU024896.371754–86395366693612Mtr72GYPSHAN2GypsyAC158209.1342449–47637409518922Mtr73-GypsyAC148657.169339–74484394514622Mtr74-GypsyCU024896.386399–92950819655222Mtr75-TRIMAC147537.35119057–119409130364974Mtr76-TRIMCR954194.112806–1612565233201738Mtr77-LARDAC155884.239928–4459521914668526Mtr78-LARDCT030253.1041580–463899215497326Mtr79-TRIMAC158173.1332092–32698190607525Mtr80-env-classAC146719.3260449–6839512397947520Mtr81-LARDCT025840.261376–73139241411764214Mtr82-TRIMCT009479.431485–345198103035710Mtr83-LARDAC140022.113262–1166014408399210Mtr84-LARDAC147201.1618119–239492595831610Mtr85-TRIMAC145219.16111276–11214824987316Phylogenetic analysis classified 74 families as Copia or Gypsy superfamily (Figure 1) and the rest 11, though quite abundant in genome, could not be categorized as either superfamily by their protein-coding domain organization or sequence similarity. We found that half of the 11 families had long ORFs in their internal domain and some ORFs showed a certain degree of homology with known TE proteins. These families were categorized into TRIMs or LARDs group (TRIMs-LARDs) by their length [3].Figure 1RT phylogenetic tree of 74 LTR families. A Bel-Pao type RT (BEL-1-I_NVp from Repbase) is used as outgroup. The 74 families are grouped into Copia or Gypsy superfamily. In the tree, each family is described by its name and a superfamily label. The superfamily label is given according to the order of domains in the POL. RT similarity and domain organization give consistent categorization. Mtr3, Mtr5 and Mtr62 lack other domains except RT, so they are categorized directly though RT similarity and are marked by the lowercase initials of the superfamilies. The 14 clades, to which the 74 families belong, are shown in the figure. The placement of Mtr3 and Mtr39 is unresolved and they are marked by grey dots [see Additional file 4].Protein domain organization and phylogenetic relationships within LTR familiesWe analyzed protein-domain organization of the 85 families (see Methods) and found 9 patterns (Table 2 and [see Additional file 3]). HMMER detected the canonical Copia and Gypsy domain structures, i.e. 5'-GAG-IN-RT-3' and 5'-GAG-RT-IN-3', in 35 and 14 families, respectively. Although it failed to detect the GAG proteins in 21 families, they could still be categorized by the order of their RT and IN in the POL. Mtr3, Mtr5 and Mtr62 had RT but not IN, and they were assigned to either superfamilies by sequence similarity of RT domain (The reason of this assignment is explained in next paragraph): Mtr3 and Mtr5 were categorized as Copia superfamily while Mtr62 as Gypsy superfamily. At last, we obtained 56 Copia-like and 18 Gypsy-like families. Using a RT of Bel-Pao element (BEL-1-I_NVp from Repbase) as outgroup, we constructed the NJ phylogenetic tree of the 74 families based on their RT similarity (Figure 1). The tree branched into two clades and this split was well supported (bootstrap value: 100%). In the tree, the superfamily label of each external node was given according to the order of domains. It is worth noting that the two clades consist of neither more nor less than members of two superfamilies, respectively. In other words, the categorization of superfamily based on RT similarity concurs with that based on domain organization. This means that RT similarity is enough to categorize LTR elements at the level of superfamily. In Figure 1, Mtr3, Mtr5, and Mtr62 are also drawn with others, but the tree topology do not change if they are deleted. These results support their earlier categorization by RT similarity.Table 2Domain organization of LTR families.SuperfamilypatternaNumber of familiesCopiaGAG-IN-RT35IN-RT19RTb2GypsyGAG-RT-IN14RT-IN3RTb1OtherGAGc3RTd1Other7a This table only shows the organization of GAG, IN and RT. See S3.1 for more information.b These elements are classified by RT similarity.c ORFs inside the three families share week similarity with some putative GAG proteins in UniProt.d This RT is from Mtr81 but can not be assigned to either superfamily.To reveal the phylogenetic relationships within the superfamilies, we collected from literatures RT domains of 158 reference elements representing known Eukaryotic LTR lineages [21,29-31] and combined these data with our 74 families to construct trees [see Additional file 4]. We found that the Copia-like families belonged to 10 clades. MTC-1, a new lineage composed of Mtr5, Mtr49 and Mtr55, was recognized with middle support (Bootstrap value: 62%). The Gypsy-like families belonged to 4 well defined lineages (Figure 1).PBS pattern of LTR elementsWe investigated TSD, PPT and PBS patterns of LTR elements. Although the TSDs and PPTs did not show significant sequence preference, we found clear tRNA usage bias through the PBS strings. The validation of a PBS sequence was to find the string which was located immediately downstream of the 5' LTR and a reverse complement to the 3' ends of a tRNA [32]. By the criterion of matching at least 14 bp, the PBSs were detected in 80 families, including all members in Copia and Gypsy superfamilies and 6 other families. We found that tRNAs corresponding to His, Phe, Tyr, Ser and Cys were never used as primer of reverse transcription and the majority of the rest 15 actually detected tRNA types occurred with low frequency. By contrast, the most-detected 4 types were used by nearly 3/4 families. Moreover, tRNAMet occurred in about 60% of the 80 families and was the most frequently used type in both superfamilies and TRIMs-LARDs. tRNAArg was the second important primer in Gypsy superfamily and was only used by this group (Table 3 and [see Additional file 3]).Table 3tRNA usage of LTR familiestRNA typeCopia superfamilyGypsy superfamilyOtherallMet346242Lys4127Leu5016Arg0404Ala1203Glu1203Val1203Asn2002Ile2002Tyr2002Asp0112Gln1001Gly1001Pro1001Thr1001SUM5618680Copia superfamilyThe present research discovered 64 novel families, including 38 Copia-like, 15 Gypsy-like and 11 other ones. We describe each group in the following three sections, with emphasis on some important families.The total copy-number of the 56 Copia-like families reached 4816 and their sequences had a total length of 7164 Kb, about 3% of all genomic sequences investigated. Full-length elements varied in length from 3 to 12.7 Kb, with an average of 5.9 Kb. The longest family Mtr2/COPIA4, which had more than 130 copies, was one of the most abundant Copia-like families. We detected 3 long ORFs in its internal domain. From 5' to 3', the first two ORFs encoded canonical GAG-POL proteins. The third ORF, located less than 1 Kb downstream of the POL region, encoded a protein longer than 790AA. Although the homology search against Uniprot [33] did not return significant match to this ORF, it showed quite high conservation among the family members (Figure 2b). Moreover, we found that a 258 bp subregion in this ORF matched the putative ENV protein of the Glycine max putative endogenous retrovirus SIRE1-8 with low significance (similarity: 22%, e-value: 0.003). These results indicated that this ORF probably encoded a protein related to putative plant ENV.Figure 2Structure of LTR families. Each sub-figure gives the structure of a family. the X-axis displays coordinates of nucleotides and the Y-axis displays average similarities among the full-length members of that family (calculated using the PLOTCON program in the EMBOSS package [46]). Grey stripes show the positions of the LTR and the domains detected. We display ORFs that are >500 bp in length. The arrow under a ORF label represents the length of that ORF. In Mtr67, the ORFs are found in both chains and their orientation is indicated by the arrows above the ORF labels. Sudden collapse of similarity (e.g. 1.8–2 Kb of Mtr8 and 2–3.5 Kb of Mtr10) is caused by the insertion or deletion events in one or two family members.Besides Mtr2, there were 6 Copia-like families longer than 10 Kb and all of them had the GAG, IN and RT domains. Mtr1 was the most abundant Copia-like family and the second largest in all the 85 families. It had 69 full-length members, more than 800 copies and its sequences reached more than 2.5 Mb in length, about 36% of the total length of the Copia-like copies. The percentage was 6.6 times as high as that of the second largest Copia-like family Mtr2, whose sequences were 392 Kb in length. The PBS of Mtr1 bounded tRNAMet and the length of its full-length members and LTRs were about 12 and 1.3 Kb, respectively. Outside Mt, the best match of its RT domain in Uniprot was from V. vinifera (Accession: A5BWH6). Similar to Mtr2, a third long ORF was also detected in its internal domain (Figure 2a). However, this ORF, matching a putative uncharacterized Mt protein (Q2HU06), was less conservative among the family members. The length of the LTRs of Copia-like families fell within the range of 126 bp to 2 Kb, with an average of 506 bp. LTRs could be roughly categorized into 4 zones: 100–500 bp, 600–850 bp, 1.2–1.45 Kb and >1.5 Kb. The number of families in these zones was 39, 8, 6 and 3, respectively [see Additional file 2]. We found that long elements tended to have long LTRs. In fact, almost all LTRs longer than 1 Kb were from elements longer than 10 Kb. Mtr3 and Mtr18 were exceptions. The length of the LTR of Mtr3 reached 1.28 Kb yet that of the full length was only 6 Kb. As one of the most abundant Copia-like retrotransposon, Mtr3 had 292 Kb sequences in the genome, It was a typical non-autonomous family since no >500 bp ORF could be detected in its internal domain and its RT degenerated to a fragment of 40AA.There were 11 Copia-like families whose LTRs and full-length sequences were shorter than 200 bp and 5 Kb, respectively. 6 of them have been deposited in Repbase. Mtr30/MTCOPIA2 and Mtr32/COP3 were quite active in the genome and their copies corresponded to more than 200 and 109 Kb sequences, respectively. Despite short LTRs, we detected 5'-IN-RT-3' domains in all of them and GAG proteins in 7.Gypsy superfamilyThe 18 Gypsy-like families corresponded to 11652 Kb sequences and constituted 5% of all genomic data. This group had a full length of 5.1 to 18.7 Kb, with an average of 9.8 Kb. Their LTRs were from 313 to 3.5 Kb in length, with an average of 1.5 Kb. Compared with Copia-like LTR retrotransposons, Gypsy-like elements were longer in general: the great majority of Gypsy-like families had a full length > 6 Kb and LTR >500 bp, while the full length and the LTR of most Copia-like families were < 6 Kb and 500bp, respectively [see Additional file 2]. The total copy-number in this superfamily was 7434, about 1.5 times more than Copia superfamily (4816). Despite fewer members, Gypsy superfamily contributed more to the Mt genome than Copia superfamily because of longer length and more active amplification in the past.Mtr57 and Mtr59 were Ogre families [20,22] and Mtr57/Ogre1 was the largest in all the 85 families. Our survey detected its 114 full-length members and more than 1000 copies in total. The length of its sequences reached 4.2 Mb. Mtr70 was closely related to Mtr57 in the phylogenetic tree, and was the second longest in all the families (The longest one is Mtr57/Ogre1). These two families used tRNAArg as primer and their internal domain encoded 5'-GAG-RT-RH-IN-3' proteins. We detected an ORF of 1527 bp located upstream of the GAG and an intron in the POL. Such phylogenetic and structural features well support that Mtr70 is a novel Ogre family. Mtr58 was the second largest family in the Gypsy group and the third largest in all the families. It had more than 600 copies in total and corresponded to 1.4 Mb sequences, about 1/3 of Mtr57/Ogre1. Its full length were about 8.8 Kb and the LTRs were 2.1 Kb in length. Its internal domain encoded 5'-GAG-PRO-RT-RH-IN-3' proteins (Figure 2f) and its PBS matched tRNAMet well. Phylogenetically, this family belonged to the Tekay clade.Mtr67, Mtr60 and Mtr64 were the other 3 families longer than >10 Kb. Similar to Copia superfamily, it was found that, when the full-length of an element was >10 Kb, the LTR was correspondingly >1 Kb. The only exceptional family Mtr67 had a LTR of 720 bp. Besides normal 5'-GAG-RT-RH-IN-3' domains, this family had 5 additional >500 bp ORFs downstream of the POL. They were all located in the complementary chain and had no match in Uniport. However, these ORFs were quite conservative among the family members (Figure 2g). We estimated that these ORFs were derived from other sources and later captured by Mtr67. The short LTR reflected short original length of this family. Mtr60 belonged to the Athila clade (Figure 1). Its PBS bound tRNAAsp and its protein-coding domains organized as 5'-GAG-RT-IN-3'. Downstream of the POL, there were two >500 bp ORFs encoding uncharacterized proteins (best match in Uniprot: A2Q2P5 and A2Q2P6). Mtr64, the sister branch of Mtr60, also had an extra ORF downstream of the POL. Its best match in Uniprot was from Garden asparagus (Q2AA44) and it shared weak similarity with the first extra ORF in Mtr60 (Similarity: 24%, e-value: 2e-06). Known elements of the Athila clade were putative plant endogenous retroviruses, thus the possibility that the extra ORFs in Mtr60 and Mtr64 encoded the putative ENVs was strong, although they did not share significant similarity with the putative ENVs of known Athila elements.Mtr65 and Mtr74 were from the CRM clade, while Mtr72/GYPSHAN and Mtr73 were from Renia. Families of these two clades were relatively inactive in Mt: each had a copy-number <50 and corresponded to <80 Kb sequences. Even so, they all showed multiple domains in their internal domain [see Additional file 3].TRIMs and LARDsBecause their internal domain lacked strong homology to any known TE proteins, the 11 families were required to have at least 5 strong hits in the genome. The total sequences of them were 3654 Kb in length, about 1.6% of all data. According to the suggestion of [3], the 5 families that had a length less than 4 Kb were classified as TRIMs and the other 6 as LARDs.We detected ORFs longer than 700 bp in 4 families: Mtr80 had 2 such ORFs. One shared weak similarity with a GAG protein in rice (Q7XRT, 35%, 4 × 10-92) and the other with a putative transposon protein in A. thaliana (Q9XH30, 27%, 2 × 10-8). Although HMMER failed to detect (e-value: 10-6) RT and IN domains in 4 of the 5 full-length members, previous analyses suggested that Mtr80/MEGY belonged to a distinct clade called env-class [21]. Mtr78, Mtr82 and Mtr76 each had only one ORF >700 bp. The best Uniprot match of the ORF in Mtr78 was again Q9XH30 (24%, 2 × 10-9), while that of the ORF in Mtr82 was an uncharacterized protein from grape (AB5PC1, 46%, 3 × 10-57). The ORF in Mtr76 was highly similar to that in Mtr82 (80.1%) and a subregion in this ORF was reported as a GAG by HMMER (Figure 2h).Mtr81 is the longest family in the 11. Although its internal domain was long, most members in this family rarely contained long ORFs. Searching the internal domain against Uniprot with BLASTX retrieved a 438 bp region homologous to some RT proteins and the best well-studied match was from the maize (Zea mays) Opie element (Q8H7T1, 60%, 1 × 10-48). The high similarity with Opie strongly supported that this ORF did encode a RT domain. However, it could not be detected by HMMER and this indicated that this RT might be not built in current RT profiles. Phylogenetic analysis further revealed that it belonged to neither Copia nor Gypsy superfamily but was close to the outgroup [see Additional file 5]. Further investigation is needed to fully resolve its position.The last family that shared homology with known TE proteins was Mtr85. It had a ~321 bp ORF encoding a fragment of RH (A4PUT7, similarity: 91%, e-value: 1 × 10-49). Each of the other 5 families had more than 10 strong hits and quite large copy-number in the genome. Mtr75 was the shortest family, of which the full-length members and the LTRs were only 364 and 130 bp in length, respectively. Instead of typical 5'-TG-3', 5' ends of its LTRs were 5'-TA-3'. Despite highly degenerated internal domain, this family had 9 full-length members and 74 strong hits. Mtr77, Mtr83 and Mtr84 were LARDs. Similar to Mtr75, Mtr77 was an abundant non-autonomous family with highly degenerated internal domain (about 200 bp).Structure of LTR retrotransposons in MtWe studied the structure of the LTR families and Figure 2 displays the structures of 8 highly abundant ones, five of which have been described above. The structure of Ogre elements is not shown because it has been reported previously [22]. Here we just point out that the LTR regions of most families tend to be less conservative among the family members in comparison with TE proteins, as well as many long ORFs. This result well supports that LTRs are the most rapidly evolving regions in LTR retrotransposons [3,34].Insertion-deletion of LTR retrotransposons in MtMtr1, Mtr2, Mtr6, Mtr10, Mtr57-59 and Mtr76 each had more than 10 full-length members. The total number of these full-length copies was 296, making up 56.3% of all the full-length elements identified. Their total copies constituted 45.3% of the LTR retrotransposon sequences.Paleontology analysis on the 296 elements revealed that they were quite young: all were inserted within the last 2 MY and 90% within 0.4 MY. Compared with others, the active period of Mtr6 (10 full-length copies) and Mtr76 (17 full-length copies) was relatively long [see Additional file 6]. Recent researches have argued that truncated LTR elements were mainly caused by unequal homologous or illegitimate recombination within genome and the result of recombination was the deletion of genomic sequences [27,32,35,36]. We estimated the deletion of LTR copies in Mt genome from two aspects: 1) the deletion of full-length structure and 2) the number of DNA loss. The deletion of a full-length structure means that mutation and recombination remove so many structural characters of a full-length element that it can not be recognized any more. Assuming that repetitive sequences are removed at a constant rate, the survival time of full-length structure obeys an exponential distribution and therefore the half-life is an index to estimate the speed of removal. With this method, [32] estimated that the half-life of the full-length elements in rice was about 0.79 MY.We calculated the insertion date of all the 526 full-length elements and found that 90% of them inserted within the last 0.52 MY (Figure 3a). Fitting of the distribution to the exponential function obtained α = -2.71, which corresponded to a half-life τ = 0.26 MY (Figure 3b). The bootstrap revealed that the half-life varied between 0.24 to 0.3 MY. To compare the speed of removal in legume and grass, we calculated the half-life for 705 full-length elements in the two sequenced rice genomes (Oryza sativa indica ssp. and japonica ssp.). Elements in rice and Mt were predicted under the same parameters (Hao Wang, unpublished data). As can be seen from Figure 3c and 3d, our data supported that the half-life of full-length structure in rice was ~0.4 MY, a lesser value than the estimation of [32] but still greater than that in Mt. Furthermore, statistical testing revealed that the insertion dates in the two species were from different distributions (Kolmogorov-Smirnov test, P-Value: 3.4 × 10-14). If the mean substitution rates of LTR elements in Mt and grass are approximately the same [37], the above results support that the full-length structures have been deleted more rapid in Mt than in rice. If the deletions have been occurred randomly in genome, the results further indicate that the removal of LTR elements in Mt has been more rapid than in rice.Figure 3Half-life of full-length LTR retrotransposons in Mt and rice. 526 Mt and 705 rice full-length elements are analyzed. Each bin represents 0.1 MY. Vertical bars under the histogram represent insertion events. a) The distribution of the insertion date of Mt elements. Fitting of this distribution to a exponential curve indicates that the insertions in the recent 0.1 MY have been significantly active. b) Fitting the dates to the exponential curve. The logarithm of the dates fits the straight line y = 0.52 - 2.71x well. Therefore the rate of the exponential curve is α = -2.71, which corresponds to a half-life of 0.26 MY. c) and d) display the fitting in rice, which gives a half-life of 0.4 MY.The total number of the strong hits was only 6% of all detected copies, but their size reached 42%. This indicated that LTR elements in Mt were highly fragmented and these truncated copies, great in number, might be generated by the removal of genomic DNA. If the truncated LTR copies were real vestiges of paralogous copies of families and if they had similar lengthes to the representative copies at the time of insertion, the difference of the length of truncated and representative copies provided the amount of deleted DNA since their insertion [36]. The estimation of the upper and lower limits of DNA loss could be as follows: we used the copies of Rset (see Methods) to estimate the lower limit. The data revealed that 5.5 Mb sequences have been deleted. Since Rset only consisted of not-so-severely truncated copies, it caused an underestimate of DNA loss. In contrast, we used all the truncated copies to estimate the upper limit and this gave more than 46 Mb of DNA has been removed. Since only 40% of the genome was analyzed here, we estimated that more than 10 Mb of LTR retrotransposon sequences have been deleted from the Mt genome.ConclusionWe have systematically identified and described LTR retrotransposons in nearly half of the Medicago truncatula genome, investigated their classification, structure, evolutionary dynamics and impact on the evolution of the host genome. The present work has provided a LTR retrotransposon landscape for this model legume. The sequencing of other species such as Lotus japonicus and Glycine max will provide great opportunity to study comparatively the evolutionary dynamics of LTR families in two or more legume organisms and further explore the interactions between these elements and their host genomes.MethodsGenomic sequences, LTR element databases and tRNA databaseThe Mt genomic data were composed of 1826 BACs, about 233 Mb in length. The data were downloaded from Medicago truncatula Sequencing Resources website (Version 1.0. Released on July, 2006) [38].A database of known legume LTR retrotransposons was constructed by extracting legume elements from literatures [21,29-31], Repbase [23] and TIGR Plant Repeat Databases [24]. This database was used to discriminate previously reported families from novel ones discovered in this research. A Mt tRNA database was also built by scanning the genome with tRNAscan-SE [39]. It was used to detect PBS of LTR elements by LTR_FINDER, our newly developed ab initio tool for the prediction of full-length LTR retrotransposons [25].Mining LTR retrotransposons in Medicago truncatula genomeWe first identified candidates of full-length element with LTR_FINDER, then annotated other copies related to them in the genome by homology search and the elimination of pseudo-copies. At last, only the candidates that had multiple copies were kept as LTR retrotransposons for further analysis.The initial LTR_FINDER scan retrieved more than 600 candidates. They were then subjected to the following steps to validate copies.1. Selected reliable LTR copies of candidates (Rset). The Mt genome was searched against each LTR candidate and all the matches longer than 100 bp were taken as the basic set of copies related to that candidate. This basic set were partitioned into two sets: Part I and Part II. Part I (called Rset) consisted of matches that covered both the LTR region and internal domain to a certain length (Figure 4), which, obviously, was a subset of all copies because it excluded severely truncated ones. Part II consisted of other matches. Since this set might contain pseudo-copies generated by unrelated sequences, it is processed by the following two steps to eliminate pseudo-copies.Figure 4Homologous matches of an candidate. The lines represent matches generated by whole-genome homology search of a reference candidate. Some matches are made of several pieces (segments on the same horizontal line). All the matches are categorized into Part I and Part II. Part I (Rset) consists of the matches that cover both the LTR region and the internal domain. They are reliable copies of the reference candidate. Part II is further classified into pseudo-copies and \"copies in part II\". Pseudo-copies are the matches that correspond to unrelated sequences. Unrelated sequences (dark grey regions) are the subregions that have significantly high matches (grey stripes) or that match some LTR elements well (not showing here). At last, \"copies in part II\" and Rset are combined to obtain the total copies of the candidate.2. Detected unrelated sequences derived from other LTR elements and eliminated their pseudo-copies. The internal domain of each candidate was searched against all the candidates to find whether it contained other LTR retrotransposons. The subregions derived from other candidates were recorded as unrelated sequences. Accordingly, the matches generated by such subregions were eliminated from Part II.3. Detected unrelated sequences derived from other TEs and eliminated their pseudo-copies. When an unrelated sequence was derived from an abundant TE, it would have many matches in the genome. Therefore, if the subregions of the internal domain had significantly high number of matches, the possibility that they were unrelated sequences was high (Figure 4). We used a sliding window to find such subregions:(a) A window of size w (say 100 bp) moved from 5' to 3' along the internal domain and stopped at the first position that at least k (say 20) hits were found in it. Here a hit was a member of Part II that covered at least 80% of the window.(b) The window extended 1 bp and was checked if it still had k (or more) hits. The extension continued until less than k hits were in the window, then the current window was marked and this completed a search cycle. the next cycle started from the next position to the 3' end of the window. The search continued until the window reached the 3' end of the internal domain.(c) All of the marked windows were checked to filter those that obviously overlapped with TE proteins. Subsequently, the remaining windows were connected if the distances between them were less than a threshold (say 10 bp). At last, the regions covered by such windows were taken as unrelated sequences and their corresponding matches were eliminated from Part II (Figure 4).Although this simple greedy algorithm might fail to deal with a few complicated situations, it identified most of the pseudo-copies efficiently.4. Obtained copies of candidates. Rset and the remaining copies in Part II were combined to obtain the copies of the candidates. If one locus in the genome matched several candidates, it was assigned to the best matched one. At last, a full-length candidate was taken as a LTR element if it had several hits covering at least 80% of it. Such hits are called strong hits or strong-hit copies. After the above validation, a total of 526 full-length elements and their 16565 related copies were selected. To be reliable, the above validation excluded the matches that were shorter than 100 bp. This might skip some severely truncated copies and thus brought some underestimation of DNA loss and the contribution of LTR retrotransposons to the genome.Subsequently, full-length elements were categorized into families by their sequence similarity [3], and the copies of each family were obtained by combining all the copies of its full-length members. The annotation process identified unrelated TE sequences in elements and discarded pseudo-copies, thus estimated the contribution of LTR elements to the genome more accurately.TE domain identificationLTR_FINDER tried to detect ORFs of RT, IN and RH in the full-length elements by calling PS_SCAN [25,40]. Besides this, we scanned them (e-value: 10-6) with the hmmsearch program in the HMMER package [41] to locate positions of important domains. The profiles were downloaded directly from Pfam (V22.0) [42]. According to the suggestion of [26], TE domains were represented by the following profiles: RT by PF00078, PF07727 and PF05380; IN by PF00665, PF00552 and PF02022; PRO by PF00026 and PF00077; RH by PF00075; GAG by PF03732 and PF00098; and ENV by PF03078. We note that the scan process may skip some domains if they are highly divergent among different retrotransposon families or not built in Pfam profiles.Phylogenetic and statistical analysisThe phylogenetic tree and multiple alignments were constructed by CLUSTALW [43]. The tree was edited with MEGA4 [44]. Statistical analyses were performed by R [45]. Following the suggestion of [37], 1.3 × 10-8/site/yr was used as the average substitution rate of Mt LTR elements to obtain the insertion date of elements.Authors' contributionsHW designed and carried out the LTR retrotransposon studies, participated in the design of LTR retrotransposon mining program and drafted the manuscript. J-SL participated in the design of the program of LTR copy validation. All authors read and approved the final manuscript.Supplementary MaterialAdditional file 1Sequences of 526 full-length LTR elements identified in this study. this file contains the sequences of all full-length LTR elements identified in this study.Click here for fileAdditional file 2Distribution of the length of Copia and Gypsy superfamilies. this file contains two figures showing the distributions of the full-length and LTR length of Mt elements.Click here for fileAdditional file 3Domain structure and PBS usage of LTR families. this file provides the information to relate the LTR families with their protein domain structures and reverse transcription primer tRNAs.Click here for fileAdditional file 4Phylogeny of Copia- and Gypsy-like LTR families. this file contains the phylogenetic analysis of Copia and Gypsy superfamilies.Click here for fileAdditional file 5Phylogenetic position of Mtr81. the RT sequence of Mtr81 do not belongs to Copia or Gypsy superfamily. It is placed as a third branch the phylogenetic tree.Click here for fileAdditional file 6Insertion dates of 8 abundant LTR families. this file contains a figure showing the distribution of the insertion time of 8 abundant LTR families.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2533035.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533035",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533035\nAUTHORS: Rosa M. Ayala-Lugo, Hemant Pawar, David M. Reed, Paul R. Lichter, Sayoko E. Moroi, Michael Page, James Eadie, Veronica Azocar, Eugenio Maul, Christine Ntim-Amponsah, William Bromley, Ebenezer Obeng-Nyarkoh, A. Tim Johnson, Theresa Guckian Kijek, Catherine A. Downs, Jenae M. Johnson, Rodolfo A. Perez-Grossmann, Maria-Luisa Guevara-Fujita, Ricardo Fujita, Margaret R. Wallace, Julia E. Richards\n\nABSTRACT:\nPurposeTo evaluate the extent to which mutations in the optineurin (OPTN) glaucoma gene play a role in glaucoma in different populations.MethodsCase-controlled study of OPTN sequence variants in individuals with or without glaucoma in populations of different ancestral origins and evaluate previous OPTN reports. We analyzed 314 subjects with African, Asian, Caucasian and Hispanic ancestries included 229 cases of primary open-angle glaucoma, 51 cases of juvenile-onset open-angle glaucoma, 33 cases of normal tension glaucoma, and 371 controls. Polymerase chain reaction-amplified OPTN coding exons were resequenced and case frequencies were compared to frequencies in controls matched for ancestry.ResultsThe E50K sequence variant was identified in one individual from Chile with normal tension glaucoma, and the 691_692insAG variant was found in one Ashkenazi Jewish individual from Russia. The R545Q variant was found in two Asian individuals with primary open-angle glaucoma; one of Filipino ancestry and one of Korean ancestry. In addition to presenting OPTN allele frequencies for Caucasian and Asian populations that have been the subject of previous reports, we also present information for populations of Hispanic and black African ancestries.ConclusionsOur study contributes additional evidence to support the previously reported association of the OPTN E50K mutation with glaucoma. After finding an additional 691_692insAG OPTN variant, we can still only conclude that this variant is rare. Combined analysis of our data with data from more than a dozen other studies indicates no association of R545Q with glaucoma in most populations. Those same studies disagree in their conclusions regarding the role of M98K in glaucoma. Our analysis of the combined data provides statistically significant evidence of association of M98K with normal tension glaucoma in Asian populations, but not in Caucasian populations; however, the validity of this conclusion is questionable because of large differences in allele frequencies between and within populations. It is currently not possible to tell how much of the underlying cause of the allele frequency difference is attributable to demographic, technical, or ascertainment differences among the studies.\n\nBODY:\nIntroductionThe optic neuropathy called glaucoma is the second most common cause of bilateral blindness in the world [1]. The most frequent form of glaucoma is open-angle glaucoma (OAG), which can occur as adult-onset primary open-angle glaucoma (POAG) or juvenile-onset primary open-angle glaucoma (JOAG). Prevalence of OAG has been measured at higher frequencies in individuals of African ancestry than in those of European or Asian ancestry [2]. Although the most common form of OAG in the US involves elevated intraocular pressure (IOP), OAG also develops in individuals whose IOP is never observed outside of the normal range (normal tension glaucoma, NTG). Associated risk factors for OAG include race (i.e., population genetic factors, particularly ethnic ancestry), a family history of glaucoma, increasing age, and an IOP elevated above the normal range.During the last decade, genetic mapping and cloning experiments have demonstrated that glaucoma has substantial genetic components [3]. Among the more than one dozen mapped glaucoma loci, 11 GLC1 loci cause OAG [3], two GLC3 loci cause congenital glaucoma [4,5], and the remaining known loci are responsible for secondary and developmental forms of glaucoma [3]. Additional genetic risk factors for differential severity of OAG have been reported [3].Mutations in the myocilin (MYOC) gene at the GLC1A locus are found in 2.6-4.3% of POAG cases [6] and up to one-third of the familial JOAG cases [7]. There has not yet been a follow-up on a recent report that mutations in WD repeat domain 36 (WDR36) that occur in cases of POAG that map to the GLC1G locus on chromosome 5 [8].Rezaie et al. [9] described mutations in the optineurin (OPTN) gene, located within the GLC1E interval at 10p15-p14 [10], in 16.7% of families from a predominantly NTG population. The original report of OPTN involvement in glaucoma presented three likely disease-causing variants designated E50K, 691_692insAG, and R545Q, and one proposed risk factor M98K [9]. Further studies find association of some OPTN alleles with OAG, but others report no evidence of association of OAG with those same alleles [11-24].In this paper, we present new data on our OPTN mutation screening of 314 open-angle glaucoma patients who have either POAG, JOAG, or NTG, from populations of Caucasian, Asian, Hispanic, and African ancestry. We present case reports of individuals with E50K and 691_692insAG mutations and discuss findings from more than a dozen studies that have carried out OPTN mutation screening.MethodsSubjectsInformed consent was obtained from each participant according to a HIPAA-compliant study protocol approved by The University of Michigan Institutional Review Board for review of human subjects studies. Ophthalmologic examinations included slit-lamp biomicroscopy, optic disk examination, IOP by applanation, gonioscopy, and refraction. Individuals with known surgical or pharmacologic risk factors for glaucoma, such as steroid use, were excluded from this study.OAG was diagnosed based on the presence of open filtration angles, glaucomatous optic discs and glaucomatous visual field changes. Individuals with elevated IOP, greater than or equal to 22 mmHg, were considered to have POAG if they showed adult-onset at 35 years of age or older, and to have JOAG if they showed onset prior to 35 years of age. They were deemed to have NTG if their highest known IOP never exceeded 21 mmHg.Table 1 lists our study case and control subjects by their diagnosis and ancestry. Our 314 OAG subjects included 51 JOAG, 230 POAG, and 33 NTG cases. Our control samples came from 371 unrelated individuals. Normal control samples were matched for race to the cases, so that a sequence variant found in a particular case population had control screening carried out only in the control population of the same ancestry. In most cases, the sample screened was the proband of the family, but sometimes a different case from that family had to be used, such as when the proband was someone with an ambiguous diagnosis. For some families in which the proband had a sequence variant of interest, additional relatives were also screened. The group of normal control samples included samples from 48 individuals of African ancestry (Corielle Institute, Camden, NJ), 19 individuals of Hispanic ancestry, and 99 individuals of Asian ancestry (Corielle Institute) who had not been characterized for ophthalmologic phenotype. These uncharacterized population controls were used in a subset of experiments as detailed in Results.Table 1Ancestry and diagnosis of subjects.PopulationJOAGPOAGNTGTotal OAGControlTotalAfrican (U.S., Ghana, Nigeria, and the Caribbean)146348188169Asian (Korea, China, and the Philippines)2125117122Caucasian (Europe and the Middle East)3215926217116333Hispanic (Mexico, Puerto Rico, Chile, Panama, and Colombia)371115061Totals5123033314371685The frequency distribution of cases and controls according to open-angle glaucoma (OAG) condition and ancestry (national) in our analysis included 314 cases with juvenile onset OAG (JOAG), primary open-angle glaucoma (POAG), and normal tension glaucoma (NTG).Mutation ScreeningOPTN was screened via sequencing of polymerase chain reaction (PCR) amplified DNA. Genomic DNA was extracted from peripheral blood samples using Puregene DNA Isolation kits (Gentra Systems, Minneapolis, MN) following the manufacurer's protocol. OPTN coding exons were amplified by PCR in a 20 μl reaction containing 50 ng of genomic DNA, 1.5 mmol/l MgCl2, 0.5 μmol/l of each primer, 0.125 mmol/l of each dNTP, and 0.5 units of Amplitaq Gold (PE Applied Biosystems, Foster City, CA) in 1X final concentration of the PCR buffer. Primers used for exon amplification are listed in Table 2. PCR conditions were 10 min at 95 °C followed by 36 cycles of 1 min at 95 °C, 1 min at 55 °C, and 1 min at 72 °C with a final extension for 10 min at 72 °C. Sequencing of PCR-amplified DNA was used to screen all 314 OAG cases for mutations in the coding sequences (i.e., exons 4 through 16) and splice sites flanking OPTN exons. When a sequence variant was detected in a patient, we screened for that specific mutation in the control population samples of the same ancestry.Table 2Primers used for sequencing the OPTN gene.ExonForward primer sequence (5'-3')Reverse primer sequence (5'-3')4TGGAGAGAAAGTGGGCAACTCACCAGCTACCACCTATGGA5GGCATCTTTCAATTCAGAGCCGACACGTAAGATTCCACTGC6TCCCAGAGCTCTGCGATTAAGCTACACTGGAATTTCCTCA7TCTGAGCCACCCCGTTTAAAGACCTCCGGTGACAAG8GGAGAATGTTCTGGAAAGCAGGGGTGAACTGTATGGTATCT9CCCCTGATCCTTTATCCCAAAATTCAGTGGCTGGACTAC10TGGTTCAGCCTGTTTTCTCCCCCCCCATCTTACAAGTATTTC11TGGCCAGGTCTAGTGAAGAATTTATCCCCCTCTCTGAGAG12GAAATGCTAGTAGGTCGTGGCCCTGACCATAGGACATTCA13CCGGCCAGAGCTGATAATAGATCCACTGAGCACTTTCC14CTAGCAGGATTGTGCATCGTGTGGCGCGAACACAGCTATT15TTTCCCCTACTTCTGTGGACGAGACTGACGGGTGCTATAT16TCATGTCCCACTACGTGTTGTGTGCCCGGCCTGTTTTCTTPrimers used in amplification of OPTN exons were also used in sequencing reactions. Primers located in introns were placed far enough away from the exon boundaries to allow visualization of the sequence of the splice sites. Exons 4 through 16 are the exons that contain coding sequence.Individuals with E50K and 691_692insAG mutations who are presented in the case reports were screened for mutations in MYOC. PCR amplification of the three MYOC exons was conducted as described in reference [7] with some modified primers as listed in Table 3. PCR products were purified with a QIAquick PCR purification kit (Qiagen, Santa Clarita, CA). Sequenced PCR products were analyzed on an ABI 377 sequencer or at the University of Michigan DNA Sequencing Core facility on either an ABI 3730 or 3700 sequencer.Table 3Primers used for sequencing the MYOC gene.ExonForward primer sequence (5'-3')Reverse primer sequence (5'-3')1AGGCTGGCTCCCCAGTATATACTGCTGAACTCAGAGTCCCC1BAGGCCAATGTCAAGTCATCCATCTCCAGAACTGACTTGTCTC2ACATAGTCAATCCTTGGGCCTAAAGACCACGTGGCACA3ACTGGCTCTGCCAAGCTTCCGCATGAGGCTGGCTCTCCCTTCAGCCTGCT3BGAGCTGAATACCGAGACAGTGAAGAGGCCTGCTTCATCCACAGCCAACPrimers used in amplification of MYOC exons were also used in sequencing reactions. Primers located in introns were placed far enough away from the exon boundaries to allow visualization of the sequence of the splice sites. Primer pairs 1A, 2, and 3A were used for PCR amplification and sequencing of exons 1, 2, and 3, respectively. Primers 1B and 3B are internal primers that were used for sequencing purposes only.Statistical AnalysisPublished reports on the frequency of OAG mutations were compared using several different statistical tests. Because some studies contained expected frequencies of less than 5, Fisher's exact test was chosen to examine the 2x2 contingency tables of individual studies. In tandem with Fisher's exact test, odds ratios and 95% confidence intervals for the odds ratio were calculated. To estimate odds ratios for whole populations based on multiple studies, fixed effect estimates were calculated using a Mantel-Haenszel (MH) model [25]. Homogeneity was evaluated with the Woolf test, in which the p value allows a determination of the appropriateness of combining studies by testing for evidence of effect modification by study group (i.e., testing whether the odds ratios are the same in all studies). For a case-control study, an odds ratio greater than 1 indicates that OAG cases are more likely to have the gene of interest than controls. For 2x2 contingency tables, independence is equivalent to an odds ratio of 1. All statistics were computed using the open source statistical program R 2.3.1 with the packages rmeta 2.12, meta 0.5, and vcd 0.9-7 [26-29].ResultsE50K OPTN mutation in a case with normal tension glaucomaCase 1, a 52 year old Chilean female (III:1; Figure 1), was diagnosed with NTG at age 42 years and her highest pretreatment IOP was 18 mmHg in both eyes. After bilateral trabeculectomies and betaxolol treatment, her IOPs were 6 mmHg in the right eye and 10 mmHg in the left eye. Gonioscopic exam revealed open angles in both eyes, with iris processes noted circumferentially in both eyes. A dilated funduscopic exam demonstrated advanced glaucomatous cupping with cup-to-disc ratios of 1.0 in both eyes and absence of hemorrhage. Sequence changes were absent in the MYOC gene coding sequence and splice sites. Her family history showed evidence of autosomal dominant inheritance (Figure 1). The E50K mutation was found in three of the proband's four affected relatives that were in the study (II:1, III:2, and III:7; Figure 1). The E50K mutation was absent in the proband's affected aunt (II:3; Figure 1) as well as five unaffected relatives that were screened (Figure 1).Figure 1Lack of complete cosegregation of E50K with glaucoma in the pedigree of a Chilean family. The arrow indicates the proband (Case 1). Filled symbols are affected individuals with NTG, open symbols are individuals who are unaffected or reported to be unaffected. Symbols with a cross indicate individuals who are glaucoma-suspect, symbols with a center dot indicate glaucoma-affected individuals according to family report, and partially filled symbols denote individuals affected with POAG. Diagonal lines mark deceased individuals. Individuals denoted with ++ have E50 alleles on both chromosomes and ones with M+ carry the E50K heterozygous change. Members of generation four are young enough that they are not expected to be affected yet.691_692insAG OPTN mutation in a case with primary open-angle glaucomaCase 2 was a female Russian Ashkenazi Jewish immigrant diagnosed with POAG at 80 years of age. Her ocular history was significant for high myopia (right eye -17.75 diopters, left eye -19.00 diopters) and myopic retinal degeneration in both eyes. The patient had a childhood history of measles, a disease that has been identified as a possible contributor to high myopia [30,31]. At the time of the POAG diagnosis, the patient already had dense, 4-quadrant visual field defects in both eyes, attributable to the retinal degeneration, or a long-standing undiagnosed glaucoma, or both. She had cup-to-disc ratios of 0.5 in both eyes and diffuse chorioretinal atrophy. Her IOPs were 22 mmHg in the right eye and 21 mmHg in the left eye. Her IOPs decreased to 17 mmHg in the right eye and 16 mmHg in the left eye at one month after treatment with betaxolol, dipivefrin, and pilocarpine. Over the next several years, her IOPs fluctuated between the low teens and mid-twenties while she underwent treatment with medication, laser treatments, and multiple trabeculectomies. Screening of the OPTN gene revealed an insertion of AG at positions 691 and 692 (691_692insAG) in one copy of the OPTN gene. There were no other sequence variants in the coding sequence or splice sites of either OPTN or MYOC. Little family history information was available. She had a maternal grandfather affected by high myopia, but she didn't know of any other cases of glaucoma in her small family. Her only living relative, a reportedly unaffected son, declined to participate in the study.R545Q and M98K OPTN sequence variantsThe R545Q sequence variant was found in two individuals with OAG. One woman with Filipino ancestry had JOAG diagnosed at 24 years of age and had a maximum known pretreatment IOP of 50 mmHg. Her mother also had POAG and her brother is unaffected. She did not know the diagnostic status of the rest of her relatives in the Philippines. The second case, a Korean woman diagnosed at 55 years of age, had a maximum known IOP of 29 mmHg and an unknown family history of glaucoma.Among 36 OAG cases with the M98K mutation, the 26 for whom we have historical IOP data had known maximum IOPs between 16 mmHg and 55 mmHg (mean=29.6 mmHg). Out of the 283 OAG cases who lacked the M98K mutation, 253 had historical IOP data available, with the maximum recorded IOPs ranging from 14 mmHg up to 77 mmHg (mean=29.9 mmHg).OPTN sequence variants in the whole study cohortAmong the 314 OAG cases, we found a total of four (1.2%) individuals who possessed any of the three sequence variants reported by Rezaie and colleagues [9] to be disease-causing variants (Table 4).Table 4Frequency of sequence variants that alter the OPTN protein sequence.Protein changeDNA changeExonAncestryCasesControlsE50Kc.458 G>A4Hispanic1/110/50I88Vc.572 A>G5Caucasian0/2171/116A99Sc.605 G>T5African0/812/88E322Kc.1274 G>A10African1/816/88E322Kc.1274 G>A10Caucasian1/2170/90691Frameshift691_692insAG6Caucasian1/2170/116R545Qc.1944 G>A16Asian2/511/117All case samples were screened and scored for each mutation listed. Absence of a listing for one of the four control groups does not imply that it was screened. Data for M98K appear in Table 5.The E50K mutation in Case 1 was the only instance of E50K among the 314 OAG cases (1/314, 0.3%) and in none of 371 controls (<2.7%), and was one of only 11 Hispanic cases screened (1/11, 9.1%; Table 4). E50K was identified in one of the 33 NTG cases (3.0%), but was not present in the 230 POAG or 51 JOAG cases. This mutation was present in 1/11 (9.1%) Hispanic cases and none of 50 Hispanic controls. It was also absent from 86 Caucasian normal controls.The single instance of 691_692insAG in Case 2 was found among the 314 OAG cases (1/314, 0.3%), and was one of 217 Caucasian OAG cases screened (0.5%; Table 4). This was one of 230 POAG cases (0.4%) and was not present in 116 Caucasian controls, including seven samples that share Ashkenazi Jewish ancestry with the case having the mutation. In addition, 691_692insAG was present in one case in the original report by Rezeai [9] but that report did not provide information on ancestry of that case, so we could not tell whether their case and our case shared ancestral origins. This mutation was also previously reported as being absent from 200 normal control chromosomes of Caucasian origin [9].R545Q was present in two of the 314 OAG cases (0.6%; Table 4). It was found once among the 51 JOAG cases (1.9%) and once among the 33 NTG cases (3.0%). Both instances of R545Q were found in individuals of Asian ancestry (2/5, 40.0%). R545Q was also present in 11 of 117 (9.4%) controls of Asian ancestry. The absence of this allele from our Caucasian cases or controls (0/333, <0.3%) concords with previous reports that failed to find it in either cases or controls of European ancestry (0/1457), suggesting that the allele frequency in European populations may be less than 0.1% [9,11,13,22,23,32,33]. Based on our data, we suggest that R545Q may be of low prevalence in African (<0.6%) or Hispanic populations (<1.6%), but our ability to estimate frequencies in these populations is limited because of sample size.The M98K sequence variant was found in JOAG, POAG, NTG, and control populations (Table 5 and Table 6). We observed statistically insignificant differences in the frequencies between cases and controls for Africans and Caucasians, the two large sample sets in the study (Table 5). Frequencies in Asian samples (32/122, 26.2%) resembled values for African samples (18/81, 22.2%), while frequencies for Hispanic samples (2/62, 3.2%) seemed more similar to frequencies for Caucasian samples (8/116, 6.9%), but sample sizes were small. Additionally, the Asian samples showed no difference between cases and controls (p value=0.112), but the sample size was small and the use of a predominantly Chinese population to control for findings from a mixed Asian case set was problematic when looking at an allele that showed considerable variation among populations. Thus, although our overall study population showed M98K in 36 of 314 OAG individuals (11.5%) and 58 of 371 controls (15.6%), such pooling of data from different racial/ethnic groups is invalid where population frequencies vary so greatly.Table 5Frequency of M98K in four populations within our cohort.AncestryWhole populationJOAGPOAGNTGTotal OAGControlsFisher’s exact test p valueFrequency of mutation in screened samples by populationAfrican38/1693/1414/631/418/8120/881.0Asian32/1221/21/11/23/529/1170.112Hispanic2/620/31/70/11/111/500.331Caucasian22/3363/3211/1590/2614/2178/1161.0Total94/6907/5127/2302/3336/31458/371Percent of mutation in screened samples by populationAfrican22.521.422.225.022.222.7Asian26.250.0100.050.060.024.8Hispanic3.20.014.30.09.12.0Caucasian6.59.46.90.06.56.9Total3.63.71.76.11.55.6The total enumeration of both cases and controls is listed in the whole population column. Cases are subdivided according to OAG type, either JOAG, POAG, or NTG. A two-sided Fisher's exact test p value indicated no statistical significance for association between cases and controls in each ancestry category. Woolf's test for homogeneity among the ancestry frequencies yielded a p value of 0.312, indicating that the ancestral subdivisions are statistically similar.Table 6Comparison of R545Q frequency in different populations.SourceOAGControlsPercentOdds ratio95%CI boundsFisher's exact test p valueR545QTotalR545QTotalOAGControlsLowerUpperChina [14]511851504.23.31.280.364.540.753 [15]27400192626.87.30.920.501.700.876Japan [11]122473894.93.41.460.405.310.767 [16]26411112186.35.01.270.622.620.596 [17]115401000.60.01.96*0.0848.69*1.000 [20]20313101966.45.11.270.582.770.700 [21]3834583.66.90.510.112.350.446Asia[This study]261111733.39.44.820.7929.360.122Europe[This study]0217--0.0----- [11]065001620.00.0---- [13]0270940.00.0---- [9]14601002.20.06.63*.26*165.80*0.315 [22]0112--0.0-----Africa[This study]0810900.00.0----India [19]620002003.00.013.40*0.75*239.49*0.030Mixed [23]0860800.00.0---- [24]111431870.91.60.540.065.281.000The asterisks denote a calculation based on adding 0.5 to each cell in cases with a zero cell frequency, otherwise the value is nonexistent. Lower and upper bounds refer to the individual study 95% confidence interval around the odds ratio for a fixed effects Mantel-Haenszel model. Information from Leung et al. [35] were omitted because it duplicated that contained in Fan et al. [15]. Data from Toda et al. [37] were omitted because it duplicated information contained in Tang et al. [20]. In the Europe category, only the data regarding Caucasians (Iowa and Australia) in Alward et al. [11] were enumerated, while the reported cases with pigmentary, developmental, and exfoliative data were omitted.Screening identified two instances of E322K in both cases and controls - a change previously reported to be associated with glaucoma [9] (Table 4). We also found silent OPTN coding sequence polymorphisms, including L32L, T34T, L41L, E63E, A134A, and S321S, as well as previously reported intronic sequence variants [4,10,11,20,24,34,35]. We did not find Ile88Val or Ala99Ser in our case populations, but did observe them among our controls (Table 4).Pooled data on R545QR545Q was present in our Asian data set, but absent from the other three populations we screened. We found no significant differences in allele frequencies between cases and controls for R545Q. This agrees with many of the other published studies, although it should be noted that Mukhopadhyay et al. [19] did report the case-control difference to be significant (Table 6). Evaluation of odds ratios and frequencies for each of the Asian studies showed no statistically significant differences between case and control values for any of the Asian data sets (Figure 2, Table 6). With the exception of the Rezaie study [9], R545Q has been reported only in Chinese, Japanese, Korean, Filipino, Indian, and mixed-ancestry populations (Table 6).Figure 2R545Q log odds ratios and allele frequences in Asian population studies. A shows the odds ratios with 95% confidence interval bars for individual Asian studies, and pooled results for Japan, China, and both in open angle glaucoma (OAG) cases versus controls. Odds ratios and confidence intervals are fixed effect estimates resulting from the Mantel-Haenszel method. B shows the case (OAG, filled circle) and control (open circle) proportion observed for each study. Total sample sizes are listed along the right-hand margin. None of the differences between case and control frequencies are statistically significant in a comparison of the odds ratios (as readily observed from the odds ratio confidence intervals) and frequencies of R545Q mutations in any of the Asian populations studied.Our Asian data set is small, so we examined the possibility of pooling data from multiple studies. Because of observed variation in allele frequencies among studies (0.6-6.8%; Table 6), we questioned whether the data could be validly combined. Using the Woolf test for heterogeneity to address this question, we found no statistically significant study-based stratification within the individual populations for R545Q (Table 7). This means that it is reasonable to take data from the studies in Table 6 and pool them for a given population from multiple studies under a model of homogeneity. When pooled, we found no statistically significant difference between case and control values for the combined Asian data set (p value=0.541), nor for the separate Chinese (p value=0.89) or Japanese (p value=0.43) subsets (Table 7). The same lack of difference was true when considering only NTG cases.Table 7Aggregate statistical summaries in Asian populations screened for R545Q.AncestryCasesControlsPercentOdds ratio95% CI boundsWoolf test p valueFisher’s exact test p valueR545QTotalR545QTotalOAGControlLowerUpperChina32518244126.25.80.980.571.700.6480.89Japan621208286615.14.21.200.761.900.8380.43Asia9417265210735.44.81.120.781.580.9250.541China-NTG710651506.63.32.040.548.41-0.244Japan-NTG40705286615.74.21.400.842.330.8480.263Results from computing the upper and lower 95% confidence interval bounds around the odds ratio indicate that none of the Asian divisions are statistically different from an odds ratio of 1. The Woolf test for homogeneity indicates that across studies within each ancestry group the odds ratios are statistically equivalent (i.e., homogeneous, because a p value less than 0.05 would indicate heterogeneity [25]). A two-sided Fisher's exact test on the pooled frequencies was computed for those instances when the Woolf test indicated homogeneity at the 0.05 level. The ancestry groups are collated for China from Chen et al. [14] and Fan et al. [15]; and for Japan from Alward et al. [11], Funayama et al. [16], Fuse et al. [17], Tang et al. [20], and Umeda et al. [21]. The Asia listing includes data pooled from the China and Japan categories. Only Fan et al. [15] report on NTG for China.Pooled data on M98KWe found no evidence of significant difference between case and control frequencies for M98K in our Asian, African, Hispanic, or Caucasian populations (Table 8). Fuse and Alward reported statistically significant evidence of association of M98K with OAG in the Japanese population, although Alward indicated that this difference becomes nonsignificant when adjusted for testing multiple times [11,17]. Other studies do not report a significant case-control difference [11-24].Table 8Frequency of M98K in individuals from different populations.SourceCasesControlsPercentOdds ratio95% CI boundsFisher’s exact test p valueM98KTotalM98KTotalOAGControlLowerUpperChina [14]261182215022.014.71.640.883.080.148 [15]1294008128132.328.81.180.841.640.355Japan [11]5124788920.69.02.631.205.800.014 [16]814113621819.716.51.240.811.910.389 [17]25154510016.25.03.681.369.970.009 [20]513132719616.313.81.220.742.020.527 [21]128315814.51.79.631.2276.310.149Asia[This study]352911760.024.84.550.7228.600.112Europe[This study]1321781166.06.90.860.342.140.814 [11]46650101627.16.21.160.572.351.000 [12]223153957.03.22.300.677.870.224 [13]2273947.43.22.430.3815.330.310 [32]9200102004.55.00.900.362.251.000 [18]1123751104.64.51.020.353.021.000 [33]1117011006.51.06.850.8753.870.036 [9]23169942213.62.17.233.2715.980.000 [22]71057936.77.50.880.32.61.000Hispanic[This study]1111509.12.04.900.2885.050.331Africa[This study]1881208822.222.70.970.472.001.000India [19]222001120011.05.52.121.004.510.068 [39]1022001004.50inf1.04inf0.034Mixed [38]28498172185.67.80.700.381.320.315 [36]1415391009.29.01.020.422.451.000 [23]8868809.310.00.920.332.591.000 [24]12115410110.44.02.830.889.060.116Under the fixed effects Mantel-Haenszel model, individual study 95% confidence interval bounds around the odds ratio are given. For compatibility with much of the published literature the two-sided Fisher's exact test p values are given for each study. Although the presence of M98K mutations in OAG cases appears to be statistically significant relative to controls, Alward et al. [11] reported that when multi-testing is taken into account their result becomes nonsignificant. Information from Leung et al. [35] was omitted because it duplicated the information contained in Fan et al. [15]. Data from Toda et al. [37] were omitted because they duplicated the observations contained in Tang et al. [20]. Pigmentary and exfoliative data were omitted from the Europe (Caucasians living in Iowa and Australia) samples reported by Alward et al. [11]. Rezaie et al. [9] reported a p value=2.18-7.Previous studies supported our finding that M98K allele frequences are much higher in populations of Asian (555/2818, 19.7%) or African (38/169, 22.5%) ancestries than in Caucasian (179/3149, 5.7%) or Hispanic (2/61, 3.3%) populations (Table 8 and Table 9). Thus, ancestry would be a significant confounding variable when attempting to analyze data pooled from different populations.Table 9Aggregate statistical summaries for populations screened for M98K.AncestryOAGControlsPercentOdds ratio95% CI boundsWoolf test p valueFisher's exact test p valueM98KTotalM98KTotalOAGControlsLowerUpperChina15551810343129.923.91.260.941.700.3540.04Japan22012087766118.211.61.651.252.190.046-Asia3751726180109221.716.51.461.191.790.0750.0006Europe131187348127673.81.871.312.660.005-Total6004871277316712.38.71.511.291.770.002-Japan-NTG1427057766120.111.61.911.402.620.2402E-05Europe-NTG28371245447.54.41.770.973.240.1490.58Total-NTG1701076101120515.88.41.751.332.310.1776E-08The ancestry groups are collated for China: Chen et al. [14] and Fan et al. [15]; Japan: Alward et al. [11], Funayama et al. [16], Fuse et al. [17], Tang et al. [20], and Umeda et al. [21]; Europe: Alward et al. [11], Aung et al. [12], Baird et al. [13], Jansson et al. [32], Melki et al. [18], Rakhmanov et al. [33], Rezaie et al. [9], and Weisschuh et al. [22]. Asia is the pooling of the China and Japan categories. The total combines all published studies from the lines above with the addition of Craig et al. [38], Hauser et al. [36], Mukhopadhyay et al. [19], Sripriya et al. [39], Wiggs et al. [23], and Willoughby et al. [24]. The Europe-NTG group consists of the data from Alward et al. [11], Aung et al. [12], Baird et al. [13], Rakhmanov et al. [33], and Weisschuh et al. [22]. Results from computing the upper and lower 95% confidence interval bounds around the odds ratio indicate that some studies are statistically different than an odds ratio of 1. The Woolf test for homogeneity indicated that across studies within each ancestry group which of the odds ratios were statistically equivalent (i.e., heterogeneity is indicated by a p value less than 0.05 [25]). A two-sided Fisher's exact test on the pooled frequencies is given for those instances when the Woolf test indicated homogeneity at the 0.05 level. When Rezaie et al. [9] data are excluded the Europe group, then the odds ratio becomes 1.33 with a 95% confidence interval of (0.90, 1.98), Woolf p value of 0.512, and a Fisher's exact test p value of 0.072.If we consider specific defined subpopulations, where there should be less concern about ancestry serving as a confounding variable, then we are left with concerns about differences observed not just between but also within populations. When we compared the different Asian studies using odds ratios (Figure 3), the aggregate Asian data, the aggregate Japanese data, the data produced by our study, and by the individual studies of Umeda et al. [21], Fuse et al. [17], and Alward et al. [11], we found each showed significant evidence of a difference between cases and controls. When we evaluated allele frequencies for the various Asian data sets, we saw dramatic differences in allele frequencies among the different studies (Figure 3, Table 8). Also, the control values showed much greater variation among studies than the case values. The control values from some studies are higher than the case values from other studies, even though within each separate study the case frequencies are always higher than control frequencies (Figure 3). With regard to our study, in which a small number of samples came from diverse Asian regions, the observed difference could be due to the limited case sample size, differential representation of M98K within the Asian population, case versus control status, or some combination of the three.Figure 3Studywise differences appear in Japanese populations when odds ratios and frequencies of M98K mutations are compared. The left-hand graph (A) shows the odds ratios with 95% confidence interval bars for individual Asian studies and pooled results for Japan, China, and both in open angle glaucoma (OAG) cases versus controls. Odds ratios and confidence intervals are fixed effect estimates resulting from the Mantel-Haenszel method. The right-hand graph (B) shows the case (OAG, filled circle) and control (open circle) proportions observed for each study. Total sample sizes are listed along the right-hand margin. Larger samples have both narrower confidence intervals and shorter distance between fractions observed for cases and controls. Studies inconsistently estimate the odds of OAG versus controls carrying an M98K mutation, with larger studies (more than 400 total cases and controls) estimating no statistically significant difference. Other population estimates are not shown, because, among the European population-based studies, only Rezaie's study [9] showed a statistically significant difference. The single study on India yielded a significant odds ratio, but no other comparable populations have been reported [19].The results of the Woolf test for heterogeneity indicated that there was something noncomparable about the M98K findings from a number of the studies that reported results for the same populations (Table 9). In the case of the Japanese data set, pooled data showed M98K frequencies of 18.2% (220/1208) in cases versus 11.6% (77/661) in controls (p value=0.0002), but results of the Woolf test led us to suspect that we may be combining noncomparable data sets (p value=0.046). In the case of the European data set, Fisher's exact test indicated that there was a statistically significant difference between cases and controls (p value=0.001), but again the Woolf test identified heterogeneity among the data sets (p value=0.005; Table 9). When we pooled the worldwide data from the entire set of published studies, we saw a significant difference between cases and controls (p value=0.00000004), but again, testing for heterogeneity indicated that it may be invalid to pool these studies (Table 9; p value=0.0018). Interestingly, if we removed data from the Rezaie study [9], which indicated Caucasian controls but did not specify its case population composition, the remaining Caucasian data sets appeared to be homogeneous (p value=0.512) and Fisher's exact test indicated no significant difference between case and control values (Table 9; p value=0.072). Fewer studies distinguish observations for NTG only. For those studies that provided data for NTG frequencies, we saw that both the Japanese and European populations showed homogeneity across studies. However, Fisher's exact testing of pooled studies showed opposing results between Japanese (p value=0.00002) and European (p value=0.58) populations.The case versus control difference for the Hispanic data are also intriguing, with M98K case values of 1/11 (9.1%) and control values of 1/50 (2.0%). Our small Hispanic data set was not well-powered for statistical testing, but this represented an initial view of a population under represented in previous studies (Table 8). Given that our Hispanic data set represented cross-continental cases from North, Central, and South America, we have to wonder whether the apparent differences in M98K allele frequences was simply due to small sample size, or rather might be attributed to differential allele frequencies correlated with geographic origins of samples rather than case-control status.DiscussionIn our study of 314 individuals with OAG, we found 42 individuals with sequence variants predicted to alter the protein coding sequence. This included three OPTN variants previously reported to be disease-causing variants-E50K, 691_692insAG, and R545Q, as well as the M98K variant previously reported as a risk factor [9].This is the second time the 691_692insAG mutation has been reported. Both times it has been identified in case populations but not seen in controls. It is the first report of 691_692insAG in an individual of Russian Ashkenazi Jewish ancestry, and ancestry is unavailable for the previously reported case. Although OPTN defects were originally reported in a population of primarily NTG cases, we found this variant in an individual with modestly elevated IOP (22 mmHg). The shift in the reading frame that it causes and the fact that it is seen among cases but not controls suggests that it could be a causative variant. However, if we combine data from all studies, we see it in 2/3677 cases (0.0005%) in the studies that used protocols that would have detected it (all studies in Table 8 except Aung [12], Melki [18], and Wiggs [23]). Only a fraction of the 2,270 controls from those studies screened the whole sequence from all controls, so they would not have been highly likely to detect this variant among the controls even if the control frequency were equal to the case frequency. Thus, while it is tempting to say that a variant seen only in two cases might be causative, the available numbers can only support the conclusion that 691_692insAG event is a rare occurence.We report here the first observation of the E50K change in a Caucasian Hispanic individual. The observation of E50K at a frequency of 0.3% in our cases is consistent with reports of frequencies in OAG populations of 0.1% by Alward et al. [11], 0.6% by Aung et al. [12], and 0.6% by Hauser et al. [36]. The NTG subset is reported to have E50K at a higher prevalence of 13.5% (7/52),1.5% (2/132), 1.5% (1/67), and 2.9% (1/34) in studies by Rezaie et al. [9], Aung et al. [12], Hauser et al. [36], and ourselves, respectively. Many other studies found no evidence of this mutation, including reports of its absence from 237 cases with Chinese ancestry [14,35] and 961 cases with Japanese ancestry [11,14,16,17,20,21,35,37], which supports the supposition that this is a polymorphism private to the Caucasian and Hispanic populations. Variation in frequencies observed among studies may be affected by the ancestry of the population, the fraction of the cohort with familial glaucoma, and differences in specific diagnoses included in the study. Failure to see complete cosegregation of E50K with glaucoma (i.e., we have one OAG case in a family lacking the E50K mutation) raises questions about whether we are observing a phenocopy or whether E50K is not the cause of the glaucoma in this family.Our data and the compiled evidence from more than a dozen other studies support the idea that R545Q may be a private polymorphism of Asian populations. Although our Asian data set provides marginal evidence for a difference in R545Q allele frequency between cases and controls, it is a small population and the results are not statistically significant. When we pooled our data with data from other studies there did not appear to be any evidence to support a role for R545Q as a disease causing variant.We found the M98K variant in all four populations screened, but evaluation of all of the published studies leaves unresolved the issue of whether or not M98K is a risk factor for glaucoma. A similar conclusion was drawn by Craig et al. [38] but they did not analyze the population (ancestry) structure of the data for the allele. Several studies find evidence for association, while others do not. Evaluation of the published data in addition to our own indicates that there is considerable variation in allele frequencies, not only among populations, but also within populations. This variant is found in Asian and African populations at more than twice the frequency seen in Caucasian and Hispanic populations. Comparison of findings from different studies indicates large variations in allele frequencies in different study populations within Japanese (MH p value=0.00038) and Chinese (MH p value=0.118) populations (Table 9). The case-control difference is much smaller within Caucasian or African populations (Figure 3), and data on Caucasian populations show less variation between studies than the data on Japanese and Chinese populations (Table 9, Figure 3).There are a number of confounding factors that might contribute to the observed variability of allele frequency between Asian populations. Differential allele frequencies within a population could result from founder effects. At this point, there is not enough information available regarding origins of the different subpopulations (Table 10) to allow for evaluation of the likelihood of a founder effect. There appears to be a correlation between total sample size and the difference between case and control frequencies (Figure 2), although the published studies seem to be adequately powered (for Fisher's exact test, the power, or probability to reject the null hypothesis when it is true, is 0.76 for the parameters π1=0.2, π2=0.1, where each sample size is 200 and α=0.05). An alternative confounding factor could be the result of the different screening techniques applied (Table 10); however, there is no obvious correlation of high allele frequencies with one screening approach and low allele frequencies with a different technique. Selective under-representation of an allele in a data set relative to the actual allele frequency could result if M98K were in linkage disequilibrium with a neighboring polymorphism contained within the sequence of a primer used in amplification or sequencing in some studies, but not others. Some of the papers do not present the primer sequences and the available primer sequence data do not provide support for this idea. Additional contributions to variability between studies could include differences in diagnostic inclusion and exclusion criteria and fraction of familial glaucoma within each cohort. Thus, the extant data do not allow us to distinguish between technical, ascertainment and demographic models for the observed differences in M98K allele frequencies between different studies of the same population.Table 10M98K population data sources and screening methods ordered by ancestry.SourceRecruitment locations (population)MethodologyNotesChina [14]China-BeijingSSCP->sequencing [15]China-Hong KongPCR and HTCSGE->sequencingJapan [11]Japan-GifuSSCP->sequencing [16]Japan-Tokyo, Kumamoto, Hamamatsu, Hiroshima, NiigataPCR-RFLP [17]Japan-MiyagiPCR->sequencing [20]Japan-YamanashiSSCP->sequencing [21]Japan-Okayama CitysequencingAsia (other than China and Japan)[This study]USA-Michigan (Korean, Chinese, Filipino)PCR->sequencingEurope[This study]USA-MichiganPCR->sequencingCaucasian [11]Australia-Melbourne, Adelaide, USA-IowaSSCP->sequencingAustralian samples Are Caucasian (D. Mackey, personal report), Iowa population >91% Caucasian according to the State Data Center of Iowa [12]England-LondonPCR-RFLPCaucasian [13]Australia-New South WalesPCR-RFLPmostly Caucasian [32]Sweden-Uppsala and TierpsDHPLC, PCR, SNaPshot [18]France-ParisPCR-RFLPFrench and Moroccan Caucasians [33]Russia-St. PetersburgSSCP, PCR [9]USA-Chicago, Connecticut, New Haven, UK-London, Canada-TorontoPCR->sequencing and SSCP->sequencingUnspecified cases with Caucasian controls [22]Germany-Tuebingen, WuerzburgPCR-RFLP, DHPLCHispanic[This study]USA-Michigan, Florida, Mexico, Panama, Peru, Chile, ParaguayPCR->sequencingAfrica[This study]USA-Michigan, Ghana- Acra, SunyaniPCR->sequencingAfrican American and AfricanIndia [19]India-Hyperabad, KolkataSSCP->sequencing, DHPLC, PCR--RFLP [39]India-ChennaiPCR->sequencing, RFLPMixed [38]Australia/TasmaniaPCR-RFLP [36]USA-New England areaPCR->sequencing, DHPLCabout 90% Caucasian [23]USA-Massachusetts, North CarolinaPCR->sequencingabout 90% Caucasian [24]Canada-TorontoPCR-RFLPThe following methodologies were used to screen samples for variants. We omitted two studies: Wang et al. [40], a Filipino population, and Forsman et al. [34], a population from south Finland, because they are small family-based studies without population data. SSCP: single-strand conformation polymorphism. DHPLC: denaturing high-performance liquid chromatography. HTCSGE: High throughput conformation sensitive gel electrophoresis. RFLP: Restriction Fragment Length Polymorphisms.An alternative approach to evaluation of whether the M98K allele is involved in glaucoma is through the study of cosegregation in families. Wiggs et al. [23] reported lack of cosegregation in families. An accompanying population-based portion of that study [23] also failed to find association of M98K with glaucoma in a population of mostly European ancestry.One alternative explanation for why some studies find association, yet others do not, might be that there is a valid statistical difference between the case and control populations but that the M98K allele is actually associated with some other variable that differs between cases and controls, or has been excluded from cases but not controls. An obvious example of this would be IOP. The Rezaie et al. study [9] looked at mostly NTG families, while there was a lot of variation in the extent to which NTG was represented in the different populations in the other studies. If M98K is actually responsible for reducing IOP, or for preventing the rise of IOP, but is not actually causing glaucoma, then we would expect exclusion of individuals with elevated IOP from the cases would bias the M98K frequency in the cases as compared to the controls even if M98K were not actually causing glaucoma. Melki et al. [18] offered the view that M98K is associated with lower IOP. In our data set we found that cases with the M98K mutation had known maximum IOP values ranging from 16 mmHg to 55 mmHg (mean=29.6 mmHg) while those who lacked the M98K mutation had maximum recorded IOP values ranging from 14 mmHg in an NTG case up to 77 mmHg (mean=29.9 mmHg). Thus, in our data set there was no obvious difference in maximum known IOP between those with and those without M98K.The original report by Rezaie provided apparently compelling statistical evidence that M98K is a glaucoma risk factor with a p value of 2.18x10-7. The other studies that found any support for M98K association with glaucoma found only weak evidence of this association. It remains unclear what the differences are between the studies that could account for the differences in study outcome. One key issue for M98K appears to be the great variability of allele frequency reported in different studies and different populations. This would be a problem if the case population in the Rezaie et al. study [9] contained multiple populations or an admixed population that self-identified as Caucasian. In the Rezaie et al. study [9], both E50K, an apparently private Caucasian polymorphism, and R545Q, an apparently private Asian polymorphism, were in the same study cohort. While this could mean that their undescribed case population was a mixed race group, it is also the kind of thing that can happen in a fairly admixed urban population when using self identification as the basis for applying racial/ethnic classification, even when setting out to identify a relatively homogeneous population. Thus, there is the possibility that the Rezaie et al. study [9] outcome differs from the others because of differences in diagnostic inclusion and exclusion criteria or other unidentified factors, but it could also be the result of studying an allele that varies significantly between and within populations - something that can happen even when making efforts to carry out adequate matching of cases and controls.Thus the findings on M98K are currently contradictory, with some studies finding association and other studies finding no support for association, and with the differences in study outcome not assorting according to population or technology used. Because there are such substantial differences in allele frequencies between the different studies and between and within populations, it is likely that a final resolution of this question will require the following: The screening to take place by technologies selected for precision rather than high throughput; the study be adequately powered; matching of cases and controls for ancestry be highly rigorous and matched for subpopulations rather than simply matching for one of a handful of racial or ethnic categories; inclusion and exclusion criteria be carefully defined, that tests for association with associated variables such as IOP be carried out in addition to tests for association with the primary glaucoma status variable; and population substructure analysis be included in the analysis to help deal with apparent differences within populations that can be difficult to control.This study has contibuted additional evidence of association of OPTN E50K with glaucoma, and reported an additional instance of the 691_692insAG sequence variant. We have also provided new information on OPTN in populations of African and Hispanic ancestry. Evaluation of data from more than a dozen studies indicated no association of R545Q with glaucoma in most populations. Combined analysis of more than a dozen studies suggests that M98K is associated with NTG in Asian, but not Caucasian study populations, but these results must be interpreted with great caution because of the large differences in allele frequencies between and within populations.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2533124.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533124",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533124\nAUTHORS: Ben Buelow, Andrew M. Scharenberg\n\nABSTRACT:\nConditional gene targeting using the Cre-loxp system is a well established technique in numerous in vitro and in vivo systems. Ligand regulated forms of Cre have been increasingly used in these applications in order to gain temporal and spatial control over conditional targeting. The tamoxifen-regulated Cre variant mer-Cre-mer (mCrem) is widely utilized because of its reputation for tight regulation in the absence of its tamoxifen ligand. In the DT40 chicken B cell line, we generated an mCrem-based reversible switch for conditional regulation of a transgene, and in contrast with previous work, observed significant constitutive activity of mCrem. This prompted us to use our system for analysis of the parameters governing tamoxifen-regulated mCrem recombination of a genomic target. We find that robust mCrem expression correlates with a high level of tamoxifen-independent Cre activity, while clones expressing mCrem at the limit of western blot detection exhibit extremely tight regulation. We also observe time and dose-dependent effects on mCrem activity which suggest limitations on the use of conditional targeting approaches for applications which require tight temporal coordination of Cre action within a cell population.\n\nBODY:\nIntroductionConditional gene targeting allows for spatial and temporal control of gene expression both in vitro and in vivo. Several systems for conditional gene targeting exist, such as Flp/Frt, Cre/Loxp, ΦC31-att, Mx-1/IFN-l·α, and Tet on/off [1], [2]. One of the most popular is the Cre-loxp system: a medline search for “Cre recombinase” yields over 2000 publications, of which about 900 represent transgenic mice [3]. For this approach, a portion of a gene is flanked by loxp sites, 34 base pair sequences consisting of two 13 base inverted repeats flanking 8 non-palindromic bases. Cre recombinase, nominally specific for these sites [4], [5], either excises or inverts flanked sequences depending on the relative orientation of the two loxp sites, leading to changes in gene function and/or expression. Historically, the popularity of the Cre/loxp system derives from its apparent high fidelity, absence of canonical recognition sites in the mammalian genome, and its high efficiency of recombination as compared to other strategies [6], [7].Most of the work using the Cre-loxp system has been predicated on the assumption that Cre, without Loxp targets, is largely inactive in vertebrate cells. This assumption may have arisen in part due to the lack of overt pathology in Cre expressing mice [8], but given the ability of mammals to tolerate significant somatic cell death this conclusion clearly requires validation. Several studies, highlighted recently in two articles [3], [9] have shown that use of the Cre-loxp system in eukaryotic cells carries with it numerous potential confounders such as variable efficiency of Cre expression and/or recombination, variability in germline recombination, and, contrary to the previously mentioned assumption, Cre mediated toxicity [10]–[14].In order to gain temporal and spatial control over conditional gene targeting, ligand regulated forms of Cre are often utilized. The tamoxifen-regulated Cre variants Cre-ERT (Cre fused to a mutated human estrogen receptor that binds tamoxifen or 4-hydroxytamoxifen (OHT) but not endogenous estrogens, [11], [15]–[18]), and mer-Cre-mer (mCrem, which fuses Cre to two mutated murine estrogen receptors responsive to tamoxifen/OHT but not endogenous estrogens, much like Cre-ERT above [19], [20]) are favored ligand-regulated Cres because of their reputation for tight regulation in the absence of their tamoxifen/OHT ligands [20]–[25]. Tamoxifen/OHT-regulatable Cre fusion proteins demonstrate titratable toxicity, suggesting that Cre mediated toxicity is dose- and exposure time-dependent. Over the course of these titration studies it was observed that even in the absence of the inducing ligand some cytotoxicity and/or gene expression could be seen in several cell systems, including DT40 B lymphocytes [11], [19], [26], [27]. Additionally, it has been suggested that ligand independent effects of inducible Cre may be altered by the expression level of the Cre protein [26]. These results call the efficiency of regulation of these Cre variants into question, and suggest a need for a better understanding of how the parameters of inducible Cre expression, Tamoxifen/OHT dose, and Tamoxifen/OHT time of exposure interact to affect tight regulation of ligand-induced Cre activity.Our lab makes extensive use of DT40 B-cells as a model system because they support facile gene targeting and are well validated for studies of lymphocyte signaling and physiology [28]–[34]. To effect conditional gene targeting in these cells, we have historically used mCrem [35]. In the process of generating a Cre-mediated reversible switching line with a fluorescent readout of Cre activity (Fig. 1A), we observed a high level of ligand-independent mCrem activity. To understand how best to implement mCrem-based conditional gene targeting, we have used this system to explore the parameters governing tamoxifen-induced mCrem activity. The results of our investigation show that the level of mCrem expression correlates with Cre activity independent of OHT treatment: high expression leads to abundant ligand-independent activity, while low mCrem expression essentially abolishes activity. Evaluation of time and dose effects in clones with variable expression of mCrem suggests important limitations on the use of mCrem for applications requiring temporally correlated Cre activity within cell populations.10.1371/journal.pone.0003264.g001Figure 1DT40 B lymphocytes transfected with a reversible switching construct and expressing mer-Cre-mer show 4-hydroxytamoxifen independent Cre activity.\nA Schematic of the “Flip” construct. Exon 4 of NUDT9 was replaced with a CMV promoter, an in-frame region (containing a NUDT9 cDNA and an inverted mCherry cDNA, each followed by a poly-A region) flanked by loxp sites, and a neomycin selection cassette. Treatment of cells stably expressing the construct with 4-hydroxytamoxifen (OHT) should lead to reversible flipping of the floxed region, as shown in the cartoon, and subsequent expression of mCherry red fluorescent protein in ∼50% of the cells. B Stable integration of the Flip construct was verified in clone #27 via genomic PCR. The positive control represents a different construct that could be amplified with the same primer set. C Following stable transfection of clone #27 with a mer-Cre-mer (mCrem) expressing vector, cells from various clones were lysed and 50 µg of protein from each were run on an 8% SDS-PAGE gel. mCrem was visualized with polyclonal rabbit anti-Cre antibody (1∶2000, Novagen), and a highly expressing clone, 27Flip/Cre20 was selected for further study (red circle). D 27Flip and 27Flip/Cre20 cells were analyzed by flow cytometry, as indicated. In contrast to the 27Flip parental line, the mCrem expressing 27Flip/Cre20 cells showed robust red fluorescence in a large subpopulation despite the absence of the mer- ligand OHT.ResultsDT40 B lymphocytes transfected with a reversible switching construct and expressing mer-Cre-mer show OHT-independent inversion of the floxed cassetteTo better understand the function of the Nudix-type ADPR-hydrolase NUDT9, our lab targeted this gene for deletion in DT40 chicken B cells. Since we were unable to target both alleles with classical knockout constructs, we reasoned that loss of NUDT9 must be lethal in this cell type, and that therefore a conditional knockout strategy would be required. We chose to proceed with a reversible switching construct (Fig. 1A), in which a chicken NUDT9 cDNA and an mCherry fluorescence reporter were present in inverted orientation to one another. This organization afforded us a fluorescent readout of Cre activity, with activity in 100% of cells correlating to roughly 50% red fluorescent cells [24]. Following selection of a clone that had stably integrated the construct (Fig. 1B), we transfected this clone, designated 27Flip, with a vector constitutively expressing mCrem under the control of the CMV promoter.Following transfection of the 27Flip cells with the mCrem containing vector, stably expressing clones were identified and a highly expressing clone, designated 27Flip/Cre20 was selected on the basis of an mCrem western blot (Fig. 1C). Based on previous work [20], [24], it was anticipated that prior to treatment with OHT, 27Flip/Cre20 cells would show no mCherry fluorescence, similar to the parental 27Flip line. However, the 27Flip/Cre20 clone showed a sub population with robust red fluorescence (Fig. 1D) prior to any tamoxifen exposure. Thus, we sought to identify the cause of mCherry expression in these cells.Sustained OHT-independent flipping of the floxed cassette in 27Flip/Cre20 cells is a stable, ongoing processWe initially hypothesized that electroporation or other stressors associated with transduction of DT40 cells with the mCrem containing construct might lead to inefficient exclusion of mCrem from the nucleus and subsequent Cre dependent switching of our construct in affected cells. To test this hypothesis, we sorted mCherry negative (“white”) and mCherry-positive (“red”) cells by flow cytometry, and monitored the fluorescence of the separated populations over time in the absence of OHT (Fig. 2). If mCrem activity were a transient event associated with electroporation, these sorted populations would be expected to remain stable over time. However, we observed that white cells showed an outgrowth of red cells and vice versa over a course of 7 weeks in culture. Interestingly, red cells converted to white cells more rapidly, a phenomenon potentially related to idiosyncratic differences in susceptibility of the flip cassette to recombination, or a survival disadvantage associated with loss of NUDT9 expression. These results suggested that in contrast to previous results, mCrem was capable of mediating significant levels of constitutive recombination activity. Because such unlicensed mCrem dependent recombination is a potential confounding factor for the reversible switching approach to conditional gene inactivation, we set out to better characterize this phenomenon.10.1371/journal.pone.0003264.g002Figure 2Sustained OHT-independent flipping of the floxed cassette in 27Flip/Cre20 cells is a stable, ongoing process.mCherry (+) and mCherry (−) 27Flip/Cre20 cells were separated by FACS and grown in culture for ∼7 weeks. During this time period, a steady decrease in the purity of the sorted populations was observed by flow cytometry. X-axis, mCherry fluorescence, Y-axis, cell number.Sustained OHT-independent Cre activity in 27Flip/Cre20 cells correlates with mer-Cre-mer expression levelSeveral previous observations suggest that control of mCrem recombination is less than perfect [11], [20], [26], [27]. Based on these observations, we were interested in clarifying the role of mCrem expression level in OHT-independent recombination. We hypothesized that low frequency OHT-independent recombination by mCrem is mCrem intrinsic, and thus that OHT-independent Cre activity should be demonstrable in other mCrem expressing clones, and furthermore that as more mCrem would provide more activity, the level of constitutive flipping should correlate with mCrem expression level. To test this hypothesis, we re-transfected 27Flip cells with the mCrem containing vector to generate a panel of 27Flip/Cre clones expressing varying amounts of mCrem by western blot (Fig. 3A). These clones were assigned to either “high” (27Flip/Cre15 and -20) or “low” (27Flip/Cre26, -31, and -35) mCrem expressing categories and compared for flipping in the presence and absence of 1 µM OHT (Fig. 3B). Consistent with mCrem possessing intrinsic OHT-independent recombination activity, clones with high expression showed OHT-independent Cre activity, while low expressing clones exhibited little to no activity. Although the apparent decrease in flipping for the high expressing clones at high OHT concentrations suggested a potential difference in efficiency of recombination depending on the orientation of the cassette, repeating these experiments revealed variation in the size of the red population ranging from 38–60% that changed from day to day (data not shown), suggesting that background fluctuations in the population occur in the context of constitutive ongoing flipping. Importantly, all clones responded well to OHT treatment based on their establishing a ∼50∶50 ratio of red to white cells over a 48 hour OHT treatment (with a 50/50 ratio indicating flipping in ∼100% of cells). Note that clone 35, which at first glance appeared to have less red cells than other clones, showed poorer separation of red and white populations than other clones, making the percentage of white vs. red cells less precise in this clone than others.10.1371/journal.pone.0003264.g003Figure 3Sustained OHT-independent flipping of the floxed cassette in 27Flip/Cre20 cells correlates with mer-Cre-mer expression level.\nA Following stable transfection of 27Flip cells with an mCrem expressing vector, 5 clones expressing various amounts of mCrem were selected. Cells from the clones were lysed and 50 µg of protein of each were run on an 8% SDS-PAGE gel. MCrem was visualized with polyclonal rabbit anti-Cre antibody (1∶2000, Novagen), with β-actin visualized with a polyclonal mouse antibody (1∶40000, Sigma) as a loading control. 27Flip/Cre15 and -20 were designated “high expressors”, and 27Flip/Cre26, -31, and -35 were designated “low expressors”. B 27Flip/Cre clones were either left untreated or treated with 1 µM OHT for 48 hours and then allowed to grow in culture for 18 days: all clones were subsequently analyzed by flow cytometry. The 27Flip parental cell line was used as a negative control.OHT-dependent mCrem activity saturates by 10 nM OHTIn a previous study [24] an OHT concentration of 50 nM was used to induce mCrem mediated recombination, significantly lower than the 1 µM used in the studies above. Thus, we wanted to investigate the interaction between mCrem expression level and OHT dose for induction of mCrem mediated recombination. Thus, we treated both the high and low expressing 27Flip/Cre clones with 100, 10, or 1 nM OHT for 48 hours, and compared the treatment groups for Cre activity at 5 days after the start of OHT treatment (Fig. 4). Since during OHT treatment the construct spends equal time in the forward and reverse orientations, all cells express mCherry during this time: as such, we included a “rest” period of 3+ days to allow white cells to lose any residual fluorescence. Clones 15 and 20 with constitutive flipping activity exhibited only minor changes in population distribution after OHT treatment - these minor changes may reflect differing efficiencies of flipping in the forward and reverse directions or differential effects of tamoxifen toxicity on the two populations at the higher doses. Clones with lower mCrem expression all exhibited dose dependent recombination, with the clone with the lowest apparent mCrem expression (based on a qualitative assessment of the western blot in Fig. 3A), 27Flip/Cre31, showing the lowest level of activity at 1 nM Cre. This result suggests that at sufficiently low OHT concentration, mCrem expression level becomes limiting for recombination efficiency. Nevertheless, although mCrem expression in this clone was near the limit of detection in our western blot, we saw flipping in 100% of cells at 10 nM OHT, suggesting that this concentration is sufficient to maximize Cre activity when applied for 48 hours. Consistent with previous work [11] we observed significant mCrem mediated toxicity at 1 µM and 100 nM OHT, which decreased at 10 nM and was essentially absent at 1 nM. In addition, the frequency of cell death was consistently higher among high mCrem expressors as compared to low expressors at 100 and 10 nm OHT, plateauing at 1 µM and being undetectable at 1 nM for all clones (data not shown).10.1371/journal.pone.0003264.g004Figure 4OHT-dependent flipping of the floxed cassette saturates by 10 nM OHT.27Flip/Cre clones were either left untreated or treated with 1, 10, or 100 nM OHT for 48 hours and then allowed to grow in culture for 5 days from the onset of OHT treatment: subsequently all clones were analyzed by flow cytometry. 27Flip parental cells were used as a negative control. X-axis, mCherry fluorescence, Y-axis, cell number.Efficiency of OHT-dependent Cre activity correlates with duration of OHT exposureWe noted that at all concentrations of OHT, the majority of treated cells became positive for red fluorescence within 24 hours (data not shown), suggesting that the majority of cells may already have initiated switching well before the 48 hours of OHT exposure typically employed by our and other labs [24]. We therefore set out to identify the effect of varying OHT exposure time in clones expressing different amounts of mCrem. We treated the 27Flip/Cre clones with 1 nM OHT for 1, 2, 5, 10, and 24 hours before removing OHT and allowing the cells to grow for five days in culture, at which point we examined the cells for red fluorescence by FACS (Fig. 5A). In all clones except the high expressor 27Flip/Cre15, a steady decline in recombination activity was evident as OHT exposure time decreased: for 27Flip/Cre15 flipping was already at ∼100% in the untreated population. Coupled with the dose response of the various clones, these results also provided further support for our conclusion from Fig. 3 that recombination efficiency decreases with decreasing mCrem expression. Flipping in 100% of the cells in all clones except 27Flip/Cre15 was not seen at exposure times less than 48 hours.10.1371/journal.pone.0003264.g005Figure 5Efficiency of OHT-dependent flipping of the floxed cassette correlates with duration of OHT exposure in low mCrem expressors.\nA 27Flip/Cre clones were either left untreated or treated with 1 nM OHT for 24, 10, 5, 2, or 1 hours, and then allowed to grow in culture for 5 days: subsequently all clones were analyzed by flow cytometry. 27Flip parental cells were used as a negative control. B 27Flip/Cre31 cells were either left untreated or treated with 100 nM OHT for 1, 2, 5, 10, 30, 60, 120, or 240 minutes and then allowed to grow in culture for 5 days: subsequently all clones were analyzed by flow cytometry. X-axis, mCherry fluorescence, Y-axis, cell number.Surprisingly, we observed low level induction of Cre activity at even 1 hour of 1 nM OHT exposure, a much shorter time than is typically used for mCrem induction. Consequently we were curious if, by using a high concentration of mCrem we might be able increase the percentage of cells with Cre activity at early time points. Since we showed that 100% of 27Flip/Cre31 cells underwent flipping at 10 nM applied for 48 hours (Fig. 4), we reasoned that a ten fold excess of this saturating concentration would be sufficient to induce rapid Cre activity. Thus we applied 100 nM OHT to 27Flip/Cre31 cells for short periods of time, allowed the cells to grow in culture for 5 days, and then analyzed flipping by FACS (Fig. 5B). Under these conditions, we detected roughly 50% of cells undergoing flipping after 4 hours of treatment with 100 nM OHT, and that roughly 1% of cells had undergone flipping within just 10 minutes of OHT treatment. These results suggest that although high concentration/short time exposure of OHT can induce detectable recombination, recombination does not occur with any degree of synchronicity in DT40 cells, even within cells of a clonal population.DiscussionAlthough some studies of the mCrem system [20], [24], [36] have suggested that mCrem mediated recombination is tightly controlled in the absence of ligand, work from other labs suggested ligand independent mCrem mediated recombination might occur, albeit at low frequency [11], [19], [26], [27]. This discrepancy has not been systematically addressed to date in vitro or in vivo. This prompted us to use a reversible switching system for analysis of the parameters governing tamoxifen/OHT-regulated mCrem recombination of a genomic target. In this report, we show that robust mCrem expression correlates with a high level of tamoxifen/OHT-independent Cre activity, while clones expressing mCrem at the limit of western blot detection exhibit extremely tight regulation. Additionally, we demonstrate that the Tamoxifen/OHT dose response of clones varies with mCrem expression level.In contrast to our results, it has been previously shown that in DT40 cells expressing a reversible switching construct and mCrem, little to no OHT-independent Cre activity occurred [24]. These differences may be due to epigenetic differences at the respective integration loci [37] or possibly the serendipitous use of a low mCrem expressing clone (Cre expression levels were not examined in those studies). Although it is difficult to predict or control the former, our results clearly show that limiting mCrem expression may be effective in controlling tamoxifen/OHT-independent mCrem mediated recombination, although at the expense of potentially requiring higher dose or duration of OHT treatment to induce maximal recombination. Thus, future studies using reversible switching should evaluate clones with a range of low but detectable mCrem expression to identify those which minimize unregulated mCrem activity but provide maximal recombination at the desired OHT does and time of exposure. Furthermore, it is important to monitor the stability of both the forward and reverse populations in any reversible switching study (particularly if an antibiotic resistance cassette is integrated in either direction, since this may mask switching out of the given orientation), for evidence of OHT-independent mCrem mediated recombination. If unlicensed Cre activity is an issue, clones that appear mCrem negative by western blot should be screened via FACS with or without OHT treatment, since such clones may express mCrem at levels low enough to be undetectable by western blot but nevertheless be capable, as our results suggest, of mediating mCrem dependent recombination.Given that we observed tamoxifen/OHT-independent Cre activity in several high expressing clones, we speculate that this phenomenon is an intrinsic property of mCrem in DT40 cells, possibly because of their small cytoplasmic volume, which may encourage mCrem nuclear translocation as expression increases. It is probably further generalizable that higher expression levels of mCrem in any cell line or in vivo model will lead to a greater risk for unlicensed activity: previous findings suggest that tamoxifen/OHT-independent Cre activity also occurs in mice and murine cells cultured in vitro\n[26], [27]. Thus, our results suggest that investigators should limit mCrem expression in any Cre-loxp based model system where tight ligand-dependent regulation of mCrem activity is crucial (our results in Figs. 3B, and 4 indicate that even minimal expression of mCrem is sufficient for 100% penetrance of Cre activity given an adequate OHT dose and time of exposure).Previous studies have examined the effects of tamoxifen/OHT dosage on inducible Cre-mediated recombination, but never in the context of variable Cre expression. Thus, our results provide insight on tamoxifen/OHT dose and/or time of exposure adjustments that must be made at varying Cre expression levels. If low mCrem expressing clones are selected for study, as we recommend above, investigators can expect that a sufficiently high dose of OHT must be applied if 100% penetrance of recombination is desired. Surprisingly, in our system a “sufficiently high dose” titrated to only 10 nM OHT applied for 48 hours for our lowest mCrem expressing clone, a dose 100 fold lower than our standard OHT dose and 5 fold lower than the dose used in a previous study [24]. Since work from our lab and others suggests that mCrem mediated toxicity correlates with OHT dose, we recommend the use of 10 nM OHT for 48 hours to maximize penetrance while minimizing toxicity. This recommendation is supported by our time course studies, since even in the highly expressing clone 27Flip/Cre20, 1 nM OHT exposure times of 24 hours or less resulted in decreased Cre activity. Nevertheless, exposure time and/or OHT dose can be adjusted considerably depending on mCrem expression and the desired level of penetrance. The finding that Cre activity is evident in some 27Flip/Cre31 cells within 10 minutes of high dose OHT (100 nM) application, but that activity in 100% of cells cannot be achieved within 4 hours was surprising, as it suggests that induction of mCrem activity within even a clonal cell population is highly asynchronous. Whether better synchronicity can be achieved by examining only cells at a similar point in the cell cycle or by applying toxic doses of ligand for short periods of time are interesting areas for future investigation.In conclusion, we show for the first time that robust mCrem expression allows Cre activity independent of ligand binding, while low mCrem expressing clones are largely free of this effect, and that tamoxifen/OHT dose response correlates with changing mCrem expression. Time and dose-dependent effects on mCrem activity suggest limitations on the use of conditional targeting approaches for applications which require tight temporal coordination of Cre action within a cell population.Materials and MethodsCell CultureDT-40 B lymphocytes were cultured at 37°C with 5% CO2 in RPMI 1640 Medium (Mediatech Inc.) supplemented with 10% fetal bovine serum (FBS; Mediatech Inc.), 2% chicken serum (Invitrogen), 10 units/ml penicillin/streptomycin (Mediatech Inc.), 2 mM glutamine (Mediatech Inc.) and beta mercaptoethanol (50 µM; Sigma).Molecular BiologyThe “Flip” Construct backbone containing the loxp sites, 2 MCS sites, and the SV40 and BGH poly-A signals was synthesized by DNA 2.0, Inc. Subsequently, the CMV promoter, chicken NUDT9 cDNA, mCherry cDNA, and a B-actin-Neomycin resistance cassette were inserted using various restriction sites. The Left and Right homology arms extended from exon 4 to exons 1 and 7 respectively. Integration of the Flip construct into the NUDT9 locus was verified using the following primers, which span the Neo-Right homology Arm and the Right homology arm-genomic locus junctions in their product: Forward: ACATAGCGTTGGCTACCCGTGATA, Reverse: ACTGCTTTGAAGGCCACACATTCC. The product of this PCR was also sequenced following gel- or PCR purification. Following transfection, clones were selected in G418 (2 mg/mL, invitrogen), and selected clones were screened for targeted integration via PCR.Mer-Cre-mer, expressed in the pcDNA5T/O vector (invitrogen), was a generous gift from Michael Reth by way of Tomo Kurosaki. Following transfection of mer-Cre-mer, clones were selected in hygromycin (2 mg/mL, Calbiochem), and selected clones were evaluated for mer-Cre-mer expression by western blot. Clones were lysed and 50 µg of protein from each clone were run on an 8% SDS-PAGE gel. Mer-Cre-mer was visualized using polyclonal rabbit anti-Cre antibody (1∶2000, Novagen, Inc.). β-actin controls were stained with a mouse polyclonal anti-β actin antibody (1∶40,000, Sigma). The secondary antibody was peroxidase conjugated donkey anti-rabbit from Amersham Pharmaceuticals.Stable tranfection of DT-40 B lymphocytes was carried out using a Bio-Rad Gene-Pulser electroporation apparatus. Cells (1×107/0.5 ml serum-free medium) were pulsed in 0.4-cm cuvettes with 50 µg plasmid DNA at 550 V and 25 µF.Flow CytometryFACS analyses were performed on either a BD FACSAria (Becton-Dickenson) or a BD LSRII (Becton-Dickenson), using a green laser (561 nm) to excite the mCherry protein (excitation max: 587 nm, emission max: 610 nm). Cells were resuspended in media as described above, without the color indicator. 10,000 events were collected for each panel. Sorting experiments were carried out on the BD FACSAria.Cre inductionCells expressing mer-Cre-mer were induced using 1 µM, 100 nM, 10 nM, or 1 nM 4-Hydroxytamoxifen (OHT) for 48, 24, 10, 5, 2, or 1 hour, or left untreated. Following tamoxifen treatment, cells were centrifuged at 500× g for 5 minutes to collect the cells, resuspended in 10 mL RPMI, and centrifuged again to wash the cells. The cells were then resuspended in standard medium and grown in culture for 3–50 days.\n\nREFERENCES:\n1. BrandaCSDymeckiSM\n2004\nTalking about a revolution: The impact of site-specific recombinases on genetic analyses in mice.\nDev Cell\n6\n7\n28\n14723844\n2. GlaserSAnastassiadisKStewartAF\n2005\nCurrent issues in mouse genome engineering.\nNat Genet\n37\n1187\n1193\n16254565\n3. Schmidt-SupprianMRajewskyK\n2007\nVagaries of conditional gene targeting.\nNat Immunol\n8\n665\n668\n17579640\n4. SempriniSTroupTJKotelevtsevaNKingKDavisJR\n2007\nCryptic loxP sites in mammalian genomes: genome-wide distribution and relevance for the efficiency of BAC/PAC recombineering techniques.\nNucleic Acids Res\n35\n1402\n1410\n17284462\n5. ThyagarajanBGuimaraesMJGrothACCalosMP\n2000\nMammalian genomes contain active recombinase recognition sites.\nGene\n244\n47\n54\n10689186\n6. RajewskyKGuHKuhnRBetzUAMullerW\n1996\nConditional gene targeting.\nJ Clin Invest\n98\n600\n603\n8698848\n7. MetzgerDChambonP\n2001\nSite- and time-specific gene targeting in the mouse.\nMethods\n24\n71\n80\n11327805\n8. LeeJYRistowMLinXWhiteMFMagnusonMA\n2006\nRIP-Cre revisited, evidence for impairments of pancreatic beta-cell function.\nJ Biol Chem\n281\n2649\n2653\n16326700\n9. 2007\nToxic alert.\nNature\n449\n378\n17898721\n10. de AlboranIMO'HaganRCGartnerFMalynnBDavidsonL\n2001\nAnalysis of C-MYC function in normal cells via conditional gene-targeted mutation.\nImmunity\n14\n45\n55\n11163229\n11. LoonstraAVooijsMBeverlooHBAllakBAvan DrunenE\n2001\nGrowth inhibition and DNA damage induced by Cre recombinase in mammalian cells.\nProc Natl Acad Sci U S A\n98\n9209\n9214\n11481484\n12. PfeiferABrandonEPKootstraNGageFHVermaIM\n2001\nDelivery of the Cre recombinase by a self-deleting lentiviral vector: efficient gene targeting in vivo.\nProc Natl Acad Sci U S A\n98\n11450\n11455\n11553794\n13. BabaYNakanoMYamadaYSaitoIKanegaeY\n2005\nPractical range of effective dose for Cre recombinase-expressing recombinant adenovirus without cell toxicity in mammalian cells.\nMicrobiol Immunol\n49\n559\n570\n15965304\n14. ForniPEScuoppoCImayoshiITaulliRDastruW\n2006\nHigh levels of Cre expression in neuronal progenitors cause defects in brain development leading to microencephaly and hydrocephaly.\nJ Neurosci\n26\n9593\n9602\n16971543\n15. BadeaTCWangYNathansJ\n2003\nA Noninvasive Genetic/Pharmacologic Strategy for Visualizing Cell Morphology and Clonal Relationships in the Mouse.\nJ Neuroscience\n23\n2314\n2322\n16. ZervasMMilletSAhnSJoynerAL\n2004\nCell Behaviors and Genetic Lineages of the Mesencephalon and Rhombomere 1.\nNeuron\n43\n345\n357\n15294143\n17. MatsudaTCepkoCL\n2007\nControlled expression of transgenes introduced by in vivo electroporation.\nPNAS\n104\n1027\n1032\n17209010\n18. AhnSJoynerLJ\n2005\n\nIn vivo analysis of quiescent adult neural stem cells responding to Sonic hedgehog.\nNature\n437\n894\n897\n16208373\n19. BreeRTLaiXYCanavanLELowndesNF\n2007\nComparisons between DT40 wildtype and DT40-Cre1 cells as suitable model systems for studying the DNA damage response.\nCell Cycle\n6\n2310\n2313\n17671434\n20. ZhangYRiestererCAyrallAMSablitzkyFLittlewoodTD\n1996\nInducible site-directed recombination in mouse embryonic stem cells.\nNucleic Acids Res\n24\n543\n548\n8604292\n21. HeidmannDLehnerCF\n2001\nReduction of Cre recombinase toxicity in proliferating Drosophila cells by estrogen-dependent activity regulation.\nDev Genes Evol\n211\n458\n465\n11685583\n22. FeilRWagnerJMetzgerDChambonP\n1997\nRegulation of Cre recombinase activity by mutated estrogen receptor ligand-binding domains.\nBiochem Biophys Res Commun\n237\n752\n757\n9299439\n23. FeilRBrocardJMascrezBLeMeurMMetzgerD\n1996\nLigand-activated site-specific recombination in mice.\nProc Natl Acad Sci U S A\n93\n10887\n10890\n8855277\n24. KanayamaNTodoKRethMOhmoriH\n2005\nReversible switching of immunoglobulin hypermutation machinery in a chicken B cell line.\nBiochem Biophys Res Commun\n327\n70\n75\n15629431\n25. SohalDSNghiemMCrackowerMAWittSAKimballTR\n2001\nTemporally Regulated and Tissue-Specific Gene Manipulations in the Adult and Embryonic Heart Using a Tamoxifen-Inducible Cre Protein.\nCirc Res\n89\n20\n25\n11440973\n26. BalordiFFishellG\n2007\nMosaic Removal of Hedgehog Signaling in the Adult SVZ Reveals That the Residual Wild-Type Stem Cells Have a Limited Capacity for Self-Renewal.\nJ Neuroscience\n27\n14248\n14259\n27. HayashiSMcMahonAP\n2002\nEfficient Recombination in Diverse Tissues by a Tamoxifen-Inducible Form of Cre: A Tool for Temporally Regulated Gene Activation/Inactivation in the Mouse.\nDevelopmental Biology\n244\n305\n318\n11944939\n28. BuersteddeJMTakedaS\n1991\nIncreased ratio of targeted to random integration after transfection of chicken B cell lines.\nCell\n67\n179\n188\n1913816\n29. LahtiJM\n1999\nUse of gene knockouts in cultured cells to study apoptosis.\nMethods\n17\n305\n312\n10196101\n30. AlinikulaJLassilaONeraKP\n2006\nDT40 mutants: a model to study transcriptional regulation of B cell development and function.\nSubcell Biochem\n40\n189\n205\n17623906\n31. ShinoharaHKurosakiT\n2006\nGenetic analysis of B cell signaling.\nSubcell Biochem\n40\n145\n187\n17623905\n32. RaineyMDZachosGGillespieDA\n2006\nAnalysing the DNA damage and replication checkpoints in DT40 cells.\nSubcell Biochem\n40\n107\n117\n17623903\n33. OkadaMHoriTFukagawaT\n2006\nThe DT40 system as a tool for analyzing kinetochore assembly.\nSubcell Biochem\n40\n91\n106\n17623902\n34. CaldwellRBKierzekAM\n2006\nGenome resources for the DT40 community.\nSubcell Biochem\n40\n25\n37\n17623898\n35. NadlerMJHermosuraMCInabeKPerraudALZhuQ\n2001\nLTRPC7 is a Mg.ATP-regulated divalent cation channel required for cell viability.\nNature\n411\n590\n595\n11385574\n36. LittlewoodTDHancockDCDanielianPSParkerMGEvanGI\n1995\nA modified oestrogen receptor ligand-binding domain as an improved switch for the regulation of heterologous proteins.\nNucleic Acids Res\n23\n1686\n1690\n7784172\n37. VooijsMJonkersJBernsA\n2001\nA highly efficient ligand-regulated Cre recombinase mouse line shows that LoxP recombination is position dependent.\nEMBO Rep\n2\n292\n297\n11306549"
+}

batch_0/PMC2533288.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533288",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533288\nAUTHORS: Rachel K Morris, Jeltsje S Cnossen, Marloes Langejans, Stephen C Robson, Jos Kleijnen, Gerben ter Riet, Ben W Mol, Joris AM van der Post, Khalid S Khan\n\nABSTRACT:\nBackgroundReliable antenatal identification of pre-eclampsia and small for gestational age is crucial to judicious allocation of monitoring resources and use of preventative treatment with the prospect of improving maternal/perinatal outcome. The purpose of this systematic review was to determine the accuracy of five serum analytes used in Down's serum screening for prediction of pre-eclampsia and/or small for gestational age.MethodsThe data sources included Medline, Embase, Cochrane library, Medion (inception to February 2007), hand searching of relevant journals, reference list checking of included articles, contact with experts. Two reviewers independently selected the articles in which the accuracy of an analyte used in Downs's serum screening before the 25th gestational week was associated with the occurrence of pre-eclampsia and/or small for gestational age without language restrictions. Two authors independently extracted data on study characteristics, quality and results.ResultsFive serum screening markers were evaluated. 44 studies, testing 169,637 pregnant women (4376 pre-eclampsia cases) and 86 studies, testing 382,005 women (20,339 fetal growth restriction cases) met the selection criteria. The results showed low predictive accuracy overall. For pre-eclampsia the best predictor was inhibin A>2.79MoM positive likelihood ratio 19.52 (8.33,45.79) and negative likelihood ratio 0.30 (0.13,0.68) (single study). For small for gestational age it was AFP>2.0MoM to predict birth weight < 10th centile with birth < 37 weeks positive likelihood ratio 27.96 (8.02,97.48) and negative likelihood ratio 0.78 (0.55,1.11) (single study). A potential clinical application using aspirin as a treatment is given as an example.There were methodological and reporting limitations in the included studies thus studies were heterogeneous giving pooled results with wide confidence intervals.ConclusionDown's serum screening analytes have low predictive accuracy for pre-eclampsia and small for gestational age. They may be a useful means of risk assessment or of use in prediction when combined with other tests.\n\nBODY:\nBackgroundPre-eclampsia (PET) and small for gestational age (SGA) remain significant causes of perinatal death and childhood disability [1-3]. PET has significant health implications for the mother with complications including adult respiratory distress syndrome, coagulopathy, renal and liver failure and stroke. Babies affected by SGA on reaching adulthood are at greater risk of developing cardiovascular disease, hypertension, and non-insulin dependent diabetes [4,5]. Both PET and SGA are characterized by a failure of the trophoblast invasion (at 16–22 weeks) into the spiral arteries.Second trimester serum screening for Down's syndrome is routinely offered to women in the United Kingdom and United States, either with the triple test (alpha-fetoprotein (AFP), human chorionic gonadotrophin (HCG) and unconjugated estriol) or with the addition of inhibin A as the quadruple test. More recently first trimester screening with fetal nuchal translucency, HCG and pregnancy associated plasma protein A (PAPP-A) has provided an earlier, more effective screening method [6]. Due to their origin and sites of metabolism these biochemical markers may be useful in the prediction of PET and SGA, there are however conflicting reports in the literature. Maternal serum levels of these analytes have been shown to be associated with adverse outcome [7,8] with low levels of PAPP-A having been suggested as a marker for impaired placental function and placentation [9]. There are studies however reporting contrasting views [10].Reliable antenatal identification of PET and SGA is crucial to judicious allocation of monitoring resources and use of preventative treatment [11] with the prospect of improving maternal and perinatal outcome. The variation in the design of research on accuracy of tests for prediction of PET and SGA, the scatter of this research across many databases and languages, and the dearth of clear collated up-to-date summaries of this literature contribute to the uncertainty about the best screening and monitoring strategies [12]. Systematic reviews of the literature can improve our ability to identify those pregnancies at increased risk of developing PET and SGA making additional use of test results already obtained for Down syndrome screening.The purpose of our review was to investigate the accuracy of serum biochemical markers used in first and second trimester Down syndrome serum screening in predicting PET and/or SGA. We systematically reviewed the available literature and meta-analysed the data.MethodsThe systematic review was based on our previously published prospective protocols [13,14] designed using widely recommended methods [15-18]. The protocols are available as Additional files 1 and 2.Data sources and searchesElectronic searches were performed by experienced clinical librarians targeting the prediction of PET and SGA. We searched Medline, Embase, the Cochrane Library (2006;4) and Medion from inception until February 2007. The search strategies are detailed in the published protocols [13,14] and in Additional file 3. The reference lists of all included primary and review articles were examined to identify cited articles not captured by electronic searches. No language restrictions were applied.Study selectionThe first stage of study selection was the scrutinizing of the database by two reviewers to identify articles from title and/or abstract. In a second stage, a search based on keywords for each of the analytes under review was performed within the Reference Manager database. The results of this search were scrutinized by a second reviewer. In the final stage of study selection the full papers of identified articles were obtained with final inclusion or exclusion decisions made after independent and duplicate examination of the papers. We included studies that reported on singleton pregnancies at any level of risk in any healthcare setting using any serum biochemical test used in Down syndrome serum screening before the 25th week of gestation. Test accuracy studies allowing generation of 2 × 2 tables were included.Data extraction and Study Quality AssessmentFurther details on inclusion and exclusion criteria and extracted clinical, methodological and statistical data can be found in the published protocols.Acceptable reference standards for PET were: persistent systolic blood pressure (SBP) ≥ 140 mmHg or diastolic blood pressure (DBP) ≥ 90 mmHg with proteinuria ≥ 0.3 g/24 hours or ≥ 1+ dipstick (= 30 mg/dl in a single urine sample), new after 20 weeks of gestation. Severe PET was defined as SBP ≥ 160 mmHg or DBP ≥ 110 mmHg with proteinuria ≥ 2.0 g/24 hours or ≥ 3+ dipstick, or of early onset < 34 weeks gestation. Superimposed PET was defined as the development of proteinuria ≥ 0.3 g/24 hours or ≥ 1+ dipstick after 20 weeks gestation in chronically hypertensive patients [19]. Acceptable reference standards for SGA included birth weight < 10th centile adjusted for gestational age and based on local population values and absolute birth weight threshold < 2500 g. Severe SGA was defined as birth weight < 5th or < 3rd centile or < 1750 g or and preterm SGA for SGA leading to delivery < 37 weeks. Neonatal ponderal index < 10th centile, skin fold thickness, and mid-arm circumference/head circumference were also assessed [20-24].Disagreements were resolved by consensus or arbitration of a third reviewer. For multiple/duplicate publication of the same data set, the most recent and/or complete study was included only.All included manuscripts were assessed by at least one reviewer for study and reporting quality using validated tools [25-30]. Methodological quality was defined as the confidence that the study design, conduct and analysis have minimized biases in addressing the research question, thereby focusing on the internal validity (i.e. the degree to which the results of an observation are correct for the patients being studied). Items considered important for a good quality paper were prospective design with consecutive recruitment, full verification of the test result with reference standard (> 90%), adequate description of the index test and use of appropriate reference standard, and application of any preventative treatments. Additional quality items were assessed for SGA papers; whether they excluded cases of PET from the results, whether fetuses with chromosomal and structural anomalies were excluded and whether stillbirths and intrauterine deaths were excluded from the results. Further explanation of the quality assessment can be found in Additional file 4.We excluded from the statistical analysis any paper with a case-control design as this type of design in diagnostic test accuracy studies has been shown to be associated with bias and over/under estimation of accuracy [29].Data synthesis and AnalysisFrom the 2 × 2 tables the following were calculated with their 95% confidence intervals for individual studies; sensitivity (true positive rate), specificity (true negative rate) and the likelihood ratios (LR, the ratio of the probability of the specific test result in people who do have the disease to the probability in people who do not). LRs indicate by how much a given test result raises or lowers the probability of having the disease and have been recommended by Evidence-based Medicine Groups [31,32]. Results were pooled among groups of studies with similar characteristics, the same threshold for the index test (PET and SGA), same reference standard threshold for (SGA) and the same trimester for testing. Where 2 × 2 tables contained zero cells, 0.5 was added to each cell to enable calculations.Sub-groups were defined at the start of the review based on clinical criteria known to affect prognosis, method of index test or study quality: level of risk of population (high or low based on authors assessment and calculated incidence rates from results); type of assay used for index test; whether babies with chromosomal anomalies were excluded from the results; use of preventative treatment; quality of study. Sub-group analyses were performed where there were at least 3 studies with similar characteristics within that group.Heterogeneity was assessed graphically by looking at the distribution of the sensitivities and specificities in the receiver operating characteristic (ROC) space and LRs as a measurement of accuracy size using a Forest plot. The loglikelihood and X2 test were used to assess for heterogeneity statistically. When X2 p value > 0.05 (homogenous data) the fixed effect pooling method was used; where there was heterogeneity random effects pooling was used. Summary ROC plots were produced (data not shown). Sensitivity analysis was performed to check the robustness of our results. A p value of < 0.05 was used throughout for statistical significance.All statistical analyses were performed using Meta-Disc software  and Statsdirect for drawing the Forest plots.Clinical applicationThe clinical impact of estimates of accuracy for a screening test depend on how the results of the test alter the patient's pre-test probability of disease, based on disease prevalence. The post-test probability can then be combined with estimates of effectiveness for known treatments [33]. From this data we can then calculate the number of women needed to be tested (number needed to test- NNTest), using a particular serum marker, to prevent one case of SGA with a particular treatment and the number needed to treat (NNTreat), the number of test positive women needed to be treated to prevent one case of SGA. In this review clinical application will be assessed using aspirin as this is the only treatment with any level of effectiveness for PET and SGA [11,34].ResultsLiterature identification, study characteristics, and qualityFigure 1 summarises the process of literature identification and selection. Tables detailing the individual study characteristics of the included studies are available in Additional file 5. There were twenty studies that reported on both PET and SGA.Figure 1Process from initial search to final inclusion for biochemical screening to predict pre-eclampsia/small for gestational age (up to February 2007). PET preeclampsia; PIH pregnancy induced hypertension; SGA small for gestational age.Pre-eclampsiaThere were 44 included studies for pre-eclampsia [7,9,35-75] reporting on 169,637 women (4376 preeclamptic women, incidence 2.6%). Among these 44 studies, there were 35 cohort studies and nine case-control studies [41,43,44,48,51,55,63,72,73]. There were nine prospective studies, 10 retrospective and 25 were unclearly designed. Calculated incidence rates of PET ranged from 0.6–44%. Incidence rates of PET correlated poorly with descriptions of \"high\" or \"low\" risk study populations. Four of the studies were in \"high-risk\" populations (one in IVF patients, one in patients with abnormal uterine artery Doppler and two in patients with chronic hypertension) and in three of these studies the incidence of PET was > 4%. However in 15 of the \"low-risk\" studies the incidence was > 4% and in one study in which the inclusion criteria were unclearly reported. The remaining 25 studies were in low risk, screening populations with a calculated incidence of PET < 4%.Ten studies were performed in the first trimester, 32 studies at a mean gestation between 15 to 20 weeks and two studies 20 to 24 weeks.The quality assessment of included studies for PET is summarized in Figure 2. There was poor reporting of patient selection criteria, description of index and reference tests and blinding of the reference test. Only two studies reported clearly whether preventative treatment had been used. The nine case control studies were excluded from the final meta-analysis, leaving 35 cohort studies for analysis.Figure 2Bar chart showing quality of evidence on biochemical screening markers to predict small for gestational age and pre-eclampsia.Small for gestational ageThere were 86 included studies for SGA [7,9,37,39,47,51,53-55,57,59-61,64-67,69,74,76-141], reporting on 382,005 women (20339 cases of SGA, incidence 5.32%). Among these studies, there were 61 cohort studies and 25 case control studies [53,55,76,77,82-85,88,89,94,96,97,104,113,116,119,123-125,130,131,133,135,140]. Thirty-one studies were prospective, 17 retrospective and 38 of unclear design. Calculated incidence rates of SGA correlated well with the threshold used in 78 of studies and poorly in 8, incidence range for birth weight < 10th centile was 1.2–63%. Three of the studies were performed in high risk populations, whereas the remainder were performed in low risk or screening populations. Due to the inclusion criteria of the studies the majority of tests were performed between 15 to 20 weeks. There were ten studies reporting on first trimester screening. Fifty studies reported on birth weight < 10th centile, 13 on birth weight < 5th centile, 27 on birth weight < 2500 g, 1 on birth weight < 1500 g, 1 on birth weight < 15th centile and 12 reported no threshold.The quality assessment of included studies for SGA revealed deficiencies (Figure 2). Only 40 studies contained an adequate description of the performance of the index test. None of the studies reported clearly on the performance of the reference standard. Blinding of the reference test was also poorly reported as was the use of any treatment in between the index test and reference standard. These items of quality of study design are important in diagnostic accuracy reviews.Four papers only distinguished between SGA with PET and SGA alone; intrauterine deaths and stillbirths were excluded from the results for SGA in only 16 papers, in the remainder it was unclear; chromosomal and structural anomalies were excluded from 62 studies, unclear in 24Twenty-five case control studies and eight studies [78,81,98,105,122,127,129,138] in which thresholds for SGA were not defined were excluded from the final meta-analysis, leaving 53 studies.Data analysisFor both analysis for PET and analysis for SGA, there was significant heterogeneity in all results. As a consequence of this the random effects model was used throughout the study.Maternal serum alpha fetoprotein (AFP)The results for AFP are summarized in Figure 3, all studies were performed in the second trimester. For PET there were sixteen studies included in the meta-analysis. Thresholds that were most commonly used were > 2.0MoM (multiples of median) (10 studies) and > 2.5MoM (6 studies). The most accurate predictor was AFP>2.0 MoM; LR+ 2.36 (1.46,3.83), LR- 0.96 (0.95,0.98). (One study had a better positive LR however this threshold was chosen from receiver operating curve analysis AFP>1.28MoM; LR+ 3.30 (2.00,5.43), LR- 0.44(0.22,0.90)).Figure 3Forest Plot showing likelihood ratio of a positive and negative test result with 95% confidence intervals (95% CI) for studies of alpha feto-protein (AFP) to predict pre-eclampsia and small for gestational age (birth weight threshold as indicated). Results with diamonds are pooled results (number of studies as indicated), results with squares are single studies. The number of women included in the studies is shown, all studies second trimester testing.For SGA there were thirty studies included in the meta-analysis. The commonest threshold used were > 2.0MoM (10 studies) and > 2.5MoM (5 studies) to predict birth weight < 10th centile. The best predictor for birth weight < 10th centile was AFP<10th centile; LR+ 8.80 (5.57,13.91), LR- 0.02 (0.00,0.34), this was a single study. For birth weight < 5th centile and birth weight < 2500 g, AFP>3.0MoM was the most accurate predictor. The most accurate predictor overall was AFP>2.0MoM to predict severe SGA (birth weight < 10th centile with birth < 37 weeks): LR+ 27.96 (8.02,97.48), LR- 0.78 (0.55, 1.11).Maternal serum human chorionic gonadotrophin (HCG)The results for HCG are summarized in Figure 4. There were forty seven studies overall evaluating HCG, nine for free β-HCG, eight total β-HCG and 30 total HCG. For PET there were 21 included studies in the meta-analysis, 3 looked at testing in the first trimester. The commonest thresholds used were HCG>2.0MoM (12 studies), HCG>2.5MoM (4 studies) and HCG>3.0MoM (3 studies). The most accurate predictor was HCG>2.0MoM with second trimester testing; LR+ 2.45 (1.57,3.84), LR- 0.89 (0.83,0.96). There was one study looking at severe PET as the outcome, results showed no improvement in prediction.Figure 4Forest Plot showing likelihood ratio of a positive and negative test result with 95% confidence intervals (95% CI) for studies of human chorionic gonadotrophin (HCG) to predict pre-eclampsia and small for gestational age (birth weight threshold as indicated). Results with diamonds are pooled results (number of studies as indicated), results with squares are single studies. The number of women included in the studies is shown. (a first trimester testing).For SGA there were 22 included studies in the meta-analysis, 5 looked at testing in the first trimester. The commonest thresholds used were HCG>2.0MoM (7 studies) and HCG>2.5MoM (4 studies) for birth weight < 10th centile. The most accurate predictor for birth weight < 10th centile was HCG>2.0MoM; LR+ 1.74 (1.48,2.04), LR- 0.95 (0.93,0.96). For birth weight < 5th centile HCG>2.0MoM in the second trimester was the most accurate and for birth weight < 2500 g HCG>2.5MoM.Maternal serum unconjugated EstriolThe results for unconjugated estriol are summarized in Figure 5, all studies were performed in the second trimester. For PET there were 4 included studies, the commonest threshold being estriol<0.5MoM (2 studies), this was also the most accurate predictor; LR+ 1.50 (1.02,2.19), LR- 0.99 (0.97,1.00).Figure 5Forest Plot showing likelihood ratio of a positive and negative test result with 95% confidence intervals (95% CI) for studies of estriol to predict pre-eclampsia and small for gestational age (birth weight threshold as indicated). Results with diamonds are pooled results (number of studies as indicated), results with squares are single studies. The number of women included in the studies is shown, all studies second trimester testing.For SGA there were 7 included studies, the commonest threshold was estriol<0.75MoM (2 studies) for birth weight < 10th centile. The most accurate predictor for birth weight < 10th centile was estriol<0.75MoM; LR+ 2.54 (1.54,4.19), LR- 0.75 (0.63,0.89). For birth weight < 5th centile there were 2 studies for estriol<0.5 MoM; LR+ 6.54 (0.98,43.91), LR- 0.59 (0.03,13.28).Maternal serum pregnancy associated plasma protein A (PAPP-A)The results for PAPP-A are summarized in Figure 6. For PET there were 16 included studies, all performed in the first trimester, the commonest threshold was PAPP-A<5th centile (5 studies) and PAPP-A<10th centile (3 studies). The most accurate predictor was PAPP-A<5th centile; LR+ 2.10 (1.57,2.81), LR- 0.95 (0.93,0.98).Figure 6Forest Plot showing likelihood ratio of a positive and negative test result with 95% confidence intervals (95% CI) for studies of pregnancy associated plasma protein A (PAPPA) to predict pre-eclampsia and small for gestational age (birth weight threshold as indicated). Results with diamonds are pooled results (number of studies as indicated), results with squares are single studies. The number of women included in the studies is shown. (a first trimester testing).For SGA there were 10 included studies, 7 were performed in the first trimester, the commonest thresholds were PAPP-A < 5th centile (4 studies), PAPP-A<10th centile (5 studies) for birth weight < 10th centile. The most accurate predictor for birth weight < 10th centile was PAPP-A<1st centile; LR+ 3.50 (2.53,4.82), LR- 0.98 (0.97,0.99). For birth weight < 5th centile, the most accurate predictor was again PAPP-A<1st centile; LR+ 4.36 (3.27,5.80), LR- 0.97 (0.96,0.98).Maternal serum inhibin AThe results for inhibin A are summarized in Figure 7. For PET there were 6 included studies, 1 performed in the first trimester, the commonest threshold being inhibin A>2.0MoM (2 studies) with a LR+ 6.00 (5.12,7.03), LR- 0.72 (0.48,1.09). The most accurate predictor for PET was inhibin A>2.79MoM; LR+ 19.52 (8.33,45.79), LR- 0.30 (0.13,0.68), however this result was derived from one study using a receiver operating characteristic curve to determine threshold.Figure 7Forest Plot showing likelihood ratio of a positive and negative test result with 95% confidence intervals (95% CI) for studies of Inhibin A to predict pre-eclampsia and small for gestational age (birth weight threshold as indicated). Results with diamonds are pooled results (number of studies as indicated), results with squares are single studies. The number of women included in the studies is shown. (a first trimester testing).For SGA there was only one study, looking at second trimester testing, using a cut-off of inhibin A>2.0MoM, the results for prediction of birth weight < 10th centile were LR+ 4.45 (3.92,5.06), LR- 0.92 (0.91,0.93) and birth weight < 5th centile; LR+ 4.91 (4.20,5.73), LR- 0.89 (0.87,0.91).Triple test (serum AFP, HCG and unconjugated estriol)There were no included studies for PET. For SGA there were 2 studies, second trimester testing, with different cut-offs for prediction of birth weight < 10th centile: triple test > 1:190 LR+ 1.07 (0.60,1.91), LR- 0.98 (0.82,1.17) and triple test>1:250 LR+ 2.71 (1.77,4.17), LR- 1.19 (0.01,2.47).Gestation of testingTable 1 shows the different results achieved where testing was performed in both the first and second trimester. Overall for HCG, testing in the second trimester was more accurate.Table 1Subgroup analyses of accuracy of biochemical screening to predict small for gestational age and pre-eclampsia (random effects pooling).Small for gestational ageAnalyte SubgroupPositive Likelihood Ratio (95% CI)Negative Likelihood Ratio (95% CI)Sensitivity (95% CI)Specificity (95% CI)HCG>90th centile (BW<10th centile)TrimesterFirst1.48 (0.57–3.81)0.92 (0.72–1.17)0.21 (0.06–0.46)0.86 (0.79–0.91)Second1.68 (0.37–7.63)0.97 (0.86–1.09)0.08 (0.01–0.26)0.95 (0.90–0.98)HCG<10th centile (BW<10th centile)TrimesterFirst1.29 (0.05–33.56)1.14 (0.53–2.43)0.13 (0.10–0.16)0.60 (0.57–0.63)Second2.35 (0.80–6.92)0.90 (0.76–1.08)0.16 (0.05–0.36)0.93 (0.88–0.97)HCG>2.0MoM (BW<5th centile)TrimesterFirst0.96 (0.55–1.68)1.01 (0.88–1.17)0.20 (0.10–0.34)0.79 (0.77–0.81)Second2.08 (1.78–2.42)0.94 (0.92–0.95)0.12 (0.10–0.14)0.94 (0.94–0.95)PAPPA<10th centile (BW<10th centile)TrimesterFirst1.68 (1.25–2.27)0.93 (0.88–0.98)0.17 (0.16–0.19)0.90 (0.89–0.90)Second1.82 (0.95–3.50)0.91 (0.75–1.05)0.20 (0.10–0.33)0.89 (0.85–0.92)Pre-eclampsiaHCG>2.0MoMTrimesterFirst1.77 (1.07–2.92)0.80 (0.60–1.07)0.37 (0.19–0.58)0.79 (0.77–0.81)Second2.45 (1.57–3.84)0.89 (0.83–0.96)0.19 (0.17–0.21)0.93 (0.93–0.93)(CI confidence intervals, AFP alpha feto-protein, HCG human chorionic gonadotrophin, BW birth weight, MoM multiple of median). Analyses according to gestation of testing of accuracy of biochemical screening to predict small for gestational age and pre-eclampsia (random effects pooling). (CI confidence intervals, HCG human chorionic gonadotrophin, BW birth weight, MoM multiple of median, PAPPA pregnancy associated plasma protein A)Sub-group and sensitivity analysisFor sub group analysis, a sub-group had to include at least three studies within each analyte and threshold and thus was only possible for calculated incidence of disease. The results for sub-group analysis are shown in Table 2. There was no significant difference between the subgroups.Table 2Subgroup analyses of accuracy of biochemical screening to predict small for gestational age and pre-eclampsia (random effects pooling)Small for gestational ageAnalyte SubgroupPositive Likelihood Ratio (95% CI)Negative Likelihood Ratio (95% CI)Sensitivity (95% CI)Specificity (95% CI)AFP>2.0MoM (BW<10th centile)Incidence>10%2.69 (1.36–5.31)0.98 (0.96–1.00)0.04 (0.02–0.08)0.98 (0.98–0.99)≤ 10%3.71 (2.66–5.16)0.93 (0.88–0.97)0.06 (0.05–0.07)0.98 (0.98–0.98)HCG>2.0MoM(BW<10th centile)Incidence>10%1.53 (1.1–2.12)0.89 (0.77–1.04)0.29 (0.22–0.37)0.79 (0.77–0.82)≤ 10%1.92 (1.72–2.13)0.95 (0.94–0.96)0.11 (0.1–0.12)0.94 (0.94–0.95)Pre-eclampsiaAFP>2.0MoMIncidence>4%0.85 (0.41–1.78)1.01 (0.97–1.06)0.06 (0.02–0.13)0.93 (0.91–0.94)≤ 4%2.98 (1.77–5.03)0.96 (0.95–0.97)0.08 (0.06–0.09)0.96 (0.96–0.96)HCG>2.0MoMIncidence>4%2.45 (0.65–2.93)0.68 (0.42–1.1)0.25 (0.21–0.3)0.92 (0.91–0.93)≤ 4%2.36 (1.81–3.08)0.89 (0.85–0.95)0.18 (0.16–0.2)0.93 (0.92–0.93)(CI confidence intervals, AFP alpha feto-protein, HCG human chorionic gonadotrophin, BW birth weight, MoM multiple of median)Most of the studies included in the review excluded fetuses with other structural or chromosomal anomalies from the results and included live births only thus subgroup analysis could not be performed in these areas. Sensitivity analysis including only those studies with these characteristics showed no significant difference. The same was true for the assessment of study quality i.e. most studies were of a similar quality to make sub-group analysis impossible but sensitivity analysis showed no difference when extremely low quality studies were excluded.Forest plots of sensitivity and specificity are shown in Additional file 6. Summary receiver operating characteristic curves are available from the authors on request.Clinical application with aspirinThe results for clinical application with aspirin for SGA are shown in Table 3 and for PET in Table 4. The results show that by testing with inhibin A for PET or SGA in a low risk population we can reduce the number of women needed to treat to prevent one case of SGA from 90 to 30 and for PET from 323 to 27, having to test 909 and 469 women respectively.Table 3Serum screening among pregnant women and number of women needed to be tested and treated with aspirin to prevent one case of SGA (birth weight < 10th centile).Test resultPrevalence SGA (%)Probability of SGA after testing positive (%)Risk of SGA after treatment*Probability of SGA after treatmentNNTest1NNTreat2No test, no treatment310.010.0-10.0--No test, treat all310.0-0.909.0-90Alpha feto-protein>2.0MoM: Sensitivity 60%; Specificity 98%Test all, treat test positives10.028.30.9025.416735Human chorionic gonadotrophin>2.0MoM: Sensitivity 12%; Specificity 94%Test all, treat test positives10.016.20.9014.683362Unconjugated estriol<0.75MoM: Sensitvity 37%; Specifcitiy 88%Test all, treat test positives10.022.00.9019.827045Pregnancy associated plasma protein A (PAPP-A)<1st centile: Sensitivity 3%; Specificity 99%Test all, treat test positives10.028.00.9025.2333336Inhibin A>2.0MoM: Sensitivity 11%; Specificity 98%.Test all, treat test positives10.033.10.9029.890930Alpha feto-protein>2.0MoM to predict severe FGR: Sensitivity 22%, Specificity 99%Test all, treat test positives1.022.00.9019.845445* RR 0.90 (95% CI 0.84–0.97) Askie et al. Antiplatelet agents for prevention of pre-eclampsia: meta-analysis of individual patient data. Lancet 2007;369:1791–98;11.1 NNTest is number needed to test and treat with aspirin to prevent one case of SGA calculated by 1/(proportion true positives (TP) – (proportion TP * RR)).2 NNTreat is number need to treat if only treat test positives with aspirin calculated by 1/(probability after testing positive – probability after treatment).3 Numbers are equal for all tests regardless of threshold, sensitivity and specificity.MoM multiples of medianSGA small for gestational ageTable 4Serum screening among pregnant women and number of women needed to be tested and treated with aspirin to prevent one case of PET.Test resultPrevalence PET (%)Probability of PET after testing positive (%)Risk of PET after treatment*Probability of PET after treatmentNNTest1NNTreat2No test, no treatment33.03.0-3.0--10.010.010.0No test, treat all33.0-0.92.8-32310.0-0.99.0-90Alpha feto-protein>2.0MoM: Sensitivity 7%; Specificity 96%Test all, treat test positives3.07.30.96.1476214710.026.20.918.6142948Human chorionic gonadotrophin>2.0MoM, second trimester: Sensitivity 19%; Specificity 93%Test all, treat test positives3.07.50.96.3175414210.027.20.919.352647Unconjugated estriol<0.5MoM: Sensitvity 6%; Specifcitiy 96%Test all, treat test positives3.04.60.94.0555622610.016.70.912.8166770Pregnancy associated plasma protein A (PAPP-A)<5th centile: Sensitivity 9%; Specificity 95%Test all, treat test positives3.06.50.95.5370416710.023.30.917.0111153Inhibin A>2.79MoM: Sensitivity 71%; Specificity 96%.Test all, treat test positives3.06.00.93.44692710.0216.90.961.214115* RR 0.90 (95% CI 0.81–1.01) Askie et al. Antiplatelet agents for prevention of pre-eclampsia: a meta-analysis of individual patient data. Lancet 2007;369:1791–9811.1 NNTest is number needed to test and treat with aspirin to prevent one case of FGR calculated by 1/(proportion true positives (TP) – (proportion TP * RR)).2 NNTreat is number need to treat if only treat test positives with aspirin calculated by 1/(probability after testing positive – probability after treatment).3 Numbers are equal for all tests regardless of threshold, sensitivity and specificity.MoM multiples of medianPET pre-eclampsiaDiscussionWe evaluated the accuracy of five serum screening markers used in Down's syndrome screening. The results showed low predictive accuracy overall. For PET the best predictor was inhibin A>2.79MoM. However, it is important to point out that this threshold was determined from a receiver operating characteristic curve and based on a single study. For SGA the best predictor overall for birth weight < 10th centile was AFP<10th centile while AFP>3.0MoM was the best predictor of birth weight < 5th centile. These results were both based on single studies. AFP and inhibin A showed improvements in predictive accuracy when looking at severe disease for SGA and PET respectively. HCG showed improved prediction when comparing second trimester to first trimester testing.The strength of our review and validity of its findings lies in the methodological strengths used. We complied with existing guidelines for the reporting of systematic reviews [18] and also guidelines specific to the reporting of systematic reviews of observational studies [142]. We performed extensive literature searches without language restrictions. We paid careful attention to assessment of quality of study design and reporting (The Quorum statement for this review is shown in Additional file 7).Previously published reviews in this area are restricted to a systematic review evaluating predictive tests for PET [143]. This review concluded that the tests investigated had a low predictive value, the methodology of this review has however been criticized [144] and was restricted in the thresholds and tests it reviewed. To our knowledge there are no previously reported systematic reviews in this area for SGA.We have primarily reported likelihood ratios in this review as they are thought to be more clinically meaningful than sensitivities and specificities, the use of likelihood ratios allowing us to determine post test probabilities of disease based on Bayes' theorem. Recent research suggests that independently pooled likelihood ratios should be interpreted with caution as positive and negative likelihood ratios are related statistics (just like sensitivity and specificity) [145]. We also pooled sensitivity and specificity and found no difference in the interpretation of the results. Bivariate meta-analysis is a new statistical technique that explicitly incorporates the correlation between sensitivity and specificity in a single model [146], its use is however not yet widespread nor is it easily interpreted.Our assessment of study quality was hindered by lack of clear reporting, which is a common problem in diagnostic reviews as standards for quality and checklists for assessing it are fairly new. It has been previously reported that poor study design and conduct can affect the estimates of diagnostic accuracy [28,29] however, it is not entirely clear how individual aspects of quality may effect this and to what magnitude particularly in the area of Obstetrics. Application of quality scores has been shown to be of little value on diagnostic reviews [147] however, due to the lack of clear reporting it was not possible to perform sub-group analysis based on individual quality criteria.One of the areas in which reporting was uniformly poor was in the details provided regarding performance of the reference standard. In PET definitions have changed over time with previous definitions including change increases in blood pressure. The measurement of blood pressure was poorly reported. It is important to record diastolic blood pressure with Korotkoff phase V as this is more reliably recorded and reflects true diastolic blood pressure [148-150]. For SGA there is still no convincing evidence as to which is the best definition of the condition at birth nor which is the best predictor of future infant and childhood morbidity and mortality for term infants. Population based birth weight standards were the most commonly used, however it is important to realize that these do not distinguish between the small healthy infant and the compromised infant. Customised growth charts that are adjusted for sex, gestation, parity, maternal weight and height and ethnicity, have been shown to improve the detection of infants at risk of stillbirth [151] while neonatal indices have been shown to identify the malnourished infant at risk of peripartum asphyxia [152]. Unfortunately these were rarely used as outcome measures in the included reviews.Confounding factors in measurement of serum screening markers but mainly AFP is its association with intrauterine death, preterm labour and chromosomal and structural anomalies [54,57,60]. Ideally all the included papers in this review should have included only women with live births and fetuses with no other chromosomal or structural anomalies, this however was not always clearly reported. Sensitivity analysis, including only studies that did report exclusion of these subjects showed no significant difference in estimates of test accuracy.In this review we have also assumed that the markers act independently but this may not be the case. The relationship between PET and SGA must also be taken into account. For HCG measurement the risk of SGA has been shown by logistic regression to be dependent on the presence of PET [99]. Ideally included cases of SGA for this review would have been those where there was no PET but this was again poorly reported.When assessing the clinical relevance of these tests it is important to look at severe disease as this causes the majority of maternal, fetal and neonatal complications and thus prediction and prevention of this form of disease would have the greatest health impact. For the studies included in the meta-analysis there were only three that had results for either severe PET or SGA and these were insufficient to make an accurate assessment of the prediction of this form of disease.The calculations of NNTreat and NNTest show that we can reduce the number of women needed to treat with aspirin to prevent one case of SGA/PET if we first test with a serum screening marker and then only treat the test positives. As aspirin is not routinely used as a treatment these calculations serve to contextualize the predictive value of these markers as individual tests. The costs of introducing aspirin as a treatment would need to be balanced against the costs of the test, costs of failing to treat the women with a false negative result that then go on to develop disease and any patient costs in terms of anxiety from screening and over treatment in the false positive category. To thus calculate the true clinical effectiveness of these tests these results would need to be incorporated in to a full cost-effectiveness analysis.As PET and SGA are diseases with relatively low prevalence a clinically useful test would need to have a high positive LR (> 10) and low negative LR (< 0.10) [153]. From the results of this review it is unlikely that any one serum screening marker in isolation will provide this. Future research should thus concentrate in two areas. The first should be to address the limitations within the primary literature as identified by this review; poor reporting, exclusion of intrauterine deaths and chromosomal and structural anomalies from the results, separation of PET and SGA, prediction of severe disease. This may not necessarily require further primary research as there are sufficient large, well designed cohort studies available but meta-analysis based on individual patient data. Secondly future research should focus on combinations of markers as predictors and combinations of tests such as serum screening markers and uterine artery Doppler [154] to improve the predictive accuracy to a clinically useful value.As Down's serum screening is routinely performed in many developed countries the cost of implementing use of these results as a predictive test for PET and SGA would be small. However as aspirin is the only preventative treatment with any proven benefit in these conditions and has minimal adverse events this cost has to be compared to that of implementing aspirin treatment to all pregnant women.ConclusionDown's serum screening analytes have low predictive accuracy for pre-eclampsia and small for gestational age. They may be a useful means of risk assessment or of use in prediction when combined with other tests.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsThe following authors were responsible for the study concept and design: RKM, JSC, GtR, BWM, JAMvdP, JK, KSK, SCR. The following authors were responsible for acquisition of data: RKM, JSC, ML, BWM.Analysis and interpretation of data was performed by the following authors: RKM, JSC, KSK. Drafting of the manuscript was performed by: RKM, JSC, ML, GtR, BWM, JAMvdP, KSK, JK, SCR. Statistical analysis was performed by: RKM, JSC, KSK. All authors read and approved the final manuscript. RKM had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.Pre-publication historyThe pre-publication history for this paper can be accessed here:Supplementary MaterialAdditional file 1\"Prediction of pre-eclampsia: a protocol for systematic reviews of test accuracy.\" Study protocol for pre-eclampsia systematic reviews.Click here for fileAdditional file 2\"The value of predicting restriction of fetal growth and compromise of its wellbeing: Systematic quantitative overviews (meta-analysis) of test accuracy literature.\" Study protocol for fetal growth restriction systematic reviews.Click here for fileAdditional file 3\"Search strategies for biochemical markers used in Down's serum screening to predict preeclampsia/small for gestational age.\" Electronic search strategies for systematic reviews.Click here for fileAdditional file 4\"Guide to QUADAS for Down's syndrome markers to predict pre-eclampsia/small for gestational age.\" Guide to quality assessment of included papers in review using QUADAS tool.Click here for fileAdditional file 5\"Study characteristics of studies of included studies for maternal serum biochemical (Down syndrome) screening to predict pre-eclampsia and small for gestational age.\"Click here for fileAdditional file 6\"Forest plots of sensitivity and specificity.\" Results of sensitivity and specificity displayed as Forest plots.Click here for fileAdditional file 7\"Improving the quality of reports of meta-analyses of randomised controlled trials: the Quorum Statement checklist.\" The Quorum statement checklist.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2533301.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533301",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533301\nAUTHORS: Eduardo Bernabé, Cesar M de Oliveira, Aubrey Sheiham\n\nABSTRACT:\nBackgroundAt present, there is no evidence on whether using condition-specific Oral Health-Related Quality of Life (OHRQoL) measures provides more reliable information than generic measures for needs assessment. Therefore, the objective was to assess the discriminative ability of one generic and one condition-specific OHRQoL measure, namely, respectively, the short form of the Oral Health Impact Profile (OHIP-14) and the Condition-Specific form of the Oral Impacts on Daily Performances (CS-OIDP) attributed to malocclusion, between adolescents with and without normative need for orthodontic treatment.Methods200 16–17-year-old adolescents were randomly selected from 957 schoolchildren attending a Sixth Form College in London, United Kingdom. The impact of their oral conditions on quality of life during the last 6 months was assessed using two OHRQoL measures; OHIP-14 and OIDP. Adolescents were also examined for normative orthodontic treatment need using the Index of Orthodontic Treatment Need (IOTN) and the Dental Aesthetic Index (DAI). Discriminative ability was assessed comparing the overall scores and prevalence of oral impacts, calculated using each OHRQoL measure, between adolescents with and without normative need. Using the prevalence of oral impacts allowed adjusting for covariates.ResultsThere were significant differences in overall scores for CS-OIDP attributed to malocclusion between adolescents with and without normative need for orthodontic treatment when IOTN or DAI were used to define need (p = 0.029 or 0.011 respectively), and in overall scores for OHIP-14 when DAI, but not IOTN was used to define need (p = 0.029 and 0.080 respectively). For the prevalence of impacts, only the prevalence of CS-OIDP attributed to malocclusion differed significantly between adolescents with and without normative need, even after adjusting for covariates (p = 0.017 and 0.049 using IOTN and DAI to define need).ConclusionCS-OIDP attributed to malocclusion was better able than OHIP-14 to discriminate between adolescents with and without normative needs for orthodontic treatment.\n\nBODY:\nBackgroundOral Health-Related Quality of Life (OHRQoL) can be assessed using either generic or specific measures [1,2]. Generic OHRQoL measures take into account numerous oral conditions, some occurring simultaneously, and thus collect information about wider effects of oral health on daily living. The main advantage of generic measures is that they allow comparison of various domains of quality of life for the condition being studied, as well as across populations and disease states [3-6]. One of the most commonly used generic OHRQoL measures is the two versions of Oral Health Impact Profile (OHIP); with 49 or 14 items [7,8]. On the other hand, specific OHRQoL measures focus on a particular disease, condition, symptom, function or population and thus are used when any of the aforementioned specific attributes needs to be assessed [1,4,5]. Condition-specific instruments are the most commonly used specific OHRQoL measures [1], probably because they provide more information on consequences of a specific untreated oral condition or disease and the corresponding benefits of its treatment [3,6]. The Oral Impacts on Daily Performances (OIDP) is the only OHRQoL measure designed to link specific oral conditions, such as malocclusion, and impacts on quality of life [9,10].It has been claimed that condition-specific OHRQoL measures may increase acceptability to subjects by including only relevant dimensions [1,3,6]. In addition, their specific focus makes them potentially more sensitive to small, but clinically important changes in oral health [1,4,5]. This may in turn increase responsiveness [1,3], which is particularly important when assessing oral health needs. Knowing whether there is an impact of the mouth on quality of life does not necessarily provide information on what specific dental condition was related to the impact. Condition-specific OHRQoL measures attempt to provide such information by attributing oral impacts to specific oral conditions, therefore indicating which conditions may require dental attention [11]. In this sense, the condition-specific form of the OIDP index (CS-OIDP) is an integral part of the socio-dental approach for oral health needs assessment [12-14].Although using condition-specific OHRQoL measures for needs assessment seems theoretically sound, some recent studies have also assessed oral health needs using generic OHRQoL measures [15,16]. Empirical evidence may cast light on whether using condition-specific OHRQoL measures provides more reliable information than generic measures. To do that, both types of OHRQoL measures must be evaluated first in terms of their ability to differentiate between groups differing in health statuses. Such an evaluation is part of construct validity assessment [4,17-19]. There is no evidence on whether generic or condition-specific OHRQoL measures are more appropriate for assessing dental needs. Therefore, the objective of this study was to assess the discriminative ability of one generic and one condition-specific OHRQoL measure, namely, respectively, the short form of the Oral Health Impact Profile (OHIP-14) and the Condition-Specific form of the Oral Impacts on Daily Performances (CS-OIDP) attributed to malocclusion, between adolescents with and without normative need for orthodontic treatment.MethodsPopulation and settingTwo hundred 16–17-year-old adolescents were randomly selected from a list containing the names of all the 957 schoolchildren attending the Havering Sixth Form College in London, United Kingdom during 2006. All the students selected agreed to take part in the study. Sample size was calculated to estimate a prevalence of 25% for the condition-specific oral impacts on daily performances attributed to malocclusion, with a maximum tolerable error of 5% [20].The Local Ethics Committee and the Research and Development Directorate of the University College London Hospitals National Health Service Trust approved this study. Participants signed a consent letter agreeing for their participation in the study.Data collectionFirst, information about demographic characteristics (sex, age and ethnicity), orthodontic treatment status and the impact of oral conditions on quality of life during the last 6 months was self-reported by the participants. Information about oral impacts was collected using OHIP-14 and OIDP. Adolescents self-completed OHIP-14 in their classrooms and were later interviewed individually with OIDP in a private room. The OHIP-14, which has been previously validated on British populations [21,22], assesses the frequency of problems associated with the mouth, teeth or dentures on 7 dimensions: functional limitation, physical pain, psychological discomfort, physical disability, psychological disability, social disability and handicap. Adolescents were asked to rate each of the 14 items on a 5-point ordinal scale coded 0 'never', 1 'hardly ever', 2 'occasionally', 3 'fairly often' and 4 'very often'. The overall score for OHIP-14 was obtained summing up all responses, thus ranging between 0 and 56 points [8,23]. The OIDP index, which has also been validated on British populations [21,24], assesses serious oral impacts on 8 daily performances namely, eating, speaking, cleaning mouth, relaxing, smiling, studying, emotion and social contact. If an adolescent reported an impact on any of the 8 performances, the frequency of the impact (scale from 1 to 3) and the severity of its effect on daily life (scale from 1 to 3) were scored. If no impact was reported, then a zero score was assigned. Thereafter, adolescents were asked to identify from a list those oral problems that, in their opinion, caused the impact. Only those Condition-Specific Oral Impacts on Daily Performances related to 'bad position of teeth', 'space between teeth', and 'deformity of mouth or face', were considered in the analysis as CS-OIDP attributed to malocclusion [14,25]. Performance scores were estimated by multiplying the corresponding frequency and severity scores. The overall score for the CS-OIDP attributed to malocclusion was the sum of the 8 performance scores (ranging from 0 to 72), multiplied by 100 and divided by 72 [9,10].Adolescents were then examined for normative orthodontic treatment need using both components of the Index of Orthodontic Treatment Need (IOTN) as well as the Dental Aesthetic Index (DAI). Both indexes have gained international acceptance because they are valid, reliable and easy to use [26-28]. For the Dental Health Component (DHC) of IOTN, 10 traits of malocclusion were assessed: overjet, reverse overjet, overbite, openbite, crossbite, crowding, impeded eruption, defects of cleft lip and palate as well as any craniofacial anomaly, Class II and Class III buccal occlusions, and hypodontia. Only the highest scoring trait is used to assess treatment need [29]. Thereafter, adolescents self-rated their dental attractiveness on the 10-point scale of the Aesthetic Component (AC) of IOTN [29,30]. Results from DHC and AC of IOTN were merged into a single classification according to the current General Dental Services regulations of the National Health Services in United Kingdom [31,32]. According to these regulations, orthodontic care can only be provided for individuals who have a DHC grade of 4 or 5, or grade 3 with an AC of 6 or above. All other cases were therefore classified as having no need. For DAI, 10 occlusal traits were assessed and a score was obtained using the equation: 6×(missing visible teeth) + crowding + spacing + 3×(diastema) + largest anterior maxillary irregularity + largest anterior mandibular irregularity + 2×(anterior maxillary overjet) + 4×(anterior mandibular overjet) + 4×(vertical anterior openbite) + 3×(anteroposterior molar relation) + 13 [33,34]. Each adolescent was then classified as having no need (score < 28) or need (score ≥ 28) [27]. Examinations were carried out by one of the authors (CMO), who had been previously trained and calibrated in the Department of Orthodontics at University of Cardiff where the IOTN was developed. According to weighted Kappa, inter- and intra-examiner reliability were 0.77 and 0.91 respectively.Data analysisDiscriminative ability was examined in terms of construct validity whereby the distributions of scores for both OHRQoL measures are compared between groups with different levels of oral health [18]. Since overall scores for OHIP-14 and CS-OIDP attributed to malocclusion were not normally distributed (Shapiro-Wilks test, p < 0.001 in all cases), Mann-Whitney tests were used to compare both overall scores between adolescents with and without normative need for orthodontic treatment. To aid comparison and interpretation, the magnitude of differences was also expressed as an effect size [35,36], which was calculated as the mean difference between groups divided by the pooled standard deviation. The widely accepted thresholds of 0.2, 0.5 and 0.8 were used to define 'small', 'moderate' and 'large' effect sizes [35].As the aforementioned method did not allow adjusting for covariates (sex, age, ethnicity and orthodontic treatment status), the prevalence of oral impacts was also compared between adolescents with and without normative need for orthodontic treatment. For that, the prevalence of oral impacts was calculated as the percentage of adolescents reporting one or more items 'fairly often' or 'very often' for OHIP-14 [23] and as the percentage of adolescents with a score higher than zero for CS-OIDP attributed to malocclusion [10]. Then, the prevalence of oral impacts was compared between adolescents with and with normative need using Poisson regression with robust estimation of variance while adjusting for covariates [37,38].ResultsThis study included 134 (67.0%) females and 66 (33.0%) males, 116 (58.0%) were aged 16 years and 84 (42.0%) aged 17 years; 170 were Caucasian (85%) and 30 (15.0%) were of other ethnic origins. One third (32.5%) had completed orthodontic treatment, 12.5% were currently undergoing orthodontic treatment and the remaining 55.0% were untreated. Based on the two measures of orthodontic need, 42 (21.0%) had a normative need for orthodontic treatment according to IOTN whereas 25 (12.5%) had a normative need using DAI.There were significant differences in the overall scores for CS-OIDP attributed to malocclusion between adolescents with and without normative need for orthodontic treatment when IOTN or DAI were used to define need (p = 0.029 or 0.011 respectively), and in the overall scores for OHIP-14 when DAI, but not IOTN was used to define need (p = 0.029 and 0.080 respectively). Using DAI, the mean difference in overall scores for OHIP-14 and CS-OIDP attributed to malocclusion between adolescents with and without normative need was 1.64 points (CI95%: -0.84; 4.12) and 2.13% (CI95%: 0.44; 3.81) respectively. The corresponding size effects for such mean differences in overall scores were 0.28 (CI95%: -0.14; 0.70) and 0.53 (CI95%: 0.11; 0.95) respectively (Table 1). Using IOTN, the mean difference in overall score for CS-OIDP attributed to malocclusion between adolescents with and without normative need was 1.35% (CI95%: -0.03; 2.72) and its corresponding size effect was 0.33 (CI95%: -0.01; 0.68).Table 1Comparison of the overall score for OHIP-14 and CS-OIDP attributed to malocclusion between adolescents with and without normative need for orthodontic treatment.OHRQoL measureNormative neednMeanSDp value*Effect size95% CI for effect sizeOHIP-14No need by IOTN1585.136.000.0800.13(-0.21; 0.47)(0–56 points)Need by IOTN425.885.49No need by DAI1755.085.950.0290.28(-0.14; 0.70)Need by DAI256.725.37CS-OIDPNo need by IOTN1581.133.630.0290.33(-0.01; 0.68)(0–100%)Need by IOTN422.485.25No need by DAI1751.153.670.0110.53(0.11; 0.95)Need by DAI253.285.84* Mann-Whitney test was used.In addition, there were significant differences in the prevalence of oral impacts between adolescents with and without normative need for orthodontic treatment only for CS-OIDP attributed to malocclusion (p = 0.032 and 0.049 respectively), but not for OHIP-14 (p = 0.799 and 0.211 respectively). This finding was independent of the index used to define normative need for orthodontic treatment (Table 2). After adjusting for covariates, adolescents with normative need for orthodontic treatment had respectively an 1.89 (CI95%: 1.12; 3.20) and 1.84-fold (CI95%: 1.00; 3.39) increase in the chance of reporting CS-OIDP attributed to malocclusion, compared to adolescents without normative need, when the IOTN and DAI were used to define need.Table 2Comparison of the prevalence of oral impacts, by OHIP-14 and CS-OIDP attributed to malocclusion, between adolescents with and without normative need for orthodontic treatment.PrevalenceOHRQoL measureNormative needn%PR*95% CI for PRp valueOHIP-14No need by IOTN2113.31.00Need by IOTN614.31.12(0.50; 2.49)0.790No need by DAI2112.01.00Need by DAI624.01.64(0.76; 3.55)0.211CS-OIDPNo need by IOTN2918.41.00Need by IOTN1433.31.89(1.12; 3.20)0.017No need by DAI3318.91.00Need by DAI1040.01.84(1.00; 3.39)0.049* Poisson regression was used to calculate prevalence ratios (PR) adjusted for sex, age, ethnicity and orthodontic treatment status.DiscussionThis study evaluated two widely used OHRQoL measures, OHIP-14 and CS-OIDP attributed to malocclusion, in terms of their ability to discriminate adolescents with, from those without normative need for orthodontic treatment. This was the first attempt to assess the discriminative ability of both OHRQoL measures.When overall scores for both OHRQoL measures were used to assess the impacts of oral conditions on everyday life, adolescents with normative need for orthodontic treatment always reported significantly higher OHRQoL scores than adolescents without normative need, except for the OHIP-14 overall score when IOTN was used to define need. One explanation for this finding relates to sample size. As this study was based on secondary analysis of a prevalence study [20], no evaluation of the statistical power for comparison purposes could be done. Though, it must be noted that the group with normative need was smaller when DAI than when IOTN was used to define need (25 versus 42 adolescents), and that there were group differences even with that smaller DAI sample. An alternative explanation may relate to well-known differences between DAI and IOTN [26,39,40]. With IOTN only the worst occlusal trait is recorded, which is not necessarily related to the participant's oral impact. In other words, occlusal traits that affect dental appearance and have an impact on participants' daily lives may not be captured by IOTN. In addition, DAI has many more measures of malocclusion affecting the anterior teeth than the IOTN. For example, DAI includes number of missing visible teeth, crowding in the incisal segments, spacing in the incisal segment, and measurement of any midline diastema that are not specifically addressed by IOTN. However, such differences could not explain why CS-OIDP attributed to malocclusion, but not OHIP-14 differentiated adequately between adolescents with and without normative need as defined by both indexes. Therefore, this finding indicates that the expected more sensitive, condition-specific OHRQoL measure better discriminated between adolescents with and without normative need for orthodontic treatment than the generic OHRQoL measure.Furthermore, when effect sizes were used to interpret the magnitude of mean differences in scores between adolescents with and without normative need for orthodontic treatment, better results were found for CS-OIDP attributed to malocclusion than for OHIP-14. Effect size for CS-OIDP attributed to malocclusion was moderate whereas effect size for OHIP-14 was nil when DAI was used to define normative need for orthodontic treatment.When the prevalence of oral impacts, calculated by each OHRQoL measure, was used to assess the impacts of oral conditions on everyday life, differences between adolescents with and without normative need for orthodontic treatment were found for CS-OIDP attributed to malocclusion but not for OHIP-14. This was independent of whether DAI or IOTN was used to define need. Generally, adolescents with normative need for orthodontic treatment had slightly more than four-fifth increase in the probability of reporting CS-OIDP attributed to malocclusions after controlling for the effects of covariates (sex, age, ethnicity and orthodontic treatment status). The comparison of prevalences between groups with different oral health statuses has been reported for other OHRQoL measures [41-43]. Unquestionably, this was an advantage over using mean differences because there is no way to control for covariates with non-parametric tests such as the Mann-Whitney test.Overall, different findings were found when comparing the discriminative ability of OHIP-14 and CS-OIDP attributed to malocclusion between groups with and without normative need for orthodontic treatment. These findings differed according to the indicator used to assess the impacts of oral conditions on participants' quality of life (the overall score or the prevalence of oral impacts) or the index used to define normative need for orthodontic treatment (IOTN or DAI). However, based on the present findings it appears that CS-OIDP attributed to malocclusion was better able than OHIP-14 to differentiate between the two groups of adolescents based on needs. Therefore, the present findings confirmed our earlier assumption that the condition-specific OHRQoL measures were better able to discriminate between sub-groups with different levels of oral health than their generic counterparts. This also provides empirical support for using condition-specific OHRQoL measures for oral health needs assessment.Our findings agree with the few previous studies comparing generic and condition-specific OHRQoL measures [42-44]. They showed that both OHRQoL measures are complementary, rather than alternative sources of information. Although this holds true for situations in which researchers are interested in assessing not only the overall profile of oral impacts but also those impacts on quality of life related to specific oral conditions, the present findings raise the important question, does using a generic or a condition-specific OHRQoL measure provide additional information for oral health needs assessment when the specific link between a specific oral condition leading to impacts on quality of life is required to prioritise need for professional attention? The findings from this study suggest that a condition-specific OHRQoL measure should be used in such situations. However, since these findings were based on distinguishing between adolescents with and without a specific type of normative need, they need further confirmation for other oral health needs.ConclusionAmong a population of 16–17-year-old British adolescents, the CS-OIDP attributed to malocclusion was better able than the more generic OHIP-14 to discriminate between different levels of normative need for orthodontic treatment. Findings differed according to the indicator used to assess the impacts of oral conditions on participants' quality of life (overall score or prevalence of oral impacts) or the index used to define normative need for orthodontic treatment (IOTN or DAI).Competing interestsThe authors declare that they have no competing interests.Authors' contributionsEB conceived of the study, performed statistical analysis and drafted the first version of the manuscript. CMdO organized and conducted the study, and has critically revised the manuscript. AS supervised the entire study and critically revised the manuscript. All authors read and approved the final version of the manuscript.\n\nREFERENCES:\nNo References"
+}

batch_0/PMC2533328.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533328",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533328\nAUTHORS: Nina Stoletzki\n\nABSTRACT:\nBackgroundEukaryotic mRNAs often contain secondary structures in their untranslated regions that are involved in expression regulation. Whether secondary structures in the protein coding regions are of functional importance remains unclear: laboratory studies suggest stable secondary structures within the protein coding sequence interfere with translation, while several bioinformatic studies indicate stable mRNA structures are more frequent than expected.ResultsIn contrast to several studies testing for unexpected structural stabilities, I directly compare the selective constraint of sites that differ in their structural importance. I.e. for each nucleotide, I identify whether it is paired with another nucleotide, or unpaired, in the predicted secondary structure. I assume paired sites are more important for the predicted secondary structure than unpaired sites. I look at protein coding yeast sequences and use optimal codons and synonymous substitutions to test for structural constraints. As expected under selection for secondary structures, paired sites experience higher constraint than unpaired sites, i.e. significantly lower numbers of conserved optimal codons and consistently lower numbers of synonymous substitutions. This is true for structures predicted by different algorithms.ConclusionThe results of this study are consistent with purifying selection on mRNA secondary structures in yeast protein coding sequences and suggest their biological importance. One should be aware, however, that accuracy of structure prediction is unknown for mRNAs and interrelated selective forces may contribute as well. Note that if selection pressures alternative to translational selection affect synonymous (and optimal) codon use, this may lead to under- or over-estimates of selective strength on optimal codon use depending on strength and direction of translational selection.\n\nBODY:\nBackgroundMessenger RNA (mRNA) sequences encode the amino acid sequence of the protein but may also bear additional information. For example, certain synonymous codons may improve translation [1-3] and a variety of motifs may regulate expression at the level of translation, cellular localization, decay or splicing [4-9]. Many of these motifs are secondary structures, and eukaryotic mRNAs contain regulatory structures in their 5' and 3' UTRs [10-15], or introns [16,17]. However, it remains unclear whether secondary structures in the coding regions are of functional importance. Laboratory studies suggest that local secondary structures within coding regions can interfere with translation [18,19], and one may therefore expect selection against structures that are too stable. Surprisingly, however, several bioinformatic studies find that RNA structures within the protein coding regions are more stable than expected by chance [20-23] (but see [24] for opposing result). These studies used various algorithms to predict the secondary structures of mRNA sequences, and then compared the free energy values of these structures to the values for randomized sequences.Here, I test for selection on mRNA secondary structure using another approach. Instead of testing for unexpected structural stabilities, I directly compare the selective constraint of sites that differ in their importance for the predicted secondary structure. I.e. I predict the secondary structure of coding yeast sequences using different algorithms, and for each nucleotide, I identify whether it is paired with another nucleotide, or unpaired. I assume paired sites are more important for the predicted secondary structure than unpaired sites. If there is selection for secondary structures, one might expect higher structural constraint at paired than at unpaired sites. Such constraint would affect synonymous codon use and substitution rates. In S. cerevisiae a relationship between codon use, tRNA abundance and expression level indicates that codon use is affected by selection for translationally optimal codons [1]. If there is selection for mRNA structure, structurally important sites may be under conflicting selection pressures: a codon might support the preferred mRNA structure that is translationally non-optimal. Under structural selection, one might expect lower numbers of optimal codons at paired than at unpaired sites. If mRNA structure is conserved across species, one might further expect lower numbers of synonymous substitutions at paired than at unpaired sites; possible compensatory substitutions however may make the latter test predictions less clear-cut. When a mutation occurs at a paired site and disrupts the pairing ability, a second compensatory mutation on the corresponding paired site may restore the pairing ability [25,26]. Compensatory mutations may increase substitution numbers at paired sites. Innan and Stephan [27] show however, that unless selection against deleterious intermediates is very small, substitutions should occur only very slowly in paired regions [27].Accurate structure prediction is obviously crucial for these tests. In several studies [20-22], mRNA structures are predicted by thermodynamic properties using the minimum free energy (MFE) algorithm [28] only although taking the whole ensemble of possible structures and comparative information into account is known to increase predictive accuracy [29-32]. I therefore predict the secondary structures by thermodynamic and comparative information (RNA- and ALIfold [33]), using the minimum free energy (MFE) algorithm and McCaskill's partition function of the thermodynamic equilibrium [34].Results of this study are consistent with selection upon mRNA structures: numbers of conserved optimal codons and synonymous substitutions are reduced at structurally important sites.MethodsChoice of study organism & dataI focus on Saccharomyces cerevisiae, as this model eukaryote is well studied, with genome sequences available for it and several related species. Importantly yeast allows using optimal codon numbers to investigate alternative selective constraints while controlling for effects of base composition. This is because (i) translational selection has been investigated extensively and supported in yeast [1-3]: certain translationally \"optimal\" codons increase in frequency with expression level and correspond to the most abundant tRNAs in the cell or to the tRNA with which they form the strongest binding. (ii) Crucially, translationally optimal codons in yeast are not biased towards GC-ending codons, as in many other Eukaryotic organisms. In yeast 12 optimal codons end with G or C (-GC), 12 with A or T (-AT). To control for base composition is important as RNA secondary structure predictions are – at least partly- based on thermodynamic properties and will therefore be affected by GC content: GC nucleotides form the most stable binding with three hydrogen bonds and will consequently more likely be paired in the structure. From the yeast alignments provided by Kellis et al. [35] comparing Saccharomyces cerevisiae with S. paradoxus, S. mikatae and S. bayanus, I use 492 genes that have start and stop codons but no premature stop codons or frame-shifting indels in all four species.Secondary structureI predict the secondary structure of the coding sequences using the below methods and identify for each nucleotide whether it is paired with another nucleotide, or unpaired. I assume paired sites are more important for the predicted secondary structure than unpaired sites. Note however, that unpaired sites may well be important for maintaining the mRNA's tertiary structure.Secondary structure prediction methodsThe thermodynamic stability of a secondary structure is measured as the amount of free energy released or used by forming base pairs. Positive free energy requires work to form a structure, negative free energy releases stored work. Free energy parameters are estimated from chemical melting experiments. The widely used Minimum Free Energy (MFE) algorithm [28] computes the one single structure with the most negative energy value, that thermodynamically is hence the most likely to be formed. The MFE algorithm seems fairly accurate for short RNA sequences, for which ~73% of paired sites are accurately predicted. mRNAs however are likely to be present in a population of structures [36,37]. Often 5–10% of structures share very similar free energy values [38], and the predicted MFE structure might just be one out of many thermodynamically similar structures. Taking all possible secondary structures of the thermodynamic equilibrium into account, McCaskill's algorithm [34] computes the most probable structure and calculates the probability that each site is paired. When taking base pairings with high probabilities, the accuracy of the prediction increases [29]. Another benefit of McCaskill's algorithm is that it is less affected by small but reasonable variations in the underlying energy parameters – while the MFE prediction is very sensitive [39,40]. I used the RNAfold (Vienna RNA Secondary Structure[33,41]) package to predict structures of the four yeasts separately using the MFE and McCaskill's algorithms. When using McCaskill's algorithm, I consider sites to be paired that pair with high probability (>2/3) across the structure ensemble; all other sites are considered as unpaired. With increasing sequence lengths predictive accuracy decreases presumably because of the enormous increase in the number of potential base pairings that can be made as sequence length increases [42]. I therefore look at both the complete set of genes, and at the subset of genes shorter than 800 bp.To predict the secondary structure, one can also assume structural conservation, and compute the one consensus structure that allows the largest amount of structural conservation across homologous sequences. Especially supportive of structural conservation are sites that vary at the sequence level but retain potential of Watson-Crick pairings in the structure (co-variations). Structures predicted with the aid of comparative data appear to be more accurate than those based on thermodynamic properties alone [30-32]. I use the ALIfold package [33,43] that integrates comparative information in the prediction made with either MFE or McCaskill's algorithm and predict the consensus structures of the four yeasts together using the ALIfold default settings for co-variation weight (Φ1 = 1, and Φ2 = 1).Optimal codon useCodon identification is based on the S. cerevisiae sequence. Optimal codons are defined as in Kliman et al. (2003) [44]. The relative frequency of optimal codons (Fop[45]) is the ratio of optimal codons to synonymous codons. I compute the relative frequency of optimal codons for each amino acid and gene separately. For amino acids with both one AT- as well as one GC-ending optimal codon (thr, val, ile, ser), I compute the relative optimal codon frequencies of the two optimal codons per amino acid separately. Throughout the paper, the terms \"optimal\" and \"suboptimal\" will refer to translational selection.TestsIf there is selection for secondary structures, one may expect higher constraint at structurally important (paired) than at structurally less important (unpaired) sites.(1) Under translational selection one may expect lower numbers of translationally optimal codons at paired compared to unpaired sites. Note that the analysis is restricted to those codons that are conserved across the four yeast species and are likely to experience stronger selection pressures. Restricting the analysis to conserved sites is crucial for the ALIfold measure, as it incorporates substitutions in its prediction: ALIfold may tend to pair conserved sites, and under translational selection conserved sites tend to have higher optimal codon use than non-optimal sites. This could generate an artificial positive correlation between optimal codon numbers and structure when considering all codons. As GC-ending optimal codons are more likely to be paired, I look at GC- and AT-ending optimal codons separately. I do this for the four yeast species separately (using RNAfold) as well as for their consensus structure (using ALIfold), using MFE as well as McCaskill's algorithm for both methods.(2) If mRNA structures are conserved across species one may further expect lower numbers of substitutions at paired compared to unpaired sites. As ALIfold incorporates comparative information, this test is only meaningful for structures predicted by RNAfold. Codons experiencing non-synonymous substitutions are excluded from this analysis as one may expect possible selection on mRNA structure will mainly affect synonymous substitutions, while non-synonymous substitutions will be more constrained for other reasons. To check the structural similarity and potential conservation of predicted structures across species, I first compute the relative number of base pairings per gene that are consistently, i.e. unambiguously, predicted to be paired or unpaired across species. To estimate structural constraint at synonymous sites, I count for each synonymous optimal and non-optimal codon how often the respective third codon position is paired and unpaired in the S. cerevisiae structure (RNAfold) and how often the codon is conserved or experiences a synonymous substitution compared to S. paravensis. Note that translational selection and structural selection may be counter-balancing with respect to synonymous substitution numbers. I.e. unpaired sites with high numbers of optimal codons may experience reduced synonymous substitution numbers due to translational selection while paired sites with high numbers of non-optimal codons may experience reduced synonymous substitution numbers due to structural selection. To disentangle structural selection from translational selection, I look at optimal and non-optimal codons separately as. I further look at GC- and AT-ending codons separately as mutational processes and gene conversion events may be compositionally biased [46].StatisticsEach of our analyses generates a set of 2 × 2 contingency tables per gene and per amino acid or codon. These are divided according to whether the site is paired or unpaired in the predicted secondary structure, and whether (1) the codon is optimal or non-optimal, and whether (2) the codon is conserved or synonymous polymorphic across the four species. To combine these independent 2 × 2 tables, I use the Mantel-Haenszel Z statistic according to Sokal and Rohlf [47]. I compute joint probabilities for all tables or certain subsets. To disentangle an effect of GC content on synonymous codon use at paired sites, I combine amino acids with AT-ending ending and amino acids with GC-ending optimal codons. I exclude contingency tables when expected values were zero, tested for homogeneity and computed the joint odds ratio (WMH) and its significance, including the continuity correction. I orient the odds ratio such that selection in favour of mRNA secondary structure is indicated by WMH <1: i.e. lower numbers of optimal codons, and lower numbers of synonymous substitutions at paired sites.Results1) Conserved optimal codon numbers are significantly lower at paired compared to unpaired sites, irrespective of the method (RNA- and ALIfold) and algorithm (MFE and McCaskill's) used to predict secondary structure. Crucially, the tendency remains whether I consider amino acids with AT- or GC-ending optimal codons (Tables 1, 2).Table 1Comparison of conserved optimal codon numbers at paired and unpaired sites.MethodAlgorithmALLGC-endingAT-endingGC leu & lysGenes shorter than 800 bpRNAfold S. cerevisiaeMFE0.646 ***0.624 **0.503 ***1.353 ***Mc0.542 ***0.567 ***0.422 ***0.910 ***RNAfold S. paravensisMFE0.667 ***0.608 ***0.542 NS1.137 ***Mc0.560 ***0.571 ***0.407 **1.119 ***RNAfold S. mikitaeMFE0.653 ***0.590 ***0.544 *1.172 ***Mc0.572 ***0.623 ***0.411 ***1.131 ***RNAfold S. bayanusMFE0.640 ***0.597 **0.502 ***1.181 ***Mc0.537 **0.557 ***0.401 ***1.003 ***ALIfoldMFE0.638 ***0.577 ***0.465 NS1.499 ***Mc0.468 **0.444 ***0.326 ***0.997 ***All genesRNAfold S. cerevisiaeMFE0.920 ***0.863 ***0.751 ***1.584 ***Mc0.841 ***0.866 ***0.676 ***1.436 ***RNAfold S. paravensisMFE0.878 ***0.826 ***0.733 ***1.497 ***Mc0.819 ***0.822 ***0.659 ***1.460 ***RNAfold S. mikitaeMFE0.912 ***0.814 ***0.790 NS1.160 ***Mc0.839 ***0.869 ***0.740 NS1.404 ***RNAfold S. bayanusMFE0.904 ***0.855 ***0.742 ***1.590 ***Mc0.833 ***0.840 ***0.675 ***1.455 ***ALIfoldMFE0.899 ***0.759 **0.739 NS1.937 ***Mc0.770 ***0.724 ***0.645 ***1.529 ***I combine contingency tables for all amino acids and genes (ALL) and subsets of amino acids with GC- and AT- ending optimal codons (leu and lys are treated separately, as these two GC-ending amino acids behave very opposing, see below Table 2). Mantel Haenzsel estimators and significances are presented, WMH <1 = lower optimal codon use at paired than at unpaired sites.* < 0.05, ** < 0.01, *** < 0.005, NS = not significant. Structure prediction is based on ALIfold and RNAfold using MFE and McCaskill's (Mc) algorithm.Table 2Comparison of conserved optimal codon numbers at paired and unpaired sites.RNAfold (S. cerevisiae)ALIfoldMFEMcMFEMcAmino acids with GC-ending optimal codonsLeuTTG1.380 ***1.273 ***1.811 ***1.456 ***LysAAG1.772 ***1.654 ***2.019 ***1.654 ***PheTTC1.075 ***1.120 ***0.786 NS0.864 **TyrTAC0.809 NS0.720 NS0.741 NS0.631 NSHisCAC0.707 NS0.600 NS0.568 NS0.657 *AspGAC0.813 NS0.838 NS0.656 ***0.746 **AsnAAC0.891 *0.831 NS0.725 **0.637 NSAmino acids with one GC- and one AT-ending optimal codonIleATC1.318 ***1.254 ***1.555 ***1.049 ***IleATT1.272 ***0.978 ***1.561 ***1.009 ***ValGTC0.585 NS0.852 ***0.734 ***0.560 ***ValGTT0.705 *0.794 NS0.806 *0.686 NSThrACC0.838 **0.921 ***0.811 ***0.750 ***ThrACT0.940 ***0.876 ***0.932 ***0.655 ***SerTCC0.614 NS0.668 NS0.583 NS0.591 NSSerTCT0.794 NS0.816 NS0.816 ***0.985 ***Amino acids with AT-ending optimal codonsAlaGCT0.968 **0.904 ***1.073 ***0.883 ***ArgAGA,CGT0.588 ***0.545 ***0.518 ***0.460 ***GlyGGT0.894 NS0.779 NS0.901 ***0.752 NSGlnCAA0.442 ***0.353 ***0.293 ***0.278 ***GluGAA0.946 ***0.350 ***0.386 ***0.315 ***ProCCA0.851 NS0.734 NS0.729 **0.689 *CysTGT0.500 NS0.382 NS0.755 ***0.403 NSSeparately for each amino acid, I combine contingency tables of the different genes. Mantel Haenzsel estimators and significances are presented, with WMH <1 = lower optimal codon use at paired than at unpaired sites. Structure prediction is based on ALIfold and RNAfold using MFE and McCaskill's (Mc) algorithm.* < 0.05, ** < 0.01, *** < 0.005, NS = not significantThe tendency is true for most amino acids separately; even GC-ending optimal codons that are more likely to be paired for thermodynamic reasons tend to be less frequent at paired sites (Table 2). Notable exceptions however are leu, lys, ile, and for RNAfold additionally phe (Table 2). One explanation for these exceptions may be that selection strength for translationally optimal codons is stronger in these amino acids, for example translational errors may be more likely or more costly. Considering prediction accuracy may decrease with gene length, I first restricted the data to genes shorter than 800 bp; including all genes however does not change the result.(2) I first check the similarity and potential conservation of structures predicted by RNAfold. The major parts of mRNAs do not seem conserved in structure across species or prediction accuracy is low: 75% of sites are ambiguous, i.e. predicted to be paired in one or more species, but predicted to be unpaired in the remaining species (Table 3). When looking pairwise on average 41% of sites are ambiguous; number of ambiguous sites is only slightly lower for short genes. The ambiguity of predicted structural status will introduce considerable noise and may cause non-significant results. Despite high ambiguity in structure prediction the numbers of synonymous substitutions are consistently lower (WMH<1) at paired sites (Table 4). The tendencies remain when restricting the data to genes shorter than 800 bp. Other species comparisons lead to similar results (data not presented). Results become significant for GC-ending (optimal and non-optimal) codons (when structure is predicted using McCaskill's algorithm). G and C nucleotides do not only form stronger bonds and are more likely to be paired and structurally important, they are also more likely to be unambiguously predicted paired than A and T nucleotides (means GC: 0.263, AT: 0.153, t = 29.8409, df = 866.05, ***). This could reduce the level of noise and cause the significance of results for GC-ending codons.Table 3Similarity of predicted structures for species pairs.Species comparisonPrediction Method(P+U)/allP/allU/allGenes shorter than 800 bpAcross all 4 yeastsMFE27% ± 0.617% ± 0.310% ± 0.5Mc27% ± 0.49% ± 0.218% ± 0.3S. cerevisiae – S. paravensisMFE63% ± 0.636% ± 0.327% ± 0.3Mc64% ± 0.522% ± 0.442% ± 0.5S. cerevisiae – S. mikitaeMFE59% ± 0.334% ± 0.225% ± 0.2Mc61% ± 0.420% ± 0.441% ± 0.6S. cerevisiae – S. bayanusMFE42% ± 0.423% ± 0.519% ± 0.1Mc59% ± 0.320% ± 0.040% ± 0.5All genesAcross all 4 yeastsMFE25% ± 0.38% ± 0.116% ± 0.2Mc25% ± 0.416% ± 0.29% ± 0.1S. cerevisiae – S. paravensisMFE61% ± 0.436% ± 0.225% ± 0.2Mc63% ± 0.322% ± 0.341% ± 0.4S. cerevisiae – S. mikitaeMFE58% ± 0.235% ± 0.223% ± 0.1Mc60% ± 0.320% ± 0.340% ± 0.4S. cerevisiae – S. bayanusMFE57% ± 0.234% ± 0.123% ± 0.1Mc58% ± 0.220% ± 0.239% ± 0.4The average percentages of sites (± variances) unambiguously predicted to be paired (P/all) and/or unpaired ((P+U)/all, U/all) for the respective species comparison using RNAfold MFE and McCaskill's (Mc) algorithm are presented.Table 4Comparison of synonymous substitution numbers at paired and unpaired sites.ATOptATNoptATGCOptGCNoptGCAll(1)MFE0.511NS0.542NS0.600NS0.488NS0.296NS0.414NS0.481NSMc0.567NS0.528NS0.546NS0.544**0.327*0.458**0.505NS(2)MFE0.255NS0.201NS0.232NS0.140NS0.257NS0.213NS0.225NSMc0.255NS0.230NS0.238NS0.264NS0.167NS0.222NS0.232NSLooking at S. cerevisiae and S. paravensis , I compare numbers of each codon in S. cerevisiae being either synonymous non-conserved or conserved at paired or unpaired sites. Structure prediction is based on RNAfold upon the S. cerevisiae sequence using MFE and McCaskill's (Mc) algorithm. Mantel Haenzsel estimators and significances are presented. WMH<1 = lower numbers of synonymous substitutions at paired sites. (1) All genes, (2) Genes that are shorter than 800 bp.* < 0.05, ** < 0.01, *** < 0.005, NS = not significantDiscussionI tested for evidence of selective constraint acting on mRNA secondary structures in protein coding yeast genes. Predicted secondary structures differ greatly according to the prediction method used and between species. Nevertheless, there are significantly fewer conserved optimal codons and consistently fewer synonymous substitutions at paired sites for all predicted secondary structures. The results of this study are consistent with purifying selection on mRNA secondary structures.Similar tendencies of codon use have been reported for Drosophila and humans: mRNA stability seems high when optimal codon use is low in Drosophila [48] and paired sites contain an excess of rare codons in humans [49]. Note that in this study, the comparison of optimal codon use is restricted to conserved sites. Besides the methodological need for ALIfold (see Material and Methods), the restriction to conserved sites restricts the analysis to sites potentially under considerable strong selection. For RNAfold structures, results become non-significant when not restricting the data to these conserved sites (data not presented). Strong conflicting selection pressures seem to act on certain sites while the remaining sites seem less constrained for structure. Selection on local and not global structures may explain these results and contribute to the low structural similarity across species. Selection on local mRNA structures in coding regions of eukaryotic genes has been suggested before [49]. Beside the low structural similarity also compensatory substitutions may contribute to the non-significant results when comparing substitution numbers at paired and unpaired sites.Previous bioinformatic studies that focussed on whether or not the thermodynamic stability of mRNA structures of various organisms is selected for or against [20-23,49,55] lead to partly inconsistent results and controversies about the accurate randomization procedure. In these studies, the observed MFE is compared to the expected MFE, which is estimated by taking the mean MFE of randomized versions of the same sequence, and a significant deviation is taken as evidence for selection for or against thermodynamic stability of the structure. The randomization of sequences can be performed in a number of different ways holding various properties of the sequence constant, while randomizing others. The properties are of biological importance; variables that are affected by forces other than selection for mRNA structure – for example the amino acid sequence – should be fixed. Which variables should remain free to vary however may not always be obvious, while the results are very sensitive to them. Di-nucleotide content for example might be selected for its effect on stability and should be allowed to vary for randomized sequences argue Chamary and Hurst [22]. However, di-nucleotides might well be affected by mutation bias, or selected for some other reason [21], in which case, di-nucleotide content should be kept fixed. The control of di-nucleotides in fact renders significant results non-significant [20-23,55].Note that in contrast to comparing observed and expected MFE values, the comparison of constraint at paired and unpaired sites does not indicate that selection acts for or against the thermodynamic stability of the structure, but that the very predicted structure is under selection. With respect to selection for or against stability of structures, ALIfold results indicate that the thermodynamically most stable global structure is not conserved across the four yeasts: ALIfold consensus energy value is much higher i.e. less stable compared to the average energy value of the single sequences [see also Washietl et al. [50] for approach]. This is conform with results of Babak et al. [24] which support selection against stability of structures in coding regions. It is reasonable to expect selection on mRNA structures may act against too stable structures because too stable and un-flexible mRNA structures may interfere for instance with translation [18] and some mRNAs flexibility may allow their specific and dynamic complexes with other factors. mRNAs lead a complex life [51] and besides thermodynamic stability, selection on mRNA structure may also exist to maintain specific local or global mRNA structures that allow binding and interaction with other factors and thus effect biological functioning. Not only structural targets may be of effect, also accessibility of sequence targets may depend on global or local mRNA structures.While results of this study are consistent with selection upon mRNA structures in coding regions and support laboratory studies that report synonymous substitutions are functionally important with respect to mRNA structure and translation in humans [52-54], two considerations should be made. First, we do not know whether thermodynamic mRNA structure predictions predict the mRNA structures that are formed in the cell. mRNAs are generally associated with other factors [51], and effects of mRNA-associated microRNAs and proteins on the structure are hard to predict. Also, kinetics of mRNA folding and pseudo-knots are not considered here. Even with the comparative method, mRNA structures may remain at best approximations of the real mRNA structures in the cell. Secondly, the predicted and also the real structure will be affected by certain DNA patterns – however whether or not the respective DNA patterns are selected for mRNA structure or another reason may be hard to judge. There are several DNA patterns one may consider. (i) Di-nucleotide content of naturally occurring sequences leads to higher than expected thermodynamic stability [e.g. [21,23,55]]. Di-nucleotide content may be selected for its effect on mRNA structure but it may also be affected by mutation bias, or selected for some other reason, for example for nucleosome positioning [56-58] or transcription pause sites [59]. (ii) Frequency of polypurine tracts is increased in exons and may affect thermodynamic structure. Again, polypurine tracts may be selected with respect to mRNA structure but also for other reasons such as enhancing splicing [60]. (iii) Translational protein folding into alpha-helix and beta sheets may affect synonymous codon use [61] and periodic DNA patterns may affect mRNA structure. If thermodynamic predictions correspond to any other force such as selection on nucleosome positioning and transcription or co-translational pause sites, the observed patterns may be a consequence of that and inference of selection acting directly upon on mRNA structure may be incorrect.Alternative selection upon mRNA structures (or any other selective target) may counterbalance translational selection and explain why the bias towards translationally optimal codon is never complete and even in highly expressed genes non-optimal codons are used. Alternative selection may also contribute to the discrepancy between expected and observed codon bias [58], and may lead to systematic underestimates of selection strength for optimal codons. As selection for mRNA structures may be acting stronger on GC-ending codons, in organisms in which potential translationally optimal codons are biased towards GC, such as Drosophila, mammals, C. elegans, estimates of selective strength for optimal codons may also be overestimated. It will be worth considering effects of alternative selection and disentangling the different targets of selection.ConclusionI tested for evidence of selective constraint acting on mRNA secondary structures in yeast. Predicted structures differ greatly according to the prediction method used and between species. Nevertheless, there are significantly fewer conserved optimal codons and consistently fewer synonymous substitutions at paired sites for all predicted secondary structures. These results are consistent with purifying selection on mRNA secondary structures in protein coding yeast sequences and suggest their biological importance. One should consider however that accuracy of structure prediction is unknown for mRNAs and interrelated selective forces may contribute. Selective pressures alternative to translational selection seem affect synonymous and optimal codon use in yeast. Depending on strength and direction of translational selection in an organism, such alternative selective forces may lead to under- or over-estimates of selective strength on optimal codon use.\n\nREFERENCES:\n1. IkemuraTReview Codon Usage and tRNA Content in Unicellular and Multicellular OrganismsMol Biol Evol1985213343916708\n2. KanayaSYamadaYKinouchiMKudoYIkemuraTCodon Usage and tRNA genes in eukarotes: correlation of codon usage diversity with translational efficiency and with CG-dinucleotide usage as assessed by multivariate analysisJ Mol Evol2001534–529029810.1007/s00239001021911675589\n3. PercudaniRPavesiAOttonelloSTransfer RNA gene redundancy and translational selection in Saccharomyces cerevisiaeJ Mol Biol199726832233010.1006/jmbi.1997.09429159473\n4. BakerKEParkerRNonsense-mediated mRNA decay: terminating erroneous gene expressionCur Opinion Cell Biol20041629329910.1016/j.ceb.2004.03.003\n5. BurattiEBaralleFEInfluence of RNA secondary structure on the Pre-mRNA splicing processMol Cell Biol200424105051051410.1128/MCB.24.24.10505-10514.200415572659\n6. GonsalvezGBUrbinatiCRLongRMRNA localization in yeast: moving towards a mechanismBiol Cell200597758610.1042/BC2004006615601259\n7. KozakMReview Regulation of translation via mRNA structure in prokaryotes and eukaryotesGene2005361133710.1016/j.gene.2005.06.03716213112\n8. StJohnstonDMoving messages: the intracellular localization of mRNAsNature Reviews Molecular Cell Biology2005636337510.1038/nrm1643\n9. TekotteHDavisIIntracellular mRNA localization: motors move messagesTrends in Genetics20021863664210.1016/S0168-9525(02)02819-612446149\n10. BabendureJBabendureJLDingJ-HTsienRYControl of mammalian translation by mRNA structure near capsRNA20061211110.1261/rna.230990616373489\n11. PelletierJSonnenbergNThe involvment of mRNA secondary structure in protein synthesisBiochem Cell Biol1987655765813322328\n12. BateyRTStructures of regulatory elements in mRNAsCur Opinion Struct Biol20061629930610.1016/j.sbi.2006.05.001\n13. RingnerMKroghMFolding Free Energies of 5'-UTRs Impact Post-Transcriptional Regulation on a Genomic Scale in YeastPLOS Comp Biol20051, 7e72\n14. MignoneFGissiCLIuniSPesoleGUntranslated regions of mRNAsGenome Biol20023Review S000410.1186/gb-2002-3-3-reviews0004\n15. ParschJStephanWTandaSA Highly Conserved Sequence in the 3'-Untranslated Region of the Drosophila Adh Gene Plays a Functional Role in Adh ExpressionGenetics19991516676749927459\n16. KirbyDAMuseSVStephanWMaintenance of pre-mRNA secondary structure by epistatic selectionProc Natl Acad Sci USA1995929047905110.1073/pnas.92.20.90477568070\n17. ChenYStephanWCompensatory evolution of a precursor messenger RNA secondary structure in the Drosophila melanogaster Adh geneProc Natl Acad Sci USA2003100114991150410.1073/pnas.193283410012972637\n18. DomaMKParkerREndonucleotic cleavage of eukaryotic mRNAs with stalls in translation elongationNature200644056156410.1038/nature0453016554824\n19. LiebhaberSACashFEshlemanSSTranslation inhibition by an mRNA coding region secondary structure is determined by its proximity to the AUG initiation codonJ Mol Biol199222660962110.1016/0022-2836(92)90619-U1507219\n20. SeffensWDigbyDmRNAs have greater negative folding free energies than shuffled or codon choice randomized sequencesNucleic Acids Res1999271578158410.1093/nar/27.7.157810075987\n21. KatzLBurgeCBWidespread selection for Local RNA secondary Structure in Coding Regions of Bacterial GenesGenome Research2003132042205110.1101/gr.125750312952875\n22. ChamaryJVHurstLDEvidence for selection on synonymous mutations affecting stability of mRNA secondary structure in mammalsGenome Biol20056R7510.1186/gb-2005-6-9-r7516168082\n23. ShabalinaSAOgurtsovAYSpiridonovNAA periodic pattern of mRNA secondary structure created by the genetic codeNucleic Acids Res2006342428243710.1093/nar/gkl28716682450\n24. BabakTBlencoweBJHughesTRConsiderations in the identification of functional RNA structural elements in genomic alignmentsBMC Bioinformatics200783310.1186/1471-2105-8-3317263882\n25. KimuraMThe role of compensatory neutral mutations in molecular evolutionJ Genet19856471910.1007/BF02923549\n26. ChenYCarliniDBBainesJFParschJBravermanJTandaSStephanWRNA secondary structure and compensatory evolutionGenes Genet Syst19997427128610.1266/ggs.74.27110791023\n27. InnanHStephanWSelection Intensity Against Deleterious Mutations in RNA Secondary Structures and Rate of Compensatory Nucleotide SubstitutionsGenetics200115938939911560913\n28. ZukerMStieglerPOptimal computer folding of large RNA sequences using thermodynamic and auxiliary informationNucleic Acids Res1981913314810.1093/nar/9.1.1336163133\n29. MathewsDHUsing an RNA secondary structure partition function to determine confidence in base pairs predicted by free energy minimizationRNA20041081178119010.1261/rna.765090415272118\n30. GutellRRJCLeeJJCannoneThe accuracy of ribosomal RNA comparative structure modelsCur Opinion Struct Biol20021230131010.1016/S0959-440X(02)00339-1\n31. GardnerPPGriegerichRA comprehensive comparison of comparative RNA structure prediction methodsBMC Bioinformatics2004514015810.1186/1471-2105-5-14015458580\n32. MathewsDHTurnerDHPrediction of RNA secondary structure by free energy minimizationCur Opinion Struct Biol200626027027810.1016/j.sbi.2006.05.010\n33. HofackerILVienna RNA secondary structure serverNucleic Acids Res2003313429343110.1093/nar/gkg59912824340\n34. McCaskillJSThe Equilibrium Partition Function and Base Pair Binding Probabilities for RNA Secondary StructureBiopolymers1990291105111910.1002/bip.3602906211695107\n35. KellisMPattersonNEndrizziMLanderESSequencing and comparison of yeast species to identify genes and regulatory elementsNature200342324125410.1038/nature0164412748633\n36. ChristoffersonREMcSwiegenJAKoningsDApplication of computational technologies to ribozyme biotechnology productsJ Mol Struct THEOCHEM1994311273284\n37. BettsLSpremulliLLAnalysis of the role of the Shine-Dalgarno sequence and mRNA secondary structure on the efficiency of translational initiation in the Euglena gracilis chloroplast atpH mRNAJ Biol Chem199426926456264637929367\n38. EddySRHow do RNA folding algorithms work?Nature Biotechnology2004221457145810.1038/nbt1104-1457\n39. ZukerMCalculating nucleic acid secondary structureCur Opinion Struct Biol20001030331010.1016/S0959-440X(00)00088-9\n40. LaytonDMBundschuhRA statistical analysis of RNA folding algorithms through thermodynamic parameter perturbationNucleic Acids Res20053351952410.1093/nar/gkh98315673712\n41. HofackerILFontanaWStadlerPFBonhoefferLSTackerMSchusterPFast folding and comparison of RNA secondary structuresMonatshefte Chemie199412516718810.1007/BF00818163\n42. MathewsDHSabinaJZukerMTurnerDHExpanded sequence dependence of thermodynamic parameters provides robust prediction of RNA secondary structureJ Mol Biol199928891194010.1006/jmbi.1999.270010329189\n43. HofackerILFaketeMStadlerPFSecondary Structure Prediction for Aligned RNA SequencesJ Mol Biol200231910596610.1016/S0022-2836(02)00308-X12079347\n44. KlimanRMNaheelahISantiagoMSelection conflicts, gene expression, and codon usage trends in yeastJ Mol Evol2003579810910.1007/s00239-003-2459-912962310\n45. IkemuraTCorrelation between the abundance of Escherichia coli transfer RNAs and the occurrence of the respective codons in its protein genes: a proposal for a synonymous codon choice that is optimal for the E. coli translational systemJ Mol Biol198115138940910.1016/0022-2836(81)90003-66175758\n46. BirdsellJAIntegrating Genomics, Bioinformatics, and Classical Genetics to Study the Effects of Recombination on Genome EvolutionMol Biol Evol2002191181119712082137\n47. SokalRRRohlfJFBiometry19953Freeman and Company. New York\n48. CarliniDBChenYStephanWThe relationship between third-codon position nucleotide content, codon bias, mRNA secondary structure and gene expression in the Drosophilid Alcohol Dehydrogenase genes Adh and AdhrGenetics200115962363311606539\n49. MeyerIMMiklosIStatistical evidence for conserved local secondary structure in the coding regions of eukaryotic mRNAs and pre-mRNAsNucleic Acids Res2005336338634810.1093/nar/gki92316275783\n50. WashietlSHofackerILStadlerPFFast and reliable prediction of non-coding RNAsProc Natl Acad Sci20051072454245910.1073/pnas.040916910215665081\n51. MooreMJFrom Birth to Death: The complex lifes of eukaryotic mRNAsScience2005309\n52. ShenLXBasilionJPStantonVPJrSingle-nucleotide polymorphisms can cause different structural folds of mRNAProc Natl Acad Sci1999967871787610.1073/pnas.96.14.787110393914\n53. DuanJWainwrightMSComeronJMSaitouNSandersARGelernterJGejmanPVSynonymous mutation in the human dopamine receptor D2 (DrD2) affect mRNA stability and synthesis of the receptorHum Mol Genetics20031220521610.1093/hmg/ddg055\n54. NackleyAGShabalinaSATichilevaIESatterfieldKKorchynskyiOMakarovSSMaixnerWDiatchenkoLHuman Catechol-O-Methyltransferase Haplotypes modulate protein expression by altering secondary structureScience20063141930193310.1126/science.113126217185601\n55. WorkmanCKroghANo evidence that mRNAs have lower free energies than random sequences with the same di-nucleotide distributionNucleic Acids Res1999274816482210.1093/nar/27.24.481610572183\n56. SegalEFondufe-MittendorfYChenLThastromACFieldYMooreIKWangJ-PZWidomJA genomic code for nucleosome positioningNature200644277277810.1038/nature0497916862119\n57. IoshikhesIPAlbertIZantonSJPughFNucleosome positions predicted through comparative genomicsNat Genet2006381104510.1038/ng187817006463\n58. IoshikhesIBolshoyADerenshteynKBorodovskyMTrifonovENNucleosome DNA Sequence Pattern Revealed by Multiple Alignment of Experimentally Mapped SequencesJ Mol Biol199626212913910.1006/jmbi.1996.05038831784\n59. HerbertKMLa PortaAWongBJMooneyRANeumanKCLandickRBlockSMSequence-Resolved Detection of Pausing by Single RNA Polymerase MoleculesCell20061251083109410.1016/j.cell.2006.04.03216777599\n60. PazAMesterDNevoEKarolALooking for Organization Patterns of Highly Extressed Genes: Purine-Pyrimidine Composition of Precursor mRNAsJ Mol Evol20076424826010.1007/s00239-006-0135-617211550\n61. Kimichi-SarfatyCOhMMKimI-WSaunaZECalcangnoAMAmbudkarSVA \"Silent\" Polymorphism in the MDR1 Gene changes substrate specificityScience200731552552810.1126/science.113530817185560\n62. BulmerMThe selection-mutation-drift theory of synonymous codon usageGenetics19911298979071752426"
+}